All Patches for PowerMAX 4.3 are:
1553v5drv-001 cnd-008 egl-002 inet-011 nsu-004
1553v5drv-002 crosslibs-001 egl-003 inet-012 nsu-005
1553v5drv-003 crosslibs-002 egl-004 inet-013 nsu-006
1553v5drv-004 crosslibs-003 egl-005 inet-014 oam-001
1553v5lib-001 crosslibs-004 egl-006 ip-001 pg-001
1553v5lib-002 crosslibs-005 fbs-001 ip-002 pg-002
base-001 crosslibs-006 fbs-002 ip-003 pg-003
base-002 crosslibs-007 fbs-003 ip-004 pg-004
base-003 crypt-001 fbs-004 ip-005 pg-005
base-004 crypt-002 fbs-005 ip-006 pg-006
base-005 crypt-003 fbs-006 is-001 pg-007
base-006 crypt-004 fbs-007 ise-001 pg-008
base-007 crypt-005 fbsman-001 kdb-001 rmxf-001
base-008 crypt-006 fbsman-002 librt-001 rpc-001
base-009 crypt-007 fbsman-003 man-001 rpc-002
base-010 crypt-008 fibre-001 man-002 rpc-003
base-011 crypt-int-001 fibre-002 man-003 rpc-005
base-012 crypt-int-002 fibre-003 man-004 rvsrv-001
base-013 crypt-int-003 fibre-004 man-005 softint-001
base-014 crypt-int-004 fibre-005 man-006 softint-002
base-015 crypt-int-005 gf-001 man-007 softint-003
base-016 crypt-int-006 gpib-001 man-008 softint-004
cld-001 crypt-int-007 hsde-001 man-009 trace-001
cmds-001 dec-001 hsde-002 man-010 trace-002
cmds-002 dec-002 ide-001 man-011 trace-003
cmds-003 dec-003 ide-002 man-012 trace-004
cmds-004 dec-004 ide-003 man-013 trace-005
cmds-005 dec-005 ie-001 mvc-001 trace-006
cmds-006 dec-006 ie-002 mvc-002 trace-007
cmds-007 dec-007 ie-003 mvc-003 trace-008
cmds-008 dec-008 ie-004 ncr-001 trace-009
cmds-009 diskless-001 ie-005 ncr-002 via-001
cmds-010 diskless-002 inet-001 ncr-003 via-002
cmds-011 diskless-003 inet-002 netcmds-001 via-003
cmds-012 diskless-004 inet-003 nfs-001 via-004
cnd-001 diskless-005 inet-004 nfs-002 vmet-001
cnd-002 diskless-006 inet-005 nfs-003 vp-001
cnd-003 diskless-007 inet-006 nfs-004 xfsd-001
cnd-004 diskless-008 inet-007 ngpib-001 xfsd-002
cnd-005 diskless-009 inet-008 nsu-001
cnd-006 diskless-010 inet-009 nsu-002
cnd-007 egl-001 inet-010 nsu-003
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: 1553v5drv-001
Date Issued: 02/02/2000 13:21:42
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: 1553v5lib-001
Related SARs: #HM12681
Brief Description:
PowerMAX OS 4.3 1553v5drv package release updates
##############################################################################
Problem Description:
1. HM12681: The SBS ABI1553V5 driver was not allowing the use of 32 RTs
(remote terminals) and 31 SAs (subaddresses).
Problem Resolution:
1. For one channel board, do not split the on-board memory anymore, so
the channel can use all the memory. For two-channel board, divide the
on-board memory in such a way that the first channel will have enough
memory to define all 32 RTs.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/1553v5drv/abi_type.h
/usr/lib/lib1553v5drv.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: 1553v5drv-001
Date Issued: 02/02/2000 11:30:50
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: 1553v5lib-001
Related SARs: #HM12681
Brief Description:
PowerMAX OS 4.3 1553v5drv package release updates
##############################################################################
Problem Description:
1. HM12681: The SBS ABI1553V5 driver was not allowing the use of 32 RTs
(remote terminals) and 31 SAs (subaddresses).
Problem Resolution:
1. For one channel board, do not split the on-board memory anymore, so
the channel can use all the memory. For two-channel board, divide the
on-board memory in such a way that the first channel will have enough
memory to define all 32 RTs.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/1553v5drv/abi_type.h
/usr/lib/lib1553v5drv.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: 1553v5drv-002
Date Issued: 09/21/2001 16:44:37
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #533
Defect Tickets: none
Brief Description:
PowerMAX_OS 4.3 1553v5drv package release updates
Problem Description:
1. When running the same application, the SBS 1553 controller
"Rev R" board would hang intermittantly in the middle of a BCRT
transfer whereas the "Rev P" board would not. The bus had to be
reset after each hang.
The problem would show up when using the Multi-function Display
Unit (MFD) as the RT...not when simulating the MFD as an RT on
a second board device.
Problem Resolution:
1. Upgraded the 1553 firmware from f008c.dat and f009c.dat to
f008f.dat and f009f.dat, respectively.
Enhancements:
None
Object(s) To Be Replaced:
/usr/src/drivers/1553v5drv/firmware/f008f.dat
/usr/src/drivers/1553v5drv/firmware/f009f.dat
/usr/src/drivers/1553v5drv/firmware/firmware.cfg
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: 1553v5drv-002
Date Issued: 09/28/2001 15:26:51
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: #533
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 1553v5drv package release updates
Problem Description:
1. When running the same application, the SBS 1553 controller
"Rev R" board would hang intermittantly in the middle of a BCRT
transfer whereas the "Rev P" board would not. The bus had to be
reset after each hang.
The problem would show up when using the Multi-function Display Unit
(MFD) as the RT...not when simulating the MFD as an RT on a second
board device.
Problem Resolution:
1. Upgraded the 1553 firmware from f008c.dat and f009c.dat to
f008f.dat and f009f.dat, respectively.
Enhancements:
None
Object(s) To Be Replaced:
/usr/src/drivers/1553v5drv/firmware/f008f.dat
/usr/src/drivers/1553v5drv/firmware/f009f.dat
/usr/src/drivers/1553v5drv/firmware/firmware.cfg
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: 1553v5drv-003
Date Issued: 07/10/2002 14:43:32
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #644
Brief Description:
PowerMAX OS 4.3 1553v5drv package release updates
##############################################################################
P1: SAR #644: The 1553v5drv(7) manpage gave incorrect information with
respect to the VME address to supply to the abiconfig utility for
PowerHawk systems other than the Model 610.
R1: The 1553v5drv(7) manpage is modified to give the correct information
with respect to the VME address to supply to the abiconfig utility for
PowerHawk systems other than the Model 610.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat7/1553v5drv.7.z
/usr/share/man/man7/1553v5drv.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: 1553v5drv-003
Date Issued: 07/24/2002 14:35:44
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #644
Brief Description:
PowerMAX OS 4.3 1553v5drv package release updates
###############################################################################
P1: SAR #644: The 1553v5drv(7) manpage gave incorrect information with
respect to the VME address to supply to the abiconfig utility for
PowerHawk systems other than the Model 610.
R1: The 1553v5drv(7) manpage is modified to give the correct information
with respect to the VME address to supply to the abiconfig utility for
PowerHawk systems other than the Model 610.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat7/1553v5drv.7.z
/usr/share/man/man7/1553v5drv.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: 1553v5drv-004
Date Issued: 01/07/2004 15:56:01
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 1553v5drv package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1549: The C++ compiler produced many warnings when compiling
the 1553V5 example programs, while the C compiler did not. Working
executables were produced.
R1: Added -w, supress warnings, to the CFLAGS variable in the
examples Makefile.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/1553v5drv/abi.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: 1553v5drv-004
Date Issued: 01/07/2004 15:56:22
Software Package: 1553v5drv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 1553v5drv package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1549: The C++ compiler produced many warnings when compiling
the 1553V5 example programs, while the C compiler did not. Working
executables were produced.
R1: Added -w, supress warnings, to the CFLAGS variable in the
examples Makefile.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/1553v5drv/abi.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: 1553v5lib-001
Date Issued: 02/02/2000 13:22:31
Software Package: 1553v5lib pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: 1553v5drv-001
Related SARs: #25, HM12681
Brief Description:
PowerMAX OS 4.3 1553v5lib package release updates
##############################################################################
Problem Description:
1. #25: The 48bit, 1 micro second resolution timer of CE1092 1553
Interface board was cleared (reset) every 10 - 20 msec.
2. HM12681: The SBS ABI1553V5 driver was not allowing the use of 32 RTs
(remote terminals) and 31 SAs (subaddresses).
Problem Resolution:
1. Revised the SBS library code to allow user application program to
clear the timer.
2. For one channel boards, do not split the on-board memory anymore, so
the channel can use all the memory. For two-channel boards, divide
the on-board memory in such a way that the first channel will have
enough memory to define all 32 RTs.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/lib1553v5lib.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: 1553v5lib-001
Date Issued: 02/02/2000 11:31:58
Software Package: 1553v5lib pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: 1553v5drv-001
Related SARs: #25, HM12681
Brief Description:
PowerMAX OS 4.3 1553v5lib package release updates
##############################################################################
Problem Description:
1. #25: The 48bit, 1 micro second resolution timer of CE1092 1553
Interface board was cleared (reset) every 10 - 20 msec.
2. HM12681: The SBS ABI1553V5 driver was not allowing the use of 32 RTs
(remote terminals) and 31 SAs (subaddresses).
Problem Resolution:
1. Revised the SBS library code to allow user application program to
clear the timer.
2. For one channel boards, do not split the on-board memory anymore, so
the channel can use all the memory. For two-channel boards, divide
the on-board memory in such a way that the first channel will have
enough memory to define all 32 RTs.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/lib1553v5lib.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: 1553v5lib-002
Date Issued: 01/07/2004 15:56:11
Software Package: 1553v5lib pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description:
PowerMAX OS 4.3 1553v5lib package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1549: The C++ compiler produced many warnings when compiling
the 1553V5 example programs, while the C compiler did not. Working
executables were produced.
R1: Added -w, supress warnings, to the CFLAGS variable in the
examples Makefile.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/src/drivers/1553v5lib/examples/Makefile
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: 1553v5lib-002
Date Issued: 01/07/2004 15:56:34
Software Package: 1553v5lib pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 1553v5lib package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1549: The C++ compiler produced many warnings when compiling
the 1553V5 example programs, while the C compiler did not. Working
executables were produced.
R1: Added -w, supress warnings, to the CFLAGS variable in the
examples Makefile.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/src/drivers/1553v5lib/examples/Makefile
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-001
Date Issued: 09/15/1999 10:41:57
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: diskless-001, trace-001, ip (4.3)
Related SARs: HM12641, HM12685, HM12673
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
PowerMAX OS 4.3 Patch Set 1
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P1" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 1".
New PowerMAX OS 4.3 Software Patch(es):
base-001 Base System Patch 001
cmds-001 Advanced Commands Patch 001
cnd-001 Condor Ethernet Driver Patch 001
diskless-001 Diskless Systems Package Patch 001
fbs-001 Frequency Based Scheduler Patch 001
man-001">fbsman-001 Frequency Based Scheduler On-line Manual Pages Patch 001
fibre-001 Fibre Channel Driver Patch 001
ide-001 Internal IDE/ATA Disk Controller Patch 001
inet-001 Internet Utilities Patch 001
kdb-001 Kernel Debugger Patch 001
man-001 On-line Manual Pages Patch 001
nsu-001 Network Support Utilities Patch 001
trace-001 KernelTrace Utilities Patch 001
New PowerMAX OS 4.3 Software Package(s):
ip Interphase 4511 PMC FDDI Driver
Note: The "ip" package is currently only supported on the Power Hawk 620.
##############################################################################
Problem Description:
1. HM12641: Setting the PATH environment variable to NULL prior to
invocation of the idbuild(1M) or idtune(1M) utilities will result in
failure of those utilities.
2. The config(1M) utility aborts when searching for a kernel module
using the "Find by Keyword" operation.
3. HM12685: The config(1M) utility aborts when displaying all tunables
using the "All Tunables" operation.
4. HM12673: The IPC maximum number of semaphores per id is a hard-coded
value. This should be modifiable as a system tunable.
5. The metrics calculations done in the kma and kmadbg modules during
allocations and deallocations of kernel virtual space were not
protected by locks or atomic operations. Updates of these counters
could be lost if the same counter were being updated on different
CPUs or the update was interrupted by a higher priority function that
also requests an allocation or deallocation. This could cause the
data returned by sar(1M) (`sar -k`) to be inaccurate.
6. The debug version of kernel space allocator was calling cmn_err()
with the kmacorrupt_lock locked. cmn_err() was called to add an
informational message to the syslog that it was turning the detection
of "use-after-free" problems back on after it had previously been
turned off upon reaching the limits specified by the tunables
KMEM_LIMIT_UPPER and KMEM_LIMIT_LOWER. Because cmn_err() also calls
the allocator, the system would hang forever trying to take the same
lock again.
7. strip(1) removed the symbol table of shared libraries.
8. pthread_self(3pthread), pthread_cleanup_push(3pthread), and
pthread_cleanup_pop(3pthread) were not visible to c++ standard users.
9. There were instances when linking c++ programs where the .bss section
would end up with an odd number size.
10. Some snmp trap definitions were missing from <snmp/snmp_trap.h>.
11. The syscx(INT_VECTOR, ..) service would always return failure
condition on Power Hawk 640's even when passed a valid interrupt
vector.
The result is that it would be unable to determine which CPU an
interrupt vector is assigned to. It would also result in a failure
when using the mpadvise(MPA_CPU_INTVEC) library routine.
12. Small timing windows in RCIM (real-time clock and interrupt module)
driver. Evident when doing substantial simultaneous RCIM ETI
activity.
13. When ktrace(1) enables timing using RCIM clock, it could have a bad
effect on various system timing routines leading to failures.
14. Minor problems in clock_synchronize(1M) when using "interactive" mode.
15. On MTX604-070 systems, the PowerStack II system with 7 PCI slots, PCI
slots 0, 1, and 2 are functional but PCI slots 3, 4, 5, and 6 are not.
Problem Resolution:
1. The idbuild(1M) and idtune(1M) utilities were modified to define and
export the minimum PATH needed to execute successfully upon
invocation, returning the PATH to its original setting upon
completion.
2. Made config(1M) more robust in its handling of ill-formed module
descriptions and tunable descriptions.
3. Increased the number of tunables that config can display using
"All Tunables" operation. Also, fixed so that utility doesn't abort
when this tunable limit is reached.
4. Added system tunable SEMMSL.
5. The macros used by the kma and kmadbg modules to increment and
decrement the metrics counters were not protected by locks or by the
use of the atomic increment/decrement functions. The macros used by
these modules in <sys/metrics.h> were modified to use the atomic
increment and decrement functions.
6. kmem_alloc_debug() was changed to unlock the kmacorrupt_lock before
it calls cmn_err().
7. Fixed strip so that it does not remove the symbol table when
stripping a shared library.
8. Fixed header file <pthread.h> so that pthread_self,
pthread_cleanup_push, pthread_cleanup_pop are also visible to c++
standard users.
9. Fixed 'ld' so that in such occurences .bss is always rounded to next
16 bytes.
10. Added additional snmp trap definitions to <snmp/snmp_trap.h>.
11. Corrected Power Hawk 640 version of the system service that returns
CPU assignment for an interrupt vector.
12. Closed timing windows in RCIM driver.
13. Fixed so that RCIM clock is only used for trace purposes and not
other internal kernel timing functions.
14. Corrected clock_synchronize(1M) utility.
15. The kernel was modified to support the interrupt map defined by the
MTX604-070's Raven MPIC.
Enhancements:
1. Added new attributes and tags to DWARF information used to provide
debug information for applications:
a) Added new attribute for phase3.2 of the MAXAda compiler:
DW_AT_child_unit
b) Added new tag for phase4 of the MAXAda compiler:
DW_TAG_inst_package_param
c) Added new attributes to provide additional information about
inline calls in both the MAXAda and c++ compilers.
NightView will use this information to provide "virtual
walkback" information. The attributes are:
DW_AT_call_column
DW_AT_call_file
DW_AT_call_line
2. Added definition to <sys/adapter_pci.h> for ADAPTER_IP, which is
required to support the "ip" software package in PowerMAX OS 4.3.
The "ip" driver supports the Interphase PMC/PCI based 4511/5511 FDDI
adapters. The "ip" package is currently only supported on the
Power Hawk 620 platform.
Object(s) To Be Replaced:
/etc/conf/bin/idbuild
/etc/conf/bin/idtune
/etc/conf/mtune.d/ipc
/etc/conf/pack.d/bsp4600/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/bspmtx/Driver.o
/etc/conf/pack.d/ipc/space.c
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/rcim/Driver.o
/etc/conf/pack.d/svc/Driver.o
/sbin/config
/usr/ccs/bin/dump
/usr/ccs/bin/ld
/usr/ccs/bin/strip
/usr/ccs/lib/libdwarf.a
/usr/include/dwarfwrite.h
/usr/include/libdwarf.h
/usr/include/pthread.h
/usr/include/snmp/snmp_trap.h
/usr/include/sys/adapter_pci.h
/usr/include/sys/metrics.h
/usr/sbin/clock_synchronize
Special Conditions for Installation:
1. The following additional object will be replaced by base-001:
* /usr/bin/idld
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-001
Date Issued: 09/15/1999 12:53:54
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: none
Related SARs: HM12641, HM12685, HM12673
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
PowerMAX OS 4.3 Patch Set 1
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P1" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 1".
New PowerMAX OS 4.3 Software Patch(es):
base-001 Base System Patch 001
cmds-001 Advanced Commands Patch 001
cnd-001 Condor Ethernet Driver Patch 001
fbs-001 Frequency Based Scheduler Patch 001
man-001">fbsman-001 Frequency Based Scheduler On-line Manual Pages Patch 001
inet-001 Internet Utilities Patch 001
kdb-001 Kernel Debugger Patch 001
man-001 On-line Manual Pages Patch 001
nsu-001 Network Support Utilities Patch 001
trace-001 KernelTrace Utilities Patch 001
##############################################################################
Problem Description:
1. HM12641: Setting the PATH environment variable to NULL prior to
invocation of the idbuild(1M) or idtune(1M) utilities will result in
failure of those utilities.
2. The config(1M) utility aborts when searching for a kernel module
using the "Find by Keyword" operation.
3. HM12685: The config(1M) utility aborts when displaying all tunables
using the "All Tunables" operation.
4. HM12673: The IPC maximum number of semaphores per id is a hard-coded
value. This should be modifiable as a system tunable.
5. The metrics calculations done in the kma and kmadbg modules during
allocations and deallocations of kernel virtual space were not
protected by locks or atomic operations. Updates of these counters
could be lost if the same counter were being updated on different
CPUs or the update was interrupted by a higher priority function that
also requests an allocation or deallocation. This could cause the
data returned by sar(1M) (`sar -k`) to be inaccurate.
6. The debug version of kernel space allocator was calling cmn_err()
with the kmacorrupt_lock locked. cmn_err() was called to add an
informational message to the syslog that it was turning the detection
of "use-after-free" problems back on after it had previously been
turned off upon reaching the limits specified by the tunables
KMEM_LIMIT_UPPER and KMEM_LIMIT_LOWER. Because cmn_err() also calls
the allocator, the system would hang forever trying to take the same
lock again.
7. strip(1) removed the symbol table of shared libraries.
8. pthread_self(3pthread), pthread_cleanup_push(3pthread), and
pthread_cleanup_pop(3pthread) were not visible to c++ standard users.
9. There were instances when linking c++ programs where the .bss section
would end up with an odd number size.
10. Some snmp trap definitions were missing from <snmp/snmp_trap.h>.
11. Due to chip erratas, 'sync' instructions are required in the internal
kernel lock and atomic operations. The TurboHawk uses the 'Mach5'
processor and does not require these instructions. 'sync'
instructions can have a substantial effect on overall system
performance.
12. The NCR SCSI controller was generating stray interrupts on TurboHawk
systems with 8 processors and 2 global memory boards. The stray
interrupts were seen under the system level I/O test, PowerIO, when
the edge trigger interrupts were being exercised. This problem
prevents configuring system disks on the global memory boards.
Problem Resolution:
1. The idbuild(1M) and idtune(1M) utilities were modified to define and
export the minimum PATH needed to execute successfully upon
invocation, returning the PATH to its original setting upon
completion.
2. Made config(1M) more robust in its handling of ill-formed module
descriptions and tunable descriptions.
3. Increased the number of tunables that config can display using
"All Tunables" operation. Also, fixed so that utility doesn't abort
when this tunable limit is reached.
4. Added system tunable SEMMSL.
5. The macros used by the kma and kmadbg modules to increment and
decrement the metrics counters were not protected by locks or by the
use of the atomic increment/decrement functions. The macros used by
these modules in <sys/metrics.h> were modified to use the atomic
increment and decrement functions.
6. kmem_alloc_debug() was changed to unlock the kmacorrupt_lock before
it calls cmn_err().
7. Fixed strip so that it does not remove the symbol table when
stripping a shared library.
8. Fixed header file <pthread.h> so that pthread_self,
pthread_cleanup_push, pthread_cleanup_pop are also visible to c++
standard users.
9. Fixed 'ld' so that in such occurences .bss is always rounded to next
16 bytes.
10. Added additional snmp trap definitions to <snmp/snmp_trap.h>.
11. At start-up time, the unnecessary 'sync' instructions are patched to
'nop' instructions if running on a TurboHawk.
12. The interrupt request function used by PowerMAX OS was not being
implemented properly. In this case the edge trigger interrupt tests
were using the interrupt request function to activate an edge trigger
interrupt. The interrupt request function was performing a
read-modify-write operation to the interrupt controller's request
register. The function should have been performing a single bit
write to the request register.
Enhancements:
1. Added new attributes and tags to DWARF information used to provide
debug information for applications:
a) Added new attribute for phase3.2 of the MAXAda compiler:
DW_AT_child_unit
b) Added new tag for phase4 of the MAXAda compiler:
DW_TAG_inst_package_param
c) Added new attributes to provide additional information about
inline calls in both the MAXAda and c++ compilers.
NightView will use this information to provide "virtual
walkback" information. The attributes are:
DW_AT_call_column
DW_AT_call_file
DW_AT_call_line
2. Added definition to <sys/adapter_pci.h> for ADAPTER_IP.
Object(s) To Be Replaced:
/etc/conf/bin/idbuild
/etc/conf/bin/idtune
/etc/conf/mtune.d/ipc
/etc/conf/pack.d/bsp6800t/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/ipc/space.c
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/util/Driver.o
/sbin/config
/usr/ccs/bin/dump
/usr/ccs/bin/ld
/usr/ccs/bin/strip
/usr/ccs/lib/libdwarf.a
/usr/include/dwarfwrite.h
/usr/include/libdwarf.h
/usr/include/pthread.h
/usr/include/snmp/snmp_trap.h
/usr/include/sys/adapter_pci.h
/usr/include/sys/metrics.h
Special Conditions for Installation:
1. The following additional object will be replaced by base-001:
* /usr/bin/idld
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-002
Date Issued: 02/02/2000 13:36:17
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: cnd-002, crosslibs-001, crypt-001, crypt-int-001,
dec-001, egl-001, fbs-002, man-002">fbsman-002, inet-002, ip-001,
librt-001, man-002, mvc-001, pg-001
Related SARs: #39, #45, #59, #72, #98, #103, #114, #116, #117, #166
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
PowerMAX OS 4.3 Patch Set 2
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P2" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 2".
New PowerMAX OS 4.3 Software Patch(es):
1553v5drv-001 1553v5-ABI User Level Driver Patch 001 (4.3P2)
1553v5lib-001 1553v5-ABI Library Interfaces Patch 001 (4.3P2)
base-002 Base System Patch 002 (4.3P2)
cld-001 MPEG-2 Decoder Driver Patch 001 (4.3P2)
cmds-002 Advanced Commands Patch 002 (4.3P2)
cnd-002 Condor Ethernet Driver Patch 002 (4.3P2)
crosslibs-001 Libraries for Cross Compiling Patch 001 (4.3P2)
crypt-001 Domestic Encryption Utilities Patch 001 (4.3P2)
crypt-int-001 International Encryption Utilities Patch 001 (4.3P2)
dec-001 DEC Ethernet Driver Patch 001 (4.3P2)
diskless-002 Diskless Systems Package Patch 002 (4.3P2)
egl-001 Eagle Ethernet Driver Patch 001 (4.3P2)
fbs-002 Frequency Based Scheduler Patch 002 (4.3P2)
man-002">fbsman-002 Frequency Based Scheduler On-line Manual Pages
Patch 002 (4.3P2)
hsde-001 High Speed Data Enhanced Channel Driver
Patch 001 (4.3P2)
inet-002 Internet Utilities Patch 002 (4.3P2)
ip-001 Interphase 4511 PMC FDDI Driver Patch 001 (4.3P2)
librt-001 Real-time Libraries Patch 001 (4.3P2)
man-002 On-line Manual Pages Patch 002 (4.3P2)
mvc-001 Multiplexor VME Controller Driver Patch 001 (4.3P2)
pg-001 Peregrine FDDI Driver Patch 001 (4.3P2)
trace-002 KernelTrace Utilities Patch 002 (4.3P2)
%#############################################################################
Problem Description:
1. Use of more expensive sync instructions affected the virtual memory
system performance.
2. The vi(1) tag facility was not handling duplicate tags properly.
vi is supposed to treat duplicate tags as a circular list.
Successive `:ta' commands (with no argument) takes the user to the
next duplicate in the list in a roundrobin fashion. Instead, once
the end of the list was reached, error messages would be printed for
all following `:ta' commands.
3. #116: Using standard mvc ports, reads return immediately if
conditions satisfy the read (i.e., adequate data are available to be
returned). Using mvcrt ports, the initial read would stall when the
read should have been satisfied.
Additionally, if VMIN=0 and VTIME>0, a read would always wait for the
timeout before returning, even when data were available to satisfy
the read. To be correct, VTIME is the time for completion of the
read, rather than an intercharacter timer, if the read cannot
otherwise be satisfied. The read should complete if either a single
character is read or if the timer expires (no characters returned).
4. #59: The xdr_double() function contained in libnsl will return an
Arithmetic Exception error and dump core.
5. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
6. #98: The system(3S) I/O function should use fork1(2) in
multi-threaded applications instead of fork(2).
7. In UFS filesystems, the modification times of two different files
might show that the clock was running backwards, which means that the
file modified earlier might have the more recent timestamp.
8. #114: ksh(1) can't address directory when its size exceeds 2 Gbytes.
When using ksh to do integer arithmetic operation with command
expr(1), the upper limit for the integer is 2^31 + 1.
9. User-level interrupt routines are allowed to do only a limited number
of system calls. Two are server_wake and server_wake1. However, in
some cases, this can cause a process preemption which would panic the
system. Preemption is not allowed while in a user-level interrupt
routine.
10. Confusing debug messages were being printed when a configured drive
was not found.
11. User-level interrupts with shared vector devices could cause memory
corruption. Counters associated with shared vectors was being used
incorrectly.
12. #72: Threads may returned too early from pthread_once(3pthread).
When more than one thread calls pthread_once() at roughly the same
time, the threads that are not the first caller do not wait until the
first caller thread completes its execution of the initialization
routine.
13. If a parent process creates one or more POSIX timers and then calls
fork(2) or fork1(2), the child process will terminate with an
internal libthread error if the child attempts to also create a POSIX
timer with timer_create(3C).
14. If a parent process creates a POSIX timer with timer_create(3C) and
specifies the SIGEV_CALLBACK option and then calls fork(2), the
background callback thread will incorrectly exit and remove that
POSIX timer from the parent process.
15. #117: sigtimedwait(2) may cause other threads within a process to
terminate. In a multi-threaded process, if one of the threads uses
sigtimedwait(2) to catch certain signals, and one of these signals is
already pending by the time that thread enters the kernel's
sigtimedwait() routine, then the kernel will destroy all the other
LWPs in this process if there is no signal catching routine setup to
catch this signal and if the default action for the pending signal is
to terminate the process.
16. It is possible to create a symbolic link on an XFS filesystem in a
directory to which the user does not have write permission.
17. The kernel may panic with the functions page_sortadd() and
xfscr_getpage() in the stack trace under certain XFS file overwrite
conditions. The failure conditions depend on the length and offset
of the overwrite, the file disk space allocation boundries and the
state of the page cache. The problem occurs when the overwrite
requires the physical read of part of the overwritten data and is due
to a fault in the read optimization code.
18. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
19. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
20. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
21. #39: DL_TEST_REQ messages do not work correctly.
22. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
23. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
24. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
25. #45: The system would hang with an out of memory condition when
using SOCK_RAW sockets with RAWIP under the following conditions:
1) The destination host for the sender is the local system and
there is no receiver currently running.
2) The destination host for both the sender and receiver is
the local system and both tests are currently running.
If the destination host(s) is not the local system (data is going
over the wire), the hang does not occur.
26. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
27. Setting the RCIM_ETI_LEVEL tunable causes system to hang with stray
interrupts.
28. The rcimconfig(1M) utility reports polarity incorrectly.
29. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
30. #103: The console could not initialize VGA compatible mode on Matrox
G200 SD PCI card. The console program (resident monitor/boot loader)
includes a VGATTY driver designed to use keyboard and VGA compatible
video card for console I/O during system initialization. However,
this function relies on a small library of device-specific init
routines that can place the selected video card into a VGA compatible
mode.
Problem Resolution:
1. The more expensive sync instruction was replaced in some places with
the cheaper eieio instruction.
2. When the end of the list is reached, an error message is printed.
The following `:ta' then goes back to the first tage in the duplicate
tag list.
3. mvcrt ports, when opened for reading, now immediately issue a read
command, permitting the first read issued by the user to be
immediately satisfied when data is available.
The VMIN/VTIME code now correctly checks for available data before
waiting for the timeout when VMIN=0 and VTIME>0.
4. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
5. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
6. The system(3S) I/O function has been modified to use fork1(2) instead
of fork(2) for all applications.
7. A global timer is used to keep the more recent time between the
time-of-the-day clock and the last file timestamp assigned to any
file. A new file timestamp is assigned based on the global timer.
8. Long long integer support has been added to command expr. Now the
command can do integer arithmetic operation beyond 2^32 (up to
2^63+1).
9. When returning from user-level interrupt routine, it will no longer
check for and take requested preemptions. If there is a requested
preemption it will be handled after the UI routine completes.
10. Removed the messages.
11. Fix shared-vector code in user-level interrupt logic.
12. Block the other threads so that they do not return from
pthread_once() until the initialization routine has been completely
executed by the first caller thread.
13. The thread library's internal POSIX timer structures are now
re-initialized by the first thread that calls timer_create(3C) or
timer_delete(3C) within the child process.
14. The kernel and thread library code was modified so that fork
operations no longer cause the background callback thread to call
thr_exit(3thread) upon an interrupted return while waiting on a
POSIX timer expiration internally within the thread library.
15. Change the kernel's sigtimedwait() routine so that when a signal is
already pending, the calling LWP will simply return to the caller
with the selected signal number, without interfering with the other
LWPs in the same process.
16. A check of the user write permission added to the symbolic link
create code path.
17. The filesystem page read optimization is fixed to request the correct
pages for the overwrite boundary conditions.
18. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
19. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
20. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
21. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
22. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
23. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
24. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
25. There were a large number of outstanding kernel memory allocations by
icmp_pkt() and pullupmsg() that were not being deallocated until the
writer socket was closed. If the test issued a large number of write
(sendto) requests to the socket, the system would eventually hang
waiting for kernel memory to become available.
The outstanding allocations were message blocks that contained
ICMP_SOURCE_QUENCH error messages. These had been sent by ip because
the original data message could not be echoed to the read (RD) side
of the writer socket because it's queue became full.
icmp_inbound_error() was not checking for flow control before passing
the ICMP error message upstream to the next module. It continued
sending the ICMP errors up through the rawip module even after the
stream head's queue was full. In addition, rawip should have been
discarding these error messages. Because the application never read
any of the echo or ICMP messages from the read side of the writer
socket, flow control was never being relieved and eventually the
machine ran out of memory.
icmp_inbound_error() was changed to call canputnext() to check for
flow control before an ICMP error message is passed upstream.
rawip_rput() was modified to discard all M_CTL type message blocks.
The memory leak detection debug code was also enhanced in the kma
and kmadbg kernel drivers.
26. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
27. Fix code associated with RCIM_ETI_LEVEL. It was setting the wrong
register.
28. Fixed rcimconfig.
29. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
30. Added support to console program to initialize MGA G200 SD card's PCI
resources, SDRAM memory, PLLs, DAC, and then place the MGA chip into
a VGA compatible state. After this initialization, the VGATTY driver
in console functions properly with the new G200 cards.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/dmac/Driver.o
/etc/conf/pack.d/fbs/stubs.c
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/rcim/Driver.o
/etc/conf/pack.d/rtserial/Driver.o
/etc/conf/pack.d/sbc/stubs.c
/etc/conf/pack.d/sfs/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/config
/stand/cp1
/usr/bin/expr
/usr/bin/vi
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libc.so.1
/usr/include/sys/fs/sfs_inode.h
/usr/include/sys/bus.h
/usr/include/sys/pin.h
/usr/include/sys/resource.h
/usr/sbin/rcimconfig
Special Conditions for Installation:
During installation of base-002, an updated version of /stand/cp1 will be
installed and the new console overlay will be written to partition 6 on
the system disk. Following successful installation, the system must be
shut down and reset to load the new console before rebooting. Failure
to do so may result in unexpected or unreliable behavior.
The following object(s) will also be replaced by base-002:
* /usr/bin/edit
* /usr/bin/ex
* /usr/bin/vedit
* /usr/bin/view
* /usr/ccs/lib/libp/libc.so
* /usr/lib/libposix1c.a
* /usr/lib/libposix1c.so
* /usr/lib/libnsl.a
* /usr/lib/libsocket.a
* /usr/lib/libxti.a
* /usr/lib/libnsl.so
* /usr/lib/libxti.so
Possible Side Effects:
If the base-002 package is removed (pkgrm(1M)), the original console will
be re-written to partition 6 on the system disk. To ensure there are no
adverse side effects, the kernel should be rebuilt (idbuild(1M)) and the
system shut down and reset to load the original console before rebooting
the system.
Suggested procedure for removing base-002:
# pkgrm base-002
# /etc/conf/bin/idbuild -B
# sync
# init 0
# Reset system
Follow the normal boot procedure detailed in the PowerMAX OS Product
Release Notes to load the new console and boot the new kernel.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-002
Date Issued: 02/02/2000 11:23:28
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: cnd-002, crosslibs-001, crypt-001, crypt-int-001,
dec-001 egl-001, fbs-002, man-002">fbsman-002, ie-001, inet-002,
librt-001, man-002, mvc-001, pg-001
Related SARs: #39, #45, #47, #59, #72, #98, #114, #116, #117, #166
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
PowerMAX OS 4.3 Patch Set 2
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P2" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 2".
New PowerMAX OS 4.3 Software Patch(es):
1553v5drv-001 1553v5-ABI User Level Driver Patch 001 (4.3P2)
1553v5lib-001 1553v5-ABI Library Interfaces Patch 001 (4.3P2)
base-002 Base System Patch 002 (4.3P2)
cmds-002 Advanced Commands Patch 002 (4.3P2)
cnd-002 Condor Ethernet Driver Patch 002 (4.3P2)
crosslibs-001 Libraries for Cross Compiling Patch 001 (4.3P2)
crypt-001 Domestic Encryption Utilities Patch 001 (4.3P2)
crypt-int-001 International Encryption Utilities Patch 001 (4.3P2)
dec-001 DEC Ethernet Driver Patch 001 (4.3P2)
egl-001 Eagle Ethernet Driver Patch 001 (4.3P2)
fbs-002 Frequency Based Scheduler Patch 002 (4.3P2)
man-002">fbsman-002 Frequency Based Scheduler On-line Manual Pages
Patch 002 (4.3P2)
hsde-001 High Speed Data Enhanced Channel Driver
Patch 001 (4.3P2)
ie-001 Night Hawk ISE Ethernet Interface Module
Patch 001 (4.3P2)
inet-002 Internet Utilities Patch 002 (4.3P2)
librt-001 Real-time Libraries Patch 001 (4.3P2)
man-002 On-line Manual Pages Patch 002 (4.3P2)
mvc-001 Multiplexor VME Controller Driver Patch 001 (4.3P2)
pg-001 Peregrine FDDI Driver Patch 001 (4.3P2)
trace-002 KernelTrace Utilities Patch 002 (4.3P2)
##############################################################################
Problem Description:
1. Use of more expensive sync instructions affected the virtual memory
system performance.
2. The vi(1) tag facility was not handling duplicate tags properly.
vi is supposed to treat duplicate tags as a circular list.
Successive `:ta' commands (with no argument) takes the user to the
next duplicate in the list in a roundrobin fashion. Instead, once
the end of the list was reached, error messages would be printed for
all following `:ta' commands.
3. #116: Using standard mvc ports, reads return immediately if
conditions satisfy the read (i.e., adequate data are available to be
returned). Using mvcrt ports, the initial read would stall when the
read should have been satisfied.
Additionally, if VMIN=0 and VTIME>0, a read would always wait for the
timeout before returning, even when data were available to satisfy
the read. To be correct, VTIME is the time for completion of the
read, rather than an intercharacter timer, if the read cannot
otherwise be satisfied. The read should complete if either a single
character is read or if the timer expires (no characters returned).
4. #59: The xdr_double() function contained in libnsl will return an
Arithmetic Exception error and dump core.
5. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
6. #98: The system(3S) I/O function should use fork1(2) in
multi-threaded applications instead of fork(2).
7. In UFS filesystems, the modification times of two different files
might show that the clock was running backwards, which means that the
file modified earlier might have the more recent timestamp.
8. #114: ksh(1) can't address directory when its size exceeds 2 Gbytes.
When using ksh to do integer arithmetic operation with command
expr(1), the upper limit for the integer is 2^31 + 1.
9. User-level interrupt routines are allowed to do only a limited number
of system calls. Two are server_wake and server_wake1. However, in
some cases, this can cause a process preemption which would panic the
system. Preemption is not allowed while in a user-level interrupt
routine.
10. Confusing debug messages were being printed when a configured drive
was not found.
11. User-level interrupts with shared vector devices could cause memory
corruption. Counters associated with shared vectors were being used
incorrectly.
12. #72: Threads may returned too early from pthread_once(3pthread).
When more than one thread calls pthread_once() at roughly the same
time, the threads that are not the first caller do not wait until the
first caller thread completes its execution of the initialization
routine.
13. If a parent process creates one or more POSIX timers and then calls
fork(2) or fork1(2), the child process will terminate with an
internal libthread error if the child attempts to also create a POSIX
timer with timer_create(3C).
14. If a parent process creates a POSIX timer with timer_create(3C) and
specifies the SIGEV_CALLBACK option and then calls fork(2), the
background callback thread will incorrectly exit and remove that
POSIX timer from the parent process.
15. #117: sigtimedwait(2) may cause other threads within a process to
terminate. In a multi-threaded process, if one of the threads uses
sigtimedwait(2) to catch certain signals, and one of these signals is
already pending by the time that thread enters the kernel's
sigtimedwait() routine, then the kernel will destroy all the other
LWPs in this process if there is no signal catching routine setup to
catch this signal and if the default action for the pending signal is
to terminate the process.
16. It is possible to create a symbolic link on an XFS filesystem in a
directory to which the user does not have write permission.
17. The kernel may panic with the functions page_sortadd() and
xfscr_getpage() in the stack trace under certain XFS file overwrite
conditions. The failure conditions depend on the length and offset
of the overwrite, the file disk space allocation boundries and the
state of the page cache. The problem occurs when the overwrite
requires the physical read of part of the overwritten data and is due
to a fault in the read optimization code.
18. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
19. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
20. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
21. #39: DL_TEST_REQ messages do not work correctly.
22. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
23. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
24. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
25. #45: The system would hang with an out of memory condition when
using SOCK_RAW sockets with RAWIP under the following conditions:
1) The destination host for the sender is the local system and
there is no receiver currently running.
2) The destination host for both the sender and receiver is
the local system and both tests are currently running.
If the destination host(s) is not the local system (data is going
over the wire), the hang does not occur.
26. Probes to non-existant VME addresses can result in a hardware
deadlock on TurboHawk platforms.
27. #47: Release on Register Access (RORA) mode was not supported for
devices in expansion VME I/O on TurboHawk platforms.
28. On TurboHawk platforms only one expansion VME bus was supported and
this bus was supported only by the PMC connection on the first
processor board.
29. TurboHawk systems would hard hang while booting if the new TurboHawk
local memory with interval timer was installed.
30. When NightTrace encounters an "int on no int" interrupt event, it
displays the incorrect interrupt name for the event. There are
several possible causes of this:
a. An interrupt vector is not reserved for the interrupt.
Thus, the vector can be allocated later by a device. When
this happens, it becomes impossible to distinguish an
"int on no int" interrupt from a valid interrupt for the
device which allocated the vector.
b. The vector the kernel uses to indicate the "int on no int"
interrupt is not the same on all machines.
Problem Resolution:
1. The more expensive sync instruction was replaced in some places with
the cheaper eieio instruction.
2. When the end of the list is reached, an error message is printed.
The following `:ta' then goes back to the first tage in the duplicate
tag list.
3. mvcrt ports, when opened for reading, now immediately issue a read
command, permitting the first read issued by the user to be
immediately satisfied when data is available.
The VMIN/VTIME code now correctly checks for available data before
waiting for the timeout when VMIN=0 and VTIME>0.
4. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
5. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
6. The system(3S) I/O function has been modified to use fork1(2) instead
of fork(2) for all applications.
7. A global timer is used to keep the more recent time between the
time-of-the-day clock and the last file timestamp assigned to any
file. A new file timestamp is assigned based on the global timer.
8. Long long integer support has been added to command expr. Now the
command can do integer arithmetic operation beyond 2^32 (up to
2^63+1).
9. When returning from user-level interrupt routine, it will no longer
check for and take requested preemptions. If there is a requested
preemption it will be handled after the UI routine completes.
10. Removed the messages.
11. Fix shared-vector code in user-level interrupt logic.
12. Block the other threads so that they do not return from
pthread_once() until the initialization routine has been completely
executed by the first caller thread.
13. The thread library's internal POSIX timer structures are now
re-initialized by the first thread that calls timer_create(3C) or
timer_delete(3C) within the child process.
14. The kernel and thread library code was modified so that fork
operations no longer cause the background callback thread to call
thr_exit(3thread) upon an interrupted return while waiting on a
POSIX timer expiration internally within the thread library.
15. Change the kernel's sigtimedwait() routine so that when a signal is
already pending, the calling LWP will simply return to the caller
with the selected signal number, without interfering with the other
LWPs in the same process.
16. A check of the user write permission added to the symbolic link
create code path.
17. The filesystem page read optimization is fixed to request the correct
pages for the overwrite boundary conditions.
18. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
19. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
20. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
21. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
22. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
23. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
24. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
25. There were a large number of outstanding kernel memory allocations by
icmp_pkt() and pullupmsg() that were not being deallocated until the
writer socket was closed. If the test issued a large number of write
(sendto) requests to the socket, the system would eventually hang
waiting for kernel memory to become available.
The outstanding allocations were message blocks that contained
ICMP_SOURCE_QUENCH error messages. These had been sent by ip because
the original data message could not be echoed to the read (RD) side
of the writer socket because it's queue became full.
icmp_inbound_error() was not checking for flow control before passing
the ICMP error message upstream to the next module. It continued
sending the ICMP errors up through the rawip module even after the
stream head's queue was full. In addition, rawip should have been
discarding these error messages. Because the application never read
any of the echo or ICMP messages from the read side of the writer
socket, flow control was never being relieved and eventually the
machine ran out of memory.
icmp_inbound_error() was changed to call canputnext() to check for
flow control before an ICMP error message is passed upstream.
rawip_rput() was modified to discard all M_CTL type message blocks.
The memory leak detection debug code was also enhanced in the kma
and kmadbg kernel drivers.
26. Following a VME bus timeout on TurboHawk platforms the kernel relies
on the hardware to clear the error state of the PLX9060 bridge.
27. RORA is now supported for devices in expansion VME I/O on TurboHawk
platforms. This feature is implemented as a system tunable named
TURBO_VME_INT_ROR. This tunable supports VME devices on the primary
VME bus as well as on the expansion VME bus.
28. The TurboHawk now supports up to four expansion VME busses. These
busses can be configured via the PMC connectors on any of the four
processor boards.
29. Remove reference to TurboHawk tick timer. This feature is not
available.
30. NightTrace problems resolved as follows:
a. An interrupt vector is statically reserved in the kernel's
interrupt vector table.
b. Interrupt vector 127 is reserved on the NightHawk 6800.
Interrupt vector 64 is reserved on the PowerMAXION and
TurboHawk.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/cf.d/ivt.s
/etc/conf/mtune.d/pci
/etc/conf/mtune.d/vme
/etc/conf/pack.d/bsp6800t/Driver.o
/etc/conf/pack.d/fbs/stubs.c
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/io/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/pci/Driver.o
/etc/conf/pack.d/pci/space.c
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/rtserial/Driver.o
/etc/conf/pack.d/sfs/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/vme/space.c
/etc/conf/pack.d/xfs/Driver.o
/sbin/config
/sbin/expr
/usr/bin/expr
/usr/bin/vi
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libc.so.1
/usr/include/sys/fs/sfs_inode.h
/usr/include/sys/bus.h
/usr/include/sys/pin.h
/usr/include/sys/resource.h
/usr/include/sys/syscx.h
Special Conditions for Installation:
The following object(s) will also be replaced by base-002:
* /usr/bin/edit
* /usr/bin/ex
* /usr/bin/vedit
* /usr/bin/view
* /usr/ccs/lib/libp/libc.so
* /usr/lib/libposix1c.a
* /usr/lib/libposix1c.so
* /usr/lib/libnsl.a
* /usr/lib/libsocket.a
* /usr/lib/libxti.a
* /usr/lib/libnsl.so
* /usr/lib/libxti.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-003
Date Issued: 05/30/2000 12:58:11
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: crosslibs-002
Related SARs: 106, 187, 197, 203, 225, 226, 230
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 3
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P3" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 3".
New PowerMAX OS 4.3 Software Patch(es):
base-003 Base System Patch 003 (4.3P3)
cnd-003 Condor Ethernet Driver Patch 003 (4.3P3)
crosslibs-002 Libraries for Cross Compiling Patch 002 (4.3P3)
dec-002 DEC Ethernet Driver Patch 002 (4.3P3)
diskless-003 Diskless Systems Package Patch 003 (4.3P3)
egl-002 Eagle Ethernet Driver Patch 002 (4.3P3)
fibre-002 Fibre Channel Driver Patch 002 (4.3P3)
inet-003 Internet Utilities Patch 003 (4.3P3)
ip-002 Interphase 4511 PMC FDDI Driver Patch 002 (4.3P3)
nfs-001 Network File System Utilities Patch 001 (4.3P3)
nsu-002 Network Support Utilities Patch 002 (4.3P3)
pg-002 Peregrine FDDI Driver Patch 002 (4.3P3)
rvsrv-001 Residential Video Server Tools Patch 001 (4.3P3)
trace-003 KernelTrace Utilities Patch 003 (4.3P3)
xfsd-001 Distributed XFS File System Patch 001 (4.3P3)
New PowerMAX OS 4.3 Software Package(s):
gpib NI PCI-GPIB Kernel Level Driver
Note: The "gpib" package is supported on the Power Hawk 620, Power Hawk 640,
and PowerStack II.
##############################################################################
Problem Description:
1. SAR 106: On a closely-coupled system (CCS) where a Matrox graphics
card is configured on each of the CPU systems the target systems
could hang.
2. SAR 187: Target closely-coupled systems (CCS) intermittently fail to
scan disks at boot time. A disk that is configured a multi-initiator
arrangement on a CCS system will sometimes fail to be scanned on the
target systems.
3. SAR 197: iconnect(3C) doesn't work with the Z8536 clocks on Motorola
SBCs.
4. SAR 203: The rtserial read (receive) code could walk off the end of
a page and cause a kernel fault.
5. SAR 225: Multi-threaded and/or multi-LWP applications can cause the
system to hang if a user-level debugger stop operation is attempted
while one of the other threads/LWPs is currently blocked in the
kernel on a sigtimedwait(2) or sigwait(2) call.
6. SAR 226: Trying to fork() a multi-threaded process may hang the
process. If one thread is currently in
thr_join(3thread)/pthread_join(3pthread) when another thread in the
same process calls fork(2), then the thread blocked in thr_join()
causes the fork() operation to never complete. Additionally, if the
thread inside thr_join() call is a multiplexing (MUX, not bound)
thread, then it will run continuously, consuming CPU cycles.
7. SAR 230: The lwp_steal() function attempts to find work for an idle
cpu in a multi-processor system. It does this by taking a waiting
and ready LWP from a busy processor's queue and placing it on the
idle processor's queue. The manipulation of these queues was done
mostly without the protection of a lock and could result in mangled
processor queues and a hung system.
8. xfsrestore(1M) may coredump when trying to restore filesystems with
inode numbers higher than 65535. The dummy entry at the end of a
directory block in the dump is not handled correctly in xfsrestore
when the inode number is 64K or larger, resulting in an alignment
fault.
9. The kernel would panic in page_alloc_blind(). The kernel keeps
various lists of free pages to give to applications as they need
them. In one rare circumstance during the walking of this list,
looking for a page of the appropriate type was not protected by a
semaphore for the entire length of the walk. This opened the
possibility of a corrupted freelist which would cause the panic.
10. Applications sometimes stall waiting for memory forever. On systems
where most of memory was locked down in applications, applications
not so locked down would sometimes stall out forever, waiting for
memory even though there was some memory available for them to use.
The cause: locked pages not only lock down a physical page, but also
a PTE in the system page table. If most of memory is locked down in
applications, then most of the PTEs will be also locked down. Since
any one virtual address has only 16 PTEs into which it can be placed,
if all 16 happened to be in use by other locked pages then the new
page being faulted in would never be able to complete until some
application owning one of the 16 locked pages would release the lock.
11. nanosleep(3C) was failing in certain configurations.
12. There were failures where processor run queues were getting
destroyed. The problem was that a single lwp was added to the
run queues twice. This is a problem that leads to the run queues
becoming corrupted. Failures occurred between the interaction with
selfblock() (from nanosleep(), etc.) and a process that becomes
vm-seized.
13. Motorola SBCs made approximately 120 pages of memory unavailable for
any use on the typical system. The PowerPC architecture requires
that the PTE page table be allocated on a memory boundary aligned
with its size. This allocation algorithm, under certain
circumstances, would allocate memory for a PTE table almost twice the
required size in order to achieve this alignment.
14. Certain custom configurations allow third-party VME boards to serve
as the VME bus controller. By default, the Motorola SBC serves as
the VME bus controller which may conflict with the third-party VME
bus controller.
Problem Resolution:
1. The cons driver has been modified to correct this problem.
2. The scsi_scan() function of the gd driver would sometimes return
IMERR_DEVICE_BUSY during the boot scan for disks on the target CCS
systems. This was possible because the same disk may actually have
been busy on the master CCS system at the time of the scan. The
scsi_scan() function was modified such that if the IMERR_DEVICE_BUSY
was returned initially then up to five more retries would be
attempted.
3. The Z8536 clocks share an interrupt vector with 2 secondary UARTs.
It is not structured as a standard shared interrupt (although it
probably should be). The interrupt vector is artificially marked as
"shared". This causes a problem with iconnect because it is not
recognized and organized like a standard "shared" interrupt. This
was changed so there is a unique interrupt vector for each of the
Z8536 clocks. This is invoked indirectly like the RCIM driver does.
4. The rtserialinput() routine now checks the write pointer value when
first entered, and if it is beyond the end of the physical queue, it
is adjusted to the start of the queue before any of the data is
written to the queue.
5. The kernel's sigtimedwait() routine was ignoring stop events when
checking for signals/events for the current LWP. This would cause
the LWP to loop/run forever inside the sigtimedwait() kernel routine.
The sigtimedwait() routine now checks for EVF_PL_PRSTOP events
(debugger stop events), and returns back out of this sigtimedwait()
routine so that the kernel can now successfully stop the LWP.
6. The thread library's thr_join() routine (which handles both
thr_join(3thread) and pthread_join(3pthread) calls) was re-coded so
that it can now process pending signals while waiting to join with
another thread. This change lets the fork(2) processing (which
requires all other threads in the same process to receive a signal
and subsequently 'rendezvous' in the kernel while a duplicate image
of the process is made in order to create the child process) to
proceed successfully. The thr_join() routine was previously blocking
signals/events that were posted to this LWP/thread, thus preventing
the fork(2) operation from completing.
7. Serialize the operation of lwp_steal() so that multiple processors do
not walk over each other should two or more processors be running in
lwp_steal() at the same time.
8. Change made to xfsrestore to correct the processing of dummy entries
at the end of a directory block.
9. The freelist walking algorithm was modified to be fully protected by
the appropriate semaphore.
10. Added a new tunable to PowerMAX OS, PTE_TABLESZ_MULT, which when >1
specifies a multiple of how much larger the page table is to be made
beyond its default size.
11. Corrected the nanosleep() failure.
12. The fix is to totally remove the "wait" state information when an
lwp becomes seized.
13. The PTE table allocator is now more intelligent about PTE table
placement in the kernel address space.
14. The tunable, VME_UNIV_SYSCON, was added to allow disabling of the
Motorola SBC as the VME bus controller.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/mem
/etc/conf/mtune.d/vme
/etc/conf/pack.d/cons/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/rtserial/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/vme/space.c
/etc/conf/pack.d/xfs/Driver.o
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/sys/fs/xfs_types.h
/usr/lib/libc.so.1
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/fs/xfs/xfsrestore
Special Conditions for Installation:
The following object(s) will also be replaced by base-003:
* /usr/ccs/lib/libp/libc.so
* /usr/lib/libposix1c.a
* /usr/lib/libposix1c.so
* /usr/sbin/xfsrestore
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-003
Date Issued: 05/30/2000 13:24:44
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: crosslibs-002
Related SARs: 187, 189, 196, 203, 224, 225, 226, 230, 250
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 3
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P3" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 3".
New PowerMAX OS 4.3 Software Patch(es):
base-003 Base System Patch 003 (4.3P3)
cnd-003 Condor Ethernet Driver Patch 003 (4.3P3)
crosslibs-002 Libraries for Cross Compiling Patch 002 (4.3P3)
dec-002 DEC Ethernet Driver Patch 002 (4.3P3)
egl-002 Eagle Ethernet Driver Patch 002 (4.3P3)
ie-002 Night Hawk ISE Ethernet Interface Module Patch 002 (4.3P3)
inet-003 Internet Utilities Patch 003 (4.3P3)
nfs-001 Network File System Utilities Patch 001 (4.3P3)
nsu-002 Network Support Utilities Patch 002 (4.3P3)
pg-002 Peregrine FDDI Driver Patch 002 (4.3P3)
trace-003 KernelTrace Utilities Patch 003 (4.3P3)
xfsd-001 Distributed XFS File System Patch 001 (4.3P3)
##############################################################################
Problem Description:
1. SAR 187: There is a possibility that disk drives could intermittently
fail to be scanned at boot time.
2. SAR 189: User-locked pages with anchored NUMA bindings can panic the
system. The kernel was incorrectly assuming that hard/soft anchored
user pages would never be replicated/duplicated in more than one
memory pool. As a result, the kernel would panic when it detected
this situation.
3. SAR 196: Frontplane Invalid Address Error sysfaults may panic the
system. The embedded NCR controller hardware performs look-ahead
memory reads during I/O write operations for optimization reasons.
This look-ahead read request may extend beyond the end of a local
memory pool's last physical page, into a 'hole' where no memory
exists. When this occurs, an Frontplane Invalid Address Error
results, causing the system to panic.
4. SAR 203: The rtserial read (receive) code could walk off the end of
a page and cause a kernel fault.
5. SAR 224: The ALLINTSONCPU0 kernel tunable was not functioning
properly. It was not targeting non-assigned interrupts to CPU 0 as
it should.
6. SAR 225: Multi-threaded and/or multi-LWP applications can cause the
system to hang if a user-level debugger stop operation is attempted
while one of the other threads/LWPs is currently blocked in the
kernel on a sigtimedwait(2) or sigwait(2) call.
7. SAR 226: Trying to fork() a multi-threaded process may hang the
process. If one thread is currently in
thr_join(3thread)/pthread_join(3pthread) when another thread in the
same process calls fork(2), then the thread blocked in thr_join()
causes the fork() operation to never complete. Additionally, if the
thread inside thr_join() call is a multiplexing (MUX, not bound)
thread, then it will run continuously, consuming CPU cycles.
8. SAR 230: The lwp_steal() function attempts to find work for an idle
cpu in a multi-processor system. It does this by taking a waiting
and ready LWP from a busy processor's queue and placing it on the
idle processor's queue. The manipulation of these queues was done
mostly without the protection of a lock and could result in mangled
processor queues and a hung system.
9. SAR 250: Under VME bus traffic load generated by three Peregrine
FDDI adapters running at full speed, a PowerMAXION system would panic
with a VME BUS timeout.
10. xfsrestore(1M) may coredump when trying to restore filesystems with
inode numbers higher than 65535. The dummy entry at the end of a
directory block in the dump is not handled correctly in xfsrestore
when the inode number is 64K or larger, resulting in an alignment
fault.
11. The kernel would panic in page_alloc_blind(). The kernel keeps
various lists of free pages to give to applications as they need
them. In one rare circumstance during the walking of this list,
looking for a page of the appropriate type was not protected by a
semaphore for the entire length of the walk. This opened the
possibility of a corrupted freelist which would cause the panic.
12. Applications sometimes stall waiting for memory forever. On systems
where most of memory was locked down in applications, applications
not so locked down would sometimes stall out forever, waiting for
memory even though there was some memory available for them to use.
The cause: locked pages not only lock down a physical page, but also
a PTE in the system page table. If most of memory is locked down in
applications, then most of the PTEs will be also locked down. Since
any one virtual address has only 16 PTEs into which it can be placed,
if all 16 happened to be in use by other locked pages then the new
page being faulted in would never be able to complete until some
application owning one of the 16 locked pages would release the lock.
13. nanosleep(3C) was failing when fast interval timers were installed on
TurboHawk systems.
14. There were failures where processor run queues were getting
destroyed. The problem was that a single lwp was added to the
runqueues twice. This is a problem that leads to the run queues
becoming corrupted. Failures occurred between the interaction with
selfblock() (from nanosleep(), etc.) and a process that becomes
vm-seized.
15. Certain third-party VME boards can not be set to the TurboHawk's
default A32 address range on expansion VME busses. This prevented
the boards from being used on the TurboHawk platform.
Problem Resolution:
1. The scsi_scan() function of the gd driver would sometimes return
IMERR_DEVICE_BUSY during the boot scan for disks. The scsi_scan()
function was modified such that if the IMERR_DEVICE_BUSY was returned
initially then up to five more retries would be attempted.
2. The kernel now properly handles anchored user pages with replicated
versions existing in more than one memory pool.
3. During system initialization, the last physical page of the last
local memory pool in the system is marked unavailable, thus
preventing any I/O operations from occurring on this last page.
Additionally, any local memory pool's last physical page will also be
marked unavailable if there is a hole/space between it and the start
of the next local memory pool.
4. The rtserialinput() routine now checks the write pointer value when
first entered, and if it is beyond the end of the physical queue, it
is adjusted to the start of the queue before any of the data are
written to the queue.
5. The kernel initialization code that assigns interrupts to CPUs now
pays attention to the ALLINTSONCPU0 kernel tunable when assigning
interrupts to CPUs. Interrupts that are not limited to specific CPUs
and that are not explicitly assigned are now setup to interrupt CPU 0.
6. The kernel's sigtimedwait() routine was ignoring stop events when
checking for signals/events for the current LWP. This would cause
the LWP to loop/run forever inside the sigtimedwait() kernel routine.
The sigtimedwait() routine now checks for EVF_PL_PRSTOP events
(debugger stop events), and returns back out of this sigtimedwait()
routine so that the kernel can now successfully stop the LWP.
7. The thread library's thr_join() routine (which handles both
thr_join(3thread) and pthread_join(3pthread) calls) was re-coded so
that it can now process pending signals while waiting to join with
another thread. This change lets the fork(2) processing (which
requires all other threads in the same process to receive a signal
and subsequently 'rendezvous' in the kernel while a duplicate image
of the process is made in order to create the child process) to
proceed successfully. The thr_join() routine was previously blocking
signals/events that were posted to this LWP/thread, thus preventing
the fork(2) operation from completing.
8. Serialize the operation of lwp_steal() so that multiple processors do
not walk over each other should two or more processors be running in
lwp_steal() at the same time.
9. PowerMAXIONs did not get the VME BUS timeout fix that all other Night
Hawk platforms received in PowerMAX OS 4.3. This update extends the
fix to that platform. In addition, the timing windows that the fix
opened up for VME bus traffic to use were too small and too widely
spaced apart in time to protect against VME bus timeouts in large
configurations.
10. Change made to xfsrestore to correct the processing of dummy entries
at the end of a directory block.
11. The freelist walking algorithm was modified to be fully protected by
the appropriate semaphore.
12. Added a new tunable to PowerMAX OS, PTE_TABLESZ_MULT, which when >1
specifies a multiple of how much larger the page table is to be made
beyond its default size.
13. nanosleep() now correctly reference fast interval timers.
14. The fix is to totally remove the "wait" state information when an
lwp becomes seized.
15. Three variables contained in the file /etc/conf/pack.d/vme/space.c
can be modified to specify the start address, address range, and an
offset for A32 addressing on expansion VME busses on TurboHawks.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/mem
/etc/conf/mtune.d/vme
/etc/conf/pack.d/bsp6408/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/rtserial/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/vme/space.c
/etc/conf/pack.d/xfs/Driver.o
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/sys/fs/xfs_types.h
/usr/lib/libc.so.1
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/fs/xfs/xfsrestore
Special Conditions for Installation:
The following object(s) will also be replaced by base-003:
* /usr/ccs/lib/libp/libc.so
* /usr/lib/libposix1c.a
* /usr/lib/libposix1c.so
* /usr/sbin/xfsrestore
Possible Side Effects:
1. Reference Problem #3: The system may mark the last physical page of
one or more local memory pools unavailable at system initialization
time. This may cause a reserved memory section error to occur during
boot up if an application is also attempting to use/reserve the last
page in the local memory pool that has already been marked
unavailable. In this case, the application's last reserved memory
entry should be modified so that it no longer uses the last page of
that local memory pool.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-004
Date Issued: 10/05/2000 13:39:21
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, man-003, pg-003
Related SARs: 194, 230, 257, 277, 319, 322, 327, 338, 348, 369, 374,
392, 397
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 4
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P4" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 4".
New PowerMAX OS 4.3 Software Patch(es):
base-004 Base System Patch 004 (4.3P4)
cmds-003 Advanced Commands Patch 003 (4.3P4)
cnd-004 Condor Ethernet Driver Patch 004 (4.3P4)
crosslibs-003 Libraries for Cross Compiling Patch 003 (4.3P4)
crypt-002 Domestic Encryption Utilities Patch 002 (4.3P4)
crypt-int-002 International Encryption Utilities Patch 002 (4.3P4)
dec-003 DEC Ethernet Driver Patch 003 (4.3P4)
diskless-004 Diskless Systems Package Patch 004 (4.3P4)
egl-003 Eagle Ethernet Driver Patch 003 (4.3P4)
fibre-003 Fibre Channel Driver Patch 003 (4.3P4)
inet-004 Internet Utilities Patch 004 (4.3P4)
ip-003 Interphase 4511 PMC FDDI Driver Patch 003 (4.3P4)
man-003 On-line Manual Pages Patch 003 (4.3P4)
pg-003 Peregrine FDDI Driver Patch 003 (4.3P4)
New PowerMAX OS 4.3 Software Package(s):
rmxf RAMiX PMC665 FOB Ethernet Driver
Note: The "rmxf" package is supported on the Power Hawk 620 and
Power Hawk 640.
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 230: Multi-CPU systems may hang, especially if interrupt daemons
are enabled. The kernel's lwp_steal() routine was dropping IPL to
zero while still holding a light-weight process (lwp) lock. This
routine grabs ready to run lwps on an engine that is otherwise about
to enter the idle loop. This could cause the system to hang in an
interrupt routine when it tries to wakeup an interrupt daemon lwp, if
the l_mutex was already locked in lwp_steal().
3. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new
message/datagram.
4. SAR 277: When a NFS file system is unmounted, all file privileges
set in all the NFS-mounted file system are cleared from the
privileges system table. The privileges system table is searched
for a matching device entry. Since all NFS file systems are
assigned the same device value (zero), when one is unmounted all the
privileges for all the NFS-mounted file systems are cleared.
5. SAR 319: The system or application may hang when debugging an Ada
program. The rescheduling variable's rv_priority field is used by
ADA applications (in the 'ad' scheduling class) to temporarily raise
the scheduling priority of an lwp. This raised priority is used to
obtain exclusive access to a resource, such as a user-space spin
lock.
After a lwp with a raised rv_priority hits a debugger breakpoint,
upon continuing the application the original priority of this lwp
(not its raised temporary rv_priority) is used as the basis for
resuming this lwp in the kernel.
Additionally, when the application is continued, the 'wrong' lwp may
run before the lwp with the raised priority does, and a deadlock in
user-space can occur if the 'wrong' lwp also temporarily raises its
scheduling priority and attempts to grab the same resource in user
space.
6. SAR 322: All threads are duplicated when popen(3c) is called. When
a thread within a multi-threaded process calls popen(3C), all of the
threads are recreated in the new child process with a fork(2) call.
This can cause various potential problems for all of the other
threads in the parent process, depending upon what they were doing
at the time when the one thread called popen(3c). The fork(2) call
inside popen(3c) causes the parent process's lwps to all be pulled
into the kernel and put into a quiet state (rendezvoused) before the
fork operation proceeds. This activity disrupts the parent lwp's
operations, and also creates child process lwps that attempt to
execute the same code that the parent lwps where executing at the
time of the fork.
7. SAR 327: The PTIMER_MAX tunable was not working properly with
programs that linked with libthread.
8. SAR 338: Some processes were not being pre-empted by real-time
processes. A case wherein the fsflush() process was not being
pre-empted by some real-time process that was triggered through
either pgmtrigger() or fbstrig() was found in a NightTrace
evaluation. This may have been one of many possible cases.
9. SAR 348: The RTCIOCMODCNT rtc ioctl(2) was not functioning properly
for integral real-time clocks (the Raven MPIC tick timers).
10. SAR 369: The idtune(1M) utility would not properly process the
autotune tunables from the /etc/conf/autotune.d directory.
11. SAR 374: Physical memory range starting at 0x02000000 could no
longer be reserved for shmbind(2).
12. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
13. The Data Address Breakpoint register (DABR) was being corrupted.
On systems with user-level interrupts enabled, if a debugger loaded
a value into a lwp's DABR register via the /procfs interface, then
this value would get corrupted as the lwp being debugged exited the
kernel.
14. Kernel sometimes panics during boot when enabling all interrupt
daemons. The resume() routine in the kernel may sometimes call the
save_fp_kcontext() routine with an invalid u-area structure argument
(passed in the r5 register). This may cause the kernel to write to
random locations in the kernel, or to panic when reading or writing
an invalid kernel address.
The passing of an invalid u-area address occurs when the old lwp has
a kernel continuation routine address, which is the case for some
kernel interrupt daemons when they start back up after waiting for
new work to do.
15. The PCI subsystem does not currently handle multi-function PCI
devices where one of the functions is a PCI-to-PCI bridge. This is
a valid configuration and should be supported.
16. strncpy(3C) overruns zero fill for short lengths. strncpy(s1,s2,n)
for small values of n will fill an extra position in the s1 output
buffer with a 0.
17. The "Save Tunables" option in config(1M) only saved the tunable name
and abstract, not the values, which was inconsistent with the
"Display Tunables" option.
18. Autoload of the xfsd driver in a closely-coupled (CCS) environment
hangs all systems hard. When a distributed xfs file system is
mounted on a target system in a CCS environment the xfsmountd daemon
is invoked on the server system. This daemon will attempt to load
the xfsd driver on the server system if it isn't already loaded and
if the kernel is not statically linked. After the modld() is
performed the mod_lock is not released so all systems in the CCS
environment are hard hung requiring a power cycle to clear.
19. Under certain configurations, it was undesirable to have the console
initialization code probe for VME devices.
20. Intermittent hangs of the PCI bus would occur related to the PLX9080.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. Fix the IPL handling in the lwp_steal() routine so that IPL is held
high until the lwp l_mutex lock is unlocked.
3. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
4. For the special case of NFS-mounted file systems, an additional check
is now made to match the mount point number.
5. When a lwp with a temporarily raised scheduling priority is stopped
inside the kernel, the temporary priority is now used until that lwp
exits the kernel and returns back to user space.
When a debugger continues a process, the lwp that hit the breakpoint
is chosen to continue its execution first, before any of the other
lwps in the process resume their executions.
6. The the popen(3C) routine was modified to now use fork1(2) instead
of fork(2) when creating the new child process. The fork1(2) call
will create a child process with only the one calling lwp duplicated
in the new child process.
7. The _thr_posix_timer_init() routine in libthread was changed to pick
up _SC_POSIX_TIMER_MAX instead of _SC_TIMER_MAX. The sysconf()
routine in libc was changed to get the value of
_sysconfig(_CONFIG_POSIX_TIMER_MAX) instead of _POSIX_TIMER_MAX.
A new define was added to <sys/sysconfig.h> for
_CONFIG_POSIX_TIMER_MAX. A new case for _CONFIG_POSIX_TIMER_MAX was
added to get the current value of PTIMER_MAX from the kernel.
8. The problem was by the l_kpreempt flag not being set in the lwp that
was scheduled via sched_pgmadd() in the initial process. To fix this
the l_kpreempt flag setting from the parent lwp is copied to the
child lwp during lwp_setup().
9. The kernel rtc driver's routine that starts up a real-time clock
requires additional code in order to get around a glitch in the tick
timer hardware support. The cpu destination register for that tick
timer must temporarily be cleared while the new count value is
loaded into the timer's base count register.
10. Fixed the idtune script to call idreadauto with a real argument to
the -r option.
11. The problem occurred when the PowerPC PTE table was moved to address
0x02000000 in 4.3P3 (PowerMAX OS 4.3 Patch Set 3). This was a side
effect of a new boot-time kernel memory allocator which has a
different allocation algorithm designed to reduce memory waste.
The allocator has been changed to avoid the address range 0x020000000
whenever possible.
12. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
13. The kernel low-level code that loads up the DABR register before
returning to user-mode was modified so that it now properly loads the
correct value into the DABR register for the currently executing lwp.
14. Changed the resume() routine so that a valid u-area address is always
passed to the save_fp_kcontext() routine.
15. Add support for multi-function PCI devices where one of the devices
is a PCI-to-PCI bridge.
16. Corrected strncpy routine in C library.
17. Added code to include the current tunable value, either from the
mtune file or stune file if it had been changed, in the line that
gets written to the save file.
18. In the xfs driver's xfsps_ioctl() routine the xfsd mod_lock was
released after a successful call to modld(). This allows the
xfsps_ioctl() to complete with no hang of any of the systems in the
CCS environment.
19. VME configuration and probing has been suppressed within the console.
20. Suppressed Delayed Transaction Mode on the PLX9080. This
accommodated a shortcoming in Motorola's bus interface logic related
to PCI protocol.
Enhancements:
1. Enhance the ability for some blocking routines to use CLOCK_REALTIME.
CLOCK_REALTIME or CLOCK_UNIX can now be specified to use during the
timeout periods for the following routines:
aio_suspend(3)
aio_suspend64(3)
pthread_cond_timedwait(3pthread)
select(2)
Two new C Library routines and associated manual pages were added:
set_timeout_resolution(3C)
get_timeout_resolution(3C)
which allow setting and retrieving of the specific clock resolution
(CLOCK_UNIX or CLOCK_REALTIME) for any or all of the routines
identified above.
2. A new command (TMOUT_RESMSK) was added to syscx(2) for the two new
C library routines.
3. The libthread library was modified for the aio_suspend(3),
aio_suspend64(3), and pthread_cond_timedwait(3pthread) routines. The
svc, proc, and fs drivers were modified to accommodate the
selectivity of the CLOCK_REALTIME and CLOCK_UNIX clocks in the actual
kernel blocking routines.
Object(s) To Be Replaced:
/etc/conf/bin/idtune
/etc/conf/cf.d/cswtch.s
/etc/conf/cf.d/except.s
/etc/conf/pack.d/ad/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/pci/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/procfs/Driver.o
/etc/conf/pack.d/rcim/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/sum/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/config
/stand/cp1
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/include/sys/adapter_vme.h
/usr/include/sys/lwp.h
/usr/include/sys/posix_timers.h
/usr/include/sys/sysconfig.h
/usr/include/sys/syscx.h
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libsocket.so
/usr/lib/libthread.a
/usr/lib/libthread.so
Special Conditions for Installation:
1. Reference Problem #19: A new console image will be installed on
disk partition 6 of the boot device.
2. The following link(s) will also be updated by base-004:
/usr/ccs/lib/libp/libc.so --> /usr/ccs/lib/libc.so
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libposix1c.a --> /usr/lib/libthread.a
/usr/lib/libposix1c.so --> /usr/lib/libthread.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
1. Reference Problem #9: Also note that in order for a non-repetitive
integral rtc interrupt to work properly, the kernel's rtc interrupt
routine must be executed so that the rtc tick timer can be stopped
after the first occurrence of the interrupt. This requirement is due
to the fact that there is no hardware support in the tick timers for
a 'once only' count down. (This is not true for RCIM real-time
clocks.)
Therefore, if a user-level interrupt routine is used for receiving
the one-time rtc interrupt, the IC_KROUTINE flag must be set in the
ic_flags field of the icon_conn structure in order for the kernel's
rtc interrupt routine to also be executed after the user-level
interrupt routine processing returns back into the kernel.
(See iconnect(3C) for details.)
2. Reference Problem #19: Boot devices on VME SCSI controllers
(i.e. Concurrent's VIA controller) are no longer supported.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-004
Date Issued: 10/05/2000 13:56:42
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, inet-004,
man-003, pg-003
Related SARs: 194, 230, 256, 257, 277, 282, 319, 322, 327, 338, 340,
344, 369, 374, 392, 397
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
PowerMAX OS 4.3 Patch Set 4
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P4" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 4".
New PowerMAX OS 4.3 Software Patch(es):
base-004 Base System Patch 004 (4.3P4)
cmds-003 Advanced Commands Patch 003 (4.3P4)
cnd-004 Condor Ethernet Driver Patch 004 (4.3P4)
crosslibs-003 Libraries for Cross Compiling Patch 003 (4.3P4)
crypt-002 Domestic Encryption Utilities Patch 002 (4.3P4)
crypt-int-002 International Encryption Utilities Patch 002 (4.3P4)
dec-003 DEC Ethernet Driver Patch 003 (4.3P4)
egl-003 Eagle Ethernet Driver Patch 003 (4.3P4)
ie-003 Night Hawk ISE Ethernet Interface Module Patch 003
(4.3P4)
inet-004 Internet Utilities Patch 004 (4.3P4)
man-003 On-line Manual Pages Patch 003 (4.3P4)
pg-003 Peregrine FDDI Driver Patch 003 (4.3P4)
New PowerMAX OS 4.3 Software Package(s):
rmxf RAMiX PMC665 FOB Ethernet Driver
vmet VME to PCI/PMC Carrier Card
Note: The "rmxf" and "vmet" packages are supported on the PowerMAXION.
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 230: Multi-CPU systems may hang, especially if interrupt daemons
are enabled. The kernel's lwp_steal() routine was dropping IPL to
zero while still holding a light-weight process (lwp) lock. This
routine grabs ready to run lwps on an engine that is otherwise about
to enter the idle loop. This could cause the system to hang in an
interrupt routine when it tries to wakeup an interrupt daemon lwp, if
the l_mutex was already locked in lwp_steal().
3. SAR 256, 340: Under certain hardware configurations, TurboHawk
systems would hard hang during bcopy(3C), memcpy(3C), etc. on the
VME to global memory.
4. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new
message/datagram.
5. SAR 277: When a NFS file system is unmounted, all file privileges
set in all the NFS-mounted file system are cleared from the
privileges system table. The privileges system table is searched
for a matching device entry. Since all NFS file systems are
assigned the same device value (zero), when one is unmounted all the
privileges for all the NFS-mounted file systems are cleared.
6. SAR 282: The interrupt vectors for the ise driver are not reserved.
7. SAR 319: The system or application may hang when debugging an Ada
program. The rescheduling variable's rv_priority field is used by
ADA applications (in the 'ad' scheduling class) to temporarily raise
the scheduling priority of an lwp. This raised priority is used to
obtain exclusive access to a resource, such as a user-space spin
lock.
After a lwp with a raised rv_priority hits a debugger breakpoint,
upon continuing the application the original priority of this lwp
(not its raised temporary rv_priority) is used as the basis for
resuming this lwp in the kernel.
Additionally, when the application is continued, the 'wrong' lwp may
run before the lwp with the raised priority does, and a deadlock in
user-space can occur if the 'wrong' lwp also temporarily raises its
scheduling priority and attempts to grab the same resource in user
space.
8. SAR 322: All threads are duplicated when popen(3c) is called. When
a thread within a multi-threaded process calls popen(3C), all of the
threads are recreated in the new child process with a fork(2) call.
This can cause various potential problems for all of the other
threads in the parent process, depending upon what they were doing
at the time when the one thread called popen(3c). The fork(2) call
inside popen(3c) causes the parent process's lwps to all be pulled
into the kernel and put into a quiet state (rendezvoused) before the
fork operation proceeds. This activity disrupts the parent lwp's
operations, and also creates child process lwps that attempt to
execute the same code that the parent lwps where executing at the
time of the fork.
9. SAR 327: The PTIMER_MAX tunable was not working properly with
programs that linked with libthread.
10. SAR 338: Some processes were not being pre-empted by real-time
processes. A case wherein the fsflush() process was not being
pre-empted by some real-time process that was triggered through
either pgmtrigger() or fbstrig() was found in a NightTrace
evaluation. This may have been one of many possible cases.
11. SAR 344: Processes which are scheduled on an FBS will run one more
clock tick than they are scheduled with because the clock on a
TurboHawk actually counts down to zero instead of one.
12. SAR 369: The idtune(1M) utility would not properly process the
autotune tunables from the /etc/conf/autotune.d directory.
13. SAR 374: Physical memory range starting at 0x02000000 could no
longer be reserved for shmbind(2).
14. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
15. The Data Address Breakpoint register (DABR) was being corrupted.
On systems with user-level interrupts enabled, if a debugger loaded
a value into a lwp's DABR register via the /procfs interface, then
this value would get corrupted as the lwp being debugged exited the
kernel.
16. Kernel sometimes panics during boot when enabling all interrupt
daemons. The resume() routine in the kernel may sometimes call the
save_fp_kcontext() routine with an invalid u-area structure argument
(passed in the r5 register). This may cause the kernel to write to
random locations in the kernel, or to panic when reading or writing
an invalid kernel address.
The passing of an invalid u-area address occurs when the old lwp has
a kernel continuation routine address, which is the case for some
kernel interrupt daemons when they start back up after waiting for
new work to do.
17. The PCI subsystem does not currently handle multi-function PCI
devices where one of the functions is a PCI-to-PCI bridge. This is
a valid configuration and should be supported.
18. strncpy(3C) overruns zero fill for short lengths. strncpy(s1,s2,n)
for small values of n will fill an extra position in the s1 output
buffer with a 0.
19. The "Save Tunables" option in config(1M) only saved the tunable name
and abstract, not the values, which was inconsistent with the
"Display Tunables" option.
20. The <sys/ioacc.h> header file is now required to successfully build
the kernel if the dec driver is configured.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. Fix the IPL handling in the lwp_steal() routine so that IPL is held
high until the lwp l_mutex lock is unlocked.
3. Provide a new library /usr/lib/libvmecopy.a for TurboHawk systems
that provides the 6 routines that used floating point loads and
stores with newer slower routines that do not. These routines, if
-lvmecopy is supplied on the command line by the user (cc or ld),
will override the standard routines in libc. Also included in this
library are the routines:
extern void xbcopy (char *, char *, int);
extern void xbzero (char *, int);
extern void *xmemcpy(void *, const void *, size_t);
extern void *xmemmove(void *, const void *, size_t);
extern void *xmemset(void *, int, size_t);
extern char *xstrncpy(char *, const char *, size_t);
which have the original code, if the TurboHawk user desires to use
the original fast transfers for moves not involving VME transfers.
4. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
5. For the special case of NFS-mounted file systems, an additional check
is now made to match the mount point number.
6. Specific vectors for the ise driver are now reserved.
7. When a lwp with a temporarily raised scheduling priority is stopped
inside the kernel, the temporary priority is now used until that lwp
exits the kernel and returns back to user space.
When a debugger continues a process, the lwp that hit the breakpoint
is chosen to continue its execution first, before any of the other
lwps in the process resume their executions.
8. The the popen(3C) routine was modified to now use fork1(2) instead
of fork(2) when creating the new child process. The fork1(2) call
will create a child process with only the one calling lwp duplicated
in the new child process.
9. The _thr_posix_timer_init() routine in libthread was changed to pick
up _SC_POSIX_TIMER_MAX instead of _SC_TIMER_MAX. The sysconf()
routine in libc was changed to get the value of
_sysconfig(_CONFIG_POSIX_TIMER_MAX) instead of _POSIX_TIMER_MAX.
A new define was added to <sys/sysconfig.h> for
_CONFIG_POSIX_TIMER_MAX. A new case for _CONFIG_POSIX_TIMER_MAX was
added to get the current value of PTIMER_MAX from the kernel.
10. The problem was by the l_kpreempt flag not being set in the lwp that
was scheduled via sched_pgmadd() in the initial process. To fix this
the l_kpreempt flag setting from the parent lwp is copied to the
child lwp during lwp_setup().
11. The LOAD_PRTC_CLOCK_COUNT and GET_CLOCK_COUNT macros defined in
<sys/rtcmacros.h> have been changed such that they will subtract or
add one to the specified delay, respectively.
12. Fixed the idtune script to call idreadauto with a real argument to
the -r option.
13. The problem occurred when the PowerPC PTE table was moved to address
0x02000000 in 4.3P3 (PowerMAX OS 4.3 Patch Set 3). This was a side
effect of a new boot-time kernel memory allocator which has a
different allocation algorithm designed to reduce memory waste.
The allocator has been changed to avoid the address range 0x020000000
whenever possible.
14. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
15. The kernel low-level code that loads up the DABR register before
returning to user-mode was modified so that it now properly loads the
correct value into the DABR register for the currently executing lwp.
16. Changed the resume() routine so that a valid u-area address is always
passed to the save_fp_kcontext() routine.
17. Add support for multi-function PCI devices where one of the devices
is a PCI-to-PCI bridge.
18. Corrected strncpy routine in C library.
19. Added code to include the current tunable value, either from the
mtune file or stune file if it had been changed, in the line that
gets written to the save file.
20. <sys/ioacc.h> is now included in the "base" package for PowerMAX OS
on the Night Hawk family of systems.
Enhancements:
1. Enhance the ability for some blocking routines to use CLOCK_REALTIME.
CLOCK_REALTIME or CLOCK_UNIX can now be specified to use during the
timeout periods for the following routines:
aio_suspend(3)
aio_suspend64(3)
pthread_cond_timedwait(3pthread)
select(2)
Two new C Library routines and associated manual pages were added:
set_timeout_resolution(3C)
get_timeout_resolution(3C)
which allow setting and retrieving of the specific clock resolution
(CLOCK_UNIX or CLOCK_REALTIME) for any or all of the routines
identified above.
2. A new command (TMOUT_RESMSK) was added to syscx(2) for the two new
C library routines.
3. The libthread library was modified for the aio_suspend(3),
aio_suspend64(3), and pthread_cond_timedwait(3pthread) routines. The
svc, proc, and fs drivers were modified to accommodate the
selectivity of the CLOCK_REALTIME and CLOCK_UNIX clocks in the actual
kernel blocking routines.
4. Enhancements were made for TurboHawks to turn off prefetch reads from
the local bus and allow cache reads of two lines (64 bytes), thereby
improving data throughput.
Object(s) To Be Replaced:
/etc/conf/bin/idtune
/etc/conf/cf.d/cswtch.s
/etc/conf/cf.d/except.s
/etc/conf/cf.d/ivt.s
/etc/conf/pack.d/ad/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/pci/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/procfs/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/sum/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/vme/Driver.o
/sbin/config
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/include/sys/adapter_vme.h
/usr/include/sys/ioacc.h
/usr/include/sys/lwp.h
/usr/include/sys/posix_timers.h
/usr/include/sys/rtcmacros.h
/usr/include/sys/sysconfig.h
/usr/include/sys/syscx.h
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libsocket.so
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/libvmecopy.a
Special Conditions for Installation:
1. The following link(s) will also be updated by base-004:
/usr/ccs/lib/libp/libc.so --> /usr/ccs/lib/libc.so
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libposix1c.a --> /usr/lib/libthread.a
/usr/lib/libposix1c.so --> /usr/lib/libthread.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
1. Reference Problem #3: The use of the move routines provided in
libvmecopy.a will slow down system performance, since they also
replace the routines that are called in the libraries
(e.g. printf(3S)).
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-005
Date Issued: 11/20/2000 15:52:23
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Related SARs: 383, 423
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 5
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P5" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 5".
New PowerMAX OS 4.3 Software Patch(es):
base-005 Base System Patch 005 (4.3P5)
crypt-003 Domestic Encryption Utilities Patch 003 (4.3P5)
crypt-int-003 International Encryption Utilities Patch 003 (4.3P5)
inet-005 Internet Utilities Patch 005 (4.3P5)
rpc-001 Remote Procedure Calls Utilities Patch 001 (4.3P5)
via-001 VIA SCSI Adapter Interface Patch 001 (4.3P5)
#############################################################################
Problem Description:
1. SAR #383: System sometimes hangs when using Power Hawk Graphics
display. The problem turned out to be a generic problem with the
kernel's callout queue processing; this problem is not specific to
or limited to the Power Hawk graphics display support.
The hangs were due to a deadlock between the processing of a callout
queue entry timeout expiration versus the untimeout processing of
this same callout queue entry on the same CPU.
The problem was that the dispcallout() routine was allowing
preemptions to occur at a preemption point in the routine after the
callout queue entry had already been marked as "in progress".
If the preemption resulted in the execution of a untimeout call on
the same callout queue entry, then a deadlock/hang would occur where
the untimeout_r() routine would spin forever, waiting for the callout
queue entry to be marked as no longer "in progress".
2. SAR #423: A previous enhancement for use of CLOCK_REALTIME timeout
accuracy in pthread_cond_timedwait() turned out not to be so
accurate for millisecond timeout values. An initial call would
incorrectly return 0. A second subsequent call would return after
the appropriate timeout occurred. This overhead was unacceptable.
3. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
4. A change introduced in libthread in Patch Set 4 (4.3P4) for
PTIMER_MAX resulted in a backward compatibility problem for objects
statically linked with libthread.
5. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
6. cprs(1) changes the offsets in the PUBNAME pairs incorrectly; it
results in having the offsets point 11 bytes short of the
corresponding DIE.
Problem Resolution:
1. The dispcallout() routine was modified so that it doesn't mark the
current callout queue entry "in progress" until after it checks for
and allows any preemptions to occur.
The ipl level of the callout queue entry is held up at the callout
queue's ipl level until the expiration is processed and the entry is
marked as no longer "in progress".
2. The accuracy of the timeout values for CLOCK_REALTIME has been tuned
so that it is more accurate for millisecond timeout values. The
wakeup routine placed on the hrt_callout queue during block() was
changed to hrt_server_timedsetrun() which avoids the problem of
the incorrect return on the first call to pthread_cond_timedwait().
3. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
4. libthread has been modified such that if an object statically linked
with the libthread from 4.3P5 can be run on a system that is not up
to 4.3P5 yet.
5. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
6. Make the PUBNAME offsets relative to the DIEs and not the compilation
unit.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/deflist
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/usr/ccs/bin/cprs
/usr/include/altivec.h
/usr/include/cassert
/usr/include/cctype
/usr/include/cerrno
/usr/include/cfloat
/usr/include/ciso646
/usr/include/climits
/usr/include/clocale
/usr/include/cmath
/usr/include/csetjmp
/usr/include/csignal
/usr/include/cstdarg
/usr/include/cstddef
/usr/include/cstdio
/usr/include/cstdlib
/usr/include/cstring
/usr/include/ctime
/usr/include/cwchar
/usr/include/cwctype
/usr/include/dwarf.h
/usr/include/dwarfwrite.h
/usr/include/errno.h
/usr/include/float.h
/usr/include/if_mls.h
/usr/include/libdwarf.h
/usr/include/libgen.h
/usr/include/limits.h
/usr/include/lwpsynch.h
/usr/include/machlock.h
/usr/include/macros.h
/usr/include/malloc.h
/usr/include/math.h
/usr/include/rpcsvc/rstat.h
/usr/include/stdarg.h
/usr/include/stdlib.h
/usr/include/synch.h
/usr/include/sys/cmn_err.h
/usr/include/sys/lwp_synch.h
/usr/include/sys/mman.h
/usr/include/sys/ncr53c700.h
/usr/include/sys/typedefs/typedef__simplelock_t.h
/usr/include/sys/typedefs/typedef_id_t.h
/usr/include/sys/typedefs/typedef_llong_t.h
/usr/include/sys/typedefs/typedef_lwp_mutex_t.h
/usr/include/sys/typedefs/typedef_mutex_t.h
/usr/include/sys/typedefs/typedef_rmutex_t.h
/usr/include/sys/typedefs/typedef_thread_t.h
/usr/include/sys/typedefs/typedef_thrq_elt_t.h
/usr/include/sys/typedefs/typedef_ullong_t.h
/usr/include/sys/types.h
/usr/include/tdesc.h
/usr/include/thread.h
/usr/include/ttysrv.h
/usr/include/unistd.h
/usr/include/values.h
/usr/include/varargs.h
/usr/include/wchar.h
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libsocket.so
/usr/lib/libthread.a
/usr/lib/libthread.so
Special Conditions for Installation:
The following link(s) will also be updated by base-005:
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libposix1c.a --> /usr/lib/libthread.a
/usr/lib/libposix1c.so --> /usr/lib/libthread.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-005
Date Issued: 11/20/2000 16:00:21
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: 383, 423
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 5
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P5" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 5".
New PowerMAX OS 4.3 Software Patch(es):
base-005 Base System Patch 005 (4.3P5)
crypt-003 Domestic Encryption Utilities Patch 003 (4.3P5)
crypt-int-003 International Encryption Utilities Patch 003 (4.3P5)
inet-005 Internet Utilities Patch 005 (4.3P5)
rpc-001 Remote Procedure Calls Utilities Patch 001 (4.3P5)
via-001 VIA SCSI Adapter Interface Patch 001 (4.3P5)
#############################################################################
Problem Description:
1. SAR #383: System sometimes hangs when using Power Hawk Graphics
display. The problem turned out to be a generic problem with the
kernel's callout queue processing; this problem is not specific to
or limited to the Power Hawk graphics display support.
The hangs were due to a deadlock between the processing of a callout
queue entry timeout expiration versus the untimeout processing of
this same callout queue entry on the same CPU.
The problem was that the dispcallout() routine was allowing
preemptions to occur at a preemption point in the routine after the
callout queue entry had already been marked as "in progress".
If the preemption resulted in the execution of a untimeout call on
the same callout queue entry, then a deadlock/hang would occur where
the untimeout_r() routine would spin forever, waiting for the callout
queue entry to be marked as no longer "in progress".
2. SAR #423: A previous enhancement for use of CLOCK_REALTIME timeout
accuracy in pthread_cond_timedwait() turned out not to be so
accurate for millisecond timeout values. An initial call would
incorrectly return 0. A second subsequent call would return after
the appropriate timeout occurred. This overhead was unacceptable.
3. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
4. A change introduced in libthread in Patch Set 4 (4.3P4) for
PTIMER_MAX resulted in a backward compatibility problem for objects
statically linked with libthread.
5. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
6. cprs(1) changes the offsets in the PUBNAME pairs incorrectly; it
results in having the offsets point 11 bytes short of the
corresponding DIE.
Problem Resolution:
1. The dispcallout() routine was modified so that it doesn't mark the
current callout queue entry "in progress" until after it checks for
and allows any preemptions to occur.
The ipl level of the callout queue entry is held up at the callout
queue's ipl level until the expiration is processed and the entry is
marked as no longer "in progress".
2. The accuracy of the timeout values for CLOCK_REALTIME has been tuned
so that it is more accurate for millisecond timeout values. The
wakeup routine placed on the hrt_callout queue during block() was
changed to hrt_server_timedsetrun() which avoids the problem of
the incorrect return on the first call to pthread_cond_timedwait().
3. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
4. libthread has been modified such that if an object statically linked
with the libthread from 4.3P5 can be run on a system that is not up
to 4.3P5 yet.
5. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
6. Make the PUBNAME offsets relative to the DIEs and not the compilation
unit.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/deflist
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/usr/ccs/bin/cprs
/usr/include/altivec.h
/usr/include/cassert
/usr/include/cctype
/usr/include/cerrno
/usr/include/cfloat
/usr/include/ciso646
/usr/include/climits
/usr/include/clocale
/usr/include/cmath
/usr/include/csetjmp
/usr/include/csignal
/usr/include/cstdarg
/usr/include/cstddef
/usr/include/cstdio
/usr/include/cstdlib
/usr/include/cstring
/usr/include/ctime
/usr/include/cwchar
/usr/include/cwctype
/usr/include/dwarf.h
/usr/include/dwarfwrite.h
/usr/include/errno.h
/usr/include/float.h
/usr/include/if_mls.h
/usr/include/libdwarf.h
/usr/include/libgen.h
/usr/include/limits.h
/usr/include/lwpsynch.h
/usr/include/machlock.h
/usr/include/macros.h
/usr/include/malloc.h
/usr/include/math.h
/usr/include/rpcsvc/rstat.h
/usr/include/stdarg.h
/usr/include/stdlib.h
/usr/include/synch.h
/usr/include/sys/cmn_err.h
/usr/include/sys/lwp_synch.h
/usr/include/sys/mman.h
/usr/include/sys/ncr53c700.h
/usr/include/sys/typedefs/typedef__simplelock_t.h
/usr/include/sys/typedefs/typedef_id_t.h
/usr/include/sys/typedefs/typedef_llong_t.h
/usr/include/sys/typedefs/typedef_lwp_mutex_t.h
/usr/include/sys/typedefs/typedef_mutex_t.h
/usr/include/sys/typedefs/typedef_rmutex_t.h
/usr/include/sys/typedefs/typedef_thread_t.h
/usr/include/sys/typedefs/typedef_thrq_elt_t.h
/usr/include/sys/typedefs/typedef_ullong_t.h
/usr/include/sys/types.h
/usr/include/tdesc.h
/usr/include/thread.h
/usr/include/ttysrv.h
/usr/include/unistd.h
/usr/include/values.h
/usr/include/varargs.h
/usr/include/wchar.h
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libsocket.so
/usr/lib/libthread.a
/usr/lib/libthread.so
Special Conditions for Installation:
The following link(s) will also be updated by base-005:
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libposix1c.a --> /usr/lib/libthread.a
/usr/lib/libposix1c.so --> /usr/lib/libthread.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: base-006
Date Issued: 03/29/2001 11:50:43
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: cmds-004, crosslibs-004, fibre-001, fbs-003, inet-006,
man-004, ncr-001, trace-004
Related SARS: #167, #273, #395, #401, #433, #444,
#459, #464, #476, #499, #500
Brief Description:
PowerMAX OS 4.3 base package release updates.
############################################################################
PowerMAX OS 4.3 Patch Set 6
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P6" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 6".
New PowerMAX OS 4.3 Software Patch(es):
base-006 Base System Patch 006 (4.3P6)
cmds-004 Advanced Commands Patch 004 (4.3P6)
crosslibs-004 Libraries for Cross Compiling Patch 004 (4.3P6)
fibre-004 Fibre Channel Driver Patch 004 (4.3P6)
fbs-003 Frequency Based Scheduler Patch 003 (4.3P6)
inet-006 Internet Utilities Patch 006 (4.3P6)
man-004 On-line Manual Pages Patch 004 (4.3P6)
ncr-001 NCR Internal SCSI Controller Patch 001 (4.3P6)
trace-004 KernelTrace Utilities Patch 004 (4.3P6)
#############################################################################
Problem Description:
1. SAR #433: CLOCK_REALTIME for aio_suspend()/aio_suspend64()
/pthread_cond_timedwait() is not accurate at millisecond timeout
values.
2. There is no support for vme_address(3C) on moto systems.
3. SAR #464: The 5.2 compiler is more strict and generates an error if
there are 2 of the same "using std::name", declarations in the same
scope. This happens for sigaction, one in <signal.h> and one in
<sys/signal.h>.
4. SAR #441: The client fbsd daemon sometimes failed to start up
successfully. Under a customer's closely-coupled configuration,
the client SBC's fbs daemon (fbsd) would frequently fail to start
up successfully. When this problem occurred, it would still be
possible to manually restart the fbsd daemon after the client SBC
finished booting.
5. SAR #476: Nightview sometimes hangs while attempting to attach to
a running application while it is currently creating Ada tasks.
Sometimes a Nightview dynamic attach to a process that contains
some LWPs currently executing _lwp_create(2) calls may never
complete the attach operation. Nightview will appear to be hung.
Similarly, if a program is started up with Nightview, and a
breakpoint is hit while some of the other LWPs in the process are
currently executing a _lwp_create(2) operation, Nightview will
never complete the breakpoint operation, and will appear to be hung.
6. SAR #342: pthread.h - when using c++ gives an error for multiple
declarations of using std::pthread_once;
7. SARs #167, #273, #395, and #444:
The TurboHawk (6800t) and PowerMAXION (6400 & 6408)
systems failed to properly detect and handle VME bus faults.
8. SAR #401: Strange behaviour was observed with the real-time
clocks when using the rtc(7) interfaces. In count down mode:
0c#: read count value (by ioctl RTCIOCGETCNT) is incorrect after
every second expiration, but the expirations are correct. In
non-repeat mode, the ioctl RTCIOCWAIT fails with "I/O error"
after every second ioctl RTCIOCSET (when it fails the read count
value is incorrect). 1c#: never expires after a RTCIOCGETCNT
ioctl has been called. The timer has stopped and the ioctl
RTCIOCSTART must be called again which reset(!) the counter for
1c# timers. 2c#: ok. In count up mode: 0c#: counts DOWN (doesn't
count up) and expires. 1c#: RTCIOCSET fails with "Invalid argument".
2c#: counts up, but never expires.
9. Kernel trace data did not include full (64-bit) timestamps. Comb-
ining kernel and user trace data logged from significantly disjoint
time periods caused erroneous kernel event times when displayed with
NightTrace.
10. When doing a system backup with xfsdump using the -M option,
erroneous blocksizes were returned. This problem did not occur
using the -B option.
11. The kernel did not save and restore the PowerPC Data Address Break-
point Register (DABR) on entry and exit from the kernel.
Problem Resolution:
1. The CLOCK_REALTIME functionality for aio_suspend()/aio_suspend64()
has been implemented.
2. The vme_address(3C) routine in libc has been enhanced for Night Hawk
systems and support has been provided for Power Hawk systems.
3. To prevent this each using declaration is protected.
4. The problem was that a sbcmsg(2) call that the client fbsd makes was
returning ENOMEM. The client fbsd daemon makes this call to check
that the fbsd server mailbox has already been reserved by the server
SBC fbsd daemon. This error can occur when the server SBC receives
this sbc message and the processing attempts to process this server
mailbox reservation check at a point in time when the server's kernel
is temporarily out of kernel heap freelist blocks. The fbsd code was
re-coded so that it will now recognize the ENOMEM error as a non-fatal
error, and instead of giving up, it will now retry the server mailbox
reservation sbcmsg(2) call until it succeeds or until a more serious
fatal error is returned.
5. The _lwp_create() kernel routine calls spawn_lwp() to create the new
lwp structure. The spawn_lwp() routine was modified so that it now
pulls over the EVF_PL_PRSTOP event flag from the creating lwp into
the new lwp's lwp structure, if it is currently set in the creating
lwp's lwp structure.
The reason that this change is necessary is due to the fact that
when another lwp in the same process hits a breakpoint and sub-
sequently attempts to set the debug stop event flag in all of the
other lwp structures belonging to that process, it may miss setting
the EVF_PL_PRSTOP flag in a new lwp's structure if that new lwp is
not yet completely created, and is thus not in the process's lwp
linked list. Therefore, if the EVF_PL_PRSTOP event flag is not
inherited from creating lwp's lwp structure, then the new lwp does
not see the debug stop request, and Nightview ends up waiting
forever for all lwps to enter the debug stopped state.
Note that a Nightview dynamic attach operation also makes use of an
internally generated breakpoint during the attach processing. This
is reason that dynamic attach operations have the same debug stop
event flag problem.
6. Removed the extra reference.
7. Provided VME bus sysfault support for the expansion VME bus and
refined the same support for the primary VME bus. This was done
for TurboHawk (6800t) and PowerMAXION (6400 and 6408) architectures.
8. The rtc/Driver.o was corrected for all the various problems
listed in the above description.
9. The kernel was modified to mark each kernel trace buffer with a full
(64-bit) timestamp via a new event, TR_HO_TIME (37). The ntfilter
utilitiy was modified to consume the timestamp and attribute the
header of the filtered trace data file accordingly, thus indicating
the absolute time frame for the kernel data.
10. This problem was rooted in an overflow produced by the MBtoB macro,
which was attempting to store a value larger than a long in a long
variable. That variable was asize, which is used only to report the
size in a printf statement. Changes were made to dump.h in xfsdump
and fsdump, and dumpmain.c in fsdump. Because these are size values
I made all variables unsigned, which eliminates the overflow.
11. The kernel now appropriate saves and restores the DABR register on
entry and exit from the kernel. This changes is required to support
the "Watchpoint" feature which was recently added to the NightView
debugger, version 5.3.
Object(s) To Be Replaced:
/usr/include/pthread.h
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/usr/include/signal.h
/usr/include/sys/signal.h
/usr/include/sys/iobus_err.h
/etc/conf/cf.d/sc.s
/etc/conf/cf.d/intr.s
/etc/conf/cf.d/except.s
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/lib/libc.so.1
/usr/lib/fs/xfs/xfsdump
/usr/lib/fs/ufs/fsdump
Special Conditions for Installation:
The following objects are linked to some of the objects being
replaced by this patch. These objects will also be re-linked:
/usr/lib/libposix1c.a
/usr/lib/libposix1c.so
/usr/sbin/xfsdump
/usr/sbin/fsdump
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-006
Date Issued: 03/28/2001 18:56:12
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Patch Set Number: 4.3P6
Related Patches: cmds-004. crosslibs-004, fibre-001, inet-006,
man-004, ncr-001, trace-004
Related SARS: #167, #273, #395, #401, #433, #444,
#342, #464, $342
Brief Description:
PowerMAX OS 4.3 base package release updates.
#############################################################################
PowerMAX OS 4.3 Patch Set 6
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P6" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 6".
New PowerMAX OS 4.3 Software Patch(es):
base-006 Base System Patch 006 (4.3P6)
cmds-004 Advanced Commands Patch 004 (4.3P6)
crosslibs-004 Libraries for Cross Compiling Patch 004 (4.3P6)
fibre-001 Fibre Channel Driver Patch 001 (4.3P6)
inet-006 Internet Utilities Patch 006 (4.3P6)
man-004 On-line Manual Pages Patch 004 (4.3P6)
ncr-001 NCR Internal SCSI Controller Patch 001 (4.3P6)
trace-004 KernelTrace Utilities Patch 004 (4.3P6)
#############################################################################
Problem Description:
1. SAR #433: CLOCK_REALTIME for aio_suspend()/aio_suspend64()
/pthread_cond_timedwait() is not accurate at millisecond timeout
values.
2. There is no support for vme_address(3C) on moto systems.
3. SAR #464: The 5.2 compiler is more strict and generates an error if
there are 2 of the same "using std::name", declarations in the same
scope. This happens for sigaction, one in <signal.h> and one in
<sys/signal.h>.
4. SAR #476: Nightview sometimes hangs while attempting to attach to
a running application while it is currently creating Ada tasks.
Sometimes a Nightview dynamic attach to a process that contains
some LWPs currently executing _lwp_create(2) calls may never
complete the attach operation. Nightview will appear to be hung.
Similarly, if a program is started up with Nightview, and a
breakpoint is hit while some of the other LWPs in the process are
currently executing a _lwp_create(2) operation, Nightview will
never complete the breakpoint operation, and will appear to be hung.
5. SAR #342: pthread.h - when using c++ gives an error for multiple
declarations of using std::pthread_once;
6. SARs #167, #273, #395, and #444:
The TurboHawk (6800t) and PowerMAXION (6400 & 6408)
systems failed to properly detect and handle VME bus faults.
7. SAR #401: Strange behaviour was observed with the real-time
clocks when using the rtc(7) interfaces. In count down mode:
0c#: read count value (by ioctl RTCIOCGETCNT) is incorrect after
every second expiration, but the expirations are correct. In
non-repeat mode, the ioctl RTCIOCWAIT fails with "I/O error"
after every second ioctl RTCIOCSET (when it fails the read count
value is incorrect). 1c#: never expires after a RTCIOCGETCNT
ioctl has been called. The timer has stopped and the ioctl
RTCIOCSTART must be called again which reset(!) the counter for
1c# timers. 2c#: ok. In count up mode: 0c#: counts DOWN (doesn't
count up) and expires. 1c#: RTCIOCSET fails with "Invalid argument".
2c#: counts up, but never expires.
8. Kernel trace data did not include full (64-bit) timestamps. Comb-
ining kernel and user trace data logged from significantly disjoint
time periods caused erroneous kernel event times when displayed with
NightTrace.
9. When doing a system backup with xfsdump using the -M option,
erroneous blocksizes were returned. This problem did not occur
using the -B option.
10. The kernel did not save and restore the PowerPC Data Address Break-
point Register (DABR) on entry and exit from the kernel.
Problem Resolution:
1. The CLOCK_REALTIME functionality for aio_suspend()/aio_suspend64()
has been implemented.
2. The vme_address(3C) routine in libc has been enhanced for Night Hawk
systems and support has been provided for Power Hawk systems.
3. To prevent this each using declaration is protected.
4. The _lwp_create() kernel routine calls spawn_lwp() to create the new
lwp structure. The spawn_lwp() routine was modified so that it now
pulls over the EVF_PL_PRSTOP event flag from the creating lwp into
the new lwp's lwp structure, if it is currently set in the creating
lwp's lwp structure.
The reason that this change is necessary is due to the fact that
when another lwp in the same process hits a breakpoint and sub-
sequently attempts to set the debug stop event flag in all of the
other lwp structures belonging to that process, it may miss setting
the EVF_PL_PRSTOP flag in a new lwp's structure if that new lwp is
not yet completely created, and is thus not in the process's lwp
linked list. Therefore, if the EVF_PL_PRSTOP event flag is not
inherited from creating lwp's lwp structure, then the new lwp does
not see the debug stop request, and Nightview ends up waiting
forever for all lwps to enter the debug stopped state.
Note that a Nightview dynamic attach operation also makes use of an
internally generated breakpoint during the attach processing. This
is reason that dynamic attach operations have the same debug stop
event flag problem.
5. Removed the extra reference.
6. Provided VME bus sysfault support for the expansion VME bus and
refined the same support for the primary VME bus. This was done
for TurboHawk (6800t) and PowerMAXION (6400 and 6408) architectures.
7. The rtc/Driver.o was corrected for all the various problems
listed in the above description.
8. The kernel was modified to mark each kernel trace buffer with a full
(64-bit) timestamp via a new event, TR_HO_TIME (37). The ntfilter
utilitiy was modified to consume the timestamp and attribute the
header of the filtered trace data file accordingly, thus indicating
the absolute time frame for the kernel data.
9. This problem was rooted in an overflow produced by the MBtoB macro,
which was attempting to store a value larger than a long in a long
variable. That variable was asize, which is used only to report the
size in a printf statement. Changes were made to dump.h in xfsdump
and fsdump, and dumpmain.c in fsdump. Because these are size values
I made all variables unsigned, which eliminates the overflow.
10. The kernel now appropriate saves and restores the DABR register on
entry and exit from the kernel. This changes is required to support
the "Watchpoint" feature which was recently added to the NightView
debugger, version 5.3.
Object(s) To Be Replaced:
/usr/include/pthread.h
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/lib/libc.so.1
/usr/include/signal.h
/usr/include/sys/signal.h
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/bsp6400/Driver.o
/etc/conf/pack.d/bsp6408/Driver.o
/etc/conf/pack.d/bsp6800t/Driver.o
/etc/conf/pack.d/bsp6800/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/iobus_err/Driver.o
/etc/conf/pack.d/vme/Driver.o
/usr/include/sys/errid.h
/usr/include/sys/iobus_err.h
/etc/conf/cf.d/sc.s
/etc/conf/cf.d/intr.s
/etc/conf/cf.d/except.s
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/util/Driver.o
/usr/lib/fs/xfs/xfsdump
/usr/lib/fs/ufs/fsdump
Special Conditions for Installation:
The following objects are linked to some of the objects being
replaced by this patch. These objects will also be re-linked:
/usr/lib/libposix1c.a
/usr/lib/libposix1c.so
/usr/sbin/xfsdump
/usr/sbin/fsdump
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: base-007
Date Issued: 06/11/2001 06:49:52
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: fibre-005, ide-002, ncr-002, via-002
Related SARs: #227, #258
Brief Description:
PowerMAX OS 4.3 base package release updates
############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P7" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 7".
New PowerMAX OS 4.3 Software Patch(es):
base-007 Base System Patch 007 (4.3P7)
fibre-005 Fibre Channel Driver Patch 005 (4.3P7)
ide-002 Internal IDE/ATA Disk Controller Patch 002 (4.3P7)
ncr-002 NCR Internal SCSI Controller Patch 002 (4.3P7)
via-002 VIA SCSI Adapter Interface Patch 002 (4.3P7)
############################################################################
Problem Description:
1. SAR #227 - 1 - 1 CUSTOMER: Lockheed Martin Corp.
SCSI_SCAN on NH6400 systems taking too long. Turning
off the SCSISCAN tunable causes system panics.
2. SAR #258 - 3 - 3 CUSTOMER: Lockheed Martin Corp.
Mounting of a removable hard disk sometimes fails. If
the disk is scanned at boot time, it is fine, but if
the disk is inserted into the system while it's up and
running, the first mount command fails with "drive not
ready". The second mount will work. This was NOT a
problem on 4.2. Something has changed.
3. Magneto optical(MO) SCSI drives behave poorly. hangs,
and data corruption, unable to support media sector sizes
other than 512 bytes.
Problem Resolution:
1. SAR 227: System failed to boot if SCSISCAN in gd driver is
not enabled. System will now spin up fixed media disk drives at
first open, if not already started at IPL time.
2. SAR 258: Disk drives would not start if not connected to system
at IPL time. Logic to spin up drive was only present at IPL time.
Special logic is now located in first open/last close to provide
a fresh re-initialization of core control code.
3. Enhanced Magneto optical SCSI drive support.
Added code to support drive initiated sync negotiation before
data phase to NCR driver.
Added partial sector read/write I/O support to GD driver.
Now supports, 1K and 2K MO media sector sizes.
Added first open/last close logic for reload of critical structures.
Added code to covert reads/writes smaller than a sector size into
full sector reads or RMW operations.
Added extra code to support general physical DMA lists, and as
needed, ordered I/O processing in via, ncr, and is drivers.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/gd/space.c
/etc/conf/mtune.d/gd
/etc/conf/mdevice.d/gd
/usr/include/sys/gd.h
/etc/conf/pack.d/scsi/Driver.o
/usr/include/sys/im.h
/usr/include/sys/dsk.h
/etc/conf/pack.d/name/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: base-007
Date Issued: 06/11/2001 08:33:23
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: is-001, ncr-002, via-002
Related SARs: #227, #258
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 base package release updates
#############################################################################
PowerMAX OS 4.3 Patch Set 7
---------------------------
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P7" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 7".
New PowerMAX OS 4.3 Software Patch(es):
base-007 Base System Patch 007 (4.3P7)
is-001 Night Hawk ISE SCSI Interface Module Patch 001 (4.3P7)
ncr-002 NCR scsi Driver Patch 002 (4.3P7)
via-002 VIA scsi Driver Patch 002 (4.3P7)
#############################################################################
Problem Description:
1. SAR #227 - 1 - 1 CUSTOMER: Lockheed Martin Corp.
SCSI_SCAN on NH6400 systems taking too long. Turning
off the SCSISCAN tunable causes system panics.
2. SAR #258 - 3 - 3 CUSTOMER: Lockheed Martin Corp.
Mounting of a removable hard disk sometimes fails. If
the disk is scanned at boot time, it is fine, but if
the disk is inserted into the system while it's up and
running, the first mount command fails with "drive not
ready". The second mount will work. This was NOT a
problem on 4.2. Something has changed.
3. Magneto optical(MO) SCSI drives behave poorly. hangs,
and data corruption, unable to support media sector sizes
other than 512 bytes.
Problem Resolution:
1. SAR 227: System failed to boot if SCSISCAN in gd driver is
not enabled. System will now spin up fixed media disk drives at
first open, if not already started at IPL time.
2. SAR 258: Disk drives would not start if not connected to system
at IPL time. Logic to spin up drive was only present at IPL time.
Special logic is now located in first open/last close to provide
a fresh re-initialization of core control code.
3. Enhanced Magneto optical SCSI drive support.
Added code to support drive initiated sync negotiation before
data phase to NCR driver.
Added partial sector read/write I/O support to GD driver.
Now supports, 1K and 2K MO media sector sizes.
Added first open/last close logic for reload of critical structures.
Added code to covert reads/writes smaller than a sector size into
full sector reads or RMW operations.
Added extra code to support general physcal DMA lists, and as
needed, ordered I/O processing in via, ncr, and is drivers.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/gd/space.c
/etc/conf/mtune.d/gd
/etc/conf/mdevice.d/gd
/usr/include/sys/gd.h
/etc/conf/pack.d/scsi/Driver.o
/usr/include/sys/im.h
/usr/include/sys/dsk.h
/etc/conf/pack.d/name/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: base-008
Date Issued: 08/22/2001 11:33:01
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008, cmds-005, crosslibs-005, crypt-004,
crypt-int-004, dec-004, diskless-005, ide-003,
inet-007, man-005, netcmds-001, trace-005
Related SARs: #536, #495, #503, #520, #430, #108, #565, #520,
#498, #501
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 base package release updates
############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P8" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 8".
New PowerMAX OS 4.3 Software Patch(es):
base-008 Base System Patch 008 (4.3P8)
cmds-005 Advanced Commands Patch 005 (4.3P8)
crosslibs-005 Libraries for Cross Compiling Patch 005 (4.3P8)
crypt-004 Domestic Encryption Utilities Patch 004 (4.3P8)
crypt-int-004 International Encryption Utilities Patch 004 (4.3P8)
dec-004 DEC Ethernet Driver Patch 004 (4.3P8)
diskless-005 Diskless Systems Package Patch 005 (4.3P8)
fbs-004 Frequency Based Scheduler Patch 004 (4.3P8)
ide-002 Internal IDE/ATA Disk Controller Patch 002 (4.3P8)
inet-007 Internet Utilities Patch 007 (4.3P8)
man-005 On-line Manual Pages Patch 005 (4.3P8)
netcmds-001 Commands Networking Extension Patch 001 (4.3P8)
trace-005 KernelTrace Utilities Patch 005 (4.3P8)
############################################################################
Problem Description:
1. SAR #536:
In order for additional swap devices identified in /etc/vfstab
to be added during bootup, the /sbin/mountall script must be
modified. The line: /usr/sbin/swap -a $special must be changed
to /sbin/swap -a $special
2. SAR's #495, #503:
When an application containing gethostbyname() was statically
linked with libnsl.a and the gethostbyname() followed one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
3. The function prototypes for the set_timeout_resolution()
/get_timeout_resolution() routines are not properly declared in
an associated header file. For the new ec++ compiler this causes
compilation errors.
4. SAR #520:
A previous patch introduced a possibility in the dispatcher
wherein a real-time process could not preempt. The IPL
could be left at a level other than the one the dispatcher was
entered with. This resulted in disallowing real-time process
preemption occasionally.
5. SAR #430:
The pthread_cond_timedwait() routine exposed a race condition
on PowerMAXION systems such that it could block and not be resumed.
On a PowerMAXION system if a process was running on CPU 0 and
was programmed to block, the wakeup routine scheduled on the
global callout queue could have been called via hardclk before the
process had actually reached the blocking routine.
As such, when the process did block it could not have been awakened
again because its wakeup routine had already run.
6. SAR #108:
Symptom: an application with an sbrk(2) area that incrementally
grew into the hundreds of megabytes would occasionally
self-deadlock waiting for the next chunk of virtual memory to be
allocated to the sbrk(2) area of that application.
Problem: one of the kernel data structures needed to manage
application virtual address space would grow so large that there
was danger that not enough *kernel* virtual address space could
be found to hold the next incarnation of that table.
7. Symptom: mlock(2)ing a region of application memory that had
been created with the SHM_NCACHE flag of shmget(2) would panic
the system.
Problem: code in the kernel Driver mem/Driver.o that created PTEs
differed slightly from code nearby that would re-use an already
existing PTE; the net effect of this difference was confusion on
the kernel's part with regard to the cached/not-cached state of
the pages created by shmget(2).
8. Symptom: sometimes a large application distributed among a large
number of local memory pools would self-deadlock when an attempt
was made to lock the application down in memory.
Problem: The way PTEs were allocated was defective, resulting in
all of an applications PTEs being allocated in at most one eighth
of the available PTE buckets. When a bucket was full of locked
down PTEs, and Yet Another had to go into that particular bucket,
the application would wait for a free PTE to become available in
that bucket. That wait would never complete if the application
itself was the owner of all the PTEs already in the bucket.
9. Symptom: The cpu_bias(2) system call would sometimes return with
r1 corrupted when run on nighthawks and when the process was under
the control of a debugger.
Problem: reading and writing of another process's u-area was not
cache coherent when that u-area was in remote memory. Access to
another process's u-area is something done by only a debugger.
10. Symptom: the core dumps of some applications that utilized
dynamically loaded libraries could not be processed by the
debugger.
Problem: the core dump logic in the kernel erroneously assumed
that the text of dynamically loaded libraries was accurately
represented by the corresponding text sections in the library
files, and failed to save off those regions of the address space
to the core file.
11. On a console auto-reboot operation, such as a "shutdown ... -i6",
or on an auto-reboot after a system panic, the 2nd CPU was
entering the kernel's start routine with a bad 'msr' register value.
(The 'msr' register controls CPU attributes such as whether
or not virtual memory is enabled, and whether or not
the instruction and data caches are enabled, etc.)
In the auto-reboot case, a bad 'msr' register value could
cause us to write to random locations in memory while making
some early subroutine calls in the kernel's low-level startup
routine, before the 2nd CPU's registers could be properly
initialized by the kernel.
Eventually, the kernel would sometimes end up panic()ing in a
random fashion, when it tried to use the corrupted memory
locations.
12. SAR #565:
The thread that calls pthread_cond_timedwait() does not return
when a another thread calls pthread_cond_signal(), if the timeout
resolution is set to CLOCK_REALTIME with a
set_timeout_resolution(CLOCK_REALTIME, RUTN_PTHRCONDTWAIT) call.
In this case, the caller of pthread_cond_timedwait() doesn't
return until the timeout expires.
The problem was that the sq_wakeup() kernel routine that wakes up
a light-weight process (LWP) as the result of a
pthread_cond_signal() call was not accounting for the fact that
an LWP that had blocked itself with a high-resolution timeout
pending (CLOCK_REALTIME) would be in a different sleep state than
an LWP that had blocked itself with a low-resolution time
pending (CLOCK_UNIX).
Therefore, when the sq_wakeup() routine checked the sleep state
of the LWP that was blocked with the high-resolution timeout
pending, it erroneously decided that this LWP was not currently
blocked/sleeping in the kernel on the associated sleep queue,
and it failed to wakeup the blocked LWP.
13. SAR #520:
On multi-CPU systems it is sometimes possible to read the wrong
POSIX time of day. This is due to the fact that a race
condition existed between one CPU reading the three POSIX timer
assist page variables at the same time that these same variables
were being updated on the other CPU.
The three POSIX timer assist page variables are used in the
computation of the current POSIX time of day.
When a mix of old and new POSIX timer assist variable values
are read, then the wrong POSIX time of day is computed/obtained,
and operations such as nanosleep(3C) will not function properly.
14. The following kernel panics occurred on systems with
an enabled kmadbg kernel module:
- The kernel PANICed in the fbs_srv_unreg_dev() routine
during the un-registration of a FBS Coupled timing device.
This routine was referencing the fbs data structure of the
virtual fbs scheduler after this structure had already been
kmem_free()ed.
- The kernel PANICed in the memfs_dirtrunc() routine during
a closely-coupled client boot up sequence. This routine was
referencing a kma block after the block had already been freed.
- The kernel PANICed in the sbc_msg_recv() routine when
processing a received message; typically, on the server SBC
when the server SBC had received fbsd(1M) messages from a
client SBC in a closely-coupled system. This routine was
referencing a kmem_alloc()ed sbc message structure after
it had already been kmem_free()ed.
15. When a debugger issues a stop request on a LWP that is scheduled
by a FBS scheduler, the kernel may panic if the LWP has exited
by the time that the FBS module processes the FBS scheduler stop
operation.
If the LWP structure has already been kmem_free()ed, and
re-allocated for a different purpose, then examining the
l_fbslwpp field in the previous LWP structure can cause a kernel
data access exception panic.
16. On single CPU moto-based platforms, the allocation and
use of the next available kernel trace record would
sometimes allocate the same trace record structure twice,
thus corrupting the information stored into this same trace
record for two trace events.
This corruption would occur when the tracing of a spurious
interrupt event took place just as the CPU was allocating a
kernel trace record for another event. The method used to
serialize the kernel trace record allocation was unable to
successfully hold out spurious interrupts.
17. If one of the LWPs in a multi-lwp or multi-threaded
application hit a debugger breakpoint and the entire process
was stopped, and one of the other LWPs in the same process
is currently scheduled on a running FBS scheduler, the running
scheduler was not being stopped. The FBS scheduler would
continue to run, even though it had a LWP that had been placed
into a stopped state.
18. SAR#498:
A problem with clocks /dev/rrtc/1cN. The stop command does not
always work.
SAR#501:
An accuracy problem with the RCIM real time clock.
19. SAR #540:
The pci/Driver.o provided in Patch Set #4 causes a
target system w/PCI Expansion connected to panic.
20. Adding or removing PCI bridges can dramticly renumber the SCSI
device number convention. This can result in the GD driver
being open using old NODE numbers. The 'idmknod' does not
automaticly execute during boot sequence, since it is still
using the same kernel.
While opening invalid devices using bad node information is not
harmful, certain sequences in XFS used a gdsize entry point which
did not have similar protection. The result was a kernel crash
somewhere in beyond the single user boot phase, if the 'gd' nodes
had not been rebuilt.
21. netperf was hanging because allocb() calls were failing. This
occurred on the receiving system when heavy network traffic
occurred and the sending system was much faster than the
receiving system, or when the receiving system is low on
available memory. The receiving system was not processing this
load of packets fast enough such that a backup of queued packets
occurred. Eventually there were no whole pages available which
caused allocb to hang.
22. The number of cached rpcbind addresses was low. This could
have cause performance issues if an application was using
many different addresses (host, netid).
Problem Resolution:
1. Changed /sbin/mountall to reference /sbin/swap instead of
/usr/sbin/swap
2. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
3. The function prototypes for these routines have been properly
declared in /usr/include/sys/posix_timers.h so that compilation
with the ec++ compiler succeeds.
4. The dispatcher code was corrected so that the IPL upon exit
would be the same as upon entrance.
5. The problem was a misplaced lwp flag setting designed to
prevent this race condition. The L_USYNCTIMEDOUT lwp flag
should have been set upon entry to the wakeup routine,
usync_timedsetrun(), but there was a possible exit from this
routine without the lwp flag being set. The blocking routine,
sq_block(), checks for this flag before blocking in order to
prevent blocking if the wakeup had already occurred. This
patch sets the flag upon entry to usync_timedsetrun().
6. Added a new tuneable, MAX_SEGSIZE, which specifies the maximum
size an address space segment can reach before a new, adjacent
segment is silently born. This puts an upper limit of the size of
the segment data structure the kernel uses to track address
space segments. The default value of MAX_SEGSIZE is 64 megabytes.
7. Corrected code in the kernel Driver mem/Driver.o that created PTEs
to differ slightly from code nearby that would re-use an already
existing PTE.
8. Replaced the address space VSID allocator with a more robust
algorithm that evenly scatters PTE allocations amoung all the
available PTE buckets.
9. Modified remote u-area access code in mem/Driver.o to properly
flush remote memory before and after the access.
10. Core dumps now contain the text sections of dynamically
loaded libraries.
11. The console processor CPU startup code was changed for the
auto-reboot case so that the 2nd CPU's 'msr' register is now
properly initialized before it enters the kernel's low-level
CPU startup code.
12. The sq_wakeup() kernel routine was modified so that it now
properly recognizes both high-resolution and low-resolution
timeout LWPs that are awaiting either a wakeup or timeout to
occur.
13. The coding for reading and updating the POSIX timer assist page
variables was modified so that the reading of these variables
will always result in reading a coherent set of values, even
when they are being updated at roughly the same time on the
other CPU.
14. All three of these problems were re-coded so that they no longer
reference an already kmem_free()ed memory block. Any needed
information from these kma memory areas is now stored into a
local variable before the kma block is freed.
15. The fix is to place a lwp_hold() on the LWP structure until
the FBS stop operation completes. In this way, the LWP structure
will remain allocated until after the FBS stop code can examine
the LWP structure.
16. The kernel trace record allocation code now disables interrupts
on the Power PC processor chip, rather than attempting to raise
the Interrupt Priority Level (IPL). By disabling interrupts inside
the processor, we are now able to successfully hold out spurious
interrupts while we allocate the next available ktrace record.
17. The kernel routine that is called by all LWPs in a process that
is being stopped is called prstopped(). Within this routine,
there is a call to the fbsstop_dbg() routine. This is the
routine that will additionally stop the corresponding FBS
scheduler of the calling LWP when that LWP is scheduled on a FBS
scheduler.
The fbsstop_dbg() call was located in the wrong place within the
prstopped() routine, such that only the very last LWP to stop
within the process would make the fbsstop_dbg() routine call.
The call to fbsstop_dbg() was moved forward within the
prstopped() routine so that now all LWPs calling prstopped()
will also call fbsstop_dbg(), and thus stop any corresponding
FBS scheduler that they may be currently scheduled on.
18. Changes were made to function rtc_z8536_timer_intr() and
set_time_constant_rcim().
19. Fixed code to discriminate between PCI bridges and other types
of Bridges. P4 fix added support for mixed multifunction devices
which also included PCI to PCI bridges.
20. Add invalid device protection to gdsize entry point.
21. allocb(), strioccall(), dupb(), pullupmsg(), stropen(), strioctl(),
and qattach() should not have been calling mem_resv_check() before
they call kmem_alloc() because they don't allocate whole pages.
A call to poolrefresh_outofmem() was added to mem_resv_check()
when it detects that there are no whole memory pages available.
22. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None
Object(s) To Be Replaced:
/sbin/mountall
/stand/cp1
/etc/conf/mtune.d/mem
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/memfs/Driver.o
/etc/conf/pack.d/elf/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/procfs/Driver.o
/etc/conf/pack.d/kernel/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/pci/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/io/Driver.o
/usr/include/sys/lwp.h
/usr/include/sys/ksynch.h
/usr/include/sys/swtimer.h
/usr/include/sys/ksynch_p.h
/usr/include/sys/posix_timers.h
/usr/include/vm/vm_hat.h
/usr/ccs/lib/libc.so
/usr/ccs/lib/libc.a
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following link(s) will also be updated by base-008:
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
Possible Side Effects:
1. Reference Line item 13. Statically linked applications that use
nanosleep(3C) and/or clock_gettime(3C) with CLOCK_REALTIME as the
clock_id must be relinked in order to take advantage of this fix.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: base-008
Date Issued: 09/11/2001 15:44:40
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: cmds-005, crosslibs-005, crypt-004, dec-004,
fbs-003, inet-007, netcmds-001, man-005,
trace-005, 1553v5drv-002
Related SARs: #536, #495, #503, #520, #430, #108, #565, #520,
#498, #501
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 base package release updates
############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P8" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 8".
New PowerMAX OS 4.3 Software Patch(es):
base-008 Base System Patch 008 (4.3P8)
cmds-005 Advanced Commands Patch 005 (4.3P8)
crosslibs-005 Libraries for Cross Compiling Patch 005 (4.3P8)
crypt-004 Domestic Encryption Utilities Patch 004 (4.3P8)
crypt-int-004 International Encryption Utilities Patch 004 (4.3P8)
dec-004 DEC Ethernet Driver Patch 004 (4.3P8)
fbs-003 Frequency Based Scheduler Patch 003 (4.3P8)
inet-007 Internet Utilities Patch 007 (4.3P8)
man-005 On-line Manual Pages Patch 005 (4.3P8)
netcmds-001 Commands Networking Extension Patch 001 (4.3P8)
trace-005 KernelTrace Utilities Patch 005 (4.3P8)
1553v5drv-002 1553v5-ABI User Level Driver Patch 002 (4.3P8)
############################################################################
Problem Description:
1. SAR #536:
In order for additional swap devices identified in /etc/vfstab
to be added during bootup, the /sbin/mountall script must be
modified. The line: /usr/sbin/swap -a $special must be changed
to /sbin/swap -a $special
2. SAR's #495, #503:
When an application containing gethostbyname() was statically
linked with libnsl.a and the gethostbyname() followed one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
3. The function prototypes for the set_timeout_resolution()
/get_timeout_resolution() routines are not properly declared in
an associated header file. For the new ec++ compiler this causes
compilation errors.
4. SAR #520:
A previous patch introduced a possibility in the dispatcher
wherein a real-time process could not preempt. The IPL
could be left at a level other than the one the dispatcher was
entered with. This occasionally resulted in disallowing
real-time process preemption.
5. SAR #430:
The pthread_cond_timedwait() routine exposed a race condition
on PowerMAXION systems such that it could block and not be resumed.
On a PowerMAXION system if a process was running on CPU 0 and
was programmed to block, the wakeup routine scheduled on the
global callout queue could have been called via hardclk before the
process had actually reached the blocking routine.
As such, when the process did block it could not have been awakened
again because its wakeup routine had already run.
6. SAR #108:
Symptom: an application with an sbrk(2) area that incrementally
grew into the hundreds of megabytes would occasionally
self-deadlock waiting for the next chunk of virtual memory to be
allocated to the sbrk(2) area of that application.
Problem: one of the kernel data structures needed to manage
application virtual address space would grow so large that there
was danger that not enough *kernel* virtual address space could
be found to hold the next incarnation of that table.
7. Symptom: mlock(2)ing a region of application memory that had
been created with the SHM_NCACHE flag of shmget(2) would panic
the system.
Problem: code in the kernel Driver mem/Driver.o that created PTEs
differed slightly from code nearby that would re-use an already
existing PTE; the net effect of this difference was confusion on
the kernel's part with regard to the cached/not-cached state of
the pages created by shmget(2).
8. Symptom: sometimes a large application distributed among a large
number of local memory pools would self-deadlock when an attempt
was made to lock the application down in memory.
Problem: The way PTEs were allocated was defective, resulting in
all of an applications PTEs being allocated in at most one eighth
of the available PTE buckets. When a bucket was full of locked
down PTEs, and Yet Another had to go into that particular bucket,
the application would wait for a free PTE to become available in
that bucket. That wait would never complete if the application
itself was the owner of all the PTEs already in the bucket.
9. Symptom: The cpu_bias(2) system call would sometimes return with
r1 corrupted when run on nighthawks and when the process was under
the control of a debugger.
Problem: reading and writing of another process's u-area was not
cache coherent when that u-area was in remote memory. Access to
another process's u-area is something done only by a debugger.
10. Symptom: the core dumps of some applications that utilized
dynamically loaded libraries could not be processed by the
debugger.
Problem: the core dump logic in the kernel erroneously assumed
that the text of dynamically loaded libraries was accurately
represented by the corresponding text sections in the library
files, and failed to save off those regions of the address space
to the core file.
11. SAR #565:
The thread that calls pthread_cond_timedwait() does not return
when a another thread calls pthread_cond_signal(), if the timeout
resolution is set to CLOCK_REALTIME with a
set_timeout_resolution(CLOCK_REALTIME, RUTN_PTHRCONDTWAIT) call.
In this case, the caller of pthread_cond_timedwait() doesn't
return until the timeout expires.
The problem was that the sq_wakeup() kernel routine that wakes up
a light-weight process (LWP) as the result of a
pthread_cond_signal() call was not accounting for the fact that
an LWP that had blocked itself with a high-resolution timeout
pending (CLOCK_REALTIME) would be in a different sleep state than
an LWP that had blocked itself with a low-resolution time
pending (CLOCK_UNIX).
Therefore, when the sq_wakeup() routine checked the sleep state
of the LWP that was blocked with the high-resolution timeout
pending, it erroneously decided that this LWP was not currently
blocked/sleeping in the kernel on the associated sleep queue,
and it failed to wakeup the blocked LWP.
12. SAR #520:
On multi-CPU systems it is sometimes possible to read the wrong
POSIX time of day. This is due to the fact that a race
condition existed between one CPU reading the three POSIX timer
assist page variables at the same time that these same variables
were being updated on another CPU.
The three POSIX timer assist page variables are used in the
computation of the current POSIX time of day.
When a mix of old and new POSIX timer assist variable values
are read, then the wrong POSIX time of day is computed/obtained,
and operations such as nanosleep(3C) will not function properly.
13. The following kernel panics will occur on systems with
an enabled kmadbg kernel module:
- The kernel would panic in the fbs_srv_unreg_dev() routine
during the un-registration of a FBS Coupled timing device.
This routine was referencing the fbs data structure of the
virtual fbs scheduler after this structure had already been
kmem_free()ed.
14. When a debugger issues a stop request on a LWP that is scheduled
by a FBS scheduler, the kernel may panic if the LWP has exited
by the time that the FBS module processes the FBS scheduler stop
operation.
If the LWP structure has already been kmem_free()ed, and
re-allocated for a different purpose, then examining the
l_fbslwpp field in the previous LWP structure can cause a kernel
data access exception panic.
15. On single CPU platforms, the allocation and use of the next
available kernel trace record would sometimes allocate the
same trace record structure twice, thus corrupting the
information stored into this same trace record for two
trace events.
This corruption would occur when the tracing of a spurious
interrupt event took place just as the CPU was allocating
a kernel trace record for another event. The method used
to serialize the kernel trace record allocation was unable
to successfully hold out spurious interrupts.
16. If one of the LWPs in a multi-lwp or multi-threaded application
hit a debugger breakpoint and the entire process was stopped,
and one of the other LWPs in the same process was currently
scheduled on a running FBS scheduler, the running scheduler was
not being stopped. The FBS scheduler would continue to run,
even though it had a LWP that had been placed into a stopped state.
17. SAR#498:
A problem with clocks /dev/rrtc/1cN. The stop command does not
always work.
SAR#501:
An accuracy problem with the RCIM real time clock.
18. SAR #540:
The pci/Driver.o provided in Patch Set #4 causes a
target system w/PCI Expansion connected to panic.
19. Adding or removing PCI bridges can dramticly renumber the SCSI
device number convention. This can result in the GD driver
being open using old NODE numbers. The 'idmknod' does not
automaticly execute during boot sequence, since it is still
using the same kernel.
While opening invalid devices using bad node information is not
harmful, certain sequences in XFS used a gdsize entry point which
did not have similar protection. The result was a kernel crash
somewhere in beyond the single user boot phase, if the 'gd' nodes
had not been rebuilt.
20. netperf was hanging because allocb() calls were failing. This
occurred on the receiving system when heavy network traffic
occurred and the sending system was much faster than the
receiving system, or when the receiving system was low on
available memory. The receiving system was not processing this
load of packets fast enough such that a backup of queued packets
occurred. Eventually there were no whole pages available which
caused allocb to hang.
21. The number of cached rpcbind addresses was low. This could have
caused performance issues if an application was using many
different addresses (host, netid).
Problem Resolution:
1. Changed /sbin/mountall to reference /sbin/swap instead of
/usr/sbin/swap
2. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
3. The function prototypes for these routines have been properly
declared in /usr/include/sys/posix_timers.h so that compilation
with the ec++ compiler succeeds.
4. The dispatcher code was corrected so that the IPL upon exit
would be the same as upon entrance.
5. The problem was a misplaced lwp flag setting designed to
prevent this race condition. The L_USYNCTIMEDOUT lwp flag
should have been set upon entry to the wakeup routine,
usync_timedsetrun(), but there was a possible exit from this
routine without the lwp flag being set. The blocking routine,
sq_block(), checks for this flag before blocking in order to
prevent blocking if the wakeup has already occurred. This
patch sets the flag upon entry to usync_timedsetrun().
6. Added a new tuneable, MAX_SEGSIZE, which specifies the maximum
size an address space segment can reach before a new, adjacent
segment is silently born. This puts an upper limit on the size of
the segment data structure the kernel uses to track address
space segments. The default value of MAX_SEGSIZE is 64 megabytes.
7. Corrected code in the kernel Driver mem/Driver.o that created PTEs
to differ slightly from code nearby that would re-use an already
existing PTE.
8. Replaced the address space VSID allocator with a more robust
algorithm that evenly scatters PTE allocations amoung all the
available PTE buckets.
9. Modified remote u-area access code in mem/Driver.o to properly
flush remote memory before and after the access.
10. Core dumps now contain the text sections of dynamically
loaded libraries.
11. The sq_wakeup() kernel routine was modified so that it now
properly recognizes both high-resolution and low-resolution
timeout LWPs that are waiting for either a wakeup or timeout to
occur.
12. The coding for reading and updating the POSIX timer assist page
variables was modified so that the reading of these variables
will always result in reading a coherent set of values, even
when they are being updated at roughly the same time on the
other CPU.
13. All three of these problems were re-coded so that they no longer
reference an already kmem_free()ed memory block. Any needed
information from these kma memory areas is now stored into a
local variable before the kma block is freed.
14. The fix is to place a lwp_hold() on the LWP structure until
the FBS stop operation completes. In this way, the LWP structure
will remain allocated until after the FBS stop code can examine
the LWP structure.
15. The kernel trace record allocation code now disables interrupts
on the Power PC processor chip, rather than attempting to raise
the Interrupt Priority Level (IPL). By disabling interrupts inside
the processor, we are now able to successfully hold out spurious
interrupts while we allocate the next available ktrace record.
16. The kernel routine that is called by all LWPs in a process that
is being stopped is called prstopped(). Within this routine,
there is a call to the fbsstop_dbg() routine. This is the
routine that will additionally stop the corresponding FBS
scheduler of the calling LWP when that LWP is scheduled on a FBS
scheduler.
The fbsstop_dbg() call was located in the wrong place within the
prstopped() routine such that only the very last LWP to stop
within the process would make the fbsstop_dbg() routine call.
The call to fbsstop_dbg() was moved forward within the
prstopped() routine so that now all LWPs calling prstopped()
will also call fbsstop_dbg(), and thus stop any corresponding
FBS scheduler on which they may be currently scheduled.
17. Changes were made to function rtc_z8536_timer_intr() and
set_time_constant_rcim().
18. Fixed code to discriminate between PCI bridges and other types
of Bridges. P4 fix added support for mixed multifunction devices
which also included PCI to PCI bridges.
19. Added invalid device protection to gdsize entry point.
20. allocb(), strioccall(), dupb(), pullupmsg(), stropen(), strioctl(),
and qattach() should not have been calling mem_resv_check() before
they called kmem_alloc() because they don't allocate whole pages.
A call to poolrefresh_outofmem() was added to mem_resv_check()
when it detects that there are no whole memory pages available.
21. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None
Object(s) To Be Replaced:
/sbin/mountall
/etc/conf/mtune.d/mem
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/memfs/Driver.o
/etc/conf/pack.d/elf/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/procfs/Driver.o
/etc/conf/pack.d/kernel/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/pack.d/rtc/Driver.o
/etc/conf/pack.d/pci/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/io/Driver.o
/usr/include/sys/lwp.h
/usr/include/sys/ksynch.h
/usr/include/sys/swtimer.h
/usr/include/sys/ksynch_p.h
/usr/include/sys/posix_timers.h
/usr/include/vm/vm_hat.h
/usr/ccs/lib/libc.so
/usr/ccs/lib/libc.a
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following link(s) will also be updated by base-008:
/usr/lib/libnsl.a --> /usr/lib/libnsl_i.a
/usr/lib/libxti.a --> /usr/lib/libnsl_i.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
/usr/lib/libxti.so --> /usr/lib/libnsl_i.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
Possible Side Effects:
1. Reference Line item 12: Statically linked applications that use
nanosleep(3C) and/or clock_gettime(3C) with CLOCK_REALTIME as the
clock_id must be relinked in order to take advantage of this fix.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: base-009
Date Issued: 01/02/2002 14:19:19
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: See below.
Related SARs: #566, #556, #613, #579, #564, #569, #572
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P9" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 9".
New PowerMAX OS 4.3 Software Patch(es) are:
base-009 Base System Patch 009 (4.3P9)
cmds-006 Advanced Commands Patch 006 (4.3P9)
crosslibs-006 Libraries for Cross Compiling Patch 006 (4.3P9)
gpib-001 NI PCI-GPIB Kernel Level Driver Patch 001 (4.3P9)
inet-008 Internet Utilities Patch 008 (4.3P9)
man-006 On-line Manual Pages Patch 006 (4.3P9)
mvc-002 Multiplexor VME Controller Driver Patch 002 (4.3P9)
nsu-003 Network Support Utilities Patch 003 (4.3P9)
trace-006 KernelTrace Utilities Patch 006 (4.3P9)
via-003 VIA SCSI Adapter Interface Patch 003 (4.3P9)
vp-001 Virtual Partition Patch 001 (4.3P9)
##############################################################################
Problem Description:
1. SAR #566: The kernel occasionally PANICs at lock_nodbg+0x60 when
called by sockmodwsrv+0x1040 or sockmodwsrv+0x9c8. This occurs
when using a "no-delay" socket connect.
2. SAR #556: This fixes a data corruption problem in xfs.
3. Quitting a nview debug session from within a multi-threaded or
multi-LWP (such as ADA) process that contains one or more
ienabled LWPs may cause the system to panic or hang.
This problem is due to a race in the kernel between the ienable
disconnection kernel daemon and the LWP within the process that
is executing the exit(2) kernel code. This race could cause
the ienabled LWP to be placed into the kernel's runqueue
twice, instead of just once. This problem causes either
kernel page fault panics or system hangs.
4. Multi-threaded user applications that utilize the NightTrace
API to log trace events incorrectly tagged trace events with
with an appropriate thread identifier.
5. SAR #613: When initializing the system a parallel fsck of all UFS
filesystems is performed. If more than one UFS partition of
the same disk drive has its fsckpass number set to the same
number, then the system could panic in setbackrq() after a
gdopen(). This problem occurs on systems which have Patch Set
7 or greater installed. The problem was due to the choice of
the wakeup priority used in gdopen() whan a process would
enter a SLEEP_LOCK().
6. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
7. The user of the newfs(1M) utility was never able to modify the
value of the maximum allocated inodes for an XFS filesystem.
However, this could be performed with mkfs(1M) which
is called by newfs(1M). With larger and larger XFS partitions
this has produced a serious problem since all XFS partitions
are established with the same maximum inode value regardless
of the partition size.
8. Bug in gdopen() routine which may cause partition table
to erase and reload. This can cause an observed XFS read
only error to occur, if one is using partition 6.
9. When multiple CPUs panic simultaneously, the code to deal with
this "concurrent" panic can lock up the machine.
10. Newer Tundra chips will not accept PCI address 0 for their
base register.
11. The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
12. System occasionally panics with a floating point exception around
Xintr+0x18, when KTEXTLOCALn tuneables are set.
Problem Resolution:
1. A race condition caused sockmodwsrv to misinterrupt an M_PROTO
large packet on its queue. A socket "bigmsg" was being left on
sockmod's write queue after it was supposed to have been removed
and the pointer and flag concerning it to have been cleared. The
bigmsg is used for flow control to stop write from continuing
down the queue until a connection response (good or bad) is
received. This mechanism is only used for "no delay" connects.
A race condition was occurring between sockmodwsrv and one of
the set of functions that should be removing the bigmsg. These
removers include sockmodwsrv, sockmodclose, sockmodrput and
sockmodwput.
The race condition occurred because sockmodwsrv did not lock the
so_lock for the entire duration of the getq call and subsequent
check for S_WRDISABLE. This allowed sockmodwsrv to take the
bigmsg off the queue at the same time one of the "removers" held
the so_lock and was attemping to remove the bigmsg. The bigmsg
contained random data that sockmodwsrv was trying to interpret as
a structure with pointers to kernel data structures. When
sockmodwsrv would use the random data as a pointer the system
would panic.
2. The fix was to set the SM_WRITE flag before the call to
segmap_release in xfscr_writex() which is in xfs_rdwrx.c. I also
set the SM_ASYNC flag for performance reasons and because it was
logically correct. I removed the setting of SM_DONTNEED because
it was a bit presumptuous to assume that because we just wrote
the seg's mapped pages out that they could be put at the head of
the free list for immediate reallocation. Setting the SM_WRITE
flag causes the mapped pages to be written out via the VOP_PUTPAGE
macro upon release.
3. The window in time where it was possible for both the ienable
kernel daemon and the kernel exit code to both place the same
ienabled LWP into the runqueues was closed.
The ienable disconnection processing sequence was modified so
that the exit processing code can now always successfully detect
and skip over ienabled LWPs, and thus let the ienable kernel
daemon be the only one to place this LWP into the runqueue after
it has been successfully disconnected from the connected interrupt.
4. Multi-threaded user applications that utilize the NightTrace
API to log trace events now correctly tag each trace event
with an appropriate thread identifier.
5. The choice of the wakeup priority when a process would enter
the SLEEP_LOCK() in any gd driver routines was changed from
'pldisk' to the documented and more correct 'pridisk'.
6. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
7. The newfs(1M) utility has been enhanced to allow the user to
specify a maximum inode allocation value on the command line
with a new -x <max-inode> argument for XFS filesystems only.
If this argument is not used by the user, then a default value
will be applied which is based on the size of the partition
rather than using the same default value for all partitions.
Additionally, this new argument has been documented in the
newfs(1M) manual page and the C locale file for I18N has been
updated.
8. Corrected code now checks disk status before proceeding to
first time initialization.
9. Fixed for-loop to release ALL panic'd CPUs. Removed xprintf of
"CONCURRENT PANIC" on each of the other CPUs. The print lock
gets mucked up, causing the next CPU to hang.
10. The base register is reprogrammed to an unused portion of the
256 Mbyte window [0x0f000000]
11. Problem was due to the count of pages attached to an IO
operation not being correct.
12. The method the kernel used to turn the instruction-stream
virtual memory back on at the start of interrupt, syscall,
and exception processing was not correct when KTEXTLOCAL[1-9]
was set.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/cf.d/sc.s
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/newfs
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libucbc.a
/usr/include/sys/sockmod.h
/usr/include/sys/buf.h
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/locale/C/MSGFILES/harris.str
/usr/lib/libc.so.1
Special Conditions for Installation:
1. L/I 11: The size of the b_numpages field of the internal
kernel-only data structure, buf_t, increased from a byte
to a short. This change was done in a way that the size
of the buf_t structure did not change nor did any offset
of that structure, therefore, only users of b_numpages
itself are effected. All kernel Driver.o modules that
use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays the
contents of the field b_numpages.
Possible Side Effects:
1. L/I 11: crash(1m) from earlier patch-releases of PowerMAX OS
will no longer correctly display the b_numpages field.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: base-009
Date Issued: 01/02/2002 14:19:19
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: See below.
Related SARs: #566, #556, #613, #579, #564, #569, #572
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P9" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 9".
New PowerMAX OS 4.3 Software Patch(es) are:
base-009 Base System Patch 009 (4.3P9)
cmds-006 Advanced Commands Patch 006 (4.3P9)
crosslibs-006 Libraries for Cross Compiling Patch 006 (4.3P9)
gpib-001 NI PCI-GPIB Kernel Level Driver Patch 001 (4.3P9)
inet-008 Internet Utilities Patch 008 (4.3P9)
man-006 On-line Manual Pages Patch 006 (4.3P9)
mvc-002 Multiplexor VME Controller Driver Patch 002 (4.3P9)
nsu-003 Network Support Utilities Patch 003 (4.3P9)
trace-006 KernelTrace Utilities Patch 006 (4.3P9)
via-003 VIA SCSI Adapter Interface Patch 003 (4.3P9)
vp-001 Virtual Partition Patch 001 (4.3P9)
##############################################################################
Problem Description:
1. SAR #566: The kernel occasionally PANICs at lock_nodbg+0x60 when
called by sockmodwsrv+0x1040 or sockmodwsrv+0x9c8. This occurs
when using a "no-delay" socket connect.
2. SAR #556: This fixes a data corruption problem in xfs.
3. Quitting a nview debug session from within a multi-threaded or
multi-LWP (such as ADA) process that contains one or more
ienabled LWPs may cause the system to panic or hang.
This problem is due to a race in the kernel between the ienable
disconnection kernel daemon and the LWP within the process that
is executing the exit(2) kernel code. This race could cause
the ienabled LWP to be placed into the kernel's runqueue
twice, instead of just once. This problem causes either
kernel page fault panics or system hangs.
4. Multi-threaded user applications that utilize the NightTrace
API to log trace events incorrectly tagged trace events with
with an appropriate thread identifier.
5. SAR #613: When initializing the system a parallel fsck of all UFS
filesystems is performed. If more than one UFS partition of
the same disk drive has its fsckpass number set to the same
number, then the system could panic in setbackrq() after a
gdopen(). This problem occurs on systems which have Patch Set
7 or greater installed. The problem was due to the choice of
the wakeup priority used in gdopen() whan a process would
enter a SLEEP_LOCK().
6. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
7. The user of the newfs(1M) utility was never able to modify the
value of the maximum allocated inodes for an XFS filesystem.
However, this could be performed with mkfs(1M) which
is called by newfs(1M). With larger and larger XFS partitions
this has produced a serious problem since all XFS partitions
are established with the same maximum inode value regardless
of the partition size.
8. Bug in gdopen() routine which may cause partition table
to erase and reload. This can cause an observed XFS read
only error to occur, if one is using partition 6.
9. When multiple CPUs panic simultaneously, the code to deal with
this "concurrent" panic can lock up the machine.
10. Newer Tundra chips will not accept PCI address 0 for their
base register.
11. The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
12. System occasionally panics with a floating point exception around
Xintr+0x18, when KTEXTLOCALn tuneables are set.
Problem Resolution:
1. A race condition caused sockmodwsrv to misinterrupt an M_PROTO
large packet on its queue. A socket "bigmsg" was being left on
sockmod's write queue after it was supposed to have been removed
and the pointer and flag concerning it to have been cleared. The
bigmsg is used for flow control to stop write from continuing
down the queue until a connection response (good or bad) is
received. This mechanism is only used for "no delay" connects.
A race condition was occurring between sockmodwsrv and one of
the set of functions that should be removing the bigmsg. These
removers include sockmodwsrv, sockmodclose, sockmodrput and
sockmodwput.
The race condition occurred because sockmodwsrv did not lock the
so_lock for the entire duration of the getq call and subsequent
check for S_WRDISABLE. This allowed sockmodwsrv to take the
bigmsg off the queue at the same time one of the "removers" held
the so_lock and was attemping to remove the bigmsg. The bigmsg
contained random data that sockmodwsrv was trying to interpret as
a structure with pointers to kernel data structures. When
sockmodwsrv would use the random data as a pointer the system
would panic.
2. The fix was to set the SM_WRITE flag before the call to
segmap_release in xfscr_writex() which is in xfs_rdwrx.c. I also
set the SM_ASYNC flag for performance reasons and because it was
logically correct. I removed the setting of SM_DONTNEED because
it was a bit presumptuous to assume that because we just wrote
the seg's mapped pages out that they could be put at the head of
the free list for immediate reallocation. Setting the SM_WRITE
flag causes the mapped pages to be written out via the VOP_PUTPAGE
macro upon release.
3. The window in time where it was possible for both the ienable
kernel daemon and the kernel exit code to both place the same
ienabled LWP into the runqueues was closed.
The ienable disconnection processing sequence was modified so
that the exit processing code can now always successfully detect
and skip over ienabled LWPs, and thus let the ienable kernel
daemon be the only one to place this LWP into the runqueue after
it has been successfully disconnected from the connected interrupt.
4. Multi-threaded user applications that utilize the NightTrace
API to log trace events now correctly tag each trace event
with an appropriate thread identifier.
5. The choice of the wakeup priority when a process would enter
the SLEEP_LOCK() in any gd driver routines was changed from
'pldisk' to the documented and more correct 'pridisk'.
6. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
7. The newfs(1M) utility has been enhanced to allow the user to
specify a maximum inode allocation value on the command line
with a new -x <max-inode> argument for XFS filesystems only.
If this argument is not used by the user, then a default value
will be applied which is based on the size of the partition
rather than using the same default value for all partitions.
Additionally, this new argument has been documented in the
newfs(1M) manual page and the C locale file for I18N has been
updated.
8. Corrected code now checks disk status before proceeding to
first time initialization.
9. Fixed for-loop to release ALL panic'd CPUs. Removed xprintf of
"CONCURRENT PANIC" on each of the other CPUs. The print lock
gets mucked up, causing the next CPU to hang.
10. The base register is reprogrammed to an unused portion of the
256 Mbyte window [0x0f000000]
11. Problem was due to the count of pages attached to an IO
operation not being correct.
12. The method the kernel used to turn the instruction-stream
virtual memory back on at the start of interrupt, syscall,
and exception processing was not correct when KTEXTLOCAL[1-9]
was set.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/cf.d/sc.s
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/gd/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/pack.d/vme/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/newfs
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libucbc.a
/usr/include/sys/sockmod.h
/usr/include/sys/buf.h
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/lib/locale/C/MSGFILES/harris.str
/usr/lib/libc.so.1
Special Conditions for Installation:
1. L/I 11: The size of the b_numpages field of the internal
kernel-only data structure, buf_t, increased from a byte
to a short. This change was done in a way that the size
of the buf_t structure did not change nor did any offset
of that structure, therefore, only users of b_numpages
itself are effected. All kernel Driver.o modules that
use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays the
contents of the field b_numpages.
Possible Side Effects:
1. L/I 11: crash(1m) from earlier patch-releases of PowerMAX OS
will no longer correctly display the b_numpages field.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: base-010
Date Issued: 03/18/2002 11:11:26
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: cmds-007, cnd-005, crosslibs-007, crypt-005, trace-007,
crypt-int-005, dec-005, diskless-006, egl-004, fbs-005,
inet-009, ip-004, man-007, nfs-002, nsu-004, pg-004
Related SARs: #659, #623, #666, #627, #535, #517, #638, #683, #582,
#648
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P10" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 10".
base-010 Base System Patch 010 (4.3P10)
cmds-007 Advanced Commands Patch 007 (4.3P10)
cnd-005 Condor Ethernet Driver Patch 005 (4.3P10)
crosslibs-007 Libraries for Cross Compiling Patch 007 (4.3P10)
crypt-005 Domestic Encryption Utilities Patch 005 (4.3P10)
crypt-int-005 International Encryption Utilities Patch 005 (4.3P10)
dec-005 DEC Ethernet Driver Patch 005 (4.3P10)
diskless-006 Diskless Systems Package Patch 006 (4.3P10)
egl-004 Eagle Ethernet Driver Patch 004 (4.3P10)
fbs-005 Frequency Based Scheduler Patch 005 (4.3P10)
inet-009 Internet Utilities Patch 009 (4.3P10)
ip-004 Interphase 4511 PMC FDDI Driver Patch 004 (4.3P10)
nfs-002 Network File System Utilities Patch 002 (4.3P10)
nsu-004 Network Support Utilities Patch 004 (4.3P10)
pg-004 Peregrine FDDI Driver Patch 004 (4.3P10)
trace-007 KernelTrace Utilities Patch 007 (4.3P10)
##############################################################################
Problem Description:
1. An XFS filesystem can be forced into readonly mode if xnodes are
running out and previously used xnodes are being freed up
simultaneously. This condition occurs because an already freed
xnode is being marked to be freed. Because the test for it being
in use fails, XFS thinks the filesystem is corrupted. Therefore,
it forces the filesystem into readonly mode where it can't be
used any further.
2. SAR #623: Occasional page-fault panic in user-level interrupt
driver. All application pages are locked down as they are
supposed to be. An examination of the faulting address shows
it is a valid virtual address to a properly locked-down page.
The problem is due to corrupt segment registers; hence the PTE,
though correct, could not be found by the hardware VM as the
seg registers are part of the mapping process. The corruption
consisted of some segment registers holding values belonging to
another task.
The segment register corruption appears to be a side-effect
of hat_load_tt() not locking out interrupts when the set of
13 user address space segment registers are being updated.
This is not normally a problem except when a user level
interrupt driver is invoked; in that case, hat_load_tt() is
called from interrupt level and therefore that call to
hat_load_tt() could interrupt some other base level call to
hat_load_tt. Again, this would not be a problem (due to the
clever use of a pair of hat_load_tt's with the U.I. envelope)
if it were not for a single if-stmt inside hat_load_tt()
which is getting confused.
3. SAR #659: When the highres callout is enabled for the select()
timeouts, the special case of "struct timeval timeout == NULL"
to wait forever, doesn't.
4. SAR #666: High res callout queue entries that are removed have
their function pointer set to zero and dequeued from the linked
list.
5. SAR #627: When one process is creating a file that is also being
accessed at the same time as an NFSed file with mlockall() and
mmap() calls, then a deadlock can occur between the segvn driver
and the nfs driver.
This particular scenario occurs especially (but not only) on a
Closely Coupled system, where Posix shared memory files for a
client SBC actually exist on the server SBC across an NFS/network
connection. One process can be creating the shared memory file
while another process is trying to open this shared memory file
to mmap() it and to lock it down in memory.
The problem occurs in segvn_lockop()-segvn_fault()-nfs_getpage()
when nfs_getpage() decides to either:
- purge the attr cache or
- abort all mappings due to stale filehandle,
where these operations attempt to remove all page objects
that belong to the vnode via nfs_purge_caches().
However, there are pages currently locked by segvn_fault() that
are outside the current range of the nfs_getpage() parameters, and
the nfs code waits forever on a page lock in pvn_getdirty_range() or
pvn_abort_range() when it tries to remove all the page mappings
to the current vnode. The page happens to already be locked
by this same LWP from above in the segvn_fault() routine.
6. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
7. SAR #517: The /usr/lib/acct/ckpacct and /usr/lib/acct/runacct
scripts use /sbin/devnm to determine if there is sufficient
space in /var. However, it fails with the following message:
UX:df: ERROR: FSType cannot be identified because the devnm
command would need the file system type for the
true file system.
8. SAR #638 : Data corruption on RS232 remote port of Motorola board.
Detailed description: Writing/reading to/from /dev/remote at baud
rate 19200 or higher causes some characters to be lost.
9. SAR #683: When /usr/sbin/savecore is used with the file
/var/crashfiles/minfree to determine if a crashdump should continue,
it will fail on a statfs() call. statfs() is an obsolete system
call that should have been changed to statvfs() as documented in
/usr/include/sys/statfs.h.
10. During the processing of a core dump in core() a condition
could occur where the rendezvous() of processes failed. This
is most likely due to one of the processes being a user-level
interrupt process. The error action taken in this condition was
to panic the system. This was a bit excessive for a core dump.
11. The generic mount and umount commands were not properly
validating mount point arguments. When a relative pathname
was issued as a mount point argument, the pathname was not
rejected because the function realpath() was being used to
resolve the path. This was coded such that the argument
returned by realpath() was being used to do most of the
real work of mounting and umounting, while the optarg remained
a relative path. The optarg would then fail string
comparisons which resulted in missed code (leaving the
system in an undefined state) and causing mount table corruption
(the table would not be updated). A similar bug occurred
when a malformed special name (server:/<path>) was issued
which resulted in continual retries to nfs mount until an rpc
timeout and missed code. The way realpath() was being used
it was also possible to mount on the wrong path when given a
relative pathname.
Problem Resolution:
1. This is fixed by allowing the already freed xnode to not be
considered an indication of filesystem corruption. This is based
on the fact that this problem only occurs when xnodes are almost
completely depleted anyway and the file system can't be used any
further until some xnodes are freed up. Rather than forcing the
filesystem into readonly mode it is kept writeable for the possible
event of xnodes being freed.
2. The solution is for hat_load_tt() to lock out interrupts around
the offending if-stmt and the update to the 13 segment registers.
3. select() does correctly pass a -1 timeout into poll(), but the
highres callout does not check for the -1. Now the CLOCK_REALTIME
and CLOCK_UNIX wait forever when timeout == NULL.
4. Zero function pointer entry makes it to the hrt_alarm callout
routine. This causes a panic. R16 which points to the
hrt_callout struct looks OK(except for c_func==0), like it was
just removed from the linked list.
5. The fix is to limit the range of pages that segvn_lockop()
passes to segvn_fault() so that segvn_fault() will not be
holding onto any pages outside of the current range of pages
that nfs_getpage() is operating on. This limitation is only
enforced for NFS-based segments.
6. I have implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
7. I previously fixed this problem in PowerMAX OS 5.0. This task
backfits the change to df/devnm into PowerMAX OS 4.3 to resolve
the problem. These scripts will now succeed.
8. The fifo feature of the uart is enabled to store characters in a
buffer so that we do not lose any characters when operating at
higher baud rates.
9. The statfs() call in /usr/sbin/savecore has been changed to
statvfs(). Also changed the minfree calculations to use
double instead of int to accomodate the larger file systems
that need to be supported.
10. The error action in core(), if this condition occurs, has been
changed to log a warning to /dev/console and /var/adm/syslog,
abort the rendezvous, and return without dumping core.
11. A mount point input validation routine was written which
enforces the requirement that pathnames be absolute and
that nfs mount special names be properly formed. A new
error handling routine with new exit codes was added to
compliment the input validation routine. The man pages
were updated to reflect the new exit codes and the man page
for realpath() was corrected to reflect the fact that realpath()
always returns an absolute pathname.
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Give the customer the ability to have the initial/main thread
of a multithreaded application be a bound thread instead
of the usual multiplexing thread.
The default will still be that the initial/primordial thread
that executes the main() function be a multiplexing thread.
However, if the user declares and initializes the following
global variable in their application code:
int __primordial_bound = 1;
then the thread library will create a bound primordial thread
instead of a multiplexing primordial thread.
10. sigqueue() is changed to return a good status when a signal that
is set to be ignored is queued. Previously sigqueue() failed
setting errno to EAGAIN which according to POSIX standard
specifications should occur only in the following cases:
- The signal could not be queued due to
system-wide resource limitations.
- The resource necessary to queue the signal would cause the
process-wide limit {SIGQUEUE_MAX} to be exceeded.
11. - usr/src/common/head/math.h - Removed keyword "const" from
declaration of Kernighan & Ritchie version of __huge_val.
- usr/src/common/head/stdlib.h - Spell Kernighan & Ritchie
definition as __ctype rather than _ctype.
- usr/src/common/uts/svc/posix_timers.h - Include Kernighan &
Ritchie version signatures controlled by __STDC__.
- usr/src/ppc/head/tdesc.h - Add Linux targetted tdesc
stuff. Add appropriate structure descriptions.
- usr/src/ppc/lib/libc/ppc/str/strncpy.h - When we move a string
that has actual length = 2 mod4 (ie a zero ) then it turns out
that we were zeroing 3 extra bytes after the trailing 0.
Substittute and stb r13, 2(r11) for a stw r13, 2(r11)
Object(s) To Be Replaced:
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/cons/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/serial/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/df
/sbin/mount
/sbin/umount
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/math.h
/usr/include/net/if.h
/usr/include/snmp/snmp-mib.h
/usr/include/stdlib.h
/usr/include/sys/adapter_vme.h
/usr/include/sys/autoconf.h
/usr/include/sys/mitereg_pci.h
/usr/include/sys/posix_timers.h
/usr/include/sys/termios.h
/usr/include/sys/vm_mdep.h
/usr/include/sys/vnode.h
/usr/include/tdesc.h
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/sbin/savecore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: base-010
Date Issued: 03/18/2002 11:19:27
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: cmds-007, cnd-005, crosslibs-007, crypt-005,
crypt-int-005, dec-005, egl-004, ie-004, inet-009,
man-007, nfs-002, nsu-004, pg-004, trace-007
Related SARs: #659, #623, #666, #627, #535, #517, #683, #582, #648
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
Patches for PowerMAX OS are now released officially as "Patch Sets". A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a "Patch Set" may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
"4.3P10" is the abbreviated name for "PowerMAX OS 4.3 Patch Set 10".
base-010 Base System Patch 010 (4.3P10)
cmds-007 Advanced Commands Patch 007 (4.3P10)
cnd-005 Condor Ethernet Driver Patch 005 (4.3P10)
crosslibs-007 Libraries for Cross Compiling Patch 007 (4.3P10)
crypt-005 Domestic Encryption Utilities Patch 005 (4.3P10)
crypt-int-005 International Encryption Utilities Patch 005 (4.3P10)
dec-005 DEC Ethernet Driver Patch 005 (4.3P10)
egl-004 Eagle Ethernet Driver Patch 004 (4.3P10)
ie-004 Night Hawk ISE Ethernet Interface Module Patch 004 (4.3P10)
inet-009 Internet Utilities Patch 009 (4.3P10)
man-007 On-line Manual Pages Patch 007 (4.3P10)
nfs-002 Network File System Utilities Patch 002 (4.3P10)
nsu-004 Network Support Utilities Patch 004 (4.3P10)
pg-004 Peregrine FDDI Driver Patch 004 (4.3P10)
trace-007 KernelTrace Utilities Patch 007 (4.3P10)
##############################################################################
Problem Description:
1. An XFS filesystem can be forced into readonly mode if xnodes are
running out and previously used xnodes are being freed up
simultaneously. This condition occurs because an already freed
xnode is being marked to be freed. Because the test for it being
in use fails, XFS thinks the filesystem is corrupted. Therefore,
it forces the filesystem into readonly mode where it can't be
used any further.
2. SAR #623: Occasional page-fault panic in user-level interrupt
driver. All application pages are locked down as they are
supposed to be. An examination of the faulting address shows
it is a valid virtual address to a properly locked-down page.
The problem is due to corrupt segment registers; hence the PTE,
though correct, could not be found by the hardware VM as the
seg registers are part of the mapping process. The corruption
consisted of some segment registers holding values belonging to
another task.
The segment register corruption appears to be a side-effect
of hat_load_tt() not locking out interrupts when the set of
13 user address space segment registers are being updated.
This is not normally a problem except when a user level
interrupt driver is invoked; in that case, hat_load_tt() is
called from interrupt level and therefore that call to
hat_load_tt() could interrupt some other base level call to
hat_load_tt. Again, this would not be a problem (due to the
clever use of a pair of hat_load_tt's with the U.I. envelope)
if it were not for a single if-stmt inside hat_load_tt()
which is getting confused.
3. SAR #659: When the highres callout is enabled for the select()
timeouts, the special case of "struct timeval timeout == NULL"
to wait forever, doesn't.
4. SAR #666: High res callout queue entries that are removed have
their function pointer set to zero and dequeued from the linked
list.
5. SAR #627: When one process is creating a file that is also being
accessed at the same time as an NFSed file with mlockall() and
mmap() calls, then a deadlock can occur between the segvn driver
and the nfs driver.
This particular scenario occurs especially (but not only) on a
Closely Coupled system, where Posix shared memory files for a
client SBC actually exist on the server SBC across an NFS/network
connection. One process can be creating the shared memory file
while another process is trying to open this shared memory file
to mmap() it and to lock it down in memory.
The problem occurs in segvn_lockop()-segvn_fault()-nfs_getpage()
when nfs_getpage() decides to either:
- purge the attr cache or
- abort all mappings due to stale filehandle,
where these operations attempt to remove all page objects
that belong to the vnode via nfs_purge_caches().
However, there are pages currently locked by segvn_fault() that
are outside the current range of the nfs_getpage() parameters, and
the nfs code waits forever on a page lock in pvn_getdirty_range() or
pvn_abort_range() when it tries to remove all the page mappings
to the current vnode. The page happens to already be locked
by this same LWP from above in the segvn_fault() routine.
6. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
7. SAR #517: The /usr/lib/acct/ckpacct and /usr/lib/acct/runacct
scripts use /sbin/devnm to determine if there is sufficient
space in /var. However, it fails with the following message:
UX:df: ERROR: FSType cannot be identified because the devnm
command would need the file system type for the
true file system.
8. SAR #683: When /usr/sbin/savecore is used with the file
/var/crashfiles/minfree to determine if a crashdump should continue,
it will fail on a statfs() call. statfs() is an obsolete system
call that should have been changed to statvfs() as documented in
/usr/include/sys/statfs.h.
9. During the processing of a core dump in core() a condition
could occur where the rendezvous() of processes failed. This
is most likely due to one of the processes being a user-level
interrupt process. The error action taken in this condition was
to panic the system. This was a bit excessive for a core dump.
10. The generic mount and umount commands were not properly
validating mount point arguments. When a relative pathname
was issued as a mount point argument, the pathname was not
rejected because the function realpath() was being used to
resolve the path. This was coded such that the argument
returned by realpath() was being used to do most of the
real work of mounting and umounting, while the optarg remained
a relative path. The optarg would then fail string
comparisons which resulted in missed code (leaving the
system in an undefined state) and causing mount table corruption
(the table would not be updated). A similar bug occurred
when a malformed mount special name (server:/<path>) was issued
which resulted in continual retries to nfs mount until an rpc
timeout and missed code. The way realpath() was being used
it was also possible to mount on the wrong path when given a
relative pathname.
Problem Resolution:
1. This was fixed by allowing the already freed xnode to not be
considered an indication of filesystem corruption. This is based
on the fact that this problem only occurs when xnodes are almost
completely depleted anyway and the file system can't be used any
further until some xnodes are freed up. Rather than forcing the
filesystem into readonly mode it is kept writeable for the possible
event of xnodes being freed.
2. The solution is for hat_load_tt() to lock out interrupts around
the offending if-stmt and the update to the 13 segment registers.
3. select() does correctly pass a -1 timeout into poll(), but the
highres callout does not check for the -1. Now the CLOCK_REALTIME
and CLOCK_UNIX wait forever when timeout == NULL.
4. Zero function pointer entry makes it to the hrt_alarm callout
routine. This causes a panic. R16 which points to the
hrt_callout struct looks OK(except for c_func==0), like it was
just removed from the linked list.
5. The fix is to limit the range of pages that segvn_lockop()
passes to segvn_fault() so that segvn_fault() will not be
holding onto any pages outside of the current range of pages
that nfs_getpage() is operating on. This limitation is only
enforced for NFS-based segments.
6. I have implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
7. I previously fixed this problem in PowerMAX OS 5.0. This task
backfits the change to df/devnm into PowerMAX OS 4.3 to resolve
the problem. These scripts will now succeed.
8. The statfs() call in /usr/sbin/savecore has been changed to
statvfs(). Also changed the minfree calculations to use
double instead of int to accomodate the larger file systems
that need to be supported.
9. The error action in core(), if this condition occurs, has been
changed to log a warning to /dev/console and /var/adm/syslog,
abort the rendezvous, and return without dumping core.
10. A mount point input validation routine was written which
enforces the requirement that pathnames be absolute and
that nfs mount special names be properly formed. A new
error handling routine with new exit codes was added to
compliment the input validation routine. The man pages
were updated to reflect the new exit codes and the man page
for realpath() was corrected to reflect the fact that realpath()
always returns an absolute pathname.
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Give the customer the ability to have the initial/main thread
of a multithreaded application be a bound thread instead
of the usual multiplexing thread.
The default will still be that the initial/primordial thread
that executes the main() function be a multiplexing thread.
However, if the user declares and initializes the following
global variable in their application code:
int __primordial_bound = 1;
then the thread library will create a bound primordial thread
instead of a multiplexing primordial thread.
10. sigqueue() is changed to return a good status when a signal that
is set to be ignored is queued. Previously sigqueue() failed
setting errno to EAGAIN which according to POSIX standard
specifications should occur only in the following cases:
- The signal could not be queued due to
system-wide resource limitations.
- The resource necessary to queue the signal would cause the
process-wide limit {SIGQUEUE_MAX} to be exceeded.
11. - usr/src/common/head/math.h - Removed keyword "const" from
declaration of Kernighan & Ritchie version of __huge_val.
- usr/src/common/head/stdlib.h - Spell Kernighan & Ritchie
definition as __ctype rather than _ctype.
- usr/src/common/uts/svc/posix_timers.h - Include Kernighan &
Ritchie version signatures controlled by __STDC__.
- usr/src/ppc/head/tdesc.h - Add Linux targetted tdesc
stuff. Add appropriate structure descriptions.
- usr/src/ppc/lib/libc/ppc/str/strncpy.h - When we move a string
that has actual length = 2 mod4 (ie a zero ) then it turns out
that we were zeroing 3 extra bytes after the trailing 0.
Substittute and stb r13, 2(r11) for a stw r13, 2(r11)
Object(s) To Be Replaced:
/etc/conf/pack.d/bsp6400/Driver.o
/etc/conf/pack.d/bsp6408/Driver.o
/etc/conf/pack.d/bsp6800/Driver.o
/etc/conf/pack.d/bsp6800p/Driver.o
/etc/conf/pack.d/bsp6800t/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/cons/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/serial/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/xfs/Driver.o
/sbin/df
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/math.h
/usr/include/net/if.h
/usr/include/snmp/snmp-mib.h
/usr/include/stdlib.h
/usr/include/sys/adapter_vme.h
/usr/include/sys/autoconf.h
/usr/include/sys/mitereg_pci.h
/usr/include/sys/posix_timers.h
/usr/include/sys/termios.h
/usr/include/sys/vm_mdep.h
/usr/include/sys/vnode.h
/usr/include/tdesc.h
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/sbin/savecore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: base-011
Date Issued: 07/10/2002 14:44:01
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: 1553v5drv-003, cmds-008, crypt-006, crypt-int-006,
gf-001, inet-010, man-008, nfs-003, nsu-005, oam-001,
softint-001
Related SARs: #722, #320, #698, #544, #744, #695, #99, #679, #672,
#703
Brief Description:
PowerMAX OS 4.3 base package release updates
##############################################################################
Patches for PowerMAX OS are now released officially as Patch Sets. A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P11 is the abbreviated name for PowerMAX OS 4.3 Patch Set 11.
1553v5drv-003 1553v5-ABI User Level Driver Patch 003 (4.3P11)
base-011 Base System Patch 011 (4.3P11)
cmds-008 Advanced Commands Patch 008 (4.3P11)
crypt-006 Domestic Encryption Utilities Patch 006 (4.3P11)
gf-001 Generic SCSI Floppy Driver Patch 001 (4.3P11)
inet-010 Internet Utilities Patch 010 (4.3P11)
man-008 On-line Manual Pages Patch 008 (4.3P11)
nfs-003 Network File System Utilities Patch 003 (4.3P11)
nsu-005 Network Support Utilities Patch 005 (4.3P11)
oam-001 OA&M Patch 001 (4.3P11)
softint-001 Software Packaging Tools Patch 001 (4.3P11)
##############################################################################
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
##############################################################################
P1: SAR 722: The kernel trap routine for floating point exceptions was
incrementing the program counter (PC) associated with the user process
that was executing at the time of a floating point machine exception in
order to have the new PC point to the next instruction when returning
control to the user process.
Before returning to the user process, an additional floating point
machine exception occurred on the same CPU (due to instructions
executed by that same user process before the first exception
occurred; i.e. two exceptions were pending). The PC for the
user process was incremented again, causing the new PC to be 1
instruction past the entry point for the application signal handler.
R1: We do not advance the program counter when imprecise floating point mode
is in use.
P2: SAR #320: adb in CX/UX works with both static and dynamic linking.In
PowerMAX the problem appears only when dynamic linking is used. A
program that has been linked dynamically generates a core dump. To be
able to display values of certain data variables, adb looks through
the data segments in the core file generating a starting and ending
addresses in the core file where data should be found. The beginning
of this address range was the start of the first data segment.
In all other cases this starting address was related to the data
section's starting address in the executable file but with dynamic
linking(several data segments in the core file) the first data segment
does not necessarily corelate to the data section's starting address.
R2: Data segments in the core file related to the dynamic linking process
are not used in generating the data address range.
P3: SAR #686, #689: While an application is running the %busy from mpstat
is around a 100% after running for an hour.When qlen gets large(14+)
and the application is shut down qlen does not drop below that....
R3: High values for the disk statistics were caused by incorrect value of
"qlen"(Instantaneous operation queue length).The state of qlen is now
protected by a fast spin lock.
P4: SAR 698: kernel sometimes panics with the message
'hat_allocpte: bad PTEG'
R4: Problem was due to a leftover debug statement from the days that
PowerMAX OS was originally developed.
P5: SAR 544: kernel sometimes panics with the message
'address fault from interrupt routine'
R5: Problem was due to the kernel's 'in interrupt driver?' test not being
coded up in an atomic manner.
P6: SAR 544: servicing_interrupt() not atomic on some NightHawks
R6: Fixed servicing_interrupt() macro in nh/uts/util/ksynch_p.h
P7: Occasionally stale TLBs in kernel address space.
R7: Now waits for the tlbie insn to complete on all cpus before a new tlbie
is issued.
P8: SAR #744: Idle time calculation not high precision when HIGHRESTIMING
tuneable was enabled
R8: Alternate e_idlepct formula used when HIGHRESTIMING tuneable is enabled.
P9: PTE directory (PDN) pool too small for today's large memory systems
R9: Increased MinFreePdns from 16 to 64.
P10: hat_freepdn should be a front for kmem_free when KMACORRUPT is enabled
R10: hat_freepdn passes buffer to kmem_free when KMACORRUPT is enabled.
P11: hat_freepdn not checking alignment of the to-be-freed pdn buffer.
R11: hat_freepdn will now panic the kernel if given a misaligned buffer.
P12: kmem_free not checking freed buffers for proper alignment.
R12: kmem_free will now panic the kernel if given a misaligned buffer.
P13: Debug kmem_alloc not checking if buffer is aligned.
R13: kmem_alloc now will panic if to-be-returned buffer is misaligned.
P14: Debug kmem_alloc not filling buffer with a known 'poison' value.
R14: kmem_alloc now fills the first 128 bytes of a buffer with 0xdeadbeaf.
P15: Debug kmem_alloc not checking if secondary and primary virtual address
are equivalent.
R15: kmem_alloc now test-writes through the primary virtual address and
checks if the value is returnable through the secondary virtual address.
This will catch any TLB flush failures to either address.
P16: %cpu utilization occasionally spikes towards zero even though the
application load remains constant.
R16: Replaced a TRYLOCK with a LOCK in the per-second statistics gathering
part of the kernel clock code. If the TRYLOCK failed then several
variables used in the %cpu utilization calculations would be left
uninitialized. As there was no real need to use TRYLOCK in this code,
it was replaced with a LOCK, which cannot fail.
P17: A system with process swapouts/swapins occurring would panic when:
- a lwp tried to insert itself into a sv, event, etc. sleep queue that
it was already in,
- two engines would run the same lwp and stomp over the same kernel
stack,
- or a lwp wakeup would fail due to a corrupted sv, event, etc.
sleep queue.
R17: The swtch() routine was improperly setting SSLEEP lwps to SRUN state
when they were about to switch themselves out and they had a pending
seize operation place on them.The swtch() routine now only transitions
lwps into the SRUN state for pending seize operations if that lwp is
currently in the SRUN state.
P18: SAR #695: On occasion, when network activity is high and
/usr/lib/resolv.so is specified in /etc/netconfig and there are contents
in /etc/resolv.conf, the resolution of addresses can become very slow.
This is due to a poll() timeout value being too short in libresolv and
causing multiple retries for the resolved address.
R18: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
P19: SAR #99: When compiling a C/C++ program that contains a call to fork1()
or forkall() with the c++ compiler, the compilation will fail.
R19: The unistd.h header file did not have function prototypes or namespace
information for the fork1() and forkall() system calls.This is
corrected in this task.
P20: The pkgadd(1) utility incorrectly complains about dependencies when
installing a patch to an installed instance of a package beyond the
first instance. (i.e., pkg.2, pkg.3, ..., pkg.n)
R20: The resolution is actually in the /usr/sadm/install/bin/pkginstall
command. The dockdeps() routine was modified to correctly check
dependencies on multiple instances of packages.
P21: SAR #679: System panics in mi_timer() during a tcp_xclose().
R21: The user application was calling accept with the no-delay option set
on the socket fd. The system was "panicing" during a tcp_xclose()
because the address of the queue's tcp pointer was changed while
tcp_xclose() was waiting for a lock. When the wait returned, the old
tcp was no longer valid. This caused a series of functions to be
called with bad pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
P22: SAR #672: The in.snmpd daemon was issuing a lot of getmsg calls to the
lower stack layers for each mib attribute requested. This was causing
in.snmpd to have a very high cpu utilization when a large number of
mib attributes were requested.
R22: Previously the only method for the daemon to retrieve the mib information
that it needed from the kernel was to get all 14 sections of mib data at
once. Changes were made to the ipm, tcpm, and udpm drivers and the snmpd
daemon so the daemon can now request just one section at a time instead
of having to get all 14 sections every time..
P23: The gdopen routine could corrupt memory if other data structures were
allocated immediately after its two metric structures.
R23: The gdopen routine was referencing past the end of two of its allocated
metrics data structures if the incoming cylinder number was equal to the
max number of cylinders. gdopen now verifies that the incoming cylinder
number is within bounds before updating the metrics structures.
P24: SAR #703: LWP blocking in pthread_cond_timedwait() is not unblocking
when requested by another LWP calling pthread_cond_signal(). Broken
when blocking LWP is using CLOCK_REALTIME, CLOCK_UNIX is OK.
pthread_cond_timedwait() is supposed to release the mutex and block on
the condition variable "atomically".There is an implemented with the
sequence : prepblock() -> mutex_unlock() -> block().If a different LWP
does a pthread_cond_signal() between our mutex_lock() and block(),
we are supposed to "remember" that and the block() routine should
return immediately. The problem is when we are running with
CLOCK_UNIX, block() calls sq_block(), which has code to "remember".
When running with CLOCK_REALTIME, block() calls selfblock(), which
does not "remember".
R24: In block(),if CLOCK_REALTIME,do the "remember" checking and don't
bother to call selfblock(). Safer to check in block() than to check
in selfblock() since selfblock() is called from other places that may
not want to "remember" or even set up the "remember" reminders.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fifofs/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/io/Driver.o
/etc/conf/pack.d/ipc/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/log/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/namefs/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/sad/Driver.o
/etc/conf/pack.d/sfs/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/sbin/config
/sbin/creatiadb
/sbin/mount
/sbin/mountall
/sbin/umount
/usr/bin/adb
/usr/bin/dc
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/as
/usr/ccs/bin/cprs
/usr/ccs/bin/ld
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/sys/ksynch_p.h
/usr/include/sys/strsubr.h
/usr/include/sys/xti.h
/usr/include/unistd.h
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/locale/C/MSGFILES/pkg.str
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgname
/usr/sbin/mount
/usr/sbin/mountall
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/umount
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-011
Date Issued: 07/24/2002 14:36:07
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: 1553v5drv-003, cmds-008, crypt-006, crypt-int-006,
gf-001, inet-010, man-008, nfs-003, nsu-005, oam-001,
softint-001
Related SARs: #722, #320, #698, #544, #744, #695, #99, #679, #672,
#703
Brief Description:
PowerMAX OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS are now released officially as Patch Sets. A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P11 is the abbreviated name for PowerMAX OS 4.3 Patch Set 11.
1553v5drv-003 1553v5-ABI User Level Driver Patch 003 (4.3P11)
base-011 Base System Patch 011 (4.3P11)
cmds-008 Advanced Commands Patch 008 (4.3P11)
crypt-006 Domestic Encryption Utilities Patch 006 (4.3P11)
gf-001 Generic SCSI Floppy Driver Patch 001 (4.3P11)
inet-010 Internet Utilities Patch 010 (4.3P11)
man-008 On-line Manual Pages Patch 008 (4.3P11)
nfs-003 Network File System Utilities Patch 003 (4.3P11)
nsu-005 Network Support Utilities Patch 005 (4.3P11)
oam-001 OA&M Patch 001 (4.3P11)
softint-001 Software Packaging Tools Patch 001 (4.3P11)
###############################################################################
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR 722: The kernel trap routine for floating point exceptions was
incrementing the program counter (PC) associated with the user process
that was executing at the time of a floating point machine exception in
order to have the new PC point to the next instruction when returning
control to the user process.
Before returning to the user process, an additional floating point
machine exception occurred on the same CPU (due to instructions
executed by that same user process before the first exception
occurred; i.e. two exceptions were pending). The PC for the
user process was incremented again, causing the new PC to be 1
instruction past the entry point for the application signal handler.
R1: We do not advance the program counter when imprecise floating point mode
is in use.
P2: SAR #320: adb in CX/UX works with both static and dynamic linking.In
PowerMAX the problem appears only when dynamic linking is used. A
program that has been linked dynamically generates a core dump. To be
able to display values of certain data variables, adb looks through
the data segments in the core file generating starting and ending
addresses in the core file where data should be found. The beginning
of this address range was the start of the first data segment.
In all other cases this starting address was related to the data
section's starting address in the executable file but with dynamic
linking(several data segments in the core file) the first data segment
does not necessarily corelate to the data section's starting address.
R2: Data segments in the core file related to the dynamic linking process
are not used in generating the data address range.
P3: SAR #686, #689: While an application is running the %busy from mpstat
is around 100% after running for an hour.When qlen gets large(14+)
and the application is shut down qlen does not drop below that.
R3: High values for the disk statistics were caused by incorrect value of
"qlen"(Instantaneous operation queue length).The state of qlen is now
protected by a fast spin lock.
P4: SAR 698: kernel sometimes panics with the message
'hat_allocpte: bad PTEG'
R4: Problem was due to a leftover debug statement from the days that
PowerMAX OS was originally developed.
P5: SAR 544: kernel sometimes panics with the message
'address fault from interrupt routine'
R5: Problem was due to the kernel's `in interrupt driver?' test not being
coded up in an atomic manner.
P6: SAR 544: servicing_interrupt() not atomic on some NightHawks
R6: Fixed servicing_interrupt() macro in nh/uts/util/ksynch_p.h
P7: Occasionally stale TLBs in kernel address space.
R7: Now waits for the tlbie insn to complete on all cpus before a new tlbie
is issued.
P8: SAR #744: Idle time calculation not high precision when HIGHRESTIMING
tuneable was enabled
R8: Alternate e_idlepct formula used when HIGHRESTIMING tuneable is enabled.
P9: PTE directory (PDN) pool too small for today's large memory systems
R9: Increased MinFreePdns from 16 to 64.
P10: hat_freepdn should be a front for kmem_free when KMACORRUPT is enabled
R10: hat_freepdn passes buffer to kmem_free when KMACORRUPT is enabled.
P11: hat_freepdn not checking alignment of the to-be-freed pdn buffer
R11: hat_freepdn will now panic the kernel if given a misaligned buffer.
P12: kmem_free not checking freed buffers for proper alignment.
R12: kmem_free will now panic the kernel if given a misaligned buffer.
P13: Debug kmem_alloc not checking if buffer is aligned.
R13: kmem_alloc now will panic if to-be-returned buffer is misaligned.
P14: Debug kmem_alloc not filling buffer with a known 'poison' value.
R14: kmem_alloc now fills the first 128 bytes of a buffer with 0xdeadbeaf.
P15: Debug kmem_alloc not checking if secondary and primary virtual address
are equivalent.
R15: kmem_alloc now test-writes through the primary virtual address and
checks if the value is returnable through the secondary virtual address.
This will catch any TLB flush failures to either address.
P16: %cpu utilization occasionally spikes towards zero even though the
application load remains constant.
R16: Replaced a TRYLOCK with a LOCK in the per-second statistics gathering
part of the kernel clock code. If the TRYLOCK failed then several
variables used in the %cpu utilization calculations would be left
uninitialized. As there was no real need to use TRYLOCK in this code,
it was replaced with a LOCK, which cannot fail.
P17: A system with process swapouts/swapins occurring would panic when:
- a lwp tried to insert itself into a sv, event, etc. sleep queue that
it was already in,
- two engines would run the same lwp and stomp over the same kernel
stack,
- or a lwp wakeup would fail due to a corrupted sv, event, etc.
sleep queue.
R17: The swtch() routine was improperly setting SSLEEP lwps to SRUN state
when they were about to switch themselves out and they had a pending
seize operation place on them. The swtch() routine now only transitions
lwps into the SRUN state for pending seize operations if that lwp is
currently in the SRUN state.
P18: SAR #695: On occasion, when network activity is high and
/usr/lib/resolv.so is specified in /etc/netconfig and there are contents
in /etc/resolv.conf, the resolution of addresses can become very slow.
This is due to a poll() timeout value being too short in libresolv and
causing multiple retries for the resolved address.
R18: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
P19: SAR #99: When compiling a C/C++ program that contains a call to fork1()
or forkall() with the cc++ compiler, the compilation will fail.
R19: The unistd.h header file did not have function prototypes or namespace
information for the fork1() and forkall() system calls. This is
corrected in this task.
P20: The pkgadd(1) utility incorrectly complains about dependencies when
installing a patch to an installed instance of a package beyond the
first instance. (i.e., pkg.2, pkg.3, ..., pkg.n)
R20: The resolution is actually in the /usr/sadm/install/bin/pkginstall
command. The dockdeps() routine was modified to correctly check
dependencies on multiple instances of packages.
P21: SAR #679: System panics in mi_timer() during a tcp_xclose().
R21: The user application was calling accept with the no-delay option set
on the socket fd. The system was "panicing" during a tcp_xclose()
because the address of the queue's tcp pointer was changed while
tcp_xclose() was waiting for a lock. When the wait returned,the old
tcp was no longer valid. This caused a series of functions to be
called with bad pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
P22: SAR #672: The in.snmpd daemon was issuing a lot of getmsg calls to the
lower stack layers for each mib attribute requested. This was causing
in.snmpd to have a very high cpu utilization when a large number of
mib attributes were requested.
R22: Previously the only method for the daemon to retrieve the mib information that it needed from the kernel was to get all 14 sections of mib data at
once. Changes were made to the ipm, tcpm, and udpm drivers and the snmpd
daemon so the daemon can now request just one section at a time instead
of having to get all 14 sections every time..
P23: The gdopen routine could corrupt memory if other data structures were
allocated immediately after its two metric structures.
R23: The gdopen routine was referencing past the end of two of its allocated
metrics data structures if the incoming cylinder number was equal to the
max number of cylinders. gdopen now verifies that the incoming cylinder
number is within bounds before updating the metrics structures.
P24: SAR #703: LWP blocking in pthread_cond_timedwait() is not unblocking
when requested by another LWP calling pthread_cond_signal(). Broken
when blocking LWP is using CLOCK_REALTIME, CLOCK_UNIX is OK.
pthread_cond_timedwait() is supposed to release the mutex and block on
the condition variable "atomically". There is a implemented with the
sequence : prepblock() -> mutex_unlock() -> block(). If a different LWP
does a pthread_cond_signal() between our mutex_lock() and block(),
we are supposed to "remember" that and the block() routine should
return immediately. The problem is when we are running with
CLOCK_UNIX, block() calls sq_block(), which has code to "remember".
When running with CLOCK_REALTIME, block() calls selfblock(), which
does not "remember".
R24: In block(), if CLOCK_REALTIME, do the "remember" checking and don't
bother to call selfblock(). Safer to check in block() than to check
in selfblock() since selfblock() is called from other places that may
not want to "remember" or even set up the "remember" reminders.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fifofs/Driver.o
/etc/conf/pack.d/fs/Driver.o
/etc/conf/pack.d/io/Driver.o
/etc/conf/pack.d/ipc/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/log/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/namefs/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/sad/Driver.o
/etc/conf/pack.d/sfs/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/sbin/config
/sbin/creatiadb
/sbin/mount
/sbin/mountall
/sbin/umount
/usr/bin/adb
/usr/bin/dc
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/as
/usr/ccs/bin/cprs
/usr/ccs/bin/ld
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/include/sys/ksynch_p.h
/usr/include/sys/strsubr.h
/usr/include/sys/xti.h
/usr/include/unistd.h
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
/usr/lib/locale/C/MSGFILES/pkg.str
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgname
/usr/sbin/mount
/usr/sbin/mountall
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/umount
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-012
Date Issued: 12/11/2002 13:33:07
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: cmds-009, cnd-006, crypt-007, crypt-int-007,
diskless-007, inet-011, man-009, pg-005, rmxf-001,
rpc-002, softint-002
Related SARs: #470, #827, #781, #830, #706, #697, #780, #222, #792,
#801, #815, #816, #794
Brief Description:
PowerMAX OS 4.3 base package release updates
---------------------------------------------------------------------------
Patches for PowerMAX OS are now released officially as Patch Sets. A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P12 is the abbreviated name for PowerMAX OS 4.3 Patch Set 12.
base-012 Base System Patch 012 (4.3P12)
cmds-009 Advanced Commands Patch 009 (4.3P12)
cnd-006 Condor Ethernet Driver Patch 006 (4.3P12)
crypt-007 Domestic Encryption Utilities Patch 007 (4.3P12)
crypt-int-007 International Encryption Utilities Patch 007 (4.3P12)
diskless-007 Diskless Systems Package Patch 007 (4.3P12)
inet-011 Internet Utilities Patch 011 (4.3P12)
man-009 On-line Manual Pages Patch 009 (4.3P12)
pg-005 Peregrine FDDI Driver Patch 005 (4.3P12)
rmxf-001 RAMiX PMC665 FOB Ethernet Driver Patch 001 (4.3P12)
rpc-002 Remote Procedure Calls Utilities Patch 002 (4.3P12)
softint-002 Software Packaging Tools Patch 002 (4.3P12)
###############################################################################
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: wcrtomb glitch. When first argument is a null, should return
the same as: wcrtomb(buf, L`0`, state) where buf is an internal
buffer.
va_arg fouling up on chars and shorts. The address to pick up
the chars/shorts is wrong because of incorrect parenthesization.
R1: wcrtomb source has been modified to perform the correct actions
with a NULL first parameter.
va_arg expansions have been corrected for parenthesization
in stdvar_args.h.
P2: SAR #470: Remote backup/restore of xfs file systems was not
possible using xfsdump(1M)/xfsrestore(1M).
R2: A new option (-H) was added to xfsdump(1M) and xfsrestore(1M).
It allows backups to a tape drive on a remote system. In
addition, the commands rxfsdump(1M) and rxfsrestore(1M) were
added. They correspond to xfsdump(1M) and xfsrestore(1M)
envoked with the "-H" option.
P3: SAR #827: Power Hawk 640 crashing during bootup. The problem
appeared after installation of PowerMAX OS 4.3 patchset 10. In
patchset 10 the console driver was changed to take advantage
of the uart fifo feature by enabling the receiver fifo. Data
was read into an internal array of size 8 characters. Received
data trigger level was set at 4 bytes. Since the fifo was
enabled, there were cases of reading more than 8 characters
into the internal buffer, writing beyond its limit.
R3. When data is read from the receiver fifo the number of characters
read is compared to the length of the buffer. The trigger level
was also changed from 4 bytes to 14 bytes to improve efficiency.
P4: SAR #781: __GETPL() can return the wrong cpu's IPL because the
lwp can be context switched to another cpu between the fetch of
the cpu number and the fetch of the ipl value from the
ipl_value array. Hangs and deadlocks occurred because there
was unprotected use of __GETPL() by the lock routines.
R4: Processes that have been preempted while in kernel mode are
marked "unavailable for migration to another cpu". This is
achieved by setting a flag in the lwp_t struct. The flag is
cleared when the process resumes execution.
P5: PANIC: kmem_alloc: pvaddr=0x05ACD580 and svaddr=0x05AF4580
are not equivalent.
R5: A TLB tracking a kernel PTE was not being flushed when that
PTE was destroyed by the operating system. This is a
potentially serious condition: if a new PTE for that same
kernel virtual address to a different virtual page is
created, the unflushed TLB causes all loads and stores that
should go to the new physical page to instead go to the old
physical page. This can corrupt the system, with symptoms
similar to those of wild pointers.
This particular TLB flush problem is due to a operating system
bit mask, describing which TLBs need to be flushed, not being
cleared atomically with respect to the actual flushing of
those TLBs. The solution was to add spinlock code to make
the corresponding bit-clearing and TLB-flushing atomic.
P6: An application stalling out forever on a hatbkt full of locked
PTEs is not giving any indication of why it is stalled out.
R6: The kernel now prints a warning whenever some application goes
to sleep waiting for a PTE in some HAT bucket which is full of
locked PTEs.
P7: Killing UI process under debugger crashes system
R7: The fix is to leave the userint_info structure in the ui_table[]
array, and to let the ienabled LWP removed it AFTER it has been
EVENT_SIGNAL(), and thus no longer in a STOPPED state.
P8: The ipl priority value associated with an iconnect()ed interrupt
vector remains set to the iconnect ic_ipl value even after the
ienabled connection is removed.
R8: The fix was to save off the original ipl associated with the
interrupt vector, and restore it when the interrupt vector is
no longer connected.
P9: SAR #830: System call nesting problem in OS causing problems
with ntrace. Running an application mix that causes lots of
kernel preemption context switches while ktrace(1) is running
can cause invalid exception exit ktrace records to be generated.
These records contain the incorrect engine number in them.
These invalid records can severly confuse ntrace(1).
R9: Ensure that the correct engine number is stored into these
exception exit ktrace records. This is accomplished by reading
the current engine number while still under the ktrace fast spin
lock. Also remove the redundant interrupt exit ktrace records
that are being written in the common interrupt exit code when
the rescheduling interrupt code has already written an interrupt
exit ktrace record.
P10: SAR #706: The name to address resolver in libresolv does not
work properly when utilizing the HOSTALIASES environment
variable as documented in the hostname(5) manpage. It will
fail to resolve the name.
R10: A bug was found in the _rs_hostalias() routine wherein
the full hostname to be returned from a match in the file
specified by HOSTALIASES was not being copied to the object
returned from _rs_hostalias(). Thus, it appeared that there
was never a match found in the file and name to address
resolution would fail. This bug has been removed.
P11: SAR #697: The hp.ps.net(5) printer interface script did not
support landscape capability.
R11: The hp.ps.net(5) printer interface script has been enhanced
to support landscape capability.
P12: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R12: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
P13: SAR #222: The hptcp(1M) utility for hp networked printers would
not resolve the printer name if a system was configured for NIS.
R13: By compiling this utility dyamically it will now use the
library /usr/lib/tcpip_nis.so for name to address resolution
routines when a system is configured for NIS.
P14: SAR #792: memalign(3C) with a 32 byte boundary specified can
cause a segmentation fault in a certain test case of three
memalign()s followed by three free()s followed by a fourth
memalign(). A 32 byte memalign was too small of an alignment
to allow the actual aligned block of memory to fall beyond
the block information for the initial fragment. As such, the
size of the aligned block was placed within the information
structure of the fragment. Ultimately, when trying to reclaim
the fragment, the system would segfault on the invalid address
being shown in its right child pointer.
R14: The code in memalign() which adjusts for the alignment was
modified to ensure that, for any alignment value given, enough
proper alignments would be done to move the aligned block
past the information info of the initial fragment.
P15: SAR #801: After displaying all of the menu choices pkgadd(1M)
and pkgrm(1M) expect the user to select the choices to be added
or removed. The current default is 'all' which will add or
remove all coices in the menu. It is possible to inadvertently
enter the default by pressing <ENTER> when it is not really
intended.
R15: The pkgadd(1M) and pkgrm(1M) utilities have been enhanced to
use '?' as the default response. Thus if <ENTER> is
inadvertently pressed in response to selecting choices the
prompt will be redisplayed until the user enters an acceptable
response.
P16: SAR #815: The file /etc/.osm which is used to accumulate
console messages can have a tendency to grow outside the
bounds of the root file system where it resides. This can
contribute to kernel rebuild problems later if the space
in the root file system is used up by this file.
R16: The files, /etc/.osm and /etc/.osm.old, will be moved to /var
where they really should reside and symbolic links from the old
file locations to the new file locations will be created and
maintained. This change is effected by modification of the
osm initialization script /etc/conf/rc.d/osm.
P17: SAR #816: When running pkgask(1M) and a package is selected
for installation that is already installed, then pkgask(1M)
will fail. Subsequently, the administrator is not allowed
to continue processing additional selected packages if the
input is from a datastream file. Additionally, the failure
leaves a zero-length response file that must be removed manually.
R17: The pkgask(1M) utility has been modified to allow the system
administrator to continue processing additional selected
packages if a selected package's request script fails.
Additionally, the pkginstall utility will remove the
zero-length response file created during the pkgask(1M)
failure.
P18: This task merely backfits recent changes to libpkg in 5.1SR5
which handle large pkg names. The associated tasks were:
pkgnmlen.mew and pkgnmlen2.mew
The problem was that pkgtrans would not properly handle the
pkg name (xdevc51synergy-001). This was because of a few
remaining hardcoded array lengths where the pkg name was
involved.
R18: The fix was to change these hardcoded array lengths to use
the value of PKGSIZ which is defined in /usr/include/pkgstrct.h.
P19: Performing a pkgtrans(1M) on a large number of input packages
to an output datastream specified as "stdout" (-) where stdout
is redirected to the actual datastream can result in a corrupted
datastream. This is due to transference messages normally
destined for stdout being interspersed with the packages being
transferred which are also now destined for stdout.
R19: The fix for this problem is to turn on the PT_SILENT option
during the transference of the packages so that the
transference messages will not be displayed.
P20: Both stdio.h and wchar.h have some c99 routines without
signatures in the case of standard C99 compilations.
R20: Include those signatures and optimization specifications
when compiling for c99.
P21: vfwscanf() consumed input with %n formatter. fwprintf()
wrote "260", when -0.0 was expected.
R21: In _idoscan.c, the type char was used instead of wchar_t,
where the input was read from stdin or file.
P22: snprintf() returned 1, expected > 1 (bugzilla # 164)
R22: In cvt-config, code that dealt with signed zero has
been commented.
P24: Rounding bug in printf floating format.
R24: _idoprnt.c has been modified to round the formatted
floating point expression.
P25: -Qload= option fails with empty file assigned.
R25: Appropriate check for empty load file has been introduced for
-Qload=xxx, where xxx can be an empty file.
P26: strftime creating default format date incorrectly.
R26: The formatting of the output has been changed from
%a %b %d %H:%M:%S %Y to %a %b %e %H:%M:%S %Y in LC_TIME
P27: %hh[char] conversion broken in scan routines.
R27: Added lines for 'hh' conversion that was missing in the
original source.
P28: SAR #794: When a PCI bus error occurred, the PCI bus error
interrupt was set up to come in at 0x12 in the MPIC hardware.
This corresponds to PLXCALL, which is incorrect.
R28: The line which sets this value in the moto/uts/svc/bsp*/misc_p.c
routine mpic_enable() was changed to set it to 0x11. Affected
drivers are bsp4600, bsp2600, bspp750, and bspmtx.
bspall was also included.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/cswtch.s
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/bsp1600/Driver.o
/etc/conf/pack.d/bsp2600/Driver.o
/etc/conf/pack.d/bsp4600/Driver.o
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/bspmtx/Driver.o
/etc/conf/pack.d/bspp750/Driver.o
/etc/conf/pack.d/bsppstk/Driver.o
/etc/conf/pack.d/cons/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/rc.d/osm
/sbin/creatiadb
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/ld
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libnc.a
/usr/include/stdio.h
/usr/include/stdvar_args.h
/usr/include/sys/assym_dbg.h
/usr/include/sys/assym_nodbg.h
/usr/include/sys/consreg.h
/usr/include/sys/iconnect.h
/usr/include/sys/lwp.h
/usr/include/sys/shm.h
/usr/include/wchar.h
/usr/lib/fs/xfs/xfsdump
/usr/lib/fs/xfs/xfsrestore
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/lp/bin/hptcp
/usr/lib/lp/model/hp.ps.net
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgname
/usr/sadm/install/bin/pkgremove
/usr/sbin/creatiadb
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/rxfsdump
/usr/sbin/rxfsrestore
/etc/conf/pack.d/name/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-012
Date Issued: 12/18/2002 17:39:41
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: cmds-009, cnd-006, crypt-007, crypt-int-007,
inet-011, man-009, pg-005, rmxf-001, rpc-002,
softint-002, vmet-001
Related SARs: #470, #781, #830, #706, #697, #780, #222, #792,
#801, #815, #816
Brief Description:
PowerMAX OS 4.3 base package release updates
------------------------------------------------------------------------------
Patches for PowerMAX OS are now released officially as Patch Sets. A patch
set is a collection of patches released simultaneously. Individual patches
may have dependencies on other patches included in a patch set. Not all
patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P12 is the abbreviated name for PowerMAX OS 4.3 Patch Set 12.
base-012 Base System Patch 012 (4.3P12)
cmds-009 Advanced Commands Patch 009 (4.3P12)
cnd-006 Condor Ethernet Driver Patch 006 (4.3P12)
crypt-007 Domestic Encryption Utilities Patch 007 (4.3P12)
inet-011 Internet Utilities Patch 011 (4.3P12)
man-009 On-line Manual Pages Patch 009 (4.3P12)
pg-005 Peregrine FDDI Driver Patch 005 (4.3P12)
rmxf-001 RAMiX PMC665 FOB Ethernet Driver Patch 001 (4.3P12)
rpc-002 Remote Procedure Calls Utilities Patch 002 (4.3P12)
softint-002 Software Packaging Tools Patch 002 (4.3P12)
vmet-001 VME to PCI/PMC Carrier Card Patch 001 (4.3P12)
###############################################################################
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: wcrtomb glitch. When first argument is a null, should return
the same as: wcrtomb(buf, L`0`, state) where buf is an internal
buffer.
va_arg fouling up on chars and shorts. The address to pick up
the chars/shorts is wrong because of incorrect parenthesization.
R1: wcrtomb source has been modified to perform the correct actions
with a NULL first parameter.
va_arg expansions have been corrected for parenthesization
in stdvar_args.h.
P2: SAR #470: Remote backup/restore of xfs file systems was not
possible using xfsdump(1M)/xfsrestore(1M).
R2: A new option (-H) was added to xfsdump(1M) and xfsrestore(1M).
It allows backups to a tape drive on a remote system. In
addition, the commands rxfsdump(1M) and rxfsrestore(1M) were
added. They correspond to xfsdump(1M) and xfsrestore(1M)
envoked with the "-H" option.
P3: SAR #781: __GETPL() can return the wrong cpu's IPL because the
lwp can be context switched to another cpu between the fetch of
the cpu number and the fetch of the ipl value from the
ipl_value array. Hangs and deadlocks occurred because there
was unprotected use of __GETPL() by the lock routines.
R3: Processes that have been preempted while in kernel mode are
marked "unavailable for migration to another cpu". This is
achieved by setting a flag in the lwp_t struct. The flag is
cleared when the process resumes execution.
P4: PANIC: kmem_alloc: pvaddr=0x05ACD580 and svaddr=0x05AF4580
are not equivalent.
R4: A TLB tracking a kernel PTE was not being flushed when that
PTE was destroyed by the operating system. This is a
potentially serious condition: if a new PTE for that same
kernel virtual address to a different virtual page is
created, the unflushed TLB causes all loads and stores that
should go to the new physical page to instead go to the old
physical page. This can corrupt the system, with symptoms
similar to those of wild pointers.
This particular TLB flush problem is due to a operating system
bit mask, describing which TLBs need to be flushed, not being
cleared atomically with respect to the actual flushing of
those TLBs. The solution was to add spinlock code to make
the corresponding bit-clearing and TLB-flushing atomic.
P5: An application stalling out forever on a hatbkt full of locked
PTEs is not giving any indication of why it is stalled out.
R5: The kernel now prints a warning whenever some application goes
to sleep waiting for a PTE in some HAT bucket which is full of
locked PTEs.
P6: Killing UI process under debugger crashes system
R6: The fix is to leave the userint_info structure in the ui_table[]
array, and to let the ienabled LWP removed it AFTER it has been
EVENT_SIGNAL(), and thus no longer in a STOPPED state.
P7: The ipl priority value associated with an iconnect()ed interrupt
vector remains set to the iconnect ic_ipl value even after the
ienabled connection is removed.
R7: The fix was to save off the original ipl associated with the
interrupt vector, and restore it when the interrupt vector is
no longer connected.
P8: SAR #830: System call nesting problem in OS causing problems
with ntrace. Running an application mix that causes lots of
kernel preemption context switches while ktrace(1) is running
can cause invalid exception exit ktrace records to be generated.
These records contain the incorrect engine number in them.
These invalid records can severly confuse ntrace(1).
R8: Ensure that the correct engine number is stored into these
exception exit ktrace records. This is accomplished by reading
the current engine number while still under the ktrace fast spin
lock. Also remove the redundant interrupt exit ktrace records
that are being written in the common interrupt exit code when
the rescheduling interrupt code has already written an interrupt
exit ktrace record.
P9: SAR #706: The name to address resolver in libresolv does not
work properly when utilizing the HOSTALIASES environment
variable as documented in the hostname(5) manpage. It will
fail to resolve the name.
R9: A bug was found in the _rs_hostalias() routine wherein
the full hostname to be returned from a match in the file
specified by HOSTALIASES was not being copied to the object
returned from _rs_hostalias(). Thus, it appeared that there
was never a match found in the file and name to address
resolution would fail. This bug has been removed.
P10: SAR #697: The hp.ps.net(5) printer interface script did not
support landscape capability.
R10: The hp.ps.net(5) printer interface script has been enhanced
to support landscape capability.
P11: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R11: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
P12: SAR #222: The hptcp(1M) utility for hp networked printers would
not resolve the printer name if a system was configured for NIS.
R12: By compiling this utility dyamically it will now use the
library /usr/lib/tcpip_nis.so for name to address resolution
routines when a system is configured for NIS.
P13: SAR #792: memalign(3C) with a 32 byte boundary specified can
cause a segmentation fault in a certain test case of three
memalign()s followed by three free()s followed by a fourth
memalign(). A 32 byte memalign was too small of an alignment
to allow the actual aligned block of memory to fall beyond
the block information for the initial fragment. As such, the
size of the aligned block was placed within the information
structure of the fragment. Ultimately, when trying to reclaim
the fragment, the system would segfault on the invalid address
being shown in its right child pointer.
R13: The code in memalign() which adjusts for the alignment was
modified to ensure that, for any alignment value given, enough
proper alignments would be done to move the aligned block
past the information info of the initial fragment.
P14: SAR #801: After displaying all of the menu choices pkgadd(1M)
and pkgrm(1M) expect the user to select the choices to be added
or removed. The current default is 'all' which will add or
remove all coices in the menu. It is possible to inadvertently
enter the default by pressing <ENTER> when it is not really
intended.
R14: The pkgadd(1M) and pkgrm(1M) utilities have been enhanced to
use '?' as the default response. Thus if <ENTER> is
inadvertently pressed in response to selecting choices the
prompt will be redisplayed until the user enters an acceptable
response.
P15: SAR #815: The file /etc/.osm which is used to accumulate
console messages can have a tendency to grow outside the
bounds of the root file system where it resides. This can
contribute to kernel rebuild problems later if the space
in the root file system is used up by this file.
R15: The files, /etc/.osm and /etc/.osm.old, will be moved to /var
where they really should reside and symbolic links from the old
file locations to the new file locations will be created and
maintained. This change is effected by modification of the
osm initialization script /etc/conf/rc.d/osm.
P16: SAR #816: When running pkgask(1M) and a package is selected
for installation that is already installed, then pkgask(1M)
will fail. Subsequently, the administrator is not allowed
to continue processing additional selected packages if the
input is from a datastream file. Additionally, the failure
leaves a zero-length response file that must be removed manually.
R16: The pkgask(1M) utility has been modified to allow the system
administrator to continue processing additional selected
packages if a selected package's request script fails.
Additionally, the pkginstall utility will remove the
zero-length response file created during the pkgask(1M)
failure.
P17: This task merely backfits recent changes to libpkg in 5.1SR5
which handle large pkg names. The associated tasks were:
pkgnmlen.mew and pkgnmlen2.mew
The problem was that pkgtrans would not properly handle the
pkg name (xdevc51synergy-001). This was because of a few
remaining hardcoded array lengths where the pkg name was
involved.
R17: The fix was to change these hardcoded array lengths to use
the value of PKGSIZ which is defined in /usr/include/pkgstrct.h.
P18: Performing a pkgtrans(1M) on a large number of input packages
to an output datastream specified as "stdout" (-) where stdout
is redirected to the actual datastream can result in a corrupted
datastream. This is due to transference messages normally
destined for stdout being interspersed with the packages being
transferred which are also now destined for stdout.
R18: The fix for this problem is to turn on the PT_SILENT option
during the transferrence of the packages so that the
transference messages will not be displayed.
P19: Both stdio.h and wchar.h have some c99 routines without
signatures in the case of standard C99 compilations.
R19: Include those signatures and optimization specifications
when compiling for c99.
P20: vfwscanf() consumed input with %n formatter. fwprintf()
wrote "260", when -0.0 was expected.
R20: In _idoscan.c, the type char was used instead of wchar_t,
where the input was read from stdin or file.
P21: snprintf() returned 1, expected > 1 (bugzilla # 164)
R21: In cvt-config, code that dealt with signed zero has
been commented.
P22: Rounding bug in printf floating format.
R22: _idoprnt.c has been modified to round the formatted
floating point expression.
P23: -Qload= option fails with empty file assigned.
R23: Appropriate check for empty load file has been introduced for
-Qload=xxx, where xxx can be an empty file.
P24: strftime creating default format date incorrectly.
R25: The formatting of the output has been changed from
%a %b %d %H:%M:%S %Y to %a %b %e %H:%M:%S %Y in LC_TIME
P26: %hh[char] conversion broken in scan routines.
R26: Added lines for 'hh' conversion that was missing in the
original source.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/cswtch.s
/etc/conf/cf.d/except.s
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/bspall/Driver.o
/etc/conf/pack.d/cons/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/ui/Driver.o
/etc/conf/pack.d/util/Driver.o
/etc/conf/rc.d/osm
/sbin/creatiadb
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/ld
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libnc.a
/usr/include/stdio.h
/usr/include/stdvar_args.h
/usr/include/sys/assym_dbg.h
/usr/include/sys/assym_nodbg.h
/usr/include/sys/consreg.h
/usr/include/sys/iconnect.h
/usr/include/sys/lwp.h
/usr/include/sys/shm.h
/usr/include/wchar.h
/usr/lib/libc.so.1
/usr/lib/libnsl_i.a
/usr/lib/lp/bin/hptcp
/usr/lib/lp/model/hp.ps.net
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgname
/usr/sadm/install/bin/pkgremove
/usr/sbin/creatiadb
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/rxfsdump
/usr/sbin/rxfsrestore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-013
Date Issued: 05/08/2003 10:33:58
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: cmds-010, dec-006, diskless-008, hsde-002, inet-012,
ip-005, man-010, pg-006, rmxf-001, rpc-003, xfsd-002
Brief Description:
PowerMAX OS 4.3 base package release updates
-------------------------------------------------------------------------------
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P13 is the abbreviated name for PowerMAX OS 4.3 Patch Set 13.
PowerMAX OS 4.3P13 Software Update Package(s):
base-013 Base System Patch 013 (4.3P13)
cmds-010 Advanced Commands Patch 010 (4.3P13)
dec-006 DEC Ethernet Driver Patch 006 (4.3P13)
diskless-008 Diskless Systems Package Patch 008 (4.3P13)
hsde-002 High Speed Data Enhanced Channel Driver Patch 002 (4.3P13)
inet-012 Internet Utilities Patch 012 (4.3P13)
ip-005 Interphase 4511 PMC FDDI Driver Patch 005 (4.3P13)
man-010 On-line Manual Pages Patch 010 (4.3P13)
pg-006 Peregrine FDDI Driver Patch 006 (4.3P13)
rmxf-001 RAMiX PMC665 FOB Ethernet Driver Patch 001 (4.3P13)
rpc-003 Remote Procedure Calls Utilities Patch 003 (4.3P13)
xfsd-002 Distributed XFS File System Patch 002 (4.3P13)
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Compilation of the C library,
usr/src/ppc/lib/libc/ppc/gen/mall_data.s,
reported the bogus error.
R1. Return the static definition on dofpnum.c, but rename it
s_bzero to highlight its difference from bzero(). Modified
the error message in question to report "result is underflow".
P2. The ability to set a default boot disk other than disk(0) on
Power Hawk series 600 and Power Stack systems has not been
implemented. PPC-Bug can be set to boot the console from any
disk but after our console is loaded from disk, the disk to
boot from can only be set manually if it is not disk(0).
"pboot" flag is not saved in nvram which makes autobooting
of OS difficult.
R2. The first part of the problem has been fixed by modifying
the "fd" command to save the default boot disk in nvram.
Example:"fd -sw dsk(0,1,0,0)" will save disk with id 1 in
nvram as the default disk without asking the user
for confirmation to write to nvram.
"p" command was modified as well to behave like Power Hawk series 700 console "p" command. The "boot" and "aboot" flags
can be saved in nvram.
"aboot" flag can determine the number of seconds to wait
before booting PowerMAX OS.
Example: #0>p -sw boot 82
00000080 NVRAM updated.
#0>pboot
boot = 00000082 .
#0>p aboot 7
00000009 NVRAM updated.
#0>paboot
aboot = 00000007
P3. SAR 1401: Nview would fail to start up properly on some systems.
The problem was due to the thread library improperly deciding
to make the primordial thread a bound thread instead of a
multiplexing thread.
R3. Change the thread library so that it more thoroughly checks
to see if the user actually set the __primordial_bound global
variable to a non-zero value.
P4. On PowerHawk Series 600 and PowerStack II platforms,
the vme_address(3c) routine was returning the incorrect CPU
physical address for the A24 or A16 VME address that was
specified by the caller.
R4. The vme_address(3c) routine now returns the correct CPU physical
for A24 and A16 VME addresses on Power Hawk Series 600 and
Power Stack II platforms.
P5. SAR 1415: When two or more LWPs are executing shmget(2) calls
inside the kernel at roughly the same time, it is possible for
the same kernel shared memory id structure (kshmid_ds) to be
used for more than one shared memory segment.
When multiple shared memory segments share the same underlying
kshmid_ds structure, then problems will occur when the shared
memory segment is removed (shmtcl(2) IPC_RMID) and when a
shmget(2) call is made to create or get a shared memory region
with the previous shared memory key value.
R5. Change the kernel code that allocates kshmid_ds structures so
that each kshmid_ds structure is used for only one new shared
memory segment.
P6. SAR 1429: PANIC: Address fault from interrupt routine, usually
with a non-zero kernel address displayed. At the time of the
panic, the kernel function kmem_alloc+0x870 was executing.
R6. kmem_alloc() was trying to allocate a buffer from the global
freelist. The buffer entry on the freelist was no longer
mapped to a physical page. A race condition between the
refreshpools daemon that runs every 30 minutes and several
kernel virtual space allocation and deallocation routines
caused an invalid entry to be left on the list after the
physical page was deallocated. The next reference to that
address by kmem_alloc() caused a panic. kma_shrinkpools()
(called by refreshpools) opens a preemption point window
periodically to allow other cpus access to the freelists
for allocates and frees. Other users of the list were not
properly notifying kma_shrinkpools when they added or deleted
entries that it needed to be aware of. Changes were made to
kma_shrinkpools(), kmem_alloc(), kmem_free(), and
kma_giveback_all().
P7. SAR 849: When using mount_memfs to mount a memfs filesystem
and a swapmax option is not supplied, the mount defaults to
a swapmax of INT_MAX. This essentially allocates all of your
system's swap space to the memfs filesystem just mounted and
can eventually lead to a system lockup.
R7. The default swapmax value within the memfs driver has been
changed to a more realistic limit. It is now one fourth of
the current free swap space at the time of the mount.
P8. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R8. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
P9. SAR 1452: The Max value of the ipc tunable SHMMAX is
insufficient for newer add-on memory boards such as
the VMIVME 5565.
R9. The Max value of the ipc tunable SHMMAX has been increased
from its current value of 100MB to a new value of 256MB to
accomodate newer add-on memory boards.
P10. SAR 1428: Customer upgraded his PowerMAXION with 32MB of
global memory from PowerMAX OS 4.2 to PowerMAX OS 5.1 and
was encountering determinism problems. Triggering of the
wallclock_ager daemon was changed to reflect newer systems
with large global memory pools.
R10. Add tunables to allow the user to revert back to PowerMAX OS
4.2 limits. Defaults will be the new limits.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/radl
/etc/conf/mtune.d/gd
/etc/conf/mtune.d/ipc
/etc/conf/mtune.d/mem
/etc/conf/mtune.d/memfs
/etc/conf/mtune.d/xfs
/etc/conf/mtune.d/xfsth
/etc/conf/pack.d/ipc/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/memfs/Driver.o
/etc/conf/pack.d/name/Driver.o
/stand/cp1
/usr/ccs/bin/as
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/include/sys/vmparam.h
/usr/lib/libthread.a
/usr/lib/libthread.so
/usr/sbin/rxfsdump
/usr/sbin/rxfsrestore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-013
Date Issued: 05/08/2003 10:34:22
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: cmds-010, dec-006, hsde-002, inet-012, man-010,
pg-006, rmxf-001, rpc-003, vmet-001, xfsd-002
Brief Description:
PowerMAX OS 4.3 base package release updates
-------------------------------------------------------------------------------
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P13 is the abbreviated name for PowerMAX OS 4.3 Patch Set 13.
PowerMAX OS 4.3P13 Software Update Package(s):
base-013 Base System Patch 013 (4.3P13)
cmds-010 Advanced Commands Patch 010 (4.3P13)
dec-006 DEC Ethernet Driver Patch 006 (4.3P13)
hsde-002 High Speed Data Enhanced Channel Driver Patch 002 (4.3P13)
inet-012 Internet Utilities Patch 012 (4.3P13)
man-010 On-line Manual Pages Patch 010 (4.3P13)
pg-006 Peregrine FDDI Driver Patch 006 (4.3P13)
rmxf-001 RAMiX PMC665 FOB Ethernet Driver Patch 001 (4.3P13)
rpc-003 Remote Procedure Calls Utilities Patch 003 (4.3P13)
vmet-001 VME to PCI/PMC Carrier Card Patch 001 (4.3P13)
xfsd-002 Distributed XFS File System Patch 002 (4.3P13)
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Compilation of the C library,
usr/src/ppc/lib/libc/ppc/gen/mall_data.s,
reported the bogus error.
R1. Return the static definition on dofpnum.c, but rename it
s_bzero to highlight its difference from bzero(). Modified
the error message in question to report "result is underflow".
P2. SAR 1401: Nview would fail to start up properly on some systems.
The problem was due to the thread library improperly deciding
to make the primordial thread a bound thread instead of a
multiplexing thread.
R2. Change the thread library so that it more thoroughly checks
to see if the user actually set the __primordial_bound global
variable to a non-zero value.
P3. SAR 1415: When two or more LWPs are executing shmget(2) calls
inside the kernel at roughly the same time, it is possible for
the same kernel shared memory id structure (kshmid_ds) to be
used for more than one shared memory segment.
When multiple shared memory segments share the same underlying
kshmid_ds structure, then problems will occur when the shared
memory segment is removed (shmtcl(2) IPC_RMID) and when a
shmget(2) call is made to create or get a shared memory region
with the previous shared memory key value.
R3. Change the kernel code that allocates kshmid_ds structures so
that each kshmid_ds structure is used for only one new shared
memory segment.
P4. SAR 1429: PANIC: Address fault from interrupt routine, usually
with a non-zero kernel address displayed. At the time of the
panic, the kernel function kmem_alloc+0x870 was executing.
R4. kmem_alloc() was trying to allocate a buffer from the global
freelist. The buffer entry on the freelist was no longer
mapped to a physical page. A race condition between the
refreshpools daemon that runs every 30 minutes and several
kernel virtual space allocation and deallocation routines
caused an invalid entry to be left on the list after the
physical page was deallocated. The next reference to that
address by kmem_alloc() caused a panic. kma_shrinkpools()
(called by refreshpools) opens a preemption point window
periodically to allow other cpus access to the freelists
for allocates and frees. Other users of the list were not
properly notifying kma_shrinkpools when they added or deleted
entries that it needed to be aware of. Changes were made to
kma_shrinkpools(), kmem_alloc(), kmem_free(), and
kma_giveback_all().
P5. SAR 849: When using mount_memfs to mount a memfs filesystem
and a swapmax option is not supplied, the mount defaults to
a swapmax of INT_MAX. This essentially allocates all of your
system's swap space to the memfs filesystem just mounted and
can eventually lead to a system lockup.
R5. The default swapmax value within the memfs driver has been
changed to a more realistic limit. It is now one fourth of
the current free swap space at the time of the mount.
P6. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R6. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
P7. SAR 1452: The Max value of the ipc tunable SHMMAX is
insufficient for newer add-on memory boards such as
the VMIVME 5565.
R7. The Max value of the ipc tunable SHMMAX has been increased
from its current value of 100MB to a new value of 256MB to
accomodate newer add-on memory boards.
P8. SAR 1428: Customer upgraded his PowerMAXION with 32MB of
global memory from PowerMAX OS 4.2 to PowerMAX OS 5.1 and
was encountering determinism problems. Triggering of the
wallclock_ager daemon was changed to reflect newer systems
with large global memory pools.
R8. Add tunables to allow the user to revert back to PowerMAX OS
4.2 limits. Defaults will be the new limits.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/radl
/etc/conf/mtune.d/gd
/etc/conf/mtune.d/ipc
/etc/conf/mtune.d/mem
/etc/conf/mtune.d/memfs
/etc/conf/mtune.d/xfs
/etc/conf/mtune.d/xfsth
/etc/conf/pack.d/ipc/Driver.o
/etc/conf/pack.d/kma/Driver.o
/etc/conf/pack.d/kmadbg/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/mem/space.c
/etc/conf/pack.d/memfs/Driver.o
/etc/conf/pack.d/name/Driver.o
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libp/libc.a
/usr/include/sys/vmparam.h
/usr/lib/libthread.a
/usr/lib/libthread.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-014
Date Issued: 01/07/2004 15:56:20
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Brief Description: PowerMAX OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P14 is the abbreviated name for PowerMAX OS 4.3 Patch Set 14.
###############################################################################
PowerMAX OS 4.3P14 Software Update Package(s):
1553v5drv-004 1553v5-ABI User Level Driver Patch 004 (4.3P14)
1553v5lib-002 1553v5-ABI Library Interfaces Patch 002 (4.3P14)
base-014 Base System Patch 014 (4.3P14)
cmds-011 Advanced Commands Patch 011 (4.3P14)
cnd-007 Condor Ethernet Driver Patch 007 (4.3P14)
dec-007 DEC Ethernet Driver Patch 007 (4.3P14)
diskless-009 Diskless Systems Package Patch 009 (4.3P14)
egl-005 Eagle Ethernet Driver Patch 005 (4.3P14)
fbs-006 Frequency Based Scheduler Patch 006 (4.3P14)
ip-006 Interphase 4511 PMC FDDI Driver Patch 006 (4.3P14)
man-011 On-line Manual Pages Patch 011 (4.3P14)
mvc-003 Multiplexor VME Controller Driver Patch 003 (4.3P14)
pg-007 Peregrine FDDI Driver Patch 007 (4.3P14)
via-004 VIA SCSI Adapter Interface Patch 004 (4.3P14)
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The logmemory() kernel routine that is used to log memory errors
into the /var/adm/errfile (errfile(4)) error log file was not
filling in the e_medr[] (error data registers) fields
of the memory error record properly.
This problem resulted in incorrect information being displayed
for memory errors via the errpt(1M) or errdead(1M) programs.
R1: Correct the logmemory() kernel routine so that the e_medr[]
fields in the error log record is filled in properly.
P2: A lwp may be switched to a different CPU during a spl_request(3X)
or spl_manage(3X) operation. This will produce indeterminate
spl level results, and thus unpredictable system behavior.
R2: Internally prevent the lwp from changing CPUs during a
spl_request(3x) or spl_manage(3x) operation.
P3: On some platforms, only the last CE_PANIC message gets output to
the console during sysfault processing. The additional CE_CONT
messages that are related to this sysfault do not make it out to
the console, and often do not make it into the putbuf[].
R3: This problem is due to console logging being enabled. In this
case, these additional messages are pushed into the consbuf[]
for delayed processing via do_conslog()/strlog().
So in the trap() sysfault code, temporarily disable console
logging around those cmn_err() or printf() messages that need to
get pushed out directly to the console and putbuf[].
P4: The autotune variables were being set based on the total amount
of memory instead of just the amount of global memory. On Night
Hawk systems that have a small amount of global memory and a
large amount of local memory, too much global memory was being
used for fs buffers.
R4: The tunemem variable used by the autotune calculations was changed
to be based on globalmem instead of totalmem.
P5: pciscan -b causes a PCI bus error and panics on Night Hawk systems.
R5: The Night Hawk version of a system call that was used by pciscan
to obtain base register mapping information was destructive.
It does not check to see if a device has already been configured.
This caused the mapping for the PCI device to be changed and the
next reference to the old mapping generated a PCI bus error and
the system panic'd with a sysfault. Changed the Night Hawk version
of pciscan so it no longer makes the destructive system call.
P6: physmemcntl(2) fails with an EFAULT when either:
- only a size is given
- a begin address is given but not an end address
AND there is an empty entry in the memextent list.
R6: The physmemcntl code was not checking for empty memextent entries
(where me_floor == me_ceiling). When this situation occurred,
physmemcntl attempted to find the page table address for an invalid
me_floor or me_ceiling address. Changed the code that loops through
the memextent entries so it bypasses all empty entries.
P7: SAR 1544: A previous change to pkginstall caused any response file
created to be eliminated if there was a non-zero return code. This
should have eliminated only zero-length response files.
R7: Added the additional check for the size of the response file
before eliminating it.
P8: SAR 1535: The savecore(1M) utility cannot save a dump on a system
that has 2GB or more of physical memory or has a dumpsize 2GB or
greater. The crash(1M) utility cannot read a crashfile that is 2GB
or greater.
R8: Savecore(1M) was written such that it had a limit of 2GB - 1.
To resolve this involved changing many data types from off_t
or int into size_t. An additional change involved setting
RLIMIT_FSIZE to (2 * RLIM_INFINITY). This will now allow
savecore(1M) to save a dump on a system that has nearly 4GB of
physical memory. The crash(1M utility was enhanced to read
crashfiles that are 2GB or greater.
P9: When in input mode (-i <file>), ktrace can print out
"vector does not match" over and over and over again.
R9: It is possible to have a rescheduling interrupt exit event without
a rescheduling interrupt entry event. This can happen when we enter
the rescheduling code not via a rescheduling interrupt. This
"unmatched" state confuses ktrace and thus the message. Resolve by
not printing out the message and ignoring the event, just as ntfilter
and the newer ntrace/autofiltering do.
P10: Kernel trace file would contain an event whose timestamp was
incoherent. This confused ntfilter into thinking there was a
huge time warp around this single event.
Kernel events are logged by :
Lock trace spinlock.
Allocate trace event slot in current trace buffer.
Unlock trace spinlock.
Fill in trace event slot.
Failure sequence:
1)Application gets preempted after the unlock and before the
fill in trace event slot.
2)Other trace events are logged, enough to fill current trace buffer.
3)Current trace buffer is written to disk and discarded.
4)Application resumes and fills in it's trace event slot in a
discarded trace buffer.
R10: Fill in the trace event info while trace spinlock is locked.
Enhancements:
E1: PowerMAX OS is enhanced to support the following directives:
uleb128 value [,value]
.uleb128 value [,value]
sleb128 value [,value]
.sleb128 value [,value]
The specified values are encoded using Little Endian Base 128 (LEB128)
numbers and assembled into consecutive locations, each of which is
the minimum size for the encoded LEB128 value. There is no
restriction on the divisibility of the location counter when these
directives are used. The first two forms are synonyms which use the
unsigned LEB128 encoding. The last two forms are synonyms which use
the signed LEB128 encoding. The LEB128 encodings are described in
the DWARF Debugging Information Format manual under section 7.6,
Variable Length Data. [This manual also is included as a chapter in
Compilation Systems Volume 2].
Object(s) To Be Replaced:
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/io/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/usr/ccs/bin/as
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libnc.a
/usr/include/mqueue.h
/usr/include/sys/assym_dbg.h
/usr/include/sys/assym_nodbg.h
/usr/include/sys/ktrace.h
/usr/include/sys/lwp.h
/usr/include/sys/proc.h
/usr/include/sys/syscx.h
/usr/lib/libc.so.1
/usr/lib/libud.a
/usr/lib/libud.so
/usr/sadm/install/bin/pkginstall
/usr/sbin/pciscan
/usr/sbin/savecore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-014
Date Issued: 01/07/2004 15:56:46
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: none
Brief Description: PowerMAX OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P14 is the abbreviated name for PowerMAX OS 4.3 Patch Set 14.
###############################################################################
PowerMAX OS 4.3P14 Software Update Package(s):
1553v5drv-004 1553v5-ABI User Level Driver Patch 004 (4.3P14)
1553v5lib-002 1553v5-ABI Library Interfaces Patch 002 (4.3P14)
base-014 Base System Patch 014 (4.3P14)
cmds-011 Advanced Commands Patch 011 (4.3P14)
cnd-007 Condor Ethernet Driver Patch 007 (4.3P14)
dec-007 DEC Ethernet Driver Patch 007 (4.3P14)
egl-005 Eagle Ethernet Driver Patch 005 (4.3P14)
ie-005 Night Hawk ISE Ethernet Interface Module Patch 005 (4.3P14)
man-011 On-line Manual Pages Patch 011 (4.3P14)
mvc-003 Multiplexor VME Controller Driver Patch 003 (4.3P14)
pg-007 Peregrine FDDI Driver Patch 007 (4.3P14)
via-004 VIA SCSI Adapter Interface Patch 004 (4.3P14)
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The logmemory() kernel routine that is used to log memory errors
into the /var/adm/errfile (errfile(4)) error log file was not
filling in the e_medr[] (error data registers) fields
of the memory error record properly.
This problem resulted in incorrect information being displayed
for memory errors via the errpt(1M) or errdead(1M) programs.
R1: Correct the logmemory() kernel routine so that the e_medr[]
fields in the error log record is filled in properly.
P2: A lwp may be switched to a different CPU during a spl_request(3X)
or spl_manage(3X) operation. This will produce indeterminate
spl level results, and thus unpredictable system behavior.
R2: Internally prevent the lwp from changing CPUs during a
spl_request(3x) or spl_manage(3x) operation.
P3: On some platforms, only the last CE_PANIC message gets output to
the console during sysfault processing. The additional CE_CONT
messages that are related to this sysfault do not make it out to
the console, and often do not make it into the putbuf[].
R3: This problem is due to console logging being enabled. In this
case, these additional messages are pushed into the consbuf[]
for delayed processing via do_conslog()/strlog().
So in the trap() sysfault code, temporarily disable console
logging around those cmn_err() or printf() messages that need to
get pushed out directly to the console and putbuf[].
P4: The autotune variables were being set based on the total amount
of memory instead of just the amount of global memory. On Night
Hawk systems that have a small amount of global memory and a
large amount of local memory, too much global memory was being
used for fs buffers.
R4: The tunemem variable used by the autotune calculations was changed
to be based on globalmem instead of totalmem.
P5: pciscan -b causes a PCI bus error and panics on Night Hawk systems.
R5: The Night Hawk version of a system call that was used by pciscan
to obtain base register mapping information was destructive.
It does not check to see if a device has already been configured.
This caused the mapping for the PCI device to be changed and the
next reference to the old mapping generated a PCI bus error and
the system panic'd with a sysfault. Changed the Night Hawk version
of pciscan so it no longer makes the destructive system call.
P6: physmemcntl(2) fails with an EFAULT when either:
- only a size is given
- a begin address is given but not an end address
AND there is an empty entry in the memextent list.
R6: The physmemcntl code was not checking for empty memextent entries
(where me_floor == me_ceiling). When this situation occurred,
physmemcntl attempted to find the page table address for an invalid
me_floor or me_ceiling address. Changed the code that loops through
the memextent entries so it bypasses all empty entries.
P7: SAR 1544: A previous change to pkginstall caused any response file
created to be eliminated if there was a non-zero return code. This
should have eliminated only zero-length response files.
R7: Added the additional check for the size of the response file
before eliminating it.
P8: SAR 1535: The savecore(1M) utility cannot save a dump on a system
that has 2GB or more of physical memory or has a dumpsize 2GB or
greater. The crash(1M) utility cannot read a crashfile that is 2GB
or greater.
R8: Savecore(1M) was written such that it had a limit of 2GB - 1.
To resolve this involved changing many data types from off_t
or int into size_t. An additional change involved setting
RLIMIT_FSIZE to (2 * RLIM_INFINITY). This will now allow
savecore(1M) to save a dump on a system that has nearly 4GB of
physical memory. The crash(1M utility was enhanced to read
crashfiles that are 2GB or greater.
P9: When in input mode (-i <file>), ktrace can print out
"vector does not match" over and over and over again.
R9: It is possible to have a rescheduling interrupt exit event without
a rescheduling interrupt entry event. This can happen when we enter
the rescheduling code not via a rescheduling interrupt. This
"unmatched" state confuses ktrace and thus the message. Resolve by
not printing out the message and ignoring the event, just as ntfilter
and the newer ntrace/autofiltering do.
P10: Kernel trace file would contain an event whose timestamp was
incoherent. This confused ntfilter into thinking there was a
huge time warp around this single event.
Kernel events are logged by :
Lock trace spinlock.
Allocate trace event slot in current trace buffer.
Unlock trace spinlock.
Fill in trace event slot.
Failure sequence:
1)Application gets preempted after the unlock and before the
fill in trace event slot.
2)Other trace events are logged, enough to fill current trace buffer.
3)Current trace buffer is written to disk and discarded.
4)Application resumes and fills in it's trace event slot in a
discarded trace buffer.
R10: Fill in the trace event info while trace spinlock is locked.
P11: PowerIO Edge Triggered Interupt(ETI) test can cause a bogus VME bus
error on PowerMAXION-8(NH6408) platforms.
ETI test manually generates ETI interrupt with ioctl(ETI_REQUEST),
which OR's in the ETI bit in MA_INTREQ (interrupt request register).
This is implemented as :
read MA_INTREQ
or in our ETI bit
write MA_INTREQ
Problem occurs when another CPU clears a VME interrupt, different
bit in the MA_INTREQ , between the time we read it and wrote it.
Our write to MA_INTREQ "reasserts" the VME interrupt, causing an
IAK read to the VME, which times out.
R11: Change bsp*/config_ints.c(write_int_nc()) to SET/pulse only the
requested bit in the MA_INTREQ.
Enhancements:
E1: PowerMAX OS is enhanced to support the following directives:
uleb128 value [,value]
.uleb128 value [,value]
sleb128 value [,value]
.sleb128 value [,value]
The specified values are encoded using Little Endian Base 128 (LEB128)
numbers and assembled into consecutive locations, each of which is
the minimum size for the encoded LEB128 value. There is no
restriction on the divisibility of the location counter when these
directives are used. The first two forms are synonyms which use the
unsigned LEB128 encoding. The last two forms are synonyms which use
the signed LEB128 encoding. The LEB128 encodings are described in
the DWARF Debugging Information Format manual under section 7.6,
Variable Length Data. [This manual also is included as a chapter in
Compilation Systems Volume 2].
Object(s) To Be Replaced:
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/bsp6400/Driver.o
/etc/conf/pack.d/bsp6408/Driver.o
/etc/conf/pack.d/bsp6800p/Driver.o
/etc/conf/pack.d/io/Driver.o
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/conf/pack.d/svc/Driver.o
/etc/conf/pack.d/util/Driver.o
/usr/ccs/bin/as
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libnc.a
/usr/include/mqueue.h
/usr/include/sys/assym_dbg.h
/usr/include/sys/assym_nodbg.h
/usr/include/sys/ktrace.h
/usr/include/sys/lwp.h
/usr/include/sys/proc.h
/usr/include/sys/syscx.h
/usr/lib/libc.so.1
/usr/lib/libud.a
/usr/lib/libud.so
/usr/sadm/install/bin/pkginstall
/usr/sbin/pciscan
/usr/sbin/savecore
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-015
Date Issued: 07/08/2004 10:50:53
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Brief Description: PowerMAX OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P15 is the abbreviated name for PowerMAX OS 4.3 Patch Set 15.
###############################################################################
PowerMAX OS 4.3P15 Software Update Package(s):
base-015 Base System Patch 015 (4.3P15)
cnd-008 Condor Ethernet Driver Patch 008 (4.3P15)
crypt-008 Domestic Encryption Utilities Patch 008 (4.3P15)
diskless-010 Diskless Systems Package Patch 010 (4.3P15)
egl-006 Eagle Ethernet Driver Patch 006 (4.3P15)
fbs-007 Frequency Based Scheduler Patch 007 (4.3P15)
inet-013 Internet Utilities Patch 013 (4.3P15)
man-012 On-line Manual Pages Patch 012 (4.3P15)
ncr-003 NCR Internal SCSI Controller Patch 003 (4.3P15)
nfs-004 Network File System Utilities Patch 004 (4.3P15)
softint-003 Software Packaging Tools Patch 003 (4.3P15)
trace-009 KernelTrace Utilities Patch 009 (4.3P15)
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1648: The destructions for static objects are being executed
after the routines that have been added by their constructors
to the atexit list.
R1: In the start up routine execute the finitp code before the initp
code. This way anything that gets added to the atexit list during
constructor execution will get exectuted at exit time after the
destructor code which got got added by the finitp.
P2: DT 114: Initp routines can't access the environment without it
getting stomped by the crt0.o startup routine.
R2: Set the _environ variable only once before begining to exectue
the initp routines.
P3: DT 537/1548: Incorrect names given to the fast math intrinsics.
Not all the new intrinsics were present.
R3: Fix up the old intrinsic names in conjunction with the 6.1
compiler fixes. Add all the new transcentental intrinsic
declarations.
P4: On closely coupled systems, the association of an internal data
structure to its user-level posix named semaphore can disappear
while the semaphore is not referenced by any process(es). The
observed manifestation of the problem is that when the semaphore
is again referenced, the semaphore count will not be preserved
(because the semaphore is given a new kernel structure
association).
This problem has not been reported on systems other than closely
coupled systems, but presumably it ccold occur on these other
systems.
R4: A change to proc/Driver.o preserves the assocation, by holding
the vnode for the semaphore association so long as the semaphore
is visible in the posix semaphore namespace.
P5: DT 1805: PANIC in hat_unload() after call to hat_utt_unload().
R5: A process's data pages are temporarily switched to be read-only
after a fork until either the parent or child tries to do a
write to the page. VM then creates a new copy of the page for
the writer with write attributes. In this panic, the user
process (ntraceserv) had done an MLOCK_ALL before it did a
fork. The MLOCK_ALL caused hatptep->hatpte_wired to be TRUE.
When hat_utt_unload() checked whether it was OK to unload a
locked translation, the HAT_UNLOAD flag bit was not set so it
returned an EBUSY error. When hat_unload() got the error return
from hat_utt_unload() it generated a PANIC. The flags parameter
on the call to hat_unload() in segdz_faultpage() was changed
from HAT_NOFLAGS to HAT_UNLOCK. This will let the translation
be unloaded even if the page is wired/locked into memory.
P6: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R6: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
P7: DT 1602: Due to the size of some package names now, there is
the possibility that menu displays can scroll off the top of
the screen. This is a problem on dumb terminals where one
cannot scroll backwards in that the data is lost before it
can be seen.
R7: The printmenu() routine in libadm.a has been modified to
ensure that no more than 24 lines will be displayed when
displaying the menu selections.
P8: DT 1572: The rv_priority field in a rescheduling variable
allows an lwp, in user mode, to boost it's priority without
making a system call. This is implemented in the rescheduling
interrupt code. The rescheduling interrupt checks to see if
the lwp has an rv_priority boosted priority, and uses that
boosted priority to determine if a preemption should occurr.
The problem happens when lwpA with a boosted rv_priority is
actually preempted by lwpB, with a higher priority than lwpA's
boosted priority. At this point the lwpA gets left on the
runque at his scheduled priority, not his boosted priority.
If lwpC becomes runnable, and is at a priority above lwpA's
scheduled priority and below lwpA's boosted priority, the
runques for that CPU will be in the order of lwpB lwpC lwpA
(w/ hidden boost value) When lwpB gives up the CPU, lwpC takes
over and it should not, because lwpA should be boosted over
lwpC. MAXAda assumes this rv_priority boost stuff works, so
when it fails it can cause a priority inversion deadlock in
the Ada runtime.
R8: When the rescheduling interrupts detect an lwp's boosted
priority is higher than it's current priority, we change
that lwp's current priority and runque to the boosted priority.
This gets reset when the user lowers it's rv_priority value
and sees the rv_pending flags set. By convention, the user
is required to call resched_cntl(RESCHED_SERVE) when it sees
the rv_pending flag set.
P9: DT-1614: The tar command provides for an exclude file
(-X option) which lists files that are not to be written to
the tar file (-c) or extracted from the tar file (-x). It
would be assumed that if a directory was listed, that the
directory and all files under that directory would be excluded,
however, it was expecting that each file under each directory
and subdirectory be listed separately. This was not the
intention of the -X option, nor would it be reasonable behavior.
R9: Changed two calls to strcmp in the routine that checks for the
filename in the passed-in list to strncmp and used the string
length of the directory name as the number of characters to
compare.
P10: DT 1599: When utilizing the VPATH feature of make(1) and when
an end object and an intermediate object are not present on
the initial build in make(1) they will be properly made from
the source file. However, after editing the source and invoking
make(1) again, neither the intermediate nor the final object
gets rebuilt.
R10: Corrected a bug in the doname() routine in make(1) which was
causing it not to sense the need to rebuild the objects from
the modified source and the VPATH feature was being used.
Also corrected a bug which had disabled building for implicit
single suffix rules.
P11: The last change to 'make' fixed a VPATH problem which was
showing up in the original 'make' utility. This change turned
out to not only fix that problem but to reveal another more
sinister bug where it would try to locate standard headers
prefixed by the current working directory.
R11: In sstat() a test was being improperly done to check for
whether or not to include the VPATH path as a prefix to the
file being searched for. In the case of the 0th VPATH in the
array of VPATHs it should not have prefixed the filename
with the current working directory.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/make
/usr/ccs/lib/crt0.o
/usr/include/math.h
/usr/include/stdlib.h
/usr/lib/libnsl_i.a
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgremove
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/tar
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-015
Date Issued: 07/26/2004 11:49:06
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Brief Description: PowerMAX OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P15 is the abbreviated name for PowerMAX OS 4.3 Patch Set 15.
###############################################################################
PowerMAX OS 4.3P15 Software Update Package(s):
base-015 Base System Patch 015 (4.3P15)
cnd-008 Condor Ethernet Driver Patch 008 (4.3P15)
crypt-008 Domestic Encryption Utilities Patch 008 (4.3P15)
egl-006 Eagle Ethernet Driver Patch 006 (4.3P15)
fbs-004 Frequency Based Scheduler Patch 004 (4.3P15)
inet-013 Internet Utilities Patch 013 (4.3P15)
man-012 On-line Manual Pages Patch 012 (4.3P15)
ncr-003 NCR Internal SCSI Controller Patch 003 (4.3P15)
nfs-004 Network File System Utilities Patch 004 (4.3P15)
softint-003 Software Packaging Tools Patch 003 (4.3P15)
trace-009 KernelTrace Utilities Patch 009 (4.3P15)
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1648: The destructions for static objects are being executed
after the routines that have been added by their constructors
to the atexit list.
R1: In the start up routine execute the finitp code before the initp
code. This way anything that gets added to the atexit list during
constructor execution will get exectuted at exit time after the
destructor code which got got added by the finitp.
P2: DT 114: Initp routines can't access the environment without it
getting stomped by the crt0.o startup routine.
R2: Set the _environ variable only once before begining to exectue
the initp routines.
P3: DT 537/1548: Incorrect names given to the fast math intrinsics.
Not all the new intrinsics were present.
R3: Fix up the old intrinsic names in conjunction with the 6.1
compiler fixes. Add all the new transcentental intrinsic
declarations.
P4: On closely coupled systems, the association of an internal data
structure to its user-level posix named semaphore can disappear
while the semaphore is not referenced by any process(es). The
observed manifestation of the problem is that when the semaphore
is again referenced, the semaphore count will not be preserved
(because the semaphore is given a new kernel structure
association).
This problem has not been reported on systems other than closely
coupled systems, but presumably it ccold occur on these other
systems.
R4: A change to proc/Driver.o preserves the assocation, by holding
the vnode for the semaphore association so long as the semaphore
is visible in the posix semaphore namespace.
P5: DT 1805: PANIC in hat_unload() after call to hat_utt_unload().
R5: A process's data pages are temporarily switched to be read-only
after a fork until either the parent or child tries to do a
write to the page. VM then creates a new copy of the page for
the writer with write attributes. In this panic, the user
process (ntraceserv) had done an MLOCK_ALL before it did a
fork. The MLOCK_ALL caused hatptep->hatpte_wired to be TRUE.
When hat_utt_unload() checked whether it was OK to unload a
locked translation, the HAT_UNLOAD flag bit was not set so it
returned an EBUSY error. When hat_unload() got the error return
from hat_utt_unload() it generated a PANIC. The flags parameter
on the call to hat_unload() in segdz_faultpage() was changed
from HAT_NOFLAGS to HAT_UNLOCK. This will let the translation
be unloaded even if the page is wired/locked into memory.
P6: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R6: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
P7: DT 1602: Due to the size of some package names now, there is
the possibility that menu displays can scroll off the top of
the screen. This is a problem on dumb terminals where one
cannot scroll backwards in that the data is lost before it
can be seen.
R7: The printmenu() routine in libadm.a has been modified to
ensure that no more than 24 lines will be displayed when
displaying the menu selections.
P8: DT 1572: The rv_priority field in a rescheduling variable
allows an lwp, in user mode, to boost it's priority without
making a system call. This is implemented in the rescheduling
interrupt code. The rescheduling interrupt checks to see if
the lwp has an rv_priority boosted priority, and uses that
boosted priority to determine if a preemption should occurr.
The problem happens when lwpA with a boosted rv_priority is
actually preempted by lwpB, with a higher priority than lwpA's
boosted priority. At this point the lwpA gets left on the
runque at his scheduled priority, not his boosted priority.
If lwpC becomes runnable, and is at a priority above lwpA's
scheduled priority and below lwpA's boosted priority, the
runques for that CPU will be in the order of lwpB lwpC lwpA
(w/ hidden boost value) When lwpB gives up the CPU, lwpC takes
over and it should not, because lwpA should be boosted over
lwpC. MAXAda assumes this rv_priority boost stuff works, so
when it fails it can cause a priority inversion deadlock in
the Ada runtime.
R8: When the rescheduling interrupts detect an lwp's boosted
priority is higher than it's current priority, we change
that lwp's current priority and runque to the boosted priority.
This gets reset when the user lowers it's rv_priority value
and sees the rv_pending flags set. By convention, the user
is required to call resched_cntl(RESCHED_SERVE) when it sees
the rv_pending flag set.
P9: DT-1614: The tar command provides for an exclude file
(-X option) which lists files that are not to be written to
the tar file (-c) or extracted from the tar file (-x). It
would be assumed that if a directory was listed, that the
directory and all files under that directory would be excluded,
however, it was expecting that each file under each directory
and subdirectory be listed separately. This was not the
intention of the -X option, nor would it be reasonable behavior.
R9: Changed two calls to strcmp in the routine that checks for the
filename in the passed-in list to strncmp and used the string
length of the directory name as the number of characters to
compare.
P10: DT 1599: When utilizing the VPATH feature of make(1) and when
an end object and an intermediate object are not present on
the initial build in make(1) they will be properly made from
the source file. However, after editing the source and invoking
make(1) again, neither the intermediate nor the final object
gets rebuilt.
R10: Corrected a bug in the doname() routine in make(1) which was
causing it not to sense the need to rebuild the objects from
the modified source and the VPATH feature was being used.
Also corrected a bug which had disabled building for implicit
single suffix rules.
P11: The last change to 'make' fixed a VPATH problem which was
showing up in the original 'make' utility. This change turned
out to not only fix that problem but to reveal another more
sinister bug where it would try to locate standard headers
prefixed by the current working directory.
R11: In sstat() a test was being improperly done to check for
whether or not to include the VPATH path as a prefix to the
file being searched for. In the case of the 0th VPATH in the
array of VPATHs it should not have prefixed the filename
with the current working directory.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/cf.d/intr.s
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/proc/Driver.o
/usr/bin/pkginfo
/usr/bin/pkgparam
/usr/ccs/bin/make
/usr/ccs/lib/crt0.o
/usr/include/math.h
/usr/include/stdlib.h
/usr/lib/libnsl_i.a
/usr/sadm/install/bin/pkginstall
/usr/sadm/install/bin/pkgremove
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/tar
/etc/conf/pack.d/name/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: base-016
Date Issued: 03/13/2006 13:16:53
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 base package release updates
###############################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P16 is the abbreviated name for PowerMAX OS 4.3 Patch Set 16.
###############################################################################
PowerMAX OS 4.3P16 Software Update Package(s):
base-016 Base System Patch 015 (4.3P16)
cmds-012 Advanced Commands Patch 012
dec-008 DEC Ethernet Driver Patch 008 (4.3P16)
gpib-002 NI PCI-GPIB Kernel Level Driver Patch 002 (4.3P16)
inet-014 Internet Utilities Patch 013 (4.3P16)
man-013 On-line Manual Pages Patch 012 (4.3P16)
ngpib-001 NI 1014-D IEEE-488 User-Level Driver Patch 001 (4.3P16)
nsu-006 Network Support Utilities Patch 006 (4.3P16)
pg-008 Peregrine FDDI Driver Patch 008 (4.3P16)
rpc-005 Remote Procedure Calls Utilities Patch 005 (4.3P16)
softint-004 Software Packaging Tools Patch 003 (4.3P16)
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT539: A test program sometimes fails for shared dynamic
compilations because the routine _xwcstombs() examines 1 character
beyond the input length that it is constrained to observe. If this
extra character is non-zero, the program exhibits failure symptoms.
R1: End for-loop when the number of characters examined reaches
the input length; also return the tooshort exit for this case.
P2: Last 3 lines of procedure were inerted after end bracket (related
to DT539).
R2: Move end of procedure bracket after the 3 lines.
P3: DT2761: Some c99 defines were missing.
R3: Put the va_copy define into /usr/src/ppc/head/stdvar_args.h
Put in the extra defines of LLONG_MIN, LLONG_MAX, and UULONG_MAX
into /usr/src/ppc/head/limits.h, depending on __STDC_VERSION__.
P4: DT3387: sqrtf is extremely inaccurate when dealing with denormed
input arguments.
R4: Treat denormed input arguments as calls to the double precision
sqrt.
P5: DT3626: PH740 hangs in idle with many processes on the dispatch
queues.
R5: Cache coherency problem. Add a sync and dcbf instruction after
all writes to e_npri to force e_npri to memory and invalidate
the data cache block containing e_npri so that the idle process
will see the e_npri value change.
P6: DT3054: If the rlimits_t structure was declared in a program that
had either _KERNEL or _KMEMUSER defined, then the following error
would occur: Undefined identifier <RLIM64_NLIMITS>
R6: The rlimits_t structure in ppc/uts/proc/resource.h was changed
to conditionally include the rl64_limits structure element
based on whether _LARGEFILE64_SOURCE is defined.
P7: Compilation using the -D_KMEMUSER switch would fail with a program
that includes <sys/hrtcallout.h> because it would also try to
include <sys/hrtcallout_p.h>, which is not available on a
released system.
R7: Make /usr/include/sys/hrtcallout_p.h a released header.
P8: Doing a ":mksyshead" within a task will not put the new header
file /usr/include/sys/hrtcallout_p.h into $OBJ/usr/include/sys.
R8: Changed svc.mk so that "mksyshead" will install hrtcallout_p.h in
/usr/include/sys.
P9: DT2128: Selecting "all" or greater than 120 packages from the
menu during pkgadd or pkgask will segfault.
R9: The pkgadd(1M)/pkgask(1M) commands contained a hard coded
limit of 120 on a calloc() call. This has been changed to
use dynamic allocation based on the number of packages selected.
Also, the libpkg library did not allow greater than 128 packages
to be selected in gpkglist(). This has been increased to 256.
P10: The RIQ10192 work overlooked how LOGFAILURES was handled if
MAXTRYS was less than LOGFAILURES. The passwd utility was not
reinitializing the counters in the restrict case on a failed
attempt so it allowed success on a second failed attempt.
R10: This has been corrected in this task.
P11: DT3442: 0.222507d-307 is a denormalized number
(0x000ffffe2e8159d0). C gets the words reversed for a
denormalized number.
R11: Modify the assignments so they are not reversed.
P12: DT2704: The catman program expected the string '" t to appear on
man pages with table directives which would set up preprocessing.
However, newer man pages did not adhere to this convention. Man
pages also were not always run through col post processing which
made man pages with half-line directives hard to read. catman was
modified to automatically preprocess through tbl (a no-op if no
table directives existed) and postprocess through col -x.
R12: Modified the command string generated to preprocess through tbl
and postprocess through col -x.
P13: DT3432: A call to strchr() in the is_in_table() routine dealing
with directories in the exclude (and include) files caused the
directory names to be stripped from the beginning, not the end,
of the pathname and subdirectories didn't get handled correctly.
Therefore, if the exclude file had a line to exclude a whole
directory, the directory and one file was excluded, but all
subsequent files and/or subdirectories within it were extracted.
R13: Changed strchr() to strrchr().
Enhancements:
1. RIQ010101: Provided modifications to PowerMAX OS 4.3 and 5.1 to
support an OS feature to force the user to change their password
when logging into XDM and support the need to force mix case
during password changes and enforcement of password aging.
2. RIQ10192: Enhanced passwd and yppasswd for restrictions on
mixed-case requirements. Enhanced in.login for restrictions
on remote logins.
Object(s) To Be Replaced:
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/default/login
/etc/default/passwd
/usr/bin/passwd
/usr/ccs/bin/make
/usr/ccs/lib/libM.a
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libm.a
/usr/ccs/lib/libnc.a
/usr/include/limits.h
/usr/include/stdvar_args.h
/usr/include/sys/disp.h
/usr/include/sys/hrtcallout_p.h
/usr/include/sys/resource.h
/usr/lib/libc.so.1
/usr/lib/locale/C/MSGFILES/core.abi.str
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/tar
/usr/ucb/catman
Special Conditions for Installation:
The /usr/include/sys/hrtcallout_p.h file has been added.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: base-016
Date Issued: 03/13/2006 11:37:28
Software Package: base pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 base package release updates
################################################################################
Patches for PowerMAX OS 4.3 are now released officially as Patch Sets. A
patch set is a collection of patches released simultaneously. Individual
patches may have dependencies on other patches included in a patch set.
Not all patches in a Patch Set may be available/applicable on all supported
platforms. It is important to install all applicable patches in order to
ensure proper system operation.
4.3P16 is the abbreviated name for PowerMAX OS 4.3 Patch Set 16.
################################################################################
PowerMAX OS 4.3P16 Software Update Package(s):
base-016 Base System Patch 015 (4.3P16)
cmds-012 Advanced Commands Patch 012 (4.3P16)
dec-008 DEC Ethernet Driver Patch 008 (4.3P16)
gpib-002 NI PCI-GPIB Kernel Level Driver Patch 002 (4.3P16)
inet-014 Internet Utilities Patch 014 (4.3P16)
man-013 On-line Manual Pages Patch 013 (4.3P16)
ngpib-001 NI 1014-D IEEE-488 User-Level Driver Patch 001 (4.3P16)
nsu-006 Network Support Utilities Patch 006 (4.3P16)
pg-008 Peregrine FDDI Driver Patch 008 (4.3P16)
rpc-005 Remote Procedure Calls Utilities Patch 005 (4.3P16)
softint-004 Software Packaging Tools Patch 004 (4.3P16)
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT539: A test program sometimes fails for shared dynamic
compilations because the routine _xwcstombs() examines 1 character
beyond the input length that it is constrained to observe. If this
extra character is non-zero, the program exhibits failure symptoms.
R1: End for-loop when the number of characters examined reaches
the input length; also return the tooshort exit for this case.
P2: Last 3 lines of procedure were inerted after end bracket (related
to DT539).
R2: Move end of procedure bracket after the 3 lines.
P3: DT2761: Some c99 defines were missing.
R3: Put the va_copy define into /usr/src/ppc/head/stdvar_args.h
Put in the extra defines of LLONG_MIN, LLONG_MAX, and UULONG_MAX
into /usr/src/ppc/head/limits.h, depending on __STDC_VERSION__.
P4: DT3387: sqrtf is extremely inaccurate when dealing with denormed
input arguments.
R4: Treat denormed input arguments as calls to the double precision
sqrt.
P5: DT3626: PH740 hangs in idle with many processes on the dispatch
queues.
R5: Cache coherency problem. Add a sync and dcbf instruction after
all writes to e_npri to force e_npri to memory and invalidate
the data cache block containing e_npri so that the idle process
will see the e_npri value change.
P6: DT3054: If the rlimits_t structure was declared in a program that
had either _KERNEL or _KMEMUSER defined, then the following error
would occur: Undefined identifier <RLIM64_NLIMITS>
R6: The rlimits_t structure in ppc/uts/proc/resource.h was changed
to conditionally include the rl64_limits structure element
based on whether _LARGEFILE64_SOURCE is defined.
P7: Compilation using the -D_KMEMUSER switch would fail with a program
that includes <sys/hrtcallout.h> because it would also try to
include <sys/hrtcallout_p.h>, which is not available on a
released system.
R7: Make /usr/include/sys/hrtcallout_p.h a released header.
P8: Doing a ":mksyshead" within a task will not put the new header
file /usr/include/sys/hrtcallout_p.h into $OBJ/usr/include/sys.
R8: Changed svc.mk so that "mksyshead" will install hrtcallout_p.h in
/usr/include/sys.
P9: DT2128: Selecting "all" or greater than 120 packages from the
menu during pkgadd or pkgask will segfault.
R9: The pkgadd(1M)/pkgask(1M) commands contained a hard coded
limit of 120 on a calloc() call. This has been changed to
use dynamic allocation based on the number of packages selected.
Also, the libpkg library did not allow greater than 128 packages
to be selected in gpkglist(). This has been increased to 256.
P10: The RIQ10192 work overlooked how LOGFAILURES was handled if
MAXTRYS was less than LOGFAILURES. The passwd utility was not
reinitializing the counters in the restrict case on a failed
attempt so it allowed success on a second failed attempt.
R10: This has been corrected in this task.
P11: DT3442: 0.222507d-307 is a denormalized number
(0x000ffffe2e8159d0). C gets the words reversed for a
denormalized number.
R11: Modify the assignments so they are not reversed.
P12: DT2704: The catman program expected the string '" t to appear on
man pages with table directives which would set up preprocessing.
However, newer man pages did not adhere to this convention. Man
pages also were not always run through col post processing which
made man pages with half-line directives hard to read. catman was
modified to automatically preprocess through tbl (a no-op if no
table directives existed) and postprocess through col -x.
R12: Modified the command string generated to preprocess through tbl
and postprocess through col -x.
P13: DT3432: A call to strchr() in the is_in_table() routine dealing
with directories in the exclude (and include) files caused the
directory names to be stripped from the beginning, not the end,
of the pathname and subdirectories didn't get handled correctly.
Therefore, if the exclude file had a line to exclude a whole
directory, the directory and one file was excluded, but all
subsequent files and/or subdirectories within it were extracted.
R13: Changed strchr() to strrchr().
Enhancements:
1. RIQ010101: Provided modifications to PowerMAX OS 4.3 and 5.1 to
support an OS feature to force the user to change their password
when logging into XDM and support the need to force mix case
during password changes and enforcement of password aging.
2. RIQ10192: Enhanced passwd and yppasswd for restrictions on
mixed-case requirements. Enhanced in.login for restrictions
on remote logins.
Object(s) To Be Replaced:
/etc/conf/pack.d/mem/Driver.o
/etc/conf/pack.d/name/Driver.o
/etc/conf/pack.d/proc/Driver.o
/etc/default/login
/etc/default/passwd
/usr/bin/passwd
/usr/ccs/bin/make
/usr/ccs/lib/libM.a
/usr/ccs/lib/libc.a
/usr/ccs/lib/libc.so
/usr/ccs/lib/libm.a
/usr/ccs/lib/libnc.a
/usr/include/limits.h
/usr/include/stdvar_args.h
/usr/include/sys/disp.h
/usr/include/sys/hrtcallout_p.h
/usr/include/sys/resource.h
/usr/lib/libc.so.1
/usr/lib/locale/C/MSGFILES/core.abi.str
/usr/sbin/pkgadd
/usr/sbin/pkgchk
/usr/sbin/pkgrm
/usr/sbin/tar
/usr/ucb/catman
Special Conditions for Installation:
The /usr/include/sys/hrtcallout_p.h file has been added.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cld-001
Date Issued: 02/02/2000 13:23:19
Software Package: cld pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: PowerStack II
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cld package release updates
##############################################################################
Problem Description:
1. An alternate version of the 4/8 port decoder card requires different
setup of the internal registers.
Problem Resolution:
1. A new tunable, CLD_SVGA_OUT is provided which defines whether the
board has SVGA or regular outputs. The cld driver's internal
registers were re-programmed accordingly.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cld/Driver.o
/etc/conf/pack.d/cld/space.c
/etc/conf/mtune.d/cld
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-001
Date Issued: 09/15/1999 10:47:11
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. The crash(1M) ps command would display garbage appended to the
displayed name of a lwp.
Problem Resolution:
1. Corrected crash(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
1. The following object will also be replaced by cmds-001:
* /usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-001
Date Issued: 09/15/1999 12:57:05
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. The crash(1M) ps command would display garbage appended to the
displayed name of a lwp.
Problem Resolution:
1. Corrected crash(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
1. The following object will also be replaced by cmds-001:
* /usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-002
Date Issued: 02/02/2000 13:24:54
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: man-002
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. Several crash(1M) functions were enhanced.
Problem Resolution:
1. The crash(1M) functions `vnode' and 'page' were made more
informative. The crash man page had a formatting problem with tables
(they were coming out as paragraphs). The man page descriptions of
the `page', `pte', and `vtop' functions were adjusted for accuracy.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
The following object(s) will also be replaced by cmds-002:
* /usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-002
Date Issued: 02/02/2000 11:32:36
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: man-002
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. Several crash(1M) functions were enhanced.
Problem Resolution:
1. The crash(1M) functions `vnode' and 'page' were made more
informative. The crash man page had a formatting problem with tables
(they were coming out as paragraphs). The man page descriptions of
the `page', `pte', and `vtop' functions were adjusted for accuracy.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
The following object(s) will also be replaced by cmds-002:
* /usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-003
Date Issued: 10/05/2000 13:49:00
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, man-003, pg-003
Related SARs: 355
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. SAR 355: Large numbers of common block specifications triggered a
problem in shmdefine(1) that caused it to loop writing the same
output lines repeatedly.
Problem Resolution:
1. Use calloc(3C) rather than malloc(3C) so all fields in the internal
structure are initialized to zero.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/shmdefine
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-003
Date Issued: 10/05/2000 14:03:16
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, inet-004,
man-003, pg-003
Related SARs: 355
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. SAR 355: Large numbers of common block specifications triggered a
problem in shmdefine(1) that caused it to loop writing the same
output lines repeatedly.
Problem Resolution:
1. Use calloc(3C) rather than malloc(3C) so all fields in the internal
structure are initialized to zero.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/shmdefine
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: cmds-004
Date Issued: 03/29/2001 11:57:01
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
#############################################################################
Problem Description:
1. The crash "file" command was returning a seekmem error if the
number of used file descriptor (fd) entries was large in relation
to the total number of fd entries.
Problem Resolution:
1. The crash "file" command was returning a seekmem error if the
number of used file descriptor (fd) entries was large in relation
to the total number of fd entries. The problem was caused by an
incorrect calculation of the number of bytes to be read from the
file descriptor table. It was multiplying the total number of
entries times the number of entries used, instead of the number of
entries used times the size of each entry. This caused the file
command to read past the end of the fd table into potentially
unallocated memory.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
The following objects are linked to some of the objects being
replaced by this patch. These objects will also be re-linked:
/etc/crash
/usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cmds-004
Date Issued: 03/28/2001 19:14:12
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. The crash "file" command was returning a seekmem error if the
number of used file descriptor (fd) entries was large in relation
to the total number of fd entries.
Problem Resolution:
1. The crash "file" command was returning a seekmem error if the
number of used file descriptor (fd) entries was large in relation
to the total number of fd entries. The problem was caused by an
incorrect calculation of the number of bytes to be read from the
file descriptor table. It was multiplying the total number of
entries times the number of entries used, instead of the number of
entries used times the size of each entry. This caused the file
command to read past the end of the fd table into potentially
unallocated memory.
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
The following objects are linked to some of the objects being
replaced by this patch. These objects will also be re-linked:
/etc/crash
/usr/sbin/mview
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: cmds-005
Date Issued: 08/22/2001 11:39:14
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #551
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
#############################################################################
Problem Description:
1. SAR #551:
On the monacct listing (specifically on the "TOTAL COMMAND SUMMARY")
it looks like the totals REAL-MINS and (the next column) SIZE-K are
blending together.
2. Symptom: sometimes a large application distributed among a large
number of local memory pools would self-deadlock when an attempt
was made to lock the application down in memory.
Problem: The way PTEs were allocated was defective, resulting in
all of an applications PTEs being allocated in at most one eighth
of the available PTE buckets. When a bucket was full of locked
down PTEs, and Yet Another had to go into that particular bucket,
the application would wait for a free PTE to become available in
that bucket. That wait would never complete if the application
itself was the owner of all the PTEs already in the bucket.
Problem Resolution:
1. Reformatted the output in the acctcms program to allow more space
between the REAL-MINS and SIZE-K columns.
2. Replaced the address space VSID allocator with a more robust
algorithm that evenly scatters PTE allocations among all the
available PTE buckets.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/acct/acctcms
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: cmds-005
Date Issued: 09/11/2001 16:07:01
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: #551
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
#############################################################################
Problem Description:
1. SAR #551:
On the monacct listing (specifically on the "TOTAL COMMAND SUMMARY")
it looks like the totals REAL-MINS and (the next column) SIZE-K are
blending together.
2. Symptom: sometimes a large application distributed among a large
number of local memory pools would self-deadlock when an attempt
was made to lock the application down in memory.
Problem: The way PTEs were allocated was defective, resulting in
all of an applications PTEs being allocated in at most one eighth
of the available PTE buckets. When a bucket was full of locked
down PTEs, and Yet Another had to go into that particular bucket,
the application would wait for a free PTE to become available in
that bucket. That wait would never complete if the application
itself was the owner of all the PTEs already in the bucket.
Problem Resolution:
1. Reformatted the output in the acctcms program to allow more space
between the REAL-MINS and SIZE-K columns.
2. Replaced the address space VSID allocator with a more robust
algorithm that evenly scatters PTE allocations among all the
available PTE buckets.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/acct/acctcms
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: cmds-006
Date Issued: 01/02/2002 14:22:22
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #580, #569
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. The display of the LOCKS and NUMA columns of the `seg -f'
command of crash(1m) is garbage.
2. SAR $580: A problem with ksh was identified wherein if the number
of columns specified for the terminal settings was greater than
168 and the ksh editor was placed in emacs mode, then the ksh
would get a core dump upon entering a character. If a user were
to establish their COLUMNS environment variable prior to setting
the emacs edit mode; however, then the shell would work fine.
3. SAR #569: The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
Problem Resolution:
1. Problem was due to a pair of uninitialized variables in crash(1m).
2. The manner in which window sizing was accomplished with ksh(1)
was made to be consistent with newer versions of ksh.
3. Problem was due to the count of pages attached to an IO
operation not being correct.
Enhancements:
None
Object(s) To Be Replaced:
/usr/bin/ksh
/usr/bin/rksh
/usr/sbin/crash
Special Conditions for Installation:
1. L/I 3: The size of the b_numpages field of the internal
kernel-only data structure, buf_t, increased from a byte
to a short. This change was done in a way that the size
of the buf_t structure did not change nor did any offset
of that structure, therefore, only users of b_numpages
itself are effected. All kernel Driver.o modules that
use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays
the contents of the field b_numpages.
Possible Side Effects:
1. L/I 3:crash(1m) from earlier patch-releases of PowerMAX OS
will no longer correctly display the b_numpages field.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: cmds-006
Date Issued: 01/02/2002 14:22:22
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #580, #569
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. The display of the LOCKS and NUMA columns of the `seg -f'
command of crash(1m) is garbage.
2. SAR $580: A problem with ksh was identified wherein if the number
of columns specified for the terminal settings was greater than
168 and the ksh editor was placed in emacs mode, then the ksh
would get a core dump upon entering a character. If a user were
to establish their COLUMNS environment variable prior to setting
the emacs edit mode; however, then the shell would work fine.
3. SAR #569: The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
Problem Resolution:
1. Problem was due to a pair of uninitialized variables in crash(1m).
2. The manner in which window sizing was accomplished with ksh(1)
was made to be consistent with newer versions of ksh.
3. Problem was due to the count of pages attached to an IO
operation not being correct.
Enhancements:
None
Object(s) To Be Replaced:
/usr/bin/ksh
/usr/bin/rksh
/usr/sbin/crash
Special Conditions for Installation:
1. L/I 3: The size of the b_numpages field of the internal
kernel-only data structure, buf_t, increased from a byte
to a short. This change was done in a way that the size
of the buf_t structure did not change nor did any offset
of that structure, therefore, only users of b_numpages
itself are effected. All kernel Driver.o modules that
use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays
the contents of the field b_numpages.
Possible Side Effects:
1. L/I 3:crash(1m) from earlier patch-releases of PowerMAX OS
will no longer correctly display the b_numpages field.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: cmds-007
Date Issued: 03/18/2002 11:16:31
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. Signal info in crash program skewed by 1. Detailed description:
"proc -f -f ${slot}" shows a table of signal information,
including the handlers installed for each signal. The handler
information is skewed by 1.
Problem Resolution:
1. Resolution:Crash "proc" command was modified to display signal
information correctly.
Enhancements:
1. New `ui' function to crash(1m). This displays the contents of
the kernel's user-level interrupt driver table.
2. Any crash function which displayed an address symbolically will
now display a symbolic representation for 1) a u-pointer,
2) a lwp-pointer, 3) a proc-pointer. Previously, only addresses
which were near some kernel symbol would be displayed symbolically.
3. The crash(1m) function `calc' has a new -s option:
display calculated value as a symbolic address.
4. Several (improper) uses of printf were converted to
fprintf. This error resulted in some lines of output
going to the terminal when the output of the crash
function was redirected to some log file (-w function
option).
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: cmds-007
Date Issued: 03/18/2002 11:27:18
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
Problem Description:
1. Signal info in crash program skewed by 1. Detailed description:
"proc -f -f ${slot}" shows a table of signal information,
including the handlers installed for each signal. The handler
information is skewed by 1.
Problem Resolution:
1. Resolution:Crash "proc" command was modified to display signal
information correctly.
Enhancements:
1. New `ui' function to crash(1m). This displays the contents of
the kernel's user-level interrupt driver table.
2. Any crash function which displayed an address symbolically will
now display a symbolic representation for 1) a u-pointer,
2) a lwp-pointer, 3) a proc-pointer. Previously, only addresses
which were near some kernel symbol would be displayed symbolically.
3. The crash(1m) function `calc' has a new -s option:
display calculated value as a symbolic address.
4. Several (improper) uses of printf were converted to
fprintf. This error resulted in some lines of output
going to the terminal when the output of the crash
function was redirected to some log file (-w function
option).
Object(s) To Be Replaced:
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: cmds-008
Date Issued: 07/10/2002 14:50:05
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #698, #544, #744
Brief Description:
PowerMAX OS 4.3 cmds package release updates
##############################################################################
P1: SAR 698: kernel sometimes panics with the message
'hat_allocpte: bad PTEG'
R1: Problem was due to a leftover debug statement from the days that
PowerMAX OS was originally developed.
P2: SAR 544: kernel sometimes panics with the message
'address fault from interrupt routine'
R2: Problem was due to the kernel's 'in interrupt driver?' test not being
coded up in an atomic manner.
P3: SAR 544: servicing_interrupt() not atomic on some NightHawks.
R3: Fixed servicing_interrupt() macro in nh/uts/util/ksynch_p.h.
P4: Occasionally stale TLBs in kernel address space.
R4: Now waits for the tlbie insn to complete on all cpus before a new tlbie
is issued.
P5: SAR 744: Idle time calculation not high precision when HIGHRESTIMING
tuneable was enabled
R5: Alternate e_idlepct formula used when HIGHRESTIMING tuneable is enabled.
P6: PTE directory (PDN) pool too small for today's large memory systems
R6: Increased MinFreePdns from 16 to 64.
P7: hat_freepdn should be a front for kmem_free when KMACORRUPT is enabled
R7: hat_freepdn passes buffer to kmem_free when KMACORRUPT is enabled.
P8: hat_freepdn not checking alignment of the to-be-freed pdn buffer.
R8: hat_freepdn will now panic the kernel if given a misaligned buffer.
P9: kmem_free not checking freed buffers for proper alignment.
R9: kmem_free will now panic the kernel if given a misaligned buffer.
P10: Debug kmem_alloc not checking if buffer is aligned.
R10: kmem_alloc now will panic if to-be-returned buffer is misaligned.
P11: Debug kmem_alloc not filling buffer with a known 'poison' value.
R11: kmem_alloc now fills the first 128 bytes of a buffer with 0xdeadbeaf.
P12: Debug kmem_alloc not checking if secondary and primary virtual
addresses are equivalent.
R12: kmem_alloc now test-writes through the primary virtual address and
checks if the value is returnable through the secondary virtual address.
This will catch any TLB flush failures to either address.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/csh
/usr/sbin/crash
/usr/sbin/dfsck
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-008
Date Issued: 07/24/2002 14:42:21
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #698, #544, #744
Brief Description:
PowerMAX OS 4.3 cmds package release updates
###############################################################################
P1: SAR 698: kernel sometimes panics with the message
'hat_allocpte: bad PTEG'
R1: Problem was due to a leftover debug statement from the days that
PowerMAX OS was originally developed.
P2: SAR 544: kernel sometimes panics with the message
'address fault from interrupt routine'
R2: Problem was due to the kernel's 'in interrupt driver?' test not being
coded up in an atomic manner.
P3: SAR 544: servicing_interrupt() not atomic on some NightHawks
R3: Fixed servicing_interrupt() macro in nh/uts/util/ksynch_p.h
P4: Occasionally stale TLBs in kernel address space.
R4: Now waits for the tlbie insn to complete on all cpus before a new tlbie
is issued.
P5: SAR 744: Idle time calculation not high precision when HIGHRESTIMING
tuneable was enabled
R5: Alternate e_idlepct formula used when HIGHRESTIMING tuneable is enabled.
P6: PTE directory (PDN) pool too small for today's large memory systems
R6: Increased MinFreePdns from 16 to 64.
P7: hat_freepdn should be a front for kmem_free when KMACORRUPT is enabled
R7: hat_freepdn passes buffer to kmem_free when KMACORRUPT is enabled.
P8: hat_freepdn not checking alignment of the to-be-freed pdn buffer
R8: hat_freepdn will now panic the kernel if given a misaligned buffer.
P9: kmem_free not checking freed buffers for proper alignment.
R9: kmem_free will now panic the kernel if given a misaligned buffer.
P10: Debug kmem_alloc not checking if buffer is aligned.
R10: kmem_alloc now will panic if to-be-returned buffer is misaligned.
P11: Debug kmem_alloc not filling buffer with a known 'poison' value.
R11: kmem_alloc now fills the first 128 bytes of a buffer with 0xdeadbeaf.
P12: Debug kmem_alloc not checking if secondary and primary virtual
addresses are equivalent.
R12: kmem_alloc now test-writes through the primary virtual address and
checks if the value is returnable through the secondary virtual address.
This will catch any TLB flush failures to either address.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/csh
/usr/sbin/crash
/usr/sbin/dfsck
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-009
Date Issued: 12/11/2002 14:32:37
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: shmdefine(1) had a couple options that were meaningless or
confusing even on PowerMAX OS.
R1: The shmdefine(1) command line was modified to use more mnemonic
options, while still supporting the previous option names.
The -3 option, a reference to CX/UX version 3.0, was removed.
shmdefine(1) was made more robust when parsing incomplete or
incorrect definition files. In addition, a bug that caused
shmdefine(1) to dereference a NULL pointer was corrected.
P2: crash(1m) `trace -e' command sometimes aborts early without
generating a full display.
R2: `trace -e' now displays all process stacks.
P3: crash(1m) `memchunk' command displays addresses within crash(1m)
itself rather than within the kernel.
R3: 'memchunk' now displays kernel addresses.
Enhancements:
1: Two new commands were added to crash: kmaleak and strleak.
These commands are useful in finding kernel memory leaks
(locations that are allocating memory that is never freed).
To find potential memory leaks, periodic samples of the
outstanding allocations must be taken. Any location with an
increasing allocation count could potentially be the source
of a suspected memory leak.
The kernel must be configured with special debug versions of
the kma and/or streams drivers to use these commands. In
PowerMAX OS 5.0 or later, these debug drivers
(kmaleak and strleak) are provided with the system and can
be optionally configured in place of the standard drivers
(kma and stream). For earlier PowerMAX OS, specially built
debug versions of the kma and io drivers must be obtained
through Customer Support.
Object(s) To Be Replaced:
/usr/bin/shmdefine
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-009
Date Issued: 12/18/2002 17:43:33
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cmds package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: shmdefine(1) had a couple options that were meaningless or
confusing even on PowerMAX OS.
R1: The shmdefine(1) command line was modified to use more mnemonic
options, while still supporting the previous option names.
The -3 option, a reference to CX/UX version 3.0, was removed.
shmdefine(1) was made more robust when parsing incomplete or
incorrect definition files. In addition, a bug that caused
shmdefine(1) to dereference a NULL pointer was corrected.
P2: crash(1m) `trace -e' command sometimes aborts early without
generating a full display.
R2: `trace -e' now displays all process stacks.
P3: crash(1m) `memchunk' command displays addresses within crash(1m)
itself rather than within the kernel.
R3: 'memchunk' now displays kernel addresses.
Enhancements:
1: Two new commands were added to crash: kmaleak and strleak.
These commands are useful in finding kernel memory leaks
(locations that are allocating memory that is never freed).
To find potential memory leaks, periodic samples of the
outstanding allocations must be taken. Any location with an
increasing allocation count could potentially be the source
of a suspected memory leak.
The kernel must be configured with special debug versions of
the kma and/or streams drivers to use these commands. In
PowerMAX OS 5.0 or later, these debug drivers
(kmaleak and strleak) are provided with the system and can
be optionally configured in place of the standard drivers
(kma and stream). For earlier PowerMAX OS, specially built
debug versions of the kma and io drivers must be obtained
through Customer Support.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/shmdefine
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-010
Date Issued: 05/08/2003 10:36:22
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 cmds package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1418: "csh" generates a permission error when launched from
a process that has SETUID set.
R1. "csh" used to just exit when the real and effective user ids
were different or the real and effective group ids were
different. The case of invoking "csh" from a SETUID executable
file was not being handled. In the new version of csh, if the
uid!=euid or gid!egid, the appropriate steps are taken
depending on the values of the effective user and group ids
and the rest is handled by the OS.
P2. "csh" failed to execute a SETUID script.
R2. The reason for that failure was the fact that "csh" was being
passed the script's name as one of its arguments but in the
case of a SETUID script the file name "/dev/fd/NN" was being
passed and a descriptor "NN" inherited. "NN" corresponded to
the already opened script file. "csh" did not take into
consideration that case and simply closed all "unnecessary"
file descriptors before making use of fd/"NN".
P3. crash `stacktrace' output gets mangled while being passed
through microsoft mail servers.
R3. Reduced number of columns used in the stacktrace display from
79 to approximately 70 columns.
P4. SAR 1425: The whodo(1M) command gives incorrect device
information when invoked as a non-root user. This is due to
incorrect permissions on the executable.
R4. The whodo(1M) command permissions have been changed to give
it setuid access. It will now display the correct information
when invoked as a non-root user.
P5. SAR 1445: The edquota(1M) command cannot find a UFS file system
mounted on /dev/root. /dev/var, or /dev/usr. This is due to a
mismatch in the device id of the mount special and the device
id associated with the /quotas file. Any other UFS mounted
file system and the command would work as it should.
R5. In the setupfs() routine in edquota(1M) the extra condition of
matching the mount special device id (st_rdev) and the device
id associated the /quotas file (st_dev) has been removed.
This now enables edquota(1M) to work properly if a UFS file
system is mounted on /dev/root, /dev/var, or /dev/usr.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/csh
/usr/lib/fs/ufs/edquota
/usr/sbin/crash
/usr/sbin/whodo
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-010
Date Issued: 05/08/2003 10:36:40
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 cmds package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1418: "csh" generates a permission error when launched from
a process that has SETUID set.
R1. "csh" used to just exit when the real and effective user ids
were different or the real and effective group ids were
different. The case of invoking "csh" from a SETUID executable
file was not being handled. In the new version of csh, if the
uid!=euid or gid!egid, the appropriate steps are taken
depending on the values of the effective user and group ids
and the rest is handled by the OS.
P2. "csh" failed to execute a SETUID script.
R2. The reason for that failure was the fact that "csh" was being
passed the script's name as one of its arguments but in the
case of a SETUID script the file name "/dev/fd/NN" was being
passed and a descriptor "NN" inherited. "NN" corresponded to
the already opened script file. "csh" did not take into
consideration that case and simply closed all "unnecessary"
file descriptors before making use of fd/"NN".
P3. crash `stacktrace' output gets mangled while being passed
through microsoft mail servers.
R3. Reduced number of columns used in the stacktrace display from
79 to approximately 70 columns.
P4. SAR 1425: The whodo(1M) command gives incorrect device
information when invoked as a non-root user. This is due to
incorrect permissions on the executable.
R4. The whodo(1M) command permissions have been changed to give
it setuid access. It will now display the correct information
when invoked as a non-root user.
P5. SAR 1445: The edquota(1M) command cannot find a UFS file system
mounted on /dev/root. /dev/var, or /dev/usr. This is due to a
mismatch in the device id of the mount special and the device
id associated with the /quotas file. Any other UFS mounted
file system and the command would work as it should.
R5. In the setupfs() routine in edquota(1M) the extra condition of
matching the mount special device id (st_rdev) and the device
id associated the /quotas file (st_dev) has been removed.
This now enables edquota(1M) to work properly if a UFS file
system is mounted on /dev/root, /dev/var, or /dev/usr.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/csh
/usr/lib/fs/ufs/edquota
/usr/sbin/crash
/usr/sbin/whodo
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-011
Date Issued: 01/07/2004 15:59:19
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 cmds package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P8: SAR 1535: The savecore(1M) utility cannot save a dump on a system
that has 2GB or more of physical memory or has a dumpsize 2GB or
greater. The crash(1M) utility cannot read a crashfile that is 2GB
or greater.
R8: Savecore(1M) was written such that it had a limit of 2GB - 1.
To resolve this involved changing many data types from off_t
or int into size_t. An additional change involved setting
RLIMIT_FSIZE to (2 * RLIM_INFINITY). This will now allow
savecore(1M) to save a dump on a system that has nearly 4GB of
physical memory. The crash(1M utility was enhanced to read
crashfiles that are 2GB or greater.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crash/libcrash1.so
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cmds-011
Date Issued: 01/07/2004 16:00:08
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 cmds package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P8: SAR 1535: The savecore(1M) utility cannot save a dump on a system
that has 2GB or more of physical memory or has a dumpsize 2GB or
greater. The crash(1M) utility cannot read a crashfile that is 2GB
or greater.
R8: Savecore(1M) was written such that it had a limit of 2GB - 1.
To resolve this involved changing many data types from off_t
or int into size_t. An additional change involved setting
RLIMIT_FSIZE to (2 * RLIM_INFINITY). This will now allow
savecore(1M) to save a dump on a system that has nearly 4GB of
physical memory. The crash(1M utility was enhanced to read
crashfiles that are 2GB or greater.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crash/libcrash1.so
/usr/sbin/crash
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: cmds-012
Date Issued: 03/13/2006 13:17:44
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 cmds package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3202: shmdefine generates hex constants for datapool variable
sizes in the generated linker MAP file. ld cannot recognize
hex constants.
R1: Emit decimal sizes for individual datapool variables.
P2: DT3752: cksum results on PowerMAX OS differ from results produced
by cksum on Sun OS and HPUX.
R2: The CRC lookup table was replaced to match the CRC calculating
algorithm used in posix2 implementation.
P3: DT2907: The fwtmp(1M) utility should be capable of making binary
files from ascii files created by fwtmp(1M) which are essentially
duplicates of the original /var/adm/wtmp accounting file.
However, fwtmp(1M) is currently unable to do this correctly.
R3: An improper line size was used for array sizing in the local
acctdef.h header for building fwtmp(1M), acctcon(1M), acctcon1(1M),
and acctwtmp(1M). It was set to the original 56 bytes instead of
being set to the desired 12 bytes for PowerMAX OS. This caused an
improper read of an ascii input file when building an intermediate
binary file. This the intermediate binary file was corrupted when
compared to the original /var/adm/wtmp binary file.
P4: ftok() calls generated by shmdefine use an id=0. This is not
acceptable on Linux, so change the key to 0x1 for all targets.
R4: Changed ftok() call to use 0x1.
P5: DT2311: Fortran Datapools are not emitted properly in the
generated linker script. No symbols are actually relocated.
R5: Implemented a new clause: Fortan datapool "poolname", "objfile.o"
The Fortran Datapool clause reads the data dictionary from the
specified object file and auto-generates sharing directives in
the Link script for each datapool member.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/cksum
/usr/bin/shmdefine
/usr/lib/acct/acctcon
/usr/lib/acct/acctcon1
/usr/lib/acct/acctwtmp
/usr/lib/acct/fwtmp
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: cmds-012
Date Issued: 03/10/2006 16:36:55
Software Package: cmds pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 cmds package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3202: shmdefine generates hex constants for datapool variable
sizes in the generated linker MAP file. ld cannot recognize
hex constants.
R1: Emit decimal sizes for individual datapool variables.
P2: DT3752: cksum results on PowerMAX OS differ from results produced
by cksum on Sun OS and HPUX.
R2: The CRC lookup table was replaced to match the CRC calculating
algorithm used in posix2 implementation.
P3: DT2907: The fwtmp(1M) utility should be capable of making binary
files from ascii files created by fwtmp(1M) which are essentially
duplicates of the original /var/adm/wtmp accounting file.
However, fwtmp(1M) is currently unable to do this correctly.
R3: An improper line size was used for array sizing in the local
acctdef.h header for building fwtmp(1M), acctcon(1M), acctcon1(1M),
and acctwtmp(1M). It was set to the original 56 bytes instead of
being set to the desired 12 bytes for PowerMAX OS. This caused an
improper read of an ascii input file when building an intermediate
binary file. This the intermediate binary file was corrupted when
compared to the original /var/adm/wtmp binary file.
P4: ftok() calls generated by shmdefine use an id=0. This is not
acceptable on Linux, so change the key to 0x1 for all targets.
R4: Changed ftok() call to use 0x1.
P5: DT2311: Fortran Datapools are not emitted properly in the
generated linker script. No symbols are actually relocated.
R5: Implemented a new clause: Fortan datapool "poolname", "objfile.o"
The Fortran Datapool clause reads the data dictionary from the
specified object file and auto-generates sharing directives in
the Link script for each datapool member.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/cksum
/usr/bin/shmdefine
/usr/lib/acct/acctcon
/usr/lib/acct/acctcon1
/usr/lib/acct/acctwtmp
/usr/lib/acct/fwtmp
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-001
Date Issued: 09/15/1999 10:47:47
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, Power Hawk 640
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. When promiscuous mode is selected on the Condor board, the only
packets selected by the board have either the Condor board address or
are broadcast packets. In promiscuous mode all ethernet traffic
should be selected regardless of the destination address.
Problem Resolution:
1. The driver has been changed to set further control bits while setting
the board to promiscuous mode.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-001
Date Issued: 09/15/1999 12:57:32
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. When promiscuous mode is selected on the Condor board, the only
packets selected by the board have either the Condor board address or
are broadcast packets. In promiscuous mode all ethernet traffic
should be selected regardless of the destination address.
Problem Resolution:
1. The driver has been changed to set further control bits while setting
the board to promiscuous mode.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-002
Date Issued: 02/02/2000 13:25:59
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-002, dec-001, egl-001, ip-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. config(1M) does not display cnd (Condor) and pg (Peregrine)
descriptions correctly.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. The %%Extended%% fields were incorrectly labeled as %%Keyword%%
fields in the cnd and pg Mtune(4) files.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/cnd
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-002
Date Issued: 02/02/2000 11:35:05
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, dec-001, egl-001, ie-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. config(1M) does not display cnd (Condor) and pg (Peregrine)
descriptions correctly.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. The %%Extended%% fields were incorrectly labeled as %%Keyword%%
fields in the cnd and pg Mtune(4) files.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/cnd
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-003
Date Issued: 05/30/2000 13:02:02
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, Power Hawk 640
Related Patches: dec-002, egl-002, ip-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-003
Date Issued: 05/30/2000 13:45:10
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: dec-002, egl-002, ie-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-004
Date Issued: 10/05/2000 13:49:21
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-004, cmds-003, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, man-003, pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the cnd driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
3. SAR 39: The cnd board would occasionally transmit data from the
wrong buffer or place data into the wrong receive buffer.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The cnd driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
2. The flow control limit of the number of outstanding packets that can
be queued to each queue was set too small. Packets would be
discarded when this queue hiwater limit was reached. This could
occur on either the transmit or receive side. Changed the default
value for the cnd's queue flow control hiwater level from 40 to 8192
packets. Added the capability to change the queue's hiwater level
for this driver's in its space.c file.
3. The cnd board was using stale transmit/receive buffer pointers. The
cnd board control structures and transmit/receive buffer pointer
areas were changed to be allocated in cache inhibited memory.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
/usr/include/sys/dlpi_cnd.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: cnd-004
Date Issued: 10/05/2000 14:03:37
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, inet-004,
man-003, pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the cnd driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
3. SAR 39: The cnd board would occasionally transmit data from the
wrong buffer or place data into the wrong receive buffer.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The cnd driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
2. The flow control limit of the number of outstanding packets that can
be queued to each queue was set too small. Packets would be
discarded when this queue hiwater limit was reached. This could
occur on either the transmit or receive side. Changed the default
value for the cnd's queue flow control hiwater level from 40 to 8192
packets. Added the capability to change the queue's hiwater level
for this driver's in its space.c file.
3. The cnd board was using stale transmit/receive buffer pointers. The
cnd board control structures and transmit/receive buffer pointer
areas were changed to be allocated in cache inhibited memory.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
/usr/include/sys/dlpi_cnd.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: cnd-005
Date Issued: 03/18/2002 11:16:48
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: cnd-005
Date Issued: 03/18/2002 11:27:43
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 cnd package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-006
Date Issued: 12/11/2002 14:32:56
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #782
Brief Description:
PowerMAX OS 4.3 cnd package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #782: Could not open or use Condor (cnd) port 1 of a two
port board once the cnd-005 patch was installed. Port 0 works
properly.
R1: The DLconfig structure for cnd (cndconfig) did not contain the
address pointers for any ports other than port 0 of each cnd
board. The DLPI common code for DLopen had been changed in
PowerMAX OS 4.3 patchset 10 to use this structure to find the
pointers to the structures associated with the networking
device to be opened and was therefore unable to open any port
other than port 0. The cnd driver and space.c files were
changed to add pointers for all ports on a cnd board to the
DLconfig structure
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-006
Date Issued: 12/18/2002 17:43:45
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #782
Brief Description:
PowerMAX OS 4.3 cnd package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #782: Could not open or use Condor (cnd) port 1 of a two
port board once the cnd-005 patch was installed. Port 0 works
properly.
R1: The DLconfig structure for cnd (cndconfig) did not contain the
address pointers for any ports other than port 0 of each cnd
board. The DLPI common code for DLopen had been changed in
PowerMAX OS 4.3 patchset 10 to use this structure to find the
pointers to the structures associated with the networking
device to be opened and was therefore unable to open any port
other than port 0. The cnd driver and space.c files were
changed to add pointers for all ports on a cnd board to the
DLconfig structure
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
/etc/conf/pack.d/cnd/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-007
Date Issued: 01/07/2004 15:59:39
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 cnd package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-007
Date Issued: 01/07/2004 16:00:31
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 cnd package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-008
Date Issued: 07/08/2004 10:52:31
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 cnd package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1887: If you attemptted to reboot the system without first
powering down, the EGL interface wasn't functional after the
reboot.
R1: Neither of the egl or cnd drivers had any code to reset the
board during a system shutdown or a DLM unload. Added a
shutdown/halt routine to the egl and cnd drivers to perform
the reset. This stops the egl and cnd boards from sending
receive packets to the cpu before their driver is able to
initialize the software tables that are needed by the board.
P2: The ifOutQLen mib2 counter would occasionally get out of sync
with the actual number of queued output packets in times of
heavy output traffic in the cnd and egl drivers.
R2: Fixed the ifOutQLen counter logic in the queued packet
processing routines in both of the cnd and egl drivers.
This was occasionally causing a delay in the transmission
of queued output packets.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/cnd
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: cnd-008
Date Issued: 07/26/2004 11:50:20
Software Package: cnd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 cnd package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1887: If you attemptted to reboot the system without first
powering down, the EGL interface wasn't functional after the
reboot.
R1: Neither of the egl or cnd drivers had any code to reset the
board during a system shutdown or a DLM unload. Added a
shutdown/halt routine to the egl and cnd drivers to perform
the reset. This stops the egl and cnd boards from sending
receive packets to the cpu before their driver is able to
initialize the software tables that are needed by the board.
P2: The ifOutQLen mib2 counter would occasionally get out of sync
with the actual number of queued output packets in times of
heavy output traffic in the cnd and egl drivers.
R2: Fixed the ifOutQLen counter logic in the queued packet
processing routines in both of the cnd and egl drivers.
This was occasionally causing a delay in the transmission
of queued output packets.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/cnd
/etc/conf/pack.d/cnd/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-001
Date Issued: 02/02/2000 13:26:36
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, diskless-002, fbs-002, man-002">fbsman-002, inet-002,
librt-001
Related SARs: #47, #98
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. #98: The system(3S) I/O function should use fork1(2) in
multi-threaded applications instead of fork(2).
2. Library changes required for RCIM Coupled FBS support.
3. #47: Updates to the C library required for new VME support for
TurboHawks.
Problem Resolution:
1. The system(3S) I/O function has been modified to use fork1(2) instead
of fork(2) for all applications.
2. Real-time libraries (librt, libF77rt) updated to include RCIM Coupled
FBS support.
3. C library updated for new TurboHawk VME support.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libF77rt.a
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
/usr/lib/crosslibs/librt.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-001
Date Issued: 02/02/2000 11:35:43
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, fbs-002, man-002">fbsman-002, inet-002, librt-001
Related SARs: #47, #98
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. #98: The system(3S) I/O function should use fork1(2) in
multi-threaded applications instead of fork(2).
2. Library changes required for RCIM Coupled FBS support.
3. #47: Updates to the C library required for new VME support for
TurboHawks.
Problem Resolution:
1. The system(3S) I/O function has been modified to use fork1(2) instead
of fork(2) for all applications.
2. Real-time libraries (librt, libF77rt) updated to include RCIM Coupled
FBS support.
3. C library updated for new TurboHawk VME support.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libF77rt.a
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
/usr/lib/crosslibs/librt.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-002
Date Issued: 05/30/2000 13:02:49
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: base-003
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. nanosleep(3C) was failing in certain configurations.
Problem Resolution:
1. Corrected the nanosleep() failure.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-002
Date Issued: 05/30/2000 13:47:25
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: base-003
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. nanosleep(3C) was failing when fast interval timers were installed on
TurboHawk systems.
Problem Resolution:
1. nanosleep() now correctly reference fast interval timers.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-003
Date Issued: 10/05/2000 13:50:00
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crypt-002, crypt-int-002,
dec-003, diskless-004, egl-003, fibre-003, inet-004,
ip-003, man-003, pg-003
Related SARs: 322, 327
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. SAR 322: All threads are duplicated when popen(3c) is called. When
a thread within a multi-threaded process calls popen(3C), all of the
threads are recreated in the new child process with a fork(2) call.
This can cause various potential problems for all of the other
threads in the parent process, depending upon what they were doing
at the time when the one thread called popen(3c). The fork(2) call
inside popen(3c) causes the parent process's lwps to all be pulled
into the kernel and put into a quiet state (rendezvoused) before the
fork operation proceeds. This activity disrupts the parent lwp's
operations, and also creates child process lwps that attempt to
execute the same code that the parent lwps where executing at the
time of the fork.
2. SAR 327: The PTIMER_MAX tunable was not working properly with
programs that linked with libthread.
3. strncpy(3C) overruns zero fill for short lengths. strncpy(s1,s2,n)
for small values of n will fill an extra position in the s1 output
buffer with a 0.
Problem Resolution:
1. The the popen(3C) routine was modified to now use fork1(2) instead
of fork(2) when creating the new child process. The fork1(2) call
will create a child process with only the one calling lwp duplicated
in the new child process.
2. The _thr_posix_timer_init() routine in libthread was changed to pick
up _SC_POSIX_TIMER_MAX instead of _SC_TIMER_MAX. The sysconf()
routine in libc was changed to get the value of
_sysconfig(_CONFIG_POSIX_TIMER_MAX) instead of _POSIX_TIMER_MAX.
A new define was added to <sys/sysconfig.h> for
_CONFIG_POSIX_TIMER_MAX. A new case for _CONFIG_POSIX_TIMER_MAX was
added to get the current value of PTIMER_MAX from the kernel.
3. Corrected strncpy routine in C library.
Enhancements:
1. Enhance the ability for some blocking routines to use CLOCK_REALTIME.
CLOCK_REALTIME or CLOCK_UNIX can now be specified to use during the
timeout periods for the following routines:
aio_suspend(3)
aio_suspend64(3)
pthread_cond_timedwait(3pthread)
select(2)
Two new C Library routines and associated manual pages were added:
set_timeout_resolution(3C)
get_timeout_resolution(3C)
which allow setting and retrieving of the specific clock resolution
(CLOCK_UNIX or CLOCK_REALTIME) for any or all of the routines
identified above.
2. A new command (TMOUT_RESMSK) was added to syscx(2) for the two new
C library routines.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.so.1
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-003
Date Issued: 10/05/2000 14:04:10
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crypt-002, crypt-int-002,
dec-003, egl-003, ie-003, inet-004, man-003, pg-003
Related SARs: 322, 327
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. SAR 322: All threads are duplicated when popen(3c) is called. When
a thread within a multi-threaded process calls popen(3C), all of the
threads are recreated in the new child process with a fork(2) call.
This can cause various potential problems for all of the other
threads in the parent process, depending upon what they were doing
at the time when the one thread called popen(3c). The fork(2) call
inside popen(3c) causes the parent process's lwps to all be pulled
into the kernel and put into a quiet state (rendezvoused) before the
fork operation proceeds. This activity disrupts the parent lwp's
operations, and also creates child process lwps that attempt to
execute the same code that the parent lwps where executing at the
time of the fork.
2. SAR 327: The PTIMER_MAX tunable was not working properly with
programs that linked with libthread.
3. strncpy(3C) overruns zero fill for short lengths. strncpy(s1,s2,n)
for small values of n will fill an extra position in the s1 output
buffer with a 0.
Problem Resolution:
1. The the popen(3C) routine was modified to now use fork1(2) instead
of fork(2) when creating the new child process. The fork1(2) call
will create a child process with only the one calling lwp duplicated
in the new child process.
2. The _thr_posix_timer_init() routine in libthread was changed to pick
up _SC_POSIX_TIMER_MAX instead of _SC_TIMER_MAX. The sysconf()
routine in libc was changed to get the value of
_sysconfig(_CONFIG_POSIX_TIMER_MAX) instead of _POSIX_TIMER_MAX.
A new define was added to <sys/sysconfig.h> for
_CONFIG_POSIX_TIMER_MAX. A new case for _CONFIG_POSIX_TIMER_MAX was
added to get the current value of PTIMER_MAX from the kernel.
3. Corrected strncpy routine in C library.
Enhancements:
1. Enhance the ability for some blocking routines to use CLOCK_REALTIME.
CLOCK_REALTIME or CLOCK_UNIX can now be specified to use during the
timeout periods for the following routines:
aio_suspend(3)
aio_suspend64(3)
pthread_cond_timedwait(3pthread)
select(2)
Two new C Library routines and associated manual pages were added:
set_timeout_resolution(3C)
get_timeout_resolution(3C)
which allow setting and retrieving of the specific clock resolution
(CLOCK_UNIX or CLOCK_REALTIME) for any or all of the routines
identified above.
2. A new command (TMOUT_RESMSK) was added to syscx(2) for the two new
C library routines.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.so.1
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crosslibs-004
Date Issued: 03/29/2001 11:58:45
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates.
#############################################################################
Problem Description:
1. There is no support for vme_address(3C) on moto systems.
Problem Resolution:
1. The vme_address(3C) routine in libc has been enhanced for Night
Hawk systems and support has been provided for Power Hawk systems.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crosslibs-004
Date Issued: 03/29/2001 10:08:05
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates.
##############################################################################
Problem Description:
1. There is no support for vme_address(3C) on moto systems.
Problem Resolution:
1. The vme_address(3C) routine in libc has been enhanced for Night
Hawk systems and support has been provided for Power Hawk systems.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crosslibs-005
Date Issued: 08/22/2001 11:40:59
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
#############################################################################
Problem Description:
1. On multi-CPU systems it is sometimes possible to read the wrong
POSIX time of day. This is due to the fact that a race
condition existed between one CPU reading the three POSIX timer
assist page variables at the same time that these same variables
were being updated on the other CPU.
The three POSIX timer assist page variables are used in the
computation of the current POSIX time of day.
When a mix of old and new POSIX timer assist variable values
are read, then the wrong POSIX time of day is computed/obtained,
and operations such as nanosleep(3C) will not function properly.
Problem Resolution:
1. The coding for reading and updating the POSIX timer assist page
variables was modified so that the reading of these variables
will always result in reading a coherent set of values, even
when they are being updated at roughly the same time on the
other CPU.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crosslibs-005
Date Issued: 09/11/2001 16:08:24
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
#############################################################################
Problem Description:
1. On multi-CPU systems it is sometimes possible to read the wrong
POSIX time of day. This is due to the fact that a race
condition existed between one CPU reading the three POSIX timer
assist page variables at the same time that these same variables
were being updated on the other CPU.
The three POSIX timer assist page variables are used in the
computation of the current POSIX time of day.
When a mix of old and new POSIX timer assist variable values
are read, then the wrong POSIX time of day is computed/obtained,
and operations such as nanosleep(3C) will not function properly.
Problem Resolution:
1. The coding for reading and updating the POSIX timer assist page
variables was modified so that the reading of these variables
will always result in reading a coherent set of values, even
when they are being updated at roughly the same time on the
other CPU.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: crosslibs-006
Date Issued: 01/07/2002 14:26:07
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: base-009
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
Problem Resolution:
1. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: crosslibs-006
Date Issued: 01/07/2002 14:26:07
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
Problem Resolution:
1. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
/usr/lib/crosslibs/libc.so.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crosslibs-007
Date Issued: 03/19/2002 12:23:05
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. If input stream is NULL a core dump will occure.
Problem Resolution:
1. If the input stream is NULL then set error condition and return
immediately, rather than core dump.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crosslibs-007
Date Issued: 04/22/2002 16:24:58
Software Package: crosslibs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crosslibs package release updates
##############################################################################
Problem Description:
1. If input stream is NULL a core dump will occure.
Problem Resolution:
1. If the input stream is NULL then set error condition and return
immediately, rather than core dump.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/crosslibs/libc.a
/usr/lib/crosslibs/libc.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-001
Date Issued: 02/02/2000 13:27:45
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crypt-int-001, inet-002
Related SARs: #59, #166
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. #59: The xdr_double() function contained in the "nsl" library will
return an Arithmetic Exception error and dump core.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
Problem Resolution:
1. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
The following object(s) will also be replaced by crypt-001:
* /usr/lib/libnsl.a
* /usr/lib/libsocket.a
* /usr/lib/libxti.a
* /usr/lib/libnsl.so
* /usr/lib/libxti.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-001
Date Issued: 02/02/2000 11:36:48
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crypt-int-001, inet-002
Related SARs: #59, #166
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. #59: The xdr_double() function contained in the "nsl" library will
return an Arithmetic Exception error and dump core.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
Problem Resolution:
1. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
The following object(s) will also be replaced by crypt-001:
* /usr/lib/libnsl.a
* /usr/lib/libsocket.a
* /usr/lib/libxti.a
* /usr/lib/libnsl.so
* /usr/lib/libxti.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-002
Date Issued: 10/05/2000 13:50:45
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, man-003, pg-003
Related SARs: 194, 257, 392, 397
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new message/datagram.
3. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
3. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
The following link(s) will be updated by patch crypt-002:
/usr/lib/libnsl.a --> /usr/lib/libnsl_d.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_d.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_d.a
/usr/lib/libxti.so --> /usr/lib/libnsl_d.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-002
Date Issued: 10/05/2000 14:05:14
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003,
crypt-int-002, dec-003, egl-003, ie-003, inet-004,
man-003, pg-003
Related SARs: 194, 257, 392, 397
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new message/datagram.
3. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
3. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
The following link(s) will be updated by patch crypt-002:
/usr/lib/libnsl.a --> /usr/lib/libnsl_d.a
/usr/lib/libnsl.so --> /usr/lib/libnsl_d.so
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_d.a
/usr/lib/libxti.so --> /usr/lib/libnsl_d.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-003
Date Issued: 11/20/2000 15:57:34
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crypt package release updates
#############################################################################
Problem Description:
1. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
Problem Resolution:
1. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
The following link(s) will also be updated by crypt-003:
/usr/lib/libnsl.a --> /usr/lib/libnsl_d.a
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_d.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-003
Date Issued: 11/20/2000 16:06:06
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crypt package release updates
#############################################################################
Problem Description:
1. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
Problem Resolution:
1. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
The following link(s) will also be updated by crypt-003:
/usr/lib/libnsl.a --> /usr/lib/libnsl_d.a
/usr/lib/libsocket.a --> /usr/lib/libnsl.a
/usr/lib/libxti.a --> /usr/lib/libnsl_d.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-004
Date Issued: 08/22/2001 11:44:36
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #495, #503
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crypt package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
2. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using alot of different
addresses (host, netid).
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
2. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-004
Date Issued: 09/11/2001 16:18:31
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crypt package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
2. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using many different
addresses (host, netid).
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
2. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crypt-005
Date Issued: 03/18/2002 11:17:06
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #535
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
Problem Resolution:
1. Implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crypt-005
Date Issued: 03/18/2002 11:28:07
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #535
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
Problem Description:
1. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
Problem Resolution:
1. Implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crypt-006
Date Issued: 07/10/2002 14:50:36
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #695
Brief Description:
PowerMAX OS 4.3 crypt package release updates
##############################################################################
P1: SAR #695: On occasion, when network activity is high and
/usr/lib/resolv.so is specified in /etc/netconfig and there are contents
in /etc/resolv.conf, the resolution of addresses can become very slow.
This is due to a poll() timeout value being too short in libresolv and
causing multiple retries for the resolved address.
R1: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-006
Date Issued: 07/24/2002 14:42:51
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #695
Brief Description:
PowerMAX OS 4.3 crypt package release updates
###############################################################################
P1: SAR #695: On occasion, when network activity is high and
/usr/lib/resolv.so is specified in /etc/netconfig and there are contents
in /etc/resolv.conf, the resolution of addresses can become very slow.
This is due to a poll() timeout value being too short in libresolv and
causing multiple retries for the resolved address.
R1: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
/usr/lib/libnsl_d.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-007
Date Issued: 12/11/2002 14:33:23
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #780
Brief Description:
PowerMAX OS 4.3 crypt package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-007
Date Issued: 12/18/2002 17:44:01
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #780
Brief Description:
PowerMAX OS 4.3 crypt package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-008
Date Issued: 07/08/2004 10:52:46
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 crypt package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R1: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-008
Date Issued: 07/26/2004 11:50:30
Software Package: crypt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 crypt package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R1: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_d.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-int-001
Date Issued: 02/02/2000 13:29:07
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crypt-001, inet-002
Related SARs: #59, #166
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. #59: The xdr_double() function contained in the "nsl" library will
return an Arithmetic Exception error and dump core.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
Problem Resolution:
1. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following object(s) will also be replaced by crypt-int-001:
* /usr/lib/libnsl.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-int-001
Date Issued: 02/02/2000 11:37:26
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crypt-001, inet-002
Related SARs: #59, #166
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. #59: The xdr_double() function contained in the "nsl" library will
return an Arithmetic Exception error and dump core.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
Problem Resolution:
1. The problem was in casting the return value from xdr_double() to a
long when it should be cast to a long long.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following object(s) will also be replaced by crypt-int-001:
* /usr/lib/libnsl.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-int-002
Date Issued: 10/05/2000 13:51:15
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
dec-003, diskless-004, egl-003, fibre-003, inet-004,
ip-003, man-003, pg-003
Related SARs: 194, 392, 397
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following link(s) will be updated by patch crypt-int-002:
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-int-002
Date Issued: 10/05/2000 14:05:40
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
dec-003, egl-003, ie-003, inet-004, man-003, pg-003
Related SARs: 194, 392, 397
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
2. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
2. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following link(s) will be updated by patch crypt-int-002:
/usr/lib/libnsl.so --> /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-int-003
Date Issued: 11/20/2000 16:13:52
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
#############################################################################
Problem Description:
1. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
Problem Resolution:
1. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following links will also be updated:
/usr/lib/libnsl.so hard link to /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-int-003
Date Issued: 11/20/2000 16:19:31
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
#############################################################################
Problem Description:
1. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) may fail with ENOTCONN.
Problem Resolution:
1. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
The following links will also be updated:
/usr/lib/libnsl.so hard link to /usr/lib/libnsl_i.so
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-int-004
Date Issued: 08/22/2001 11:43:30
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #495, #503
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core dump
with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being
invoked in a static object for allocation of the data
structure to hold the host data.
2. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using a lot of different
addresses (host, netid).
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for
allocation of the data structure to hold the host data.
2. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: crypt-int-004
Date Issued: 09/12/2001 15:04:04
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core dump
with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being
invoked in a static object for allocation of the data
structure to hold the host data.
2. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using many different
addresses (host, netid).
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for
allocation of the data structure to hold the host data.
2. The number of cached rpcbind addresses was increased from 6 to 24.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crypt-int-005
Date Issued: 03/18/2002 11:17:36
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #535
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
Problem Resolution:
1. Implemented a test in tcpip.c and resolv.c of the nametoaddr
libraries. This test will check for the return of 127.0.0.1
from the ipm/Driver.o in the kernel and translate that into
a ND_NOHOST error to send back to the caller of netdir_options().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: crypt-int-005
Date Issued: 03/18/2002 11:28:46
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #535
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
##############################################################################
Problem Description:
1. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
Problem Resolution:
1. Implemented a test in tcpip.c and resolv.c of the nametoaddr
libraries. This test will check for the return of 127.0.0.1
from the ipm/Driver.o in the kernel and translate that into
a ND_NOHOST error to send back to the caller of netdir_options().
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: crypt-int-006
Date Issued: 08/02/2002 08:52:53
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAXOS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-011
Related SARs: none
Brief Description:
PowerMAXOS 4.3 crypt-int package release updates
##############################################################################
P1: On occasion, when network activity is high and /usr/lib/resolv.so is
specified in /etc/netconfig and there are contents in /etc/resolv.conf,
the resolution of addresses can become very slow. This is due to a
poll() timeout value being too short in libresolv and causing multiple
retries for the resolved address.
R1: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.so
/usr/lib/libnsl_i.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-int-006
Date Issued: 08/02/2002 08:33:20
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
###############################################################################
P1: On occasion, when network activity is high and /usr/lib/resolv.so is
specified in /etc/netconfig and there are contents in /etc/resolv.conf,
the resolution of addresses can become very slow. This is due to a
poll() timeout value being too short in libresolv and causing multiple
retries for the resolved address.
R1: This has been corrected by increasing the poll() timeout value in
libresolv 10 fold and to be more consistent with other UNIX vendors.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libnsl_i.a
/usr/lib/libnsl_i.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-int-007
Date Issued: 12/17/2002 11:22:41
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #780
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: crypt-int-007
Date Issued: 12/19/2002 14:44:19
Software Package: crypt-int pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #780
Brief Description:
PowerMAX OS 4.3 crypt-int package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-001
Date Issued: 02/02/2000 13:29:39
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, cnd-002, egl-001, ip-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. Incorrect chip setup results in excessive collisions resulting in the
failure to boot network clients.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. Correctly set up DEC chip in driver.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-001
Date Issued: 02/02/2000 11:37:58
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: PowerMAXION, TurboHawk
Related Patches: base-002, cnd-002, egl-001, ie-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. Incorrect chip setup results in excessive collisions resulting in the
failure to boot network clients.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. Correctly set up DEC chip in driver.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-002
Date Issued: 05/30/2000 13:04:21
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: cnd-003, egl-002, ip-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-002
Date Issued: 05/30/2000 13:48:12
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: cnd-003, egl-002, ie-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-003
Date Issued: 10/05/2000 13:51:39
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, diskless-004, egl-003, fibre-003,
inet-004, ip-003, man-003, pg-003
Related SARs: 39, 71, 237, 278
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network driver to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the dec driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
3. SARs 71, 237, 278: The 21143/QS6611 DEC/PHY chip combination
(used for the DEC 21143 PMC card) would never connect to one of the
full duplex modes (100BaseT or 10BaseT) when the DEC_n_MEDIA tunable
was set for either auto-negotiation or for one of the specific full
duplex modes. The interface would always be connected to one of the
half duplex modes even when the link partner supported full duplex.
4. SARs 71, 237, 278: A high rate of CRC errors was occurring when
using the DEC 21143 PMC card on Motorola platforms.
5. The on-board DEC chips on PowerStack III (7-slot) platforms were
non-functional.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The dec driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
- The dec driver's queued packet processing routine decsched()
routine was being called without the bd_lock locked.
Changed the code in the interrupt routine to lock the
bd_lock before it calls decsched().
2. Multiple changes were made to fix this problem:
- The flow control limit of the number of outstanding packets
that can be queued to each queue was set too small. Packets
would be discarded when this queue hiwater limit was
reached. This could occur on either the transmit or receive
side. Changed the default value for the dec driver's queue
flow control hiwater level from 40 to 8192 packets. Added
the capability to change the queue's hiwater level for this
driver in its space.c file.
- The dec chip will discard packets if the driver does not
provide a sufficient number of receive buffers. Changed the
receive buffer list replacement algorithm target level from
3/4ths full to full. Fixed the missed_frame and overflow
status processing for the DEC 21143 chip.
- The default value for the UDP tunable udp_sth_hiwat which
defines the flow control hiwater limit of the number of
outstanding packets that can be queued to the stream head
was increased from 1 MB to 2 MB. The maximum value for this
tunable was increased from 2 MB to 8 MB.
3. The dec driver had code specifically prevented using the full duplex
media types (100BaseT & 10BaseT) with DEC chips that used non-MII
type PHY chips. The DEC 21143 PMC card uses a QS6611 SYM (non-MII)
type of PHY chip. Auto-negotiation with the 21143/QS6611 DEC
chip/PHY combination would not work reliably. It would not always
linkup with its autonegotiation partner. Even when it would
successfully linkup, a lot of late collision errors occurred during
transmits. The dec_init_csrs() routine that is used for non-MII PHY
types was changed to initialize the DEC chip's registers to a
specific media type of auto-negotiating. The initialization routines
will try each of the supported media types (starting with the fastest
media type supported), until it either successfully connects to the
link partner or exhausts all of the media type possibilities.
4. Conditional code was added to dechardreset() where CSR0 is
initialized. The original code was always setting the cache align
size to 16 longwords and to 8 longwords in one of the PCI
configuration registers which caused problems on the 21143 PMC card
running on a Motorola platform. This code now sets the cache align
size to 8 longwords for Motorola platforms. This matches the
architecture checks that were already being done in the code that
sets up the PCI configuration registers.
5. The dec driver initialization code was not properly configuring the
on-board DEC 21143 chips on the PowerStack III. This board uses an
MII PHY type that has a different srom format type not previously
seen on other supported platforms. The srom parsing routine was
updated to correctly process this new supported srom type.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
/usr/include/sys/dec.h
/usr/include/sys/dlpi_dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: dec-003
Date Issued: 10/05/2000 14:06:04
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, egl-003, ie-003, inet-004, man-003, pg-003
Related SARs: 39, 71, 237, 278
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network driver to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the dec driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
3. SARs 71, 237, 278: The 21143/QS6611 DEC/PHY chip combination (used
for the on-board DEC on TurboHawks and the DEC 21143 PMC card) would
never connect to one of the full duplex modes (100BaseT or 10BaseT)
when the DEC_n_MEDIA tunable was set for either auto-negotiation or
for one of the specific full duplex modes. The interface would
always be connected to one of the half duplex modes even when the
link partner supported full duplex.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The dec driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
- The dec driver's queued packet processing routine decsched()
routine was being called without the bd_lock locked.
Changed the code in the interrupt routine to lock the
bd_lock before it calls decsched().
2. Multiple changes were made to fix this problem:
- The flow control limit of the number of outstanding packets
that can be queued to each queue was set too small. Packets
would be discarded when this queue hiwater limit was
reached. This could occur on either the transmit or receive
side. Changed the default value for the dec driver's queue
flow control hiwater level from 40 to 8192 packets. Added
the capability to change the queue's hiwater level for this
driver in its space.c file.
- The dec chip will discard packets if the driver does not
provide a sufficient number of receive buffers. Changed the
receive buffer list replacement algorithm target level from
3/4ths full to full. Fixed the missed_frame and overflow
status processing for the DEC 21143 chip.
- The default value for the UDP tunable udp_sth_hiwat which
defines the flow control hiwater limit of the number of
outstanding packets that can be queued to the stream head
was increased from 1 MB to 2 MB. The maximum value for this
tunable was increased from 2 MB to 8 MB.
3. The dec driver had code specifically prevented using the full duplex
media types (100BaseT & 10BaseT) with DEC chips that used non-MII
type PHY chips. The on-board 21143 DEC chip on the TurboHawk uses a
QS6611 SYM (non-MII) type of PHY chip. Auto-negotiation with the
21143/QS6611 DEC chip/PHY combination would not work reliably. It
would not always linkup with its autonegotiation partner. Even when
it would successfully linkup, a lot of late collision errors occurred
during transmits. The dec_init_csrs() routine that is used for
non-MII PHY types was changed to initialize the DEC chip's registers
to a specific media type of auto-negotiating. The initialization
routines will try each of the supported media types (starting with
the fastest media type supported), until it either successfully
connects to the link partner or exhausts all of the media type
possibilities.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
/usr/include/sys/dec.h
/usr/include/sys/dlpi_dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: dec-004
Date Issued: 08/22/2001 11:45:29
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #502
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 dec package release updates
#############################################################################
Problem Description:
1. SAR #502:
When a system with dec PMC card interface(s) or a TurboHawk
with embedded dec interfaces was booted and the networks were
not connected or were offline, the system would never detect
the dec interfaces even after they were connected or came
back online.
Problem Resolution:
1. Changes were made to fix problems in the dec driver for forced
speed/duplex initialization which must be used by the driver for
the dec PMC cards and the embedded dec interfaces on TurboHawks
(they do not properly autonegotiate). If the interface is found
to be inoperable during system initialization, the driver will
leave the interface initialized to the board's highest supported
media type. The cable can then be connected later. A message
will output to the console during initialization when the driver
detects that there is no media connected and will indicate the
speed/duplex setting that it was left set to. For successful
connections during system initialization, a message will be
output to the console that indicates the speed and duplex the
interface connected at.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/usr/include/sys/dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: dec-004
Date Issued: 09/11/2001 16:20:12
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 dec package release updates
#############################################################################
Problem Description:
1. SAR #502:
When a system with dec PMC card interface(s) or a TurboHawk
with embedded dec interfaces was booted and the networks were
not connected or were offline, the system would never detect
the dec interfaces even after they were connected or came
back online.
Problem Resolution:
1. Changes were made to fix problems in the dec driver for forced
speed/duplex initialization which must be used by the driver for
the dec PMC cards and the embedded dec interfaces on TurboHawks
(they do not properly autonegotiate). If the interface is found
to be inoperable during system initialization, the driver will
leave the interface initialized to the board's highest supported
media type. The cable can then be connected later. A message
will output to the console during initialization when the driver
detects that there is no media connected and will indicate the
speed/duplex setting that it was left set to. For successful
connections during system initialization, a message will be
output to the console that indicates the speed and duplex
of the interface and how it connected.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/usr/include/sys/dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: dec-005
Date Issued: 03/18/2002 11:17:54
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Added support for RFC 1643 dot3 transmission group attributes.
10. Fix several interface mib2 counters.
11. Added initialization for EtherChipSet and ifType.
12. Added a qreply call to the end of the DLbdspecioctl routine.
13. Changed "ifOutQlen" to "ifOutQLen".
14. Merged in a 5.1SR2 fix for the Znyz 4-port DEC PMC card that
defines a default link status bit to use as a default for cards
that don't provide the information in the srom.
15. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/mtune.d/dec
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
/etc/conf/sdevice.d/dec
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: dec-005
Date Issued: 03/18/2002 11:29:12
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 dec package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Added support for RFC 1643 dot3 transmission group attributes.
10. Fix several interface mib2 counters.
11. Added initialization for EtherChipSet and ifType.
12. Added a qreply call to the end of the DLbdspecioctl routine.
13. Changed "ifOutQlen" to "ifOutQLen".
14. Merged in a 5.1SR2 fix for the Znyz 4-port DEC PMC card that
defines a default link status bit to use as a default for cards
that don't provide the information in the srom.
15. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/mtune.d/dec
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
/etc/conf/sdevice.d/dec
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: dec-006
Date Issued: 05/08/2003 10:36:57
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 dec package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1456: Multicast did not work properly on DEC interfaces if
there were more than 13 user defined multicast groups.
R1. When more than 16 addresses are needed (3 of these are reserved
for the MAC address, broadcast address, and "well-known"
multicast group 224.0.0.0), the driver has to change the DEC
chip from "perfect" to "imperfect" filtering mode. Imperfect
filtering mode uses a hash table. Code to build the hash table
that should have returned the CRC of an address was incorrectly
returning the one's complement of the CRC. The wrong macro was
also being used to perform the byte swap of the hash data from
big-endian to little-endian mode. There were also places where
the code to build the address filtering tables (in either mode)
was being called twice. This caused unnecessary overhead each
time a multicast address was added or deleted.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/usr/include/sys/dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: dec-006
Date Issued: 05/08/2003 10:37:15
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 dec package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1456: Multicast did not work properly on DEC interfaces if
there were more than 13 user defined multicast groups.
R1. When more than 16 addresses are needed (3 of these are reserved
for the MAC address, broadcast address, and "well-known"
multicast group 224.0.0.0), the driver has to change the DEC
chip from "perfect" to "imperfect" filtering mode. Imperfect
filtering mode uses a hash table. Code to build the hash table
that should have returned the CRC of an address was incorrectly
returning the one's complement of the CRC. The wrong macro was
also being used to perform the byte swap of the hash data from
big-endian to little-endian mode. There were also places where
the code to build the address filtering tables (in either mode)
was being called twice. This caused unnecessary overhead each
time a multicast address was added or deleted.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/usr/include/sys/dec.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: dec-007
Date Issued: 01/07/2004 15:59:50
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 dec package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: decpromisc_off() was turning off promiscuous mode at the board
level when there was still one promiscuous mode user remaining.
R1: Changed the test in decpromisc_off() to leave promiscuous mode
ON if the new bd->promisc_cnt > 0. Previously it checked for
bd->promisc_cnt > 1.
P2: The internal function in the dec driver that checks for receive
overflows and missed packets should have been returning only the
number of overflows to the receive interrupt processing code.
R2: The internal function was returning the sum of the number of
overflows and the missed packets. This was corrected to return
only the number of overflows. Certain revs of the DEC chip have
a problem with their overflow logic. If an overflow occurs, the
driver might have to discard multiple input packets to clear the
condition. Because the return value incorrectly included missed
packet errors, valid packets might be discarded even though nothing
was wrong with them.
P3: The system would hang in the dec driver if the dec chip has the
overflow logic problem, and, the driver was unable to allocate
a replacement receive ring buffer.
R3: The overflow error processing was looping forever through the
linked list of receive buffer mblks. This list should have been
null terminated. The problem occurred because the system was low
on memory and a replacement receive ring buffer could not be
allocated. When the buffer to be reused was moved from the head
of the list to the tail of the list, the b_next pointer was not
being cleared. This caused the tail entry to point back to the
new head of the list and the overflow error processing code would
loop forever.
P4: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R4: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P5: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R5: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P6: The IP layer would switch from true hardware based multicast address
support to broadcast based multicast support if an attempt to exceed
the maximum number of multicast addresses occurred.
R6: Changed the DLPI common code to pass to the IP layer the actual error
code returned by the driver specific code. The drivers that support
true hardware based multicast addressing (dec, gte, pg, sym) were
changed to return different error codes based on the error type. The
IP layer is now able to distinguish between problems with the hardware
based multicast address support or whether other software based errors
have occurred. When a problem occurs with the hardware based multicast
address support, the IP layer will change the interface to use
broadcasts for multicasting.
P7: The size of the multicast address table for the dec driver was too
small.
R7: The maximum number of multicast addresses for the dec was changed
from 16 to 32.
P8: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R8: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: dec-007
Date Issued: 01/07/2004 16:00:46
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 dec package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: decpromisc_off() was turning off promiscuous mode at the board
level when there was still one promiscuous mode user remaining.
R1: Changed the test in decpromisc_off() to leave promiscuous mode
ON if the new bd->promisc_cnt > 0. Previously it checked for
bd->promisc_cnt > 1.
P2: The internal function in the dec driver that checks for receive
overflows and missed packets should have been returning only the
number of overflows to the receive interrupt processing code.
R2: The internal function was returning the sum of the number of
overflows and the missed packets. This was corrected to return
only the number of overflows. Certain revs of the DEC chip have
a problem with their overflow logic. If an overflow occurs, the
driver might have to discard multiple input packets to clear the
condition. Because the return value incorrectly included missed
packet errors, valid packets might be discarded even though nothing
was wrong with them.
P3: The system would hang in the dec driver if the dec chip has the
overflow logic problem, and, the driver was unable to allocate
a replacement receive ring buffer.
R3: The overflow error processing was looping forever through the
linked list of receive buffer mblks. This list should have been
null terminated. The problem occurred because the system was low
on memory and a replacement receive ring buffer could not be
allocated. When the buffer to be reused was moved from the head
of the list to the tail of the list, the b_next pointer was not
being cleared. This caused the tail entry to point back to the
new head of the list and the overflow error processing code would
loop forever.
P4: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R4: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P5: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R5: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P6: The IP layer would switch from true hardware based multicast address
support to broadcast based multicast support if an attempt to exceed
the maximum number of multicast addresses occurred.
R6: Changed the DLPI common code to pass to the IP layer the actual error
code returned by the driver specific code. The drivers that support
true hardware based multicast addressing (dec, gte, pg, sym) were
changed to return different error codes based on the error type. The
IP layer is now able to distinguish between problems with the hardware
based multicast address support or whether other software based errors
have occurred. When a problem occurs with the hardware based multicast
address support, the IP layer will change the interface to use
broadcasts for multicasting.
P7: The size of the multicast address table for the dec driver was too
small.
R7: The maximum number of multicast addresses for the dec was changed
from 16 to 32.
P8: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R8: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
/etc/conf/pack.d/dec/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: dec-008
Date Issued: 03/13/2006 13:17:56
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 dec package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3586: When cold booting a PH640 with the DECs and the network
switch forced to 100baseT full duplex, DEC 2 does not connect. It
will connect after subsequent warm boot or reset button pressed.
R1: Initialization logix for fixed speed/duplex connections was not
performing the correct reset sequence for the DEC 21140, causing
the interface to hang on the first TX packet after initial setup.
Only occurs with embedded DECs set for fixed speed/duplex using
DEC_x_MEDIA tunables. The embedded DECs work properly if
configured for auto-negotiation.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: dec-008
Date Issued: 03/13/2006 11:39:06
Software Package: dec pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 dec package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3586: When cold booting a PH640 with the DECs and the network
switch forced to 100baseT full duplex, DEC 2 does not connect. It
will connect after subsequent warm boot or reset button pressed.
R1: Initialization logix for fixed speed/duplex connections was not
performing the correct reset sequence for the DEC 21140, causing
the interface to hang on the first TX packet after initial setup.
Only occurs with embedded DECs set for fixed speed/duplex using
DEC_x_MEDIA tunables. The embedded DECs work properly if
configured for auto-negotiation.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/dec/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: diskless-001
Date Issued: 09/15/1999 10:48:18
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001, man-001">fbsman-001, man-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 diskless package release updates
##############################################################################
Problem Description:
1. mkvmebstrap(1M) requires the VME hostname as an argument but silently
ignores the error when the ethernet instead of the VME hostname is
specified. This affects closely-coupled but not loosely-coupled
systems.
2. Static routes generated on the file server and on the diskless client
during their boot phase interfered with communication targeted
outside of the cluster. This affects closely-coupled but not
loosely-coupled systems.
3. During the configuration of a diskless client via vmebootconfig(1M)
and netbootconfig(1M), the resource sharing table (dfstab(4)) was
being modified and the share(1M) command executed for clients
configured as embedded. Since embedded client configurations do not
support networking, these steps are not necessary. This affects both
closely-coupled and loosely-coupled systems.
4. The sbc_msgd(1M) daemon, used for closely-coupled IPC, may sometimes
get into an infinite loop, "hogging" a processor.
5. The idsymobj(1M) man page incorrectly refers to "Closely-Coupled"
systems as "Loosely-Coupled".
Problem Resolution:
1. An error condition was added so that if the ethernet hostname is
specified to mkvmebstrap, the command will error out.
2. Changes in the networking code have deemed most of the static routes
unnecessary and are no longer being generated.
3. The dfstab(4) table is modified and the share(1M) command executed
only for diskless clients configured with NFS support.
4. sbc_msgd(1M) may get error conditions when communicating to a remote
system. Some errors (such as "sbc module not configured") are fatal
and should result in the utility doing an exit instead of retrying
the operation.
5. Corrected "Loosely-Coupled" references idsymobj(1M) man page.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/bin/mkbstrap
/usr/etc/diskless.d/bin/nodeconfig
/usr/etc/diskless.d/sys.conf/rc.d/clusterboot
/usr/etc/diskless.d/sys.conf/rc.d/disklessboot.vmeboot
/usr/etc/diskless.d/sys.conf/rc.d/remoteboot
/usr/etc/diskless.d/sys.conf/vroot.d/vroot.uniqlist
/usr/sbin/sbc_msgd
/usr/share/man/cat1/idsymobj.1m.z
/usr/share/man/man1/idsymobj.1m
Special Conditions for Installation:
1. The following object will also be replaced by diskless-001:
* /etc/init.d/clusterboot
* /etc/rc3.d/S23clusterboot
* /etc/rc0.d/K13clusterboot
2. For Closely-Coupled systems only:
1) reboot the file server after the patch is installed to clear
unnecessary static routes.
2) a) re-create the memfs file image
b) reboot each diskless client
Steps 2a) and 2b) can be accomplished by using the following
command:
# mkvmebstrap -B -r memfs -p2 all
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: diskless-002
Date Issued: 02/02/2000 13:32:40
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001, fbs-002, man-002">fbsman-002, inet-002
Related SARs: #39, #92, #112, #152
Brief Description:
PowerMAX OS 4.3 diskless package release updates
##############################################################################
Problem Description:
1. Implement "foreign" virtual interrupts.
2. #92: VME Bus errors can cause VME networking to panic or hang the
system. In some situations, the DMA or CPU write operations through
the PCI-to-VME Bus bridge will fail due to VME Bus errors. When
these bus errors occur, the VME networking kernel code on the sending
SBC gets out of synchronization with the VME networking kernel code
on the receiving SBC, and the receiving SBC's VME networking code may
panic or hang the system.
3. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
4. #112: vmebootconfig(1M), the configuration tool in a closely-coupled
system architecture, does not parse correctly entries in the nodes
table (/etc/dtables/nodes.vmeboot) that use tabs instead of spaces as
field separators.
5. #152: In closely-coupled and loosely-coupled system architectures,
delays may be observed when attempting to remotely access the client
after a client reboot.
6. #39: When the vmemsg driver detects a VME bus error while DMAing VME
message data to another SBC across the VME bus, the error message:
WARNING: vmemsg: Failed DMA Operation on board: n status: ...
is output to the console and /var/adm/syslog. These messages can
cause the syslog to become full if this error occurs frequently.
This error is more likely to occur when a VME device board
(i.e pg FDDI) is configured at a VMEIRQ level other than level 5, and
a large amount of data is transferred over the pg FDDI interface at
high rates.
Problem Resolution:
1. Implemented "foreign" SBC virtual interrupts. This allows a
non-PowerMAX OS system to send a virtual interrupt to a PowerMAX OS
system running in the same rack (closely-coupled). A tunable is used
to enable this feature.
2. Additional validation checks are now made on the receiving SBC that
attempt to validate the data from sending SBC. When the received
data appears to be corrupted, the receiving SBC will then sync up its
receive buffer pointer with the sending SBC, and thus successfully
recover from the VME bus error.
3. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
4. Fixed scripts that used pattern matching expressions to parse the
nodes table. The tabs in these expressions [tabspace] had
accidentally been turned into spaces [spacesspace].
5. Files locked down during the boot sequence are now created locally
on the client, as the problem is related to file locking across NFS.
6. A new vmemsg tunable VMEMSG_IGN_DMAERRS was added. When enabled,
this tunable suppresses the syslog warning message that is usually
issued when the vmemsg driver encounters a VME bus error while
DMAing VME message data to another SBC across the VME bus.
A new metric (output_dmac_errors) was added to the sbc_metrics
structure. This metric is incremented each time the vmemsg driver
detects a VME bus error while DMAing VME message data, regardless of
the value of the VMEMSG_IGN_DMAERRS tunable.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/bin/sbcdbstrap
/etc/conf/mtune.d/sbc
/etc/conf/mtune.d/vmemsg
/etc/conf/pack.d/sbc/Driver.o
/etc/conf/pack.d/sbc/space.c
/etc/conf/pack.d/vmemsg/Driver.o
/etc/conf/pack.d/vmemsg/space.c
/usr/etc/diskless.d/bin/mkprofile
/usr/etc/diskless.d/bin/mkvroot
/usr/etc/diskless.d/sys.conf/rc.d/clusterboot
/usr/etc/diskless.d/sys.conf/rc.d/disklessboot.netboot
/usr/etc/diskless.d/sys.conf/rc.d/disklessboot.vmeboot
/usr/etc/diskless.d/sys.conf/subsys.d/CCS_FBS/rc.emb
/usr/etc/diskless.d/sys.conf/subsys.d/CCS_FBS/stune
/usr/etc/diskless.d/sys.conf/subsys.d/RC_FBS/config.sh
/usr/etc/diskless.d/sys.conf/subsys.d/RC_FBS/desc
/usr/etc/diskless.d/sys.conf/subsys.d/RC_FBS/memfs.nfs
/usr/etc/diskless.d/sys.conf/subsys.d/RC_FBS/modlist
/usr/include/sys/sbc.h
/usr/share/man/cat1/netbootconfig.1m.z
/usr/share/man/cat1/vmebootconfig.1m.z
/usr/share/man/man1/netbootconfig.1m
/usr/share/man/man1/vmebootconfig.1m
Special Conditions for Installation:
The following object(s) will also be replaced by diskless-002:
* /etc/init.d/clusterboot
* /etc/rc0.d/K13clusterboot
* /etc/rc3.d/S23clusterboot
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
%############################################################################
Patch Name: diskless-003
Date Issued: 06/02/2000 14:07:40
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 diskless package release updates
%############################################################################
Problem Description:
1. Loosely coupled target systems (LCS) would occasionally hang during
boot. Net or flash booted systems would hang early in the boot
process, before the point any messages from the booted PowerMAX OS
are printed but after the point all the messages from the LCS boot
loader were printed.
Problem Resolution:
1. Part of the LCS boot sequence was damaging data in the L1 data cache.
Depending on what was damaged, the system would either hang or
(apparently) boot correctly.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/bin/sbcbstrap
Special Conditions for Installation:
1. All LCS target kernels should be rebuilt and rebooted following
installation of diskless-003 in order to ensure correct functionality.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: diskless-004
Date Issued: 10/05/2000 13:52:17
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, fibre-003, inet-004,
ip-003, man-003, pg-003
Related SARs: 333
Brief Description:
PowerMAX OS 4.3 diskless package release updates
##############################################################################
Problem Description:
1. SAR 333: When a diskless client system is booted, the /var/tmp
directory is not initialized (cleared out).
Problem Resolution:
1. In a diskless client system start-up, all files under the /var/tmp
directory are now removed.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/sys.conf/rc.d/disklessboot.vmeboot
/usr/etc/diskless.d/sys.conf/rc.d/disklessboot.netboot
Special Conditions for Installation:
1. The following object(s) will also be updated by patch diskless-004:
<vroot_dir>/etc/rc3.d/S24disklessboot
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: diskless-005
Date Issued: 08/22/2001 11:46:25
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 diskless package release updates
#############################################################################
Problem Description:
1. The following kernel panics will occur on systems with
an enabled kmadbg kernel module:
- The kernel will panic in the fbs_srv_unreg_dev() routine
during the un-registration of a FBS Coupled timing device.
This routine was referencing the fbs data structure of the
virtual fbs scheduler after this structure had already been
kmem_free()ed.
- The kernel will panic in the memfs_dirtrunc() routine during
a closely-coupled client boot up sequence. This routine was
referencing a kma block after the block had already been freed.
- The kernel will panic in the sbc_msg_recv() routine when
processing a received message; typically, on the server SBC
where the server SBC has received fbsd(1M) messages from a
client SBC in a closely-coupled system. This routine was
referencing a kmem_alloc()ed sbc message structure after
it had already been kmem_free()ed.
2. The sbc_msgd inter-SBC message daemon on the server SBC of a
closely-coupled system can cause problems when the sbc kernel
module is enabled, but when the kernel is not yet configured to
be a server SBC in a CCS environment.
In this case, the sbc_msgd can end up looping continually on
sbcmsg() kernel calls.
3. The handling the sbc module's VME interrupt lock was incorrect,
and the calling LWP would end up spinning on this VME interrupt
lock in the sbc module's VME interrupt acknowledgement (IACK)
interrupt routine, when trying to acquire this spin lock.
Problem Resolution:
1. All three of these problems were re-coded so that they no longer
reference an already kmem_free()ed memory block. Any needed
information from these kma memory areas is now stored into a
local variable before the kma block is freed.
2. The sbc_msgd code was modified so that it now detects when the
kernel is not actually configured as a closely-coupled server
SBC. In this case, the sbc_msgd quietly exits.
3. The sbc module's handing of the internal VME interrupt lock
was re-coded so that this lock is now properly unlocked such
that the calling LWP is now able to acquire this lock when needed
inside the VME IACK interrupt routine.
Enhancements:
None
Object(s) To Be Replaced:
/usr/sbin/sbc_msgd
/etc/conf/pack.d/sbc/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: diskless-006
Date Issued: 03/18/2002 11:18:38
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #590
Brief Description:
PowerMAX OS 4.3 diskless package release updates
##############################################################################
Problem Description:
1. SAR #590: The command clock_synchronize was being executed only if
file /tmp/rcimclients on the server contained the name(s) of the
target(s). However,"/tmp/rcimclients" was being written to only
after the client(s) were booted. The whole idea of sleeping until
the targets were done was useless.
Problem Resolution:
1. "clusterboot" script was modified to execute clock_synchronize
at the right time.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/sys.conf/rc.d/clusterboot
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: diskless-007
Date Issued: 12/11/2002 14:33:41
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #806
Brief Description:
PowerMAX OS 4.3 diskless package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #806: NIS is not functional on a Closely Coupled Client
System. The ypinit(1M) utility fails.
R1: The vroot.uniqlist file has been changed to add
/etc/init.d/nis and /etc/rc2.d/S51domain to the unique list to
allow the diskless software to build a vroot that will support
the running of ypinit(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/sys.conf/vroot.d/vroot.uniqlist
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: diskless-008
Date Issued: 05/08/2003 10:37:34
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 diskless package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/sbc
/etc/conf/mtune.d/vmemsg
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: diskless-009
Date Issued: 01/07/2004 16:00:16
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 diskless package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1476: The utility /etc/diskless.d/bin/config_utils causes an
error when the -v option is used and comparing (range verifying)
the value of CLIENT_SLV_MMAP_SIZE to the max value of the tunable
SBC_SLAVE_MMAP_SIZE because the former is a value of 1KB chunks
and the latter is a value of 4KB chunks.
R1: Added a conversion factor into the get_valid() routine of
/etc/diskless.d/bin/config_utils so that we actually compare
values that represent similar entities.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/bin/config_utils
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: diskless-010
Date Issued: 07/13/2004 09:55:37
Software Package: diskless pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 diskless package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: Could not make a CCS/LCS client. make would fail.
R1: Added VPATH variable, "export VPATH=/:$DLESSD:$CLIENT_VROOT",
to line 451 of mkbstrap script. This search variable was needed
to be able to get /usr/ccs/bin/make to successfully build a
CCS/LCS client.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/etc/diskless.d/bin/mkbstrap
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-001
Date Issued: 02/02/2000 13:41:05
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-002, cnd-002, dec-001, ip-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-001
Date Issued: 02/02/2000 11:38:29
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, cnd-002, dec-001, ie-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-002
Date Issued: 05/30/2000 13:05:03
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, Power Hawk 640
Related Patches: cnd-003, dec-002, ip-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
/etc/conf/pack.d/egl/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-002
Date Issued: 05/30/2000 13:48:43
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: cnd-003, dec-002, ie-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
/etc/conf/pack.d/egl/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-003
Date Issued: 10/05/2000 13:52:42
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, fibre-003,
inet-004, ip-003, man-003, pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
Problem Resolution:
1. A check for the TX_QUEUED flag was added to the common DLPI code in
DLunitdata_req() that checks for TX_BUSY before it calls
DLxmit_packet(). Without the check for the TX_QUEUED flag, there was
a race condition with the driver's queued packet processing routine
that is called by the driver's transmit interrupt routine. New
packets were being transmitted to the board ahead of packets that
were already waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the bd_lock
after TX_BUSY was reset, but before the call to the driver's queued
packet processing routine.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: egl-003
Date Issued: 10/05/2000 14:06:37
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, ie-003, inet-004, man-003, pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
Problem Resolution:
1. A check for the TX_QUEUED flag was added to the common DLPI code in
DLunitdata_req() that checks for TX_BUSY before it calls
DLxmit_packet(). Without the check for the TX_QUEUED flag, there was
a race condition with the driver's queued packet processing routine
that is called by the driver's transmit interrupt routine. New
packets were being transmitted to the board ahead of packets that
were already waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the bd_lock
after TX_BUSY was reset, but before the call to the driver's queued
packet processing routine.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: egl-004
Date Issued: 03/18/2002 11:18:52
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: egl-004
Date Issued: 03/18/2002 11:30:17
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 egl package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: egl-005
Date Issued: 01/07/2004 16:00:31
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 egl package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: egl-005
Date Issued: 01/07/2004 16:01:07
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 egl package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: egl-006
Date Issued: 07/08/2004 10:52:54
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 egl package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1887: If you attemptted to reboot the system without first
powering down, the EGL interface wasn't functional after the
reboot.
R1: Neither of the egl or cnd drivers had any code to reset the
board during a system shutdown or a DLM unload. Added a
shutdown/halt routine to the egl and cnd drivers to perform
the reset. This stops the egl and cnd boards from sending
receive packets to the cpu before their driver is able to
initialize the software tables that are needed by the board.
P2: The ifOutQLen mib2 counter would occasionally get out of sync
with the actual number of queued output packets in times of
heavy output traffic in the cnd and egl drivers.
R2: Fixed the ifOutQLen counter logic in the queued packet
processing routines in both of the cnd and egl drivers.
This was occasionally causing a delay in the transmission
of queued output packets.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/egl
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: egl-006
Date Issued: 07/26/2004 11:50:37
Software Package: egl pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 egl package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1887: If you attemptted to reboot the system without first
powering down, the EGL interface wasn't functional after the
reboot.
R1: Neither of the egl or cnd drivers had any code to reset the
board during a system shutdown or a DLM unload. Added a
shutdown/halt routine to the egl and cnd drivers to perform
the reset. This stops the egl and cnd boards from sending
receive packets to the cpu before their driver is able to
initialize the software tables that are needed by the board.
P2: The ifOutQLen mib2 counter would occasionally get out of sync
with the actual number of queued output packets in times of
heavy output traffic in the cnd and egl drivers.
R2: Fixed the ifOutQLen counter logic in the queued packet
processing routines in both of the cnd and egl drivers.
This was occasionally causing a delay in the transmission
of queued output packets.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/egl
/etc/conf/pack.d/egl/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fbs-001
Date Issued: 09/15/1999 10:50:01
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
##############################################################################
Problem Description:
1. When multiple fbstrig(2) or fbswait(2) operations occur within the
same FBS scheduler at approximately the same time, there can be some
performance degradation with these operations due to locking
contention for various kernel locks that serialize these operations.
There can also be additional kernel locking contentions between
fbstrig(2) calls and FBS scheduler interrupts within the same
scheduler.
Problem Resolution:
1. The areas in the kernel that deal with the serialization of the
fbstrig(2), fbswait(2) and scheduler interrupt operations were
re-written to reduce the amount of locking contention on the locks
that are acquired by these parts of the kernel.
These changes should help to improve system performance in these
areas on multi-CPU systems.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
/usr/include/sys/fbs.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fbs-001
Date Issued: 09/15/1999 12:58:13
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
##############################################################################
Problem Description:
1. When multiple fbstrig(2) or fbswait(2) operations occur within the
same FBS scheduler at approximately the same time, there can be some
performance degradation with these operations due to locking
contention for various kernel locks that serialize these operations.
There can also be additional kernel locking contentions between
fbstrig(2) calls and FBS scheduler interrupts within the same
scheduler.
Problem Resolution:
1. The areas in the kernel that deal with the serialization of the
fbstrig(2), fbswait(2) and scheduler interrupt operations were
re-written to reduce the amount of locking contention on the locks
that are acquired by these parts of the kernel.
These changes should help to improve system performance in these
areas on multi-CPU systems.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
/usr/include/sys/fbs.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fbs-002
Date Issued: 02/02/2000 13:41:32
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001, man-002">fbsman-002, inet-002, librt-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
##############################################################################
Problem Description:
1. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
Problem Resolution:
1. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/fbs
/etc/conf/mtune.d/fbs
/etc/conf/pack.d/fbs/Driver.o
/etc/conf/pack.d/fbs/Modstub.o
/etc/conf/pack.d/fbs/space.c
/etc/conf/pack.d/rdevfs/Driver.o
/etc/conf/pack.d/rdevfs/stubs.c
/etc/init.d/fbs
/usr/bin/rtcp
/usr/include/sys/fbs.h
/usr/sbin/fbsd
/usr/sbin/fbsrcreg
Special Conditions for Installation:
The following object(s) will also be replaced by fbs-002:
* /etc/rc0.d/K91fbs
* /etc/rc3.d/S91fbs
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fbs-002
Date Issued: 02/02/2000 11:38:56
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crosslibs-001, man-002">fbsman-002, inet-002, librt-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
##############################################################################
Problem Description:
1. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
Problem Resolution:
1. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
/etc/conf/pack.d/fbs/Modstub.o
/etc/conf/pack.d/fbs/space.c
/usr/bin/rtcp
/usr/include/sys/fbs.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: fbs-003
Date Issued: 03/29/2001 12:00:02
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #441
Brief Description:
PowerMAX OS 4.3 fbs package release updates.
#############################################################################
Problem Description:
1. The client fbsd daemon sometimes failed to start up successfully.
Under a customer's closely-coupled configuration, the client SBC's
fbs daemon (fbsd) would frequently fail to start up successfully.
When this problem occurred, it would still be possible to manually
restart the fbsd daemon after the client SBC finished booting.
Problem Resolution:
1. The problem was that a sbcmsg(2) call that the client fbsd makes
was returning ENOMEM. The client fbsd daemon makes this call to
check that the fbsd server mailbox has already been reserved by
the server SBC fbsd daemon.
This error can occur when the server SBC receives this sbc message
and the processing attempts to process this server mailbox reser-
vation check at a point in time when the server's kernel is temp-
orarily out of kernel heap freelist blocks.
The fbsd code was re-coded so that it will now recognize the ENOMEM
error as a non-fatal error, and instead of giving up, it will now
retry the server mailbox reservation sbcmsg(2) call until it suc-
ceeds or until a more serious fatal error is returned.
Object(s) To Be Replaced:
/usr/sbin/fbsd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: fbs-003
Date Issued: 09/11/2001 16:28:49
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
#############################################################################
Problem Description:
1. If the RCIM rtc that is being used as the timing source for a
RCIM Coupled FBS scheduler has not been initialized since last
system boot, then when the associated FBS scheduler(s) are
started with a rtcp(1) "start" command or a
fbsintrpt(3rt) librt library routine call, the kernel will not
handle this error condition properly.
The kernel will instead disable all future use of RCIM Coupled
FBS support in the kernel. At this point, the kernel must be
rebooted before the RCIM Coupled FBS support may be used again.
2. The following kernel panics will occur on systems with
an enabled kmadbg kernel module:
- The kernel will panic in the fbs_srv_unreg_dev() routine
during the un-registration of a FBS Coupled timing device.
This routine was referencing the fbs data structure of the
virtual fbs scheduler after this structure had already been
kmem_free()ed.
3. When a debugger issues a stop request on an LWP that is scheduled
on a FBS scheduler, the kernel may panic if the LWP exits
before the FBS module processes the FBS scheduler stop operation.
If the LWP structure has already been kmem_free()ed, and
re-allocated for a different purpose, then examining the
l_fbslwpp field in the previous LWP structure can cause a kernel
data access exception panic.
4. If one of the LWPs in a multi-lwp or multi-threaded
application hits a debugger breakpoint and the entire process is
stopped when one of the other LWPs in the process is currently
scheduled on a running FBS scheduler, this scheduler was not
being stopped. The FBS scheduler would continue to run, even
though it had a LWP that had been placed into a stopped state.
Problem Resolution:
1. The reply message status reporting in the kernel's FBS network
message passing routine, fbs_rcmsg(), was broken. This caused
the client-side start scheduler code in the
fbs_clnt_start_sched() kernel routine to not properly report the
rtc initialization error back to the user on the rtcp(1) "start"
command or the fbsintrpt(3rt) library routine call.
This FBS network message kernel routine was fixed so that it now
passes the reply message status back to the calling kernel
routine. With this change, the fbs_clnt_start_sched() kernel
routine can now properly detect when the RCIM rtc has not yet been
initialized, and thus return the correct error condition back to
the caller in user-space, rather than disable all use of the
RCIM Coupled FBS kernel support.
2. All three of these problems were re-coded so that they no longer
reference an already kmem_free()ed memory block. Any needed
information from these kma memory areas is now stored into a
local variable before the kma block is freed.
3. The fix is to place a lwp_hold() on the LWP structure until
the FBS stop operation completes. In this way, the LWP structure
will remain allocated until after the FBS stop code can examine
the LWP structure.
4. The kernel routine that is called by all LWPs in a process that
is being stopped is called prstopped(). Within this routine,
there is a call to the fbsstop_dbg() routine. This is the
routine that will additionally stop the corresponding FBS
scheduler of the calling LWP when that LWP is scheduled on a FBS
scheduler.
The fbsstop_dbg() call was located in the wrong place within the
prstopped() routine such that only the very last LWP to stop
within the process would make the fbsstop_dbg() routine call.
The call to fbsstop_dbg() was moved forward within the
prstopped() routine so that now all LWPs calling prstopped()
will also call fbsstop_dbg(), and thus stop any corresponding
FBS scheduler on which they may currently be scheduled.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: fbs-004
Date Issued: 08/22/2001 11:47:26
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #538
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
#############################################################################
Problem Description:
1. If the RCIM rtc that is being used as the timing source for a
RCIM Coupled FBS scheduler has not been initialized since last
system boot, then when the associated FBS scheduler(s) are
started with a rtcp(1) "start" command or a
fbsintrpt(3rt) librt library routine call, the kernel will not
handle this error condition properly.
The kernel will instead disable all future use of RCIM Coupled
FBS support in the kernel. At this point, the kernel must be
rebooted before the RCIM Coupled FBS support may be used again.
2. SAR #538:
After roughly 57 registrations of closely-coupled FBS timing
devices have successfully completed, all future registrations
will fail. At this point, the entire cluster must be rebooted
in order to register any additional FBS coupled timing devices.
3. The following kernel panics will occur on systems with
an enabled kmadbg kernel module:
- The kernel will panic in the fbs_srv_unreg_dev() routine
during the un-registration of a FBS Coupled timing device.
This routine was referencing the fbs data structure of the
virtual fbs scheduler after this structure had already been
kmem_free()ed.
- The kernel will panic in the memfs_dirtrunc() routine during
a closely-coupled client boot up sequence. This routine was
referencing a kma block after the block had already been freed.
- The kernel will panic in the sbc_msg_recv() routine when
processing a received message; typically, on the server SBC
where the server SBC has received fbsd(1M) messages from a
client SBC in a closely-coupled system. This routine was
referencing a kmem_alloc()ed sbc message structure after
it had already been kmem_free()ed.
4. When a debugger issues a stop request on LWP that is scheduled
on a FBS scheduler, the kernel may panic if the LWP has exited
by the time that the FBS module gets around to processing the
FBS scheduler stop operation.
If the LWP structure has already been kmem_free()ed, and
re-allocated for a different purpose, then examining the
l_fbslwpp field in the previous LWP structure can cause a kernel
data access exception panic.
5. When one of the LWPs in a multi-lwp or multi-threaded
application hits a debugger breakpoint and the entire process is
stopped, if one of the other LWPs in this process is currently
scheduled on a running FBS scheduler, this scheduler was not
being stopped. The FBS scheduler would continue to run, even
though it had a LWP that had been placed into a stopped state.
Problem Resolution:
1. The reply message status reporting in the kernel's FBS network
message passing routine, fbs_rcmsg(), was broken. This caused
the client-side start scheduler code in the
fbs_clnt_start_sched() kernel routine to not properly report the
rtc initialization error back to the user on the rtcp(1) "start"
command or the fbsintrpt(3rt) library routine call.
This FBS network message kernel routine was fixed so that it now
passes the reply message status back to the calling kernel
routine. With this change, the fbs_clnt_start_sched() kernel
routine can now properly detect when the RCIM rtc has not yet been
initialized, and thus return the correct error condition back to
the caller in user-space, rather than disable all use of the
RCIM Coupled FBS kernel support.
2. The fbsd daemon open(2)s /dev/ip during the registration of a
FBS closely-coupled timing device. This file descriptor is used
to query the local system's network interfaces for the purpose
of obtaining all possible hostnames that are associated with the
local system.
3. All three of these problems were re-coded so that they no longer
reference an already kmem_free()ed memory block. Any needed
information from these kma memory areas is now stored into a
local variable before the kma block is freed.
4. The fix is to place a lwp_hold() on the LWP structure until
the FBS stop operation completes. In this way, the LWP structure
will remain allocated until after the FBS stop code can examine
the LWP structure.
5. The kernel routine that is called by all LWPs in a process that
is being stopped is called prstopped(). Within this routine,
there is a call to the fbsstop_dbg() routine. This is the
routine that will additionally stop the corresponding FBS
scheduler of the calling LWP when that LWP is scheduled on a FBS
scheduler.
The fbsstop_dbg() call was located in the wrong place within the
prstopped() routine, such that only the very last LWP to stop
within the process would make the fbsstop_dbg() routine call.
The call to fbsstop_dbg() was moved forward within the
prstopped() routine so that now all LWPs calling prstopped()
will also call fbsstop_dbg(), and thus stop any corresponding
FBS scheduler that they may be currently scheduled on.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
/usr/sbin/fbsd
/usr/sbin/fbsrcreg
Special Conditions for Installation:
1. RCIM Coupled FBS systems.
2. Closely-coupled systems that currently have the diskless and fbs
packages installed.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: fbs-004
Date Issued: 07/26/2004 11:50:48
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 fbs package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT-1457: With a running FBS and performance monitoring
enabled, if you perform a "clear performance monitoring
values", then the min cycle time for each scheduled process
becomes zero microseconds. The frame/cycle for that new 0us
min cycle is the first cycle after the PM values are cleared.
R1: The "last cycle time" is computed as the delta time
accumulated in lwp->l_*time since the last time this lwp
checked lwp->l_*time. When we clear the performance monitor
values, we set our "previous" time to the current lwp->l_*time.
When the lwp computes the delta time, it comes up with zero
time, and that zero time becomes the minimum cycle time
(and minimum frame time when #cycles/frame=1). Don't let this
happen.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: fbs-005
Date Issued: 03/18/2002 11:19:11
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbs package release updates
##############################################################################
Problem Description:
1. During initiailization of the fbs daemon,messages that are more
adequate for debugging purposes than normal startup, were logged
in the file /etc/.osm.
Problem Resolution:
1. With the change,they are only logged in /fbsd.debug if debugging
is enabled.
Enhancements:
None
Object(s) To Be Replaced:
/usr/sbin/fbsd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: fbs-006
Date Issued: 01/07/2004 16:00:46
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 fbs package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: SAR 1487: After patch installation fbs-005, the fbs debug output no
longer appeared on the console.
R1: Added a new option to fbsd(1m), -D, which will send debug output to
the console instead of /fbsd.debug.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/sbin/fbsd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: fbs-007
Date Issued: 07/08/2004 10:53:08
Software Package: fbs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 fbs package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT-1457: With a running FBS and performance monitoring
enabled, if you perform a "clear performance monitoring
values", then the min cycle time for each scheduled process
becomes zero microseconds. The frame/cycle for that new 0us
min cycle is the first cycle after the PM values are cleared.
R1: The "last cycle time" is computed as the delta time
accumulated in lwp->l_*time since the last time this lwp
checked lwp->l_*time. When we clear the performance monitor
values, we set our "previous" time to the current lwp->l_*time.
When the lwp computes the delta time, it comes up with zero
time, and that zero time becomes the minimum cycle time
(and minimum frame time when #cycles/frame=1). Don't let this
happen.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fbs/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-001">fbsman-001
Date Issued: 09/15/1999 10:50:36
Software Package: fbsman pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001, diskless-001, man-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbsman package release updates
##############################################################################
Problem Description:
1. Many of the FBS man pages that discuss the rdevfs(4) Closely-Coupled
FBS support were incorrectly referring to "Closely-Coupled" systems
as "Loosely-Coupled" systems.
Problem Resolution:
1. The FBS man pages were changed so that they now correctly refer to
Closely-Coupled systems where appropriate.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/rtcp.1.z
/usr/share/man/cat3/fbs_register_cluster_device.3f.z
/usr/share/man/cat3/fbs_register_cluster_device.3x.z
/usr/share/man/cat3/fbs_unregister_cluster_device.3f.z
/usr/share/man/cat3/fbs_unregister_cluster_device.3x.z
/usr/share/man/cat3/fbsaccess.3f.z
/usr/share/man/cat3/fbsaccess.3x.z
/usr/share/man/cat3/fbscycle.3f.z
/usr/share/man/cat3/fbscycle.3x.z
/usr/share/man/cat3/fbsdetach.3f.z
/usr/share/man/cat3/fbsdetach.3x.z
/usr/share/man/cat3/fbsgetrtc.3f.z
/usr/share/man/cat3/fbsgetrtc.3x.z
/usr/share/man/cat3/fbsinfo.3f.z
/usr/share/man/cat3/fbsinfo.3x.z
/usr/share/man/cat3/fbsinfo_cluster.3f.z
/usr/share/man/cat3/fbsinfo_cluster.3x.z
/usr/share/man/cat3/fbsintrpt.3f.z
/usr/share/man/cat3/fbsintrpt.3x.z
/usr/share/man/cat3/fbsquery.3f.z
/usr/share/man/cat3/fbsquery.3x.z
/usr/share/man/cat3/fbsremove.3f.z
/usr/share/man/cat3/fbsremove.3x.z
/usr/share/man/cat3/fbsresume.3f.z
/usr/share/man/cat3/fbsresume.3x.z
/usr/share/man/cat3/fbsrunrtc.3f.z
/usr/share/man/cat3/fbsrunrtc.3x.z
/usr/share/man/cat3/fbsschedself.3f.z
/usr/share/man/cat3/fbsschedself.3x.z
/usr/share/man/cat3/fbssetrtc.3f.z
/usr/share/man/cat3/fbssetrtc.3x.z
/usr/share/man/cat3/pgmquery.3f.z
/usr/share/man/cat3/pgmquery.3x.z
/usr/share/man/cat3/pgmremove.3f.z
/usr/share/man/cat3/pgmremove.3x.z
/usr/share/man/cat3/pgmreschedule.3f.z
/usr/share/man/cat3/pgmreschedule.3x.z
/usr/share/man/cat3/pgmschedule.3f.z
/usr/share/man/cat3/pgmschedule.3x.z
/usr/share/man/cat3/pgmstat.3f.z
/usr/share/man/cat3/pmclrpgm.3f.z
/usr/share/man/cat3/pmclrpgm.3x.z
/usr/share/man/cat3/pmclrtable.3f.z
/usr/share/man/cat3/pmclrtable.3x.z
/usr/share/man/cat3/pmmonitor.3f.z
/usr/share/man/cat3/pmmonitor.3x.z
/usr/share/man/cat3/pmprogram.3f.z
/usr/share/man/cat3/pmprogram.3x.z
/usr/share/man/cat3/pmqrycpu.3f.z
/usr/share/man/cat3/pmqrycpu.3x.z
/usr/share/man/cat3/pmqrylist.3f.z
/usr/share/man/cat3/pmqrylist.3x.z
/usr/share/man/cat3/pmqrypgm.3f.z
/usr/share/man/cat3/pmqrypgm.3x.z
/usr/share/man/cat3/sched_fbsqry.3x.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3f.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3x.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3f.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3x.z
/usr/share/man/cat3/sched_pgmadd.3x.z
/usr/share/man/cat3/sched_pgmqry.3x.z
/usr/share/man/cat3/sched_pgmresched.3x.z
/usr/share/man/cat3/schedfbsqry.3f.z
/usr/share/man/cat3/schedpgmadd.3f.z
/usr/share/man/cat3/schedpgmqry.3f.z
/usr/share/man/cat3/schedpgmresched.3f.z
/usr/share/man/cat4/rdevfs.4.z
/usr/share/man/man1/rtcp.1
/usr/share/man/man3/fbs_register_cluster_device.3f
/usr/share/man/man3/fbs_register_cluster_device.3x
/usr/share/man/man3/fbs_unregister_cluster_device.3f
/usr/share/man/man3/fbs_unregister_cluster_device.3x
/usr/share/man/man3/fbsaccess.3f
/usr/share/man/man3/fbsaccess.3x
/usr/share/man/man3/fbscycle.3f
/usr/share/man/man3/fbscycle.3x
/usr/share/man/man3/fbsdetach.3f
/usr/share/man/man3/fbsdetach.3x
/usr/share/man/man3/fbsgetrtc.3f
/usr/share/man/man3/fbsgetrtc.3x
/usr/share/man/man3/fbsinfo.3f
/usr/share/man/man3/fbsinfo.3x
/usr/share/man/man3/fbsinfo_cluster.3f
/usr/share/man/man3/fbsinfo_cluster.3x
/usr/share/man/man3/fbsintrpt.3f
/usr/share/man/man3/fbsintrpt.3x
/usr/share/man/man3/fbsquery.3f
/usr/share/man/man3/fbsquery.3x
/usr/share/man/man3/fbsremove.3f
/usr/share/man/man3/fbsremove.3x
/usr/share/man/man3/fbsresume.3f
/usr/share/man/man3/fbsresume.3x
/usr/share/man/man3/fbsrunrtc.3f
/usr/share/man/man3/fbsrunrtc.3x
/usr/share/man/man3/fbsschedself.3f
/usr/share/man/man3/fbsschedself.3x
/usr/share/man/man3/fbssetrtc.3f
/usr/share/man/man3/fbssetrtc.3x
/usr/share/man/man3/pgmquery.3f
/usr/share/man/man3/pgmquery.3x
/usr/share/man/man3/pgmremove.3f
/usr/share/man/man3/pgmremove.3x
/usr/share/man/man3/pgmreschedule.3f
/usr/share/man/man3/pgmreschedule.3x
/usr/share/man/man3/pgmschedule.3f
/usr/share/man/man3/pgmschedule.3x
/usr/share/man/man3/pgmstat.3f
/usr/share/man/man3/pmclrpgm.3f
/usr/share/man/man3/pmclrpgm.3x
/usr/share/man/man3/pmclrtable.3f
/usr/share/man/man3/pmclrtable.3x
/usr/share/man/man3/pmmonitor.3f
/usr/share/man/man3/pmmonitor.3x
/usr/share/man/man3/pmprogram.3f
/usr/share/man/man3/pmprogram.3x
/usr/share/man/man3/pmqrycpu.3f
/usr/share/man/man3/pmqrycpu.3x
/usr/share/man/man3/pmqrylist.3f
/usr/share/man/man3/pmqrylist.3x
/usr/share/man/man3/pmqrypgm.3f
/usr/share/man/man3/pmqrypgm.3x
/usr/share/man/man3/sched_fbsqry.3x
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3f
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3x
/usr/share/man/man3/sched_pgm_soft_overrun_query.3f
/usr/share/man/man3/sched_pgm_soft_overrun_query.3x
/usr/share/man/man3/sched_pgmadd.3x
/usr/share/man/man3/sched_pgmqry.3x
/usr/share/man/man3/sched_pgmresched.3x
/usr/share/man/man3/schedfbsqry.3f
/usr/share/man/man3/schedpgmadd.3f
/usr/share/man/man3/schedpgmqry.3f
/usr/share/man/man3/schedpgmresched.3f
/usr/share/man/man4/rdevfs.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-001">fbsman-001
Date Issued: 09/15/1999 12:58:45
Software Package: fbsman pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001, man-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbsman package release updates
##############################################################################
Problem Description:
1. The FBS man pages had several incorrect references.
Problem Resolution:
1. Corrected man pages.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/rtcp.1.z
/usr/share/man/cat3/fbsaccess.3f.z
/usr/share/man/cat3/fbsaccess.3x.z
/usr/share/man/cat3/fbscycle.3f.z
/usr/share/man/cat3/fbscycle.3x.z
/usr/share/man/cat3/fbsdetach.3f.z
/usr/share/man/cat3/fbsdetach.3x.z
/usr/share/man/cat3/fbsgetrtc.3f.z
/usr/share/man/cat3/fbsgetrtc.3x.z
/usr/share/man/cat3/fbsinfo.3f.z
/usr/share/man/cat3/fbsinfo.3x.z
/usr/share/man/cat3/fbsintrpt.3f.z
/usr/share/man/cat3/fbsintrpt.3x.z
/usr/share/man/cat3/fbsquery.3f.z
/usr/share/man/cat3/fbsquery.3x.z
/usr/share/man/cat3/fbsremove.3f.z
/usr/share/man/cat3/fbsremove.3x.z
/usr/share/man/cat3/fbsresume.3f.z
/usr/share/man/cat3/fbsresume.3x.z
/usr/share/man/cat3/fbsrunrtc.3f.z
/usr/share/man/cat3/fbsrunrtc.3x.z
/usr/share/man/cat3/fbsschedself.3f.z
/usr/share/man/cat3/fbsschedself.3x.z
/usr/share/man/cat3/fbssetrtc.3f.z
/usr/share/man/cat3/fbssetrtc.3x.z
/usr/share/man/cat3/pgmquery.3f.z
/usr/share/man/cat3/pgmquery.3x.z
/usr/share/man/cat3/pgmremove.3f.z
/usr/share/man/cat3/pgmremove.3x.z
/usr/share/man/cat3/pgmreschedule.3f.z
/usr/share/man/cat3/pgmreschedule.3x.z
/usr/share/man/cat3/pgmschedule.3f.z
/usr/share/man/cat3/pgmschedule.3x.z
/usr/share/man/cat3/pgmstat.3f.z
/usr/share/man/cat3/pmclrpgm.3f.z
/usr/share/man/cat3/pmclrpgm.3x.z
/usr/share/man/cat3/pmclrtable.3f.z
/usr/share/man/cat3/pmclrtable.3x.z
/usr/share/man/cat3/pmmonitor.3f.z
/usr/share/man/cat3/pmmonitor.3x.z
/usr/share/man/cat3/pmprogram.3f.z
/usr/share/man/cat3/pmprogram.3x.z
/usr/share/man/cat3/pmqrycpu.3f.z
/usr/share/man/cat3/pmqrycpu.3x.z
/usr/share/man/cat3/pmqrylist.3f.z
/usr/share/man/cat3/pmqrylist.3x.z
/usr/share/man/cat3/pmqrypgm.3f.z
/usr/share/man/cat3/pmqrypgm.3x.z
/usr/share/man/cat3/sched_fbsqry.3x.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3f.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3x.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3f.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3x.z
/usr/share/man/cat3/sched_pgmadd.3x.z
/usr/share/man/cat3/sched_pgmqry.3x.z
/usr/share/man/cat3/sched_pgmresched.3x.z
/usr/share/man/cat3/schedfbsqry.3f.z
/usr/share/man/cat3/schedpgmadd.3f.z
/usr/share/man/cat3/schedpgmqry.3f.z
/usr/share/man/cat3/schedpgmresched.3f.z
/usr/share/man/man1/rtcp.1
/usr/share/man/man3/fbsaccess.3f
/usr/share/man/man3/fbsaccess.3x
/usr/share/man/man3/fbscycle.3f
/usr/share/man/man3/fbscycle.3x
/usr/share/man/man3/fbsdetach.3f
/usr/share/man/man3/fbsdetach.3x
/usr/share/man/man3/fbsgetrtc.3f
/usr/share/man/man3/fbsgetrtc.3x
/usr/share/man/man3/fbsinfo.3f
/usr/share/man/man3/fbsinfo.3x
/usr/share/man/man3/fbsintrpt.3f
/usr/share/man/man3/fbsintrpt.3x
/usr/share/man/man3/fbsquery.3f
/usr/share/man/man3/fbsquery.3x
/usr/share/man/man3/fbsremove.3f
/usr/share/man/man3/fbsremove.3x
/usr/share/man/man3/fbsresume.3f
/usr/share/man/man3/fbsresume.3x
/usr/share/man/man3/fbsrunrtc.3f
/usr/share/man/man3/fbsrunrtc.3x
/usr/share/man/man3/fbsschedself.3f
/usr/share/man/man3/fbsschedself.3x
/usr/share/man/man3/fbssetrtc.3f
/usr/share/man/man3/fbssetrtc.3x
/usr/share/man/man3/pgmquery.3f
/usr/share/man/man3/pgmquery.3x
/usr/share/man/man3/pgmremove.3f
/usr/share/man/man3/pgmremove.3x
/usr/share/man/man3/pgmreschedule.3f
/usr/share/man/man3/pgmreschedule.3x
/usr/share/man/man3/pgmschedule.3f
/usr/share/man/man3/pgmschedule.3x
/usr/share/man/man3/pgmstat.3f
/usr/share/man/man3/pmclrpgm.3f
/usr/share/man/man3/pmclrpgm.3x
/usr/share/man/man3/pmclrtable.3f
/usr/share/man/man3/pmclrtable.3x
/usr/share/man/man3/pmmonitor.3f
/usr/share/man/man3/pmmonitor.3x
/usr/share/man/man3/pmprogram.3f
/usr/share/man/man3/pmprogram.3x
/usr/share/man/man3/pmqrycpu.3f
/usr/share/man/man3/pmqrycpu.3x
/usr/share/man/man3/pmqrylist.3f
/usr/share/man/man3/pmqrylist.3x
/usr/share/man/man3/pmqrypgm.3f
/usr/share/man/man3/pmqrypgm.3x
/usr/share/man/man3/sched_fbsqry.3x
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3f
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3x
/usr/share/man/man3/sched_pgm_soft_overrun_query.3f
/usr/share/man/man3/sched_pgm_soft_overrun_query.3x
/usr/share/man/man3/sched_pgmadd.3x
/usr/share/man/man3/sched_pgmqry.3x
/usr/share/man/man3/sched_pgmresched.3x
/usr/share/man/man3/schedfbsqry.3f
/usr/share/man/man3/schedpgmadd.3f
/usr/share/man/man3/schedpgmqry.3f
/usr/share/man/man3/schedpgmresched.3f
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-002">fbsman-002
Date Issued: 02/02/2000 13:43:20
Software Package: fbsman pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001, fbs-002, inet-002, librt-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbsman package release updates
##############################################################################
Problem Description:
1. Man page updates required for RCIM Coupled FBS support.
Problem Resolution:
1. Updated man pages for RCIM Coupled FBS support.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/fbsdaemon.1m.z
/usr/share/man/cat1/rtcp.1.z
/usr/share/man/cat3/fbsaccess.3f.z
/usr/share/man/cat3/fbsaccess.3x.z
/usr/share/man/cat3/fbsattach.3f.z
/usr/share/man/cat3/fbsattach.3x.z
/usr/share/man/cat3/fbscycle.3f.z
/usr/share/man/cat3/fbscycle.3x.z
/usr/share/man/cat3/fbsdetach.3f.z
/usr/share/man/cat3/fbsdetach.3x.z
/usr/share/man/cat3/fbsgetrtc.3f.z
/usr/share/man/cat3/fbsgetrtc.3x.z
/usr/share/man/cat3/fbs_info_rdev.3f.z
/usr/share/man/cat3/fbs_info_rdev.3x.z
/usr/share/man/cat3/fbs_reg_rdev.3f.z
/usr/share/man/cat3/fbs_reg_rdev.3x.z
/usr/share/man/cat3/fbs_register_cluster_device.3f.z
/usr/share/man/cat3/fbs_register_cluster_device.3x.z
/usr/share/man/cat3/fbs_unreg_rdev.3f.z
/usr/share/man/cat3/fbs_unreg_rdev.3x.z
/usr/share/man/cat3/fbs_unregister_cluster_device.3f.z
/usr/share/man/cat3/fbs_unregister_cluster_device.3x.z
/usr/share/man/cat3/fbsinfo.3f.z
/usr/share/man/cat3/fbsinfo.3x.z
/usr/share/man/cat3/fbsinfo_cluster.3f.z
/usr/share/man/cat3/fbsinfo_cluster.3x.z
/usr/share/man/cat3/fbsintrpt.3f.z
/usr/share/man/cat3/fbsintrpt.3x.z
/usr/share/man/cat3/fbsquery.3f.z
/usr/share/man/cat3/fbsquery.3x.z
/usr/share/man/cat3/fbsremove.3f.z
/usr/share/man/cat3/fbsremove.3x.z
/usr/share/man/cat3/fbsresume.3f.z
/usr/share/man/cat3/fbsresume.3x.z
/usr/share/man/cat3/fbsrunrtc.3f.z
/usr/share/man/cat3/fbsrunrtc.3x.z
/usr/share/man/cat3/fbsschedself.3f.z
/usr/share/man/cat3/fbsschedself.3x.z
/usr/share/man/cat3/fbssetrtc.3f.z
/usr/share/man/cat3/fbssetrtc.3x.z
/usr/share/man/cat3/pgmquery.3f.z
/usr/share/man/cat3/pgmquery.3x.z
/usr/share/man/cat3/pgmremove.3f.z
/usr/share/man/cat3/pgmremove.3x.z
/usr/share/man/cat3/pgmreschedule.3f.z
/usr/share/man/cat3/pgmreschedule.3x.z
/usr/share/man/cat3/pgmschedule.3f.z
/usr/share/man/cat3/pgmschedule.3x.z
/usr/share/man/cat3/pmclrpgm.3f.z
/usr/share/man/cat3/pmclrpgm.3x.z
/usr/share/man/cat3/pmclrtable.3f.z
/usr/share/man/cat3/pmclrtable.3x.z
/usr/share/man/cat3/pmmonitor.3f.z
/usr/share/man/cat3/pmmonitor.3x.z
/usr/share/man/cat3/pmprogram.3f.z
/usr/share/man/cat3/pmprogram.3x.z
/usr/share/man/cat3/pmqrycpu.3f.z
/usr/share/man/cat3/pmqrycpu.3x.z
/usr/share/man/cat3/pmqrylist.3f.z
/usr/share/man/cat3/pmqrylist.3x.z
/usr/share/man/cat3/pmqrypgm.3f.z
/usr/share/man/cat3/pmqrypgm.3x.z
/usr/share/man/cat3/sched_fbsqry.3x.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3f.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3x.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3f.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3x.z
/usr/share/man/cat3/sched_pgmadd.3x.z
/usr/share/man/cat3/sched_pgmqry.3x.z
/usr/share/man/cat3/sched_pgmresched.3x.z
/usr/share/man/cat4/rdevfs.4.z
/usr/share/man/man1/fbsdaemon.1m
/usr/share/man/man1/rtcp.1
/usr/share/man/man3/fbsaccess.3f
/usr/share/man/man3/fbsaccess.3x
/usr/share/man/man3/fbsattach.3f
/usr/share/man/man3/fbsattach.3x
/usr/share/man/man3/fbscycle.3f
/usr/share/man/man3/fbscycle.3x
/usr/share/man/man3/fbsdetach.3f
/usr/share/man/man3/fbsdetach.3x
/usr/share/man/man3/fbsgetrtc.3f
/usr/share/man/man3/fbsgetrtc.3x
/usr/share/man/man3/fbs_info_rdev.3f
/usr/share/man/man3/fbs_info_rdev.3x
/usr/share/man/man3/fbs_reg_rdev.3f
/usr/share/man/man3/fbs_reg_rdev.3x
/usr/share/man/man3/fbs_register_cluster_device.3f
/usr/share/man/man3/fbs_register_cluster_device.3x
/usr/share/man/man3/fbs_unreg_rdev.3f
/usr/share/man/man3/fbs_unreg_rdev.3x
/usr/share/man/man3/fbs_unregister_cluster_device.3f
/usr/share/man/man3/fbs_unregister_cluster_device.3x
/usr/share/man/man3/fbsinfo.3f
/usr/share/man/man3/fbsinfo.3x
/usr/share/man/man3/fbsinfo_cluster.3f
/usr/share/man/man3/fbsinfo_cluster.3x
/usr/share/man/man3/fbsintrpt.3f
/usr/share/man/man3/fbsintrpt.3x
/usr/share/man/man3/fbsquery.3f
/usr/share/man/man3/fbsquery.3x
/usr/share/man/man3/fbsremove.3f
/usr/share/man/man3/fbsremove.3x
/usr/share/man/man3/fbsresume.3f
/usr/share/man/man3/fbsresume.3x
/usr/share/man/man3/fbsrunrtc.3f
/usr/share/man/man3/fbsrunrtc.3x
/usr/share/man/man3/fbsschedself.3f
/usr/share/man/man3/fbsschedself.3x
/usr/share/man/man3/fbssetrtc.3f
/usr/share/man/man3/fbssetrtc.3x
/usr/share/man/man3/pgmquery.3f
/usr/share/man/man3/pgmquery.3x
/usr/share/man/man3/pgmremove.3f
/usr/share/man/man3/pgmremove.3x
/usr/share/man/man3/pgmreschedule.3f
/usr/share/man/man3/pgmreschedule.3x
/usr/share/man/man3/pgmschedule.3f
/usr/share/man/man3/pgmschedule.3x
/usr/share/man/man3/pmclrpgm.3f
/usr/share/man/man3/pmclrpgm.3x
/usr/share/man/man3/pmclrtable.3f
/usr/share/man/man3/pmclrtable.3x
/usr/share/man/man3/pmmonitor.3f
/usr/share/man/man3/pmmonitor.3x
/usr/share/man/man3/pmprogram.3f
/usr/share/man/man3/pmprogram.3x
/usr/share/man/man3/pmqrycpu.3f
/usr/share/man/man3/pmqrycpu.3x
/usr/share/man/man3/pmqrylist.3f
/usr/share/man/man3/pmqrylist.3x
/usr/share/man/man3/pmqrypgm.3f
/usr/share/man/man3/pmqrypgm.3x
/usr/share/man/man3/sched_fbsqry.3x
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3f
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3x
/usr/share/man/man3/sched_pgm_soft_overrun_query.3f
/usr/share/man/man3/sched_pgm_soft_overrun_query.3x
/usr/share/man/man3/sched_pgmadd.3x
/usr/share/man/man3/sched_pgmqry.3x
/usr/share/man/man3/sched_pgmresched.3x
/usr/share/man/man4/rdevfs.4
Special Conditions for Installation:
The following object(s) will also be replaced by man-002">fbsman-002:
* /usr/share/man/cat1/fbsd.1m.z
* /usr/share/man/cat3/fbsinfo_rdev.3f.z
* /usr/share/man/cat3/fbsinfo_rdev.3x.z
* /usr/share/man/cat3/fbs_register_rdev.3f.z
* /usr/share/man/cat3/fbs_register_rdev.3x.z
* /usr/share/man/cat3/fbs_unregister_rdev.3f.z
* /usr/share/man/cat3/fbs_unregister_rdev.3x.z
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-002">fbsman-002
Date Issued: 02/02/2000 11:39:54
Software Package: fbsman pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-001, crosslibs-001, fbs-002, inet-002, librt-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fbsman package release updates
##############################################################################
Problem Description:
1. Man page updates required for RCIM Coupled FBS support.
Problem Resolution:
1. Updated man pages for RCIM Coupled FBS support.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/rtcp.1.z
/usr/share/man/cat3/fbsaccess.3f.z
/usr/share/man/cat3/fbsaccess.3x.z
/usr/share/man/cat3/fbsattach.3f.z
/usr/share/man/cat3/fbsattach.3x.z
/usr/share/man/cat3/fbscycle.3f.z
/usr/share/man/cat3/fbscycle.3x.z
/usr/share/man/cat3/fbsdetach.3f.z
/usr/share/man/cat3/fbsdetach.3x.z
/usr/share/man/cat3/fbsgetrtc.3f.z
/usr/share/man/cat3/fbsgetrtc.3x.z
/usr/share/man/cat3/fbsinfo.3f.z
/usr/share/man/cat3/fbsinfo.3x.z
/usr/share/man/cat3/fbsintrpt.3f.z
/usr/share/man/cat3/fbsintrpt.3x.z
/usr/share/man/cat3/fbsquery.3f.z
/usr/share/man/cat3/fbsquery.3x.z
/usr/share/man/cat3/fbsremove.3f.z
/usr/share/man/cat3/fbsremove.3x.z
/usr/share/man/cat3/fbsresume.3f.z
/usr/share/man/cat3/fbsresume.3x.z
/usr/share/man/cat3/fbsrunrtc.3f.z
/usr/share/man/cat3/fbsrunrtc.3x.z
/usr/share/man/cat3/fbsschedself.3f.z
/usr/share/man/cat3/fbsschedself.3x.z
/usr/share/man/cat3/fbssetrtc.3f.z
/usr/share/man/cat3/fbssetrtc.3x.z
/usr/share/man/cat3/pgmquery.3f.z
/usr/share/man/cat3/pgmquery.3x.z
/usr/share/man/cat3/pgmremove.3f.z
/usr/share/man/cat3/pgmremove.3x.z
/usr/share/man/cat3/pgmreschedule.3f.z
/usr/share/man/cat3/pgmreschedule.3x.z
/usr/share/man/cat3/pgmschedule.3f.z
/usr/share/man/cat3/pgmschedule.3x.z
/usr/share/man/cat3/pmclrpgm.3f.z
/usr/share/man/cat3/pmclrpgm.3x.z
/usr/share/man/cat3/pmclrtable.3f.z
/usr/share/man/cat3/pmclrtable.3x.z
/usr/share/man/cat3/pmmonitor.3f.z
/usr/share/man/cat3/pmmonitor.3x.z
/usr/share/man/cat3/pmprogram.3f.z
/usr/share/man/cat3/pmprogram.3x.z
/usr/share/man/cat3/pmqrycpu.3f.z
/usr/share/man/cat3/pmqrycpu.3x.z
/usr/share/man/cat3/pmqrylist.3f.z
/usr/share/man/cat3/pmqrylist.3x.z
/usr/share/man/cat3/pmqrypgm.3f.z
/usr/share/man/cat3/pmqrypgm.3x.z
/usr/share/man/cat3/sched_fbsqry.3x.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3f.z
/usr/share/man/cat3/sched_pgm_set_soft_overrun_limit.3x.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3f.z
/usr/share/man/cat3/sched_pgm_soft_overrun_query.3x.z
/usr/share/man/cat3/sched_pgmadd.3x.z
/usr/share/man/cat3/sched_pgmqry.3x.z
/usr/share/man/cat3/sched_pgmresched.3x.z
/usr/share/man/man1/rtcp.1
/usr/share/man/man3/fbsaccess.3f
/usr/share/man/man3/fbsaccess.3x
/usr/share/man/man3/fbsattach.3f
/usr/share/man/man3/fbsattach.3x
/usr/share/man/man3/fbscycle.3f
/usr/share/man/man3/fbscycle.3x
/usr/share/man/man3/fbsdetach.3f
/usr/share/man/man3/fbsdetach.3x
/usr/share/man/man3/fbsgetrtc.3f
/usr/share/man/man3/fbsgetrtc.3x
/usr/share/man/man3/fbsinfo.3f
/usr/share/man/man3/fbsinfo.3x
/usr/share/man/man3/fbsintrpt.3f
/usr/share/man/man3/fbsintrpt.3x
/usr/share/man/man3/fbsquery.3f
/usr/share/man/man3/fbsquery.3x
/usr/share/man/man3/fbsremove.3f
/usr/share/man/man3/fbsremove.3x
/usr/share/man/man3/fbsresume.3f
/usr/share/man/man3/fbsresume.3x
/usr/share/man/man3/fbsrunrtc.3f
/usr/share/man/man3/fbsrunrtc.3x
/usr/share/man/man3/fbsschedself.3f
/usr/share/man/man3/fbsschedself.3x
/usr/share/man/man3/fbssetrtc.3f
/usr/share/man/man3/fbssetrtc.3x
/usr/share/man/man3/pgmquery.3f
/usr/share/man/man3/pgmquery.3x
/usr/share/man/man3/pgmremove.3f
/usr/share/man/man3/pgmremove.3x
/usr/share/man/man3/pgmreschedule.3f
/usr/share/man/man3/pgmreschedule.3x
/usr/share/man/man3/pgmschedule.3f
/usr/share/man/man3/pgmschedule.3x
/usr/share/man/man3/pmclrpgm.3f
/usr/share/man/man3/pmclrpgm.3x
/usr/share/man/man3/pmclrtable.3f
/usr/share/man/man3/pmclrtable.3x
/usr/share/man/man3/pmmonitor.3f
/usr/share/man/man3/pmmonitor.3x
/usr/share/man/man3/pmprogram.3f
/usr/share/man/man3/pmprogram.3x
/usr/share/man/man3/pmqrycpu.3f
/usr/share/man/man3/pmqrycpu.3x
/usr/share/man/man3/pmqrylist.3f
/usr/share/man/man3/pmqrylist.3x
/usr/share/man/man3/pmqrypgm.3f
/usr/share/man/man3/pmqrypgm.3x
/usr/share/man/man3/sched_fbsqry.3x
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3f
/usr/share/man/man3/sched_pgm_set_soft_overrun_limit.3x
/usr/share/man/man3/sched_pgm_soft_overrun_query.3f
/usr/share/man/man3/sched_pgm_soft_overrun_query.3x
/usr/share/man/man3/sched_pgmadd.3x
/usr/share/man/man3/sched_pgmqry.3x
/usr/share/man/man3/sched_pgmresched.3x
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-003">fbsman-003
Date Issued: 02/09/2004 12:47:48
Software Package: fbsman pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 fbsman package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
Enhancements:
P1: SAR 1487: Added a description for the fbsd(1m) -D option.
Object(s) To Be Replaced:
/usr/share/man/cat1/fbsdaemon.1m.z
/usr/share/man/man1/fbsdaemon.1m
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fibre-001
Date Issued: 09/15/1999 11:04:54
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: PowerStack II
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates
##############################################################################
Problem Description:
1. The fibre driver does not allow access to hot swap drives until a
reboot is performed.
Problem Resolution:
1. The fibre driver did not employ the correct disk startup procedure.
Now the driver determines the new (hot swapped) fibre drive is not
ready, and then spins it up allowing further access to the device.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fibre/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: fibre-001
Date Issued: 03/28/2001 19:10:59
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates.
#############################################################################
Problem Description:
1. Pre-fetch had been disabled in the Fibre Driver causing poor write
performance.
2. Multiple-CPU systems (for VOD) would frequently fail to locate all
fibre drives properly.
Problem Resolution:
1. Change fibre chip initialization to enable prefetch.
2. Corrected fibre driver coding errors to handle loop up/down
conditions properly during system boot.
Object(s) To Be Replaced:
/etc/conf/pack.d/fibre/Driver.o
/usr/include/sys/fibre.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fibre-002
Date Issued: 05/30/2000 13:06:00
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: PowerStack II
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates
##############################################################################
Problem Description:
1. A new fibre controller was not supported by the fibre driver.
2. The driver did not support hot-swapping drives properly.
3. Large fibre backplane systems took too long to boot.
4. Utility needed to program new fibre channel backplane.
Problem Resolution:
1. Added support for new Interphase 5526-019 controller.
2. Add ioctl() functions and other fixes to support removal and
insertion of drives.
3. Reduce interdrive wait time from 5 seconds to 1 second when scanning
drives.
4. Provide new bpconfig(1M) utility to program Concurrent Fibre Channel
Backplane.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/fibre
/etc/conf/pack.d/fibre/Driver.o
/sbin/bpconfig
/usr/include/sys/fibre.h
/usr/share/man/cat1/bpconfig.1m.z
/usr/share/man/man1/bpconfig.1m
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fibre-003
Date Issued: 10/05/2000 13:53:04
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: PowerStack II
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003, inet-004,
ip-003, man-003, pg-003
Related SARs: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates
##############################################################################
Problem Description:
1. The fibconfig(1M) utility was changed as a part of patch fibre-002
to include additional error reporting and other changes to match
corresponding driver changes. The object was not included in patch
fibre-002, making this utility inoperative.
Problem Resolution:
1. Install the updated fibconfig(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/sbin/fibconfig
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: fibre-004
Date Issued: 03/29/2001 12:01:02
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates.
#############################################################################
Problem Description:
1. Pre-fetch had been disabled in the Fibre Driver causing poor write
performance.
2. Multiple-CPU systems (for VOD) would frequently fail to locate all
fibre drives properly.
Problem Resolution:
1. Change fibre chip initialization to enable prefetch.
2. Corrected fibre driver coding errors to handle loop up/down
conditions properly during system boot.
Object(s) To Be Replaced:
/etc/conf/pack.d/fibre/Driver.o
/usr/include/sys/fibre.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: fibre-005
Date Issued: 06/11/2001 08:07:58
Software Package: fibre pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II/III, Motorola MCP75
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 fibre package release updates
##############################################################################
Problem Description:
1. Driver needed a sequential I/O que processing mechanism and
the ability to handle lists of Phys DMA addresses.
(support new partial sector capability).
Problem Resolution:
1. Added needed functions.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/fibre/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: gf-001
Date Issued: 07/10/2002 14:51:02
Software Package: gf pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 gf package release updates
##############################################################################
Enhancements:
1. Cleaned up some WARNING messages generated during the build of
the following commands.
Object(s) To Be Replaced:
/usr/bin/dosclear
/usr/bin/dosdel
/usr/bin/dosdir
/usr/bin/dosformat
/usr/bin/dosget
/usr/bin/dosinit
/usr/bin/dosmkdir
/usr/bin/dosput
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: gf-001
Date Issued: 07/24/2002 14:43:23
Software Package: gf pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 gf package release updates
###############################################################################
Enhancements:
1. Cleaned up some WARNING messages generated during the build of
the following commands.
Object(s) To Be Replaced:
/usr/bin/dosclear
/usr/bin/dosdel
/usr/bin/dosdir
/usr/bin/dosformat
/usr/bin/dosget
/usr/bin/dosinit
/usr/bin/dosmkdir
/usr/bin/dosput
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: gpib-001
Date Issued: 01/02/2002 14:23:02
Software Package: gpib pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 gpib package release updates
##############################################################################
Problem Description:
1. Support for the Night Hawk series missing. Also resolved some known
bugs on the Power Hawk series of PowerMAX OS 5.1.
Problem Resolution:
1. The changes made to this package were to provide support for the
Night Hawk series. Additionally, several bugs with respect to SRQ
handling were fixed and the following new ioctls were also
introduced:
GPIB_SET_READTOUT
GPIB_READ_REGS
GPIB_REGISTER_SIG
GPIB_SRQ_INT
Additionally a new gpib_srq_int.c test was added to validate SRQ
handling. Also provided a fix in the driver initialization to get
rid of spureous interrupt messages when doing an "init 6". This
will also hopefully get rid of the CPU lockup in the gpib-2-gpib
configuration.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/gpib/Driver.o
/usr/include/sys/gpib.h
/usr/src/drivers/gpib/test/Makefile
/usr/src/drivers/gpib/test/README
/usr/src/drivers/gpib/test/gpib_2_gpib.c
/usr/src/drivers/gpib/test/gpib_iotech.c
/usr/src/drivers/gpib/test/gpib_iotech.h
/usr/src/drivers/gpib/test/gpib_iotech_two.c
/usr/src/drivers/gpib/test/gpib_read.c
/usr/src/drivers/gpib/test/gpib_srq_int.c
/usr/src/drivers/gpib/test/gpib_write.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: hsde-001
Date Issued: 02/02/2000 14:03:11
Software Package: hsde pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: none
Related SARs: #104
Brief Description:
PowerMAX OS 4.3 hsde package release updates
##############################################################################
Problem Description:
1. #104: The HSDE would not initialize during boot on a Power Hawk
620/640 system because there was not enough delay in the code for
the probe of the controller to complete.
Problem Resolution:
1. Adjusted the drv_usecwait() delay and the while loop iterations in
H3300_RESET_CMD macro defined in <sys/h3300_channel.h>.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/hsde/Driver.o
/usr/include/sys/h3300_channel.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: hsde-001
Date Issued: 02/02/2000 12:54:05
Software Package: hsde pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: #104
Brief Description:
PowerMAX OS 4.3 hsde package release updates
##############################################################################
Problem Description:
1. #104: The HSDE would not initialize during boot on a Power Hawk
620/640 system because there was not enough delay in the code for
the probe of the controller to complete.
Problem Resolution:
1. Adjusted the drv_usecwait() delay and the while loop iterations in
H3300_RESET_CMD macro defined in <sys/h3300_channel.h>.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/hsde/Driver.o
/usr/include/sys/h3300_channel.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: hsde-002
Date Issued: 05/08/2003 10:37:50
Software Package: hsde pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 hsde package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/hsde
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: hsde-002
Date Issued: 05/08/2003 10:37:50
Software Package: hsde pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 hsde package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/hsde
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ide-001
Date Issued: 09/15/1999 11:05:21
Software Package: ide pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 ide package release updates
##############################################################################
Problem Description:
1. The IDE disk driver would hang on reads/writes which were not exact
multiples of the sector size (512 bytes).
Problem Resolution:
1. Correct the logic in the IDE driver which handles blocks that are not
a multiple of the 512-byte sector size.
Enhancements:
1. Added PIO mode to the IDE disk driver. This enhancement was made to
support SanDisk Flash IDE disks, which do not operate in DMA mode.
Object(s) To Be Replaced:
/etc/conf/pack.d/ide/Driver.o
/usr/include/sys/ide.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ide-002
Date Issued: 06/11/2001 08:08:58
Software Package: ide pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II/III, Motorola MCP750
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 ide package release updates
##############################################################################
Problem Description:
1. Driver needed the ability to handle lists of Physcal DMA addresses.
(support new partial sector capability).
Problem Resolution:
1. added needed functions.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ide/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: ide-003
Date Issued: 08/22/2001 11:50:33
Software Package: ide pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 ide package release updates
#############################################################################
Problem Description:
1. Under heavy ide disk usage, two possible problem in the ide
kernel driver may occur: 1. A logical unit busy list queue
can become corrupted, resulting in a kernel panic due to a
memory reference to location 0x4 as we were trying to place
the just completed mcb back onto the controller's freelist
of mcbs.
2. The last I/O operation would never complete, and it would
appear as if the IDE completion interrupt was never received
for this last I/O request. All additional queued and future
I/O requests would also never complete.
These problems were a result of race conditions between the IDE
driver's interrupt routine processing and the issuing of the next
IDE I/O request on the other CPU.
Problem Resolution:
1. The fix for both of these problems was to move more of the IDE
interrupt routine processing under the protection of the IDE
controller's kernel spin lock. This prevents the other CPU
from simultaneously issuing the next I/O request while the I/O
completion activities for the previous I/O request are still
being processed in the interrupt routine.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/ide/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ie-001
Date Issued: 02/02/2000 12:54:43
Software Package: ie pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800
Related Patches: base-002, cnd-002, dec-001, egl-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 ie package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ie/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ie-002
Date Issued: 05/30/2000 13:49:18
Software Package: ie pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800
Related Patches: cnd-003, dec-002, egl-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 ie package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ie/Driver.o
/etc/conf/pack.d/ie/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ie-003
Date Issued: 10/05/2000 14:06:56
Software Package: ie pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, inet-004, man-003,
pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 ie package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
Problem Resolution:
1. A check for the TX_QUEUED flag was added to the common DLPI code in
DLunitdata_req() that checks for TX_BUSY before it calls
DLxmit_packet(). Without the check for the TX_QUEUED flag, there was
a race condition with the driver's queued packet processing routine
that is called by the driver's transmit interrupt routine. New
packets were being transmitted to the board ahead of packets that
were already waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the bd_lock
after TX_BUSY was reset, but before the call to the driver's queued
packet processing routine.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ie/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: ie-004
Date Issued: 03/18/2002 11:30:38
Software Package: ie pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 ie package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/ie/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: ie-005
Date Issued: 01/07/2004 16:01:17
Software Package: ie pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 ie package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ie/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-001
Date Issued: 09/15/1999 11:05:54
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001, man-001
Related SARs: #29
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. The route(1M) command allows a netmask to be specified for an entry
in the routing table. However, this netmask is ignored by the IP
code, and the netmask is deduced from the class of the destination
address. This may not be what is required, and it can prevent routes
to network from being set up as required.
2. A timing window existed that could result in system panic. Generally
requires four or more CPUs and heavy network activity.
Problem Resolution:
1. The route command and the IP code are changed so that the netmask is
communicated in the ioctl associated with the command. The routing
daemon in.routed is also changed, to initialize the field in the
ioctl.
2. Closed timing window.
Enhancements:
1. #29: Enhanced the inetd(1M) daemon so that the parameters:
TOOMANY, CNT_INTVL, and RETRYTIME can be tuned instead of being
hardcoded. This tunable parameter capability for the inetd(1M)
daemon was implemented through the /etc/default mechanism. The three
parameters and their default values were placed in the file
"/etc/default/inetd". This file is read when inetd(1M) is invoked
and the specified values are used. These values can be modified
within a specific range. See inetd(1M) for further details.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/mi/Driver.o
/etc/default/inetd
/usr/sbin/in.routed
/usr/sbin/inetd
/usr/sbin/route
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-001
Date Issued: 09/15/1999 13:11:50
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001, man-001
Related SARs: #29
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. The route(1M) command allows a netmask to be specified for an entry
in the routing table. However, this netmask is ignored by the IP
code, and the netmask is deduced from the class of the destination
address. This may not be what is required, and it can prevent routes
to network from being set up as required.
2. A timing window existed that could result in system panic. Generally
requires four or more CPUs and heavy network activity.
Problem Resolution:
1. The route command and the IP code are changed so that the netmask is
communicated in the ioctl associated with the command. The routing
daemon in.routed is also changed, to initialize the field in the
ioctl.
2. Closed timing window.
Enhancements:
1. #29: Enhanced the inetd(1M) daemon so that the parameters:
TOOMANY, CNT_INTVL, and RETRYTIME can be tuned instead of being
hardcoded. This tunable parameter capability for the inetd(1M)
daemon was implemented through the /etc/default mechanism. The three
parameters and their default values were placed in the file
"/etc/default/inetd". This file is read when inetd(1M) is invoked
and the specified values are used. These values can be modified
within a specific range. See inetd(1M) for further details.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/mi/Driver.o
/etc/default/inetd
/usr/sbin/in.routed
/usr/sbin/inetd
/usr/sbin/route
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-002
Date Issued: 02/02/2000 14:03:43
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001, crypt-001, crypt-int-001,
diskless-002, fbs-002, man-002">fbsman-002, librt-001
Related SARs: #39, #126, #145, #166
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. It is not possible to change the multicast time-to-live value in
multicast IP packets. The time-to-live (TTL) field in a multicast
IP packet is always set to 1. Under some circumstances, it is
desirable to change this; for example, to enable the multicast
traffic to be routed outside of the local subnet.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
3. #145: If a configured network interface was not physically present
then all configured network interfaces would be disabled.
4. #126: The ftp(1) and rcp(1) utilities would not work if the size of
the file to be transferred was greater than 2GB.
5. Some inet-related changes are required for RCIM Coupled FBS support.
6. #45: The system would hang with an out of memory condition when
using SOCK_RAW sockets with RAWIP under the following conditions:
1) The destination host for the sender is the local system and
there is no receiver currently running.
2) The destination host for both the sender and receiver is
the local system and both tests are currently running.
If the destination host(s) is not the local system (data is going
over the wire), the hang does not occur.
Problem Resolution:
1. The value of the time-to-live field in multicast IP packets is made
configurable using the ndd(1M) utility. A new tuneable parameter
ip_def_mcast_ttl is introduced, and the multicast TTL is changed as
follows:
ndd -set /dev/ip ip_def_mcast_ttl <X>
where <X> is a value between 1 and 255 inclusive.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
3. Modified ifconfig(1M) and slink(1M) to allow all configured and
physically present network interfaces to remain enabled even when one
or more configured interfaces are not physically present.
4. Added large file system support to the ftp(1) and rcp(1) utilities.
5. Added required RCIM Coupled FBS support.
6. There were a large number of outstanding kernel memory allocations by
icmp_pkt() and pullupmsg() that were not being deallocated until the
writer socket was closed. If the test issued a large number of write
(sendto) requests to the socket, the system would eventually hang
waiting for kernel memory to become available.
The outstanding allocations were message blocks that contained
ICMP_SOURCE_QUENCH error messages. These had been sent by ip because
the original data message could not be echoed to the read (RD) side
of the writer socket because it's queue became full.
icmp_inbound_error() was not checking for flow control before passing
the ICMP error message upstream to the next module. It continued
sending the ICMP errors up through the rawip module even after the
stream head's queue was full. In addition, rawip should have been
discarding these error messages. Because the application never read
any of the echo or ICMP messages from the read side of the writer
socket, flow control was never being relieved and eventually the
machine ran out of memory.
icmp_inbound_error() was changed to call canputnext() to check for
flow control before an ICMP error message is passed upstream.
rawip_rput() was modified to discard all M_CTL type message blocks.
The memory leak detection debug code was also enhanced in the kma
and kmadbg kernel drivers.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/rawipm/Driver.o
/etc/inet/services
/usr/bin/ftp
/usr/bin/rcp
/usr/lib/tcpip.so
/usr/sbin/ifconfig
/usr/sbin/in.ftpd
/usr/sbin/slink
Special Conditions for Installation:
The following object(s) will also be replaced by inet-002:
* /usr/lib/libresolv.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-002
Date Issued: 02/02/2000 12:55:12
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crosslibs-001, crypt-001, crypt-int-001,
fbs-002, man-002">fbsman-002, librt-001
Related SARs: #39, #126, #145, #166
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. It is not possible to change the multicast time-to-live value in
multicast IP packets. The time-to-live (TTL) field in a multicast
IP packet is always set to 1. Under some circumstances, it is
desirable to change this; for example, to enable the multicast
traffic to be routed outside of the local subnet.
2. #166: When using gethostbyname() that is statically linked, a
hostname that is opposite in case to one in the /etc/hosts file will
fail. The dynamically linked use of gethostbyname will pass as it
should.
3. #145: If a configured network interface was not physically present
then all configured network interfaces would be disabled.
4. #126: The ftp(1) and rcp(1) utilities would not work if the size of
the file to be transferred was greater than 2GB.
5. Some inet-related changes are required for RCIM Coupled FBS support.
6. #45: The system would hang with an out of memory condition when
using SOCK_RAW sockets with RAWIP under the following conditions:
1) The destination host for the sender is the local system and
there is no receiver currently running.
2) The destination host for both the sender and receiver is
the local system and both tests are currently running.
If the destination host(s) is not the local system (data is going
over the wire), the hang does not occur.
Problem Resolution:
1. The value of the time-to-live field in multicast IP packets is made
configurable using the ndd(1M) utility. A new tuneable parameter
ip_def_mcast_ttl is introduced, and the multicast TTL is changed as
follows:
ndd -set /dev/ip ip_def_mcast_ttl <X>
where <X> is a value between 1 and 255 inclusive.
2. A strcasecmp() function that was left out of the nametoaddr library
functions from a previous NIS change was restored to its proper
location. Now a statically linked use of gethostbyname() will work
as expected.
3. Modified ifconfig(1M) and slink(1M) to allow all configured and
physically present network interfaces to remain enabled even when one
or more configured interfaces are not physically present.
4. Added large file system support to the ftp(1) and rcp(1) utilities.
5. Added required RCIM Coupled FBS support.
6. There were a large number of outstanding kernel memory allocations by
icmp_pkt() and pullupmsg() that were not being deallocated until the
writer socket was closed. If the test issued a large number of write
(sendto) requests to the socket, the system would eventually hang
waiting for kernel memory to become available.
The outstanding allocations were message blocks that contained
ICMP_SOURCE_QUENCH error messages. These had been sent by ip because
the original data message could not be echoed to the read (RD) side
of the writer socket because it's queue became full.
icmp_inbound_error() was not checking for flow control before passing
the ICMP error message upstream to the next module. It continued
sending the ICMP errors up through the rawip module even after the
stream head's queue was full. In addition, rawip should have been
discarding these error messages. Because the application never read
any of the echo or ICMP messages from the read side of the writer
socket, flow control was never being relieved and eventually the
machine ran out of memory.
icmp_inbound_error() was changed to call canputnext() to check for
flow control before an ICMP error message is passed upstream.
rawip_rput() was modified to discard all M_CTL type message blocks.
The memory leak detection debug code was also enhanced in the kma
and kmadbg kernel drivers.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/rawipm/Driver.o
/etc/inet/services
/usr/bin/ftp
/usr/bin/rcp
/usr/lib/tcpip.so
/usr/sbin/ifconfig
/usr/sbin/in.ftpd
/usr/sbin/slink
Special Conditions for Installation:
The following object(s) will also be replaced by inet-002:
* /usr/lib/libresolv.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-003
Date Issued: 05/30/2000 13:07:47
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: 39, 179
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being discarded because of the flow control
hiwater limit on the stream head when using UDP over a fast device
like FDDI. This hiwater limit was originally 64 KB.
2. SAR 179: The timedc(1M) command with 'msite' argument dumps core due
to an improper call to memset(3C).
Problem Resolution:
1. A new UDP tunable (udp_sth_hiwat) was added that defines the hiwater
limit for the read queue's stream head when using UDP. This tunable
can be modified using the ndd(1m) command. The new default value for
this tunable is 256 KB. The minimum value that it can be set to is
64 KB and the maximum is 1024 KB (1 MB).
2. The timedc(1M) command was fixed to correct the call to memset(3C).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/udpm/Driver.o
/usr/sbin/timedc
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-003
Date Issued: 05/30/2000 13:49:50
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: 39, 179
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being discarded because of the flow control
hiwater limit on the stream head when using UDP over a fast device
like FDDI. This hiwater limit was originally 64 KB.
2. SAR 179: The timedc(1M) command with 'msite' argument dumps core due
to an improper call to memset(3C).
Problem Resolution:
1. A new UDP tunable (udp_sth_hiwat) was added that defines the hiwater
limit for the read queue's stream head when using UDP. This tunable
can be modified using the ndd(1m) command. The new default value for
this tunable is 256 KB. The minimum value that it can be set to is
64 KB and the maximum is 1024 KB (1 MB).
2. The timedc(1M) command was fixed to correct the call to memset(3C).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/udpm/Driver.o
/usr/sbin/timedc
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-004
Date Issued: 10/05/2000 13:53:25
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, ip-003, man-003, pg-003
Related SARs: 39, 194, 257, 354, 392, 397
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being dropped by the cnd, dec, ip and pg
drivers when using network buffers (NBUFFs) with either DLPI
operations or datagram sockets (UDP) when transmitting at high
transfer rates.
2. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
3. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new message/datagram.
4. SAR 354: When using ftp(1) with "globbing ON", the mput command
fails under certain conditions with the error "Arguments too long".
5. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. The default value for the UDP tunable udp_sth_hiwat which defines
the flow control hiwater limit of the number of outstanding packets
that can be queued to the stream head was increased from 1 MB to
2 MB. The maximum value for this tunable was increased from 2 MB
to 8 MB.
2. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
3. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
4. Previously the 'glob' functionality had a hardcoded limit to the
arglist of 1701 characters. Therefore, the ftp command `mget *`
would only work if the total number of characters in the '*'
expansion was <=1701.
Though the ftp(1) man page explicitly states that mget should NOT be
used as a method for transferring entire directories, the limit has
been increased to 3x the current size, so the new arglist limit is
5115.
5. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/udpm/Driver.o
/usr/bin/ftp
/usr/lib/tcpip.so
/usr/sbin/in.bootpd
/usr/sbin/in.ftpd
Special Conditions for Installation:
The following link(s) will also be updated by patch inet-004:
/usr/lib/libresolv.a --> /usr/lib/libnsl.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-004
Date Issued: 10/05/2000 14:07:16
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, man-003, pg-003
Related SARs: 39, 194, 257, 354, 392, 397
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being dropped by the cnd, dec, ip and pg
drivers when using network buffers (NBUFFs) with either DLPI
operations or datagram sockets (UDP) when transmitting at high
transfer rates.
2. SAR 194: Errors using UNIX select(3C) with send(3N) writing to
non-blocking sockets. Part of the frame that was being written to
the socket was lost.
3. SAR 257: When utilizing a UDP socket (datagram) and calling the
receive calls (recvfrom, recvmsg) with fewer bytes to read than the
incoming message contains, the remaining bytes of the message are
retained in the buffer. This condition causes subsequent receive
calls to erroneously interpret these bytes as a new message/datagram.
4. SAR 354: When using ftp(1) with "globbing ON", the mput command
fails under certain conditions with the error "Arguments too long".
5. SAR 392,397: gethostbyname() when run in multiple threads will
cause a segmentation violation. getsockname() when run in multiple
threads where one thread is performing a socket read() will block
waiting for its response because the read() consumed its response.
Problem Resolution:
1. The default value for the UDP tunable udp_sth_hiwat which defines
the flow control hiwater limit of the number of outstanding packets
that can be queued to the stream head was increased from 1 MB to
2 MB. The maximum value for this tunable was increased from 2 MB
to 8 MB.
2. A change was made to an internal routine to return the number of
bytes actually written to write(), rather than the number of bytes
passed to write().
3. When MOREDATA is returned from recvfrom(), the additional data should
be discarded.
4. Previously the 'glob' functionality had a hardcoded limit to the
arglist of 1701 characters. Therefore, the ftp command `mget *`
would only work if the total number of characters in the '*'
expansion was <=1701.
Though the ftp(1) man page explicitly states that mget should NOT be
used as a method for transferring entire directories, the limit has
been increased to 3x the current size, so the new arglist limit is
5115.
5. gethostbyname() and any other functions in tcpip.so would all fail
because /usr/lib/tcpip.so was not thread safe. It has been rebuilt
with the _REENTRANT flag which now makes it thread safe and corrects
the multithread problems. The getsockname() was fixed in
/usr/lib/libsocket.so by changing its call of putmsg() to a more
appropriate call of putpmsg().
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/udpm/Driver.o
/usr/bin/ftp
/usr/lib/tcpip.so
/usr/sbin/in.bootpd
/usr/sbin/in.ftpd
Special Conditions for Installation:
The following link(s) will also be updated by patch inet-004:
/usr/lib/libresolv.a --> /usr/lib/libnsl.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: inet-005
Date Issued: 11/20/2000 15:58:12
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Related SARs: 400
Brief Description:
PowerMAX OS 4.3 inet package release updates
#############################################################################
Problem Description:
1. SAR #400: In a rare situation a system could panic during boot as it
performed its network routing initialization. An improper handling
of an ire (internet route entry) was found in the icmp_redirect()
routine.
2. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) would fail with ENOTCONN.
Problem Resolution:
1. A Mentat update was implemented to correct this problem in
icmp_redirect().
2. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
Special Conditions for Installation:
The following link(s) will also be updated by inet-005:
/usr/lib/libresolv.a --> /usr/lib/libnsl.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: inet-005
Date Issued: 11/20/2000 16:06:28
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: 400
Brief Description:
PowerMAX OS 4.3 inet package release updates
#############################################################################
Problem Description:
1. SAR #400: In a rare situation a system could panic during boot as it
performed its network routing initialization. An improper handling
of an ire (internet route entry) was found in the icmp_redirect()
routine.
2. A problem was introduced in the Patch Set 4 (4.3P4) wherein
getpeername(3N) would fail with ENOTCONN.
Problem Resolution:
1. A Mentat update was implemented to correct this problem in
icmp_redirect().
2. This patch corrects the problem with getpeername(3N) while removing
the fix for getsockname(3N) introduced in 4.3P4.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
Special Conditions for Installation:
The following link(s) will also be updated by inet-005:
/usr/lib/libresolv.a --> /usr/lib/libnsl.a
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-006
Date Issued: 03/29/2001 12:01:41
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #431
Brief Description:
PowerMAX OS 4.3 inet package release updates.
##############################################################################
Problem Description:
1. The snmp daemon would frequently crash when an smux client
disconnected from it.
Problem Resolution:
1. Corrected code in snmp daemon which was referencing uninitialized
memory.
Object(s) To Be Replaced:
/usr/sbin/in.snmpd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: inet-006
Date Issued: 03/28/2001 19:08:32
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #431
Brief Description:
PowerMAX OS 4.3 inet package release updates.
##############################################################################
Problem Description:
1. The snmp daemon would frequently crash when an smux client
disconnected from it.
Problem Resolution:
1. Corrected code in snmp daemon which was referencing uninitialized
memory.
Object(s) To Be Replaced:
/usr/sbin/in.snmpd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: inet-007
Date Issued: 08/22/2001 11:51:20
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #495, #503, #483, #520
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 inet package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
2. SAR #483:
When running 'ping -s localhost' an unexpected message: (ICMP
Fragmentation needed and DF set from gateway 127.0.0.1) would
appear every five minutes corresponding to a specific network
activity. ping(1M) was looking for and processing ICMP_REPLY
messages when every five minutes it would receive a few
ICMP_UNREACH messages. The ident was not being checked on
these ICMP_UNREACH messages as it was on the ICMP_REPLY messages
so it processed them also. Had it checked the ident it would have
found these messages did not belong to this ping(1M) session.
3. SAR #520:
A long application delay was occurring as a result of the
callout queue processing that occurs during an ARP cache
flush. The system was connected to a very large network.
The ARP cache was growing from 400 to 700 entries. Very few
of these entries were actually being referenced by the
affected system.
4. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using a lot of different
addresses (host, netid).
5. _tcpip_gethostbyname() was leaking a small amount of memory
each time it was called. However large amounts could be lost
over time if a large number of connects and disconnects were
issued.
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
2. ping(1M) was modified to check the ident on ICMP_UNREACH messages
received so that it would only process those that belonged to it.
3. An optimization in the ARP cache handling code was adding
ARP entries for the sender of broadcast requests even when
it was not the target of the broadcast request. This
optimization code was removed.
4. The number of cached rpcbind addresses was increased from 6 to 24.
5. _tcpip_gethostbyname() was allocating 3 small data structures
every time it was called without first checking to determine
if they had already been allocated. This caused the memory for
the previously allocated structures to be lost.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/tcpip.so
/usr/sbin/ping
/etc/conf/pack.d/arpm/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: inet-007
Date Issued: 09/11/2001 16:22:52
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 inet package release updates
#############################################################################
Problem Description:
1. SAR's #495, #503:
When an application containing gethostbyname() is statically
linked with libnsl.a and the gethostbyname() follows one of
the getservXXX() routines, then the application would core
dump with a segmentation violation. This was due to the tcpip
portion of libnsl accessing the same exact data structure for
its host data that was used for the services data from the
getservXXX() routine. It used the same data structure because
the initialization routine (_tcpip_init()) was not being invoked
in a static object for allocation of the data structure to hold
the host data.
2. SAR #483:
When running 'ping -s localhost' the unexpected message:
"ICMP Fragmentation needed and DF set from gateway 127.0.0.1"
would appear every five minutes corresponding to a specific
network activity. ping(1M) was looking for and processing
ICMP_REPLY messages when every five minutes it would receive
a few ICMP_UNREACH messages. The ident was not being checked
on these ICMP_UNREACH messages as it was on the ICMP_REPLY
messages so it processed them also. Had it checked the ident
it would have found these messages did not belong to this
ping(1M) session.
3. SAR #520:
A long application delay was occurring as a result of the
callout queue processing that occurs during an ARP cache
flush. The system was connected to a very large network.
The ARP cache was growing from 400 to 700 entries. Very few
of these entries were actually being referenced by the
affected system.
4. The number of cached rpcbind addresses was low. This could cause
performance issues if an application was using a lot of different
addresses (host, netid).
5. _tcpip_gethostbyname() was leaking a small amount of memory
each time it was called. However large amounts could be lost
over time if a large number of connects and disconnects were
issued.
Problem Resolution:
1. The tcpip portion of libnsl was corrected so that it would use
the proper initialization routine (_tcpip_init()) for allocation
of the data structure to hold the host data.
2. ping(1M) was modified to check the ident on ICMP_UNREACH messages
received so that it would only process those that belonged to it.
3. An optimization in the ARP cache handling code was adding
ARP entries for the sender of broadcast requests even when
it was not the target of the broadcast request. This
optimization code was removed.
4. The number of cached rpcbind addresses was increased from 6 to 24.
5. _tcpip_gethostbyname() was allocating 3 small data structures
every time it was called without first checking to determine
if they had already been allocated. This caused the memory for
the previously allocated structures to be lost.
Enhancements:
None
Object(s) To Be Replaced:
/usr/lib/tcpip.so
/usr/sbin/ping
/etc/conf/pack.d/arpm/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: inet-008
Date Issued: 01/02/2002 14:24:40
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #581,
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. Networking application hangs in ip_flush_readers() when trying
to exit on a single-cpu system.
2. Compiling the test program with: c++ -o testit testit.c -lucbc
The c++ error says that gethostname is undefined. Using cc,
it compiles.
Problem Resolution:
1. Changed the method of delaying that is used in ip_flush_readers().
The old method of delaying was causing networking applications to
hang on a single cpu system that had many active streams.
There were so many processes that were all hanging waiting for
the readers to finish that none of the lower priority processes
that needed to run were being given any CPU time. The new method
relinquishes the cpu to lower priority processes immediately
instead of wasting cpu time going through a series of delay loops
before relinquishing the cpu.
2. Added the declarations for gethostname() and sethostname() to
/usr/include/netdb.h which is referenced in the gethostname(3b)
man page.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/usr/include/netdb.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: inet-008
Date Issued: 01/02/2002 14:24:40
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #581,
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. Networking application hangs in ip_flush_readers() when trying
to exit on a single-cpu system.
2. Compiling the test program with: c++ -o testit testit.c -lucbc
The c++ error says that gethostname is undefined. Using cc,
it compiles.
Problem Resolution:
1. Changed the method of delaying that is used in ip_flush_readers().
The old method of delaying was causing networking applications to
hang on a single cpu system that had many active streams.
There were so many processes that were all hanging waiting for
the readers to finish that none of the lower priority processes
that needed to run were being given any CPU time. The new method
relinquishes the cpu to lower priority processes immediately
instead of wasting cpu time going through a series of delay loops
before relinquishing the cpu.
2. Added the declarations for gethostname() and sethostname() to
/usr/include/netdb.h which is referenced in the gethostname(3b)
man page.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/usr/include/netdb.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: inet-009
Date Issued: 03/18/2002 11:19:28
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #232, #535, #654
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR #232: ifconfig(1M) behaves strangely when configuring
'up' an interface, depending on the position of 'up'.
On PowerMAX OS 4.3 when 'up' is specified on the
ifconfig(1M) command line as the first parameter,
the command ignores the address in the rest of the command line.
In PowerMAX OS 4.1 this command line form was allowed
(although it should not have been).
2. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
3. SAR #654: Attempts to connect to a PowerMAX OS system via the
network from some system that has CONFIG_INET_ECN configured
into the kernel will eventually timeout. This is due to improper
testing of the TCP header flags field.
Problem Resolution:
1. The ifconfig(1M) command is correct but the usage information
needs to be clarified. Usage information corrected.
2. I have implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
3. Testing of the TCP header flags field has been corrected in
the ipm and tcpm drivers.
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/tcpm/Driver.o
/usr/include/net/if.h
/usr/include/netinet/mib2.h
/usr/lib/resolv.so
/usr/lib/tcpip.so
/usr/sbin/getid
/usr/sbin/getmany
/usr/sbin/getnext
/usr/sbin/getone
/usr/sbin/ifconfig
/usr/sbin/in.snmpd
/usr/sbin/setany
/usr/sbin/snmpstat
/usr/sbin/trap_rece
/usr/sbin/trap_send
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: inet-009
Date Issued: 03/18/2002 11:30:59
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #232, #535, #654
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
Problem Description:
1. SAR #232: ifconfig(1M) behaves strangely when configuring
'up' an interface, depending on the position of 'up'.
On PowerMAX OS 4.3 when 'up' is specified on the
ifconfig(1M) command line as the first parameter,
the command ignores the address in the rest of the command line.
In PowerMAX OS 4.1 this command line form was allowed
(although it should not have been).
2. SAR #535: netdir_options(ND_MERGEADDR) will succeed and return
the LOOPBACK address as the merged network address if there are
no default gateways specified or if the in.routed program is
not running. Otherwise, it will succeed with a properly
resolved network address for the appropriate network interface.
This is incorrect behaviour. The Mentat tcp stack introduced
in 4.2 will return the LOOPBACK address if it can't resolve to
a local interface address. netdir_options() needs to be able
to discern when this occurs and subsequently return an error to
the user indicating that the host address could not be resolved.
3. SAR #654: Attempts to connect to a PowerMAX OS system via the
network from some system that has CONFIG_INET_ECN configured
into the kernel will eventually timeout. This is due to improper
testing of the TCP header flags field.
Problem Resolution:
1. The ifconfig(1M) command is correct but the usage information
needs to be clarified. Usage information corrected.
2. I have implemented a test in tcpip.c and resolv.c of the
nametoaddr libraries. This test will check for the return
of 127.0.0.1 from the ipm/Driver.o in the kernel and translate
that into a ND_NOHOST error to send back to the caller of
netdir_options().
3. Testing of the TCP header flags field has been corrected in
the ipm and tcpm drivers.
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/tcpm/Driver.o
/usr/include/net/if.h
/usr/include/netinet/mib2.h
/usr/lib/resolv.so
/usr/lib/tcpip.so
/usr/sbin/getid
/usr/sbin/getmany
/usr/sbin/getnext
/usr/sbin/getone
/usr/sbin/ifconfig
/usr/sbin/in.snmpd
/usr/sbin/setany
/usr/sbin/snmpstat
/usr/sbin/trap_rece
/usr/sbin/trap_send
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: inet-010
Date Issued: 07/10/2002 14:52:16
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #679, #672
Brief Description:
PowerMAX OS 4.3 inet package release updates
##############################################################################
P1: SAR #679: System panics in mi_timer() during a tcp_xclose().
R1: The user application was calling accept with the no-delay option set on
the socket fd. The system was "panicing" during a tcp_xclose() because
the address of the queue's tcp pointer was changed while tcp_xclose()
was waiting for a lock. When the wait returned, the old tcp was no
longer valid. This caused a series of functions to be called with bad
pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
P2: SAR #672: The in.snmpd daemon was issuing a lot of getmsg calls to the
lower stack layers for each mib attribute requested. This was causing
in.snmpd to have a very high cpu utilization when a large number of mib
attributes were requested.
R2: Previously the only method for the daemon to retrieve the mib information
that it needed from the kernel was to get all 14 sections of mib data at
once. Changes were made to the ipm, tcpm, and udpm drivers and the snmpd
daemon so the daemon can now request just one section at a time instead
of having to get all 14 sections every time..
P3: The gdopen routine could corrupt memory if other data structures were
allocated immediately after its' two metric structures.
R3: The gdopen routine was referencing past the end of two of its allocated
metrics data structures if the incoming cylinder number was equal to the
max number of cylinders. gdopen now verifies that the incoming cylinder
number is within bounds before updating the metrics structures.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/tcpm/Driver.o
/etc/conf/pack.d/udpm/Driver.o
/usr/bin/rlogin
/usr/lib/libresolv.so
/usr/sbin/in.gated
/usr/sbin/in.snmpd
/usr/sbin/in.timed
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-010
Date Issued: 07/24/2002 14:44:43
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #679, #672
Brief Description:
PowerMAX OS 4.3 inet package release updates
###############################################################################
P1: SAR #679: System panics in mi_timer() during a tcp_xclose().
R1: The user application was calling accept with the no-delay option set on
the socket fd. The system was "panicing" during a tcp_xclose() because
the address of the queue's tcp pointer was changed while tcp_xclose()
was waiting for a lock. When the wait returned, the old tcp was no
longer valid. This caused a series of functions to be called with bad
pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
P2: SAR #672: The in.snmpd daemon was issuing a lot of getmsg calls to the
lower stack layers for each mib attribute requested. This was causing
in.snmpd to have a very high cpu utilization when a large number of mib
attributes were requested.
R2: Previously the only method for the daemon to retrieve the mib information
that it needed from the kernel was to get all 14 sections of mib data at
once. Changes were made to the ipm, tcpm, and udpm drivers and the snmpd
daemon so the daemon can now request just one section at a time instead
of having to get all 14 sections every time..
P3: The gdopen routine could corrupt memory if other data structures were
allocated immediately after its' two metric structures.
R3: The gdopen routine was referencing past the end of two of its allocated
metrics data structures if the incoming cylinder number was equal to the
max number of cylinders. gdopen now verifies that the incoming cylinder
number is within bounds before updating the metrics structures.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ipm/Driver.o
/etc/conf/pack.d/tcpm/Driver.o
/etc/conf/pack.d/udpm/Driver.o
/usr/bin/rlogin
/usr/lib/libresolv.so
/usr/sbin/in.gated
/usr/sbin/in.snmpd
/usr/sbin/in.timed
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-011
Date Issued: 12/11/2002 14:33:58
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #753, #780
Brief Description:
PowerMAX OS 4.3 inet package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #753: The in.telnetd daemon on our PowerMAX OS systems
could be vulnerable to CERT Advisory CA-2001-21. This
advisory describes the possible vulnerability due to buffer
overflow in the telrcv() function.
R1: The telrcv() function has been modified to perform bounds
checking on the buffer that was susceptible to overflow.
P2: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R2: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libresolv.so
/usr/lib/resolv.so
/usr/lib/tcpip.so
/usr/sbin/in.telnetd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-011
Date Issued: 12/18/2002 17:44:12
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #753, #780
Brief Description:
PowerMAX OS 4.3 inet package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #753: The in.telnetd daemon on our PowerMAX OS systems
could be vulnerable to CERT Advisory CA-2001-21. This
advisory describes the possible vulnerability due to buffer
overflow in the telrcv() function.
R1: The telrcv() function has been modified to perform bounds
checking on the buffer that was susceptible to overflow.
P2: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R2: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libresolv.so
/usr/lib/resolv.so
/usr/lib/tcpip.so
/usr/sbin/in.telnetd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-012
Date Issued: 05/08/2003 10:38:00
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 inet package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1473: The trpt(1M) utility has not functioned properly
since the introduction of the Mentat TCP/IP stack. It
cannot find required kernel symbols from the new TCP/IP
stack that were available in the old stack.
R1. A total re-write of the utility would be needed to make it
work with the new TCP/IP stack so it was decided to remove
this utility and it's associated man page from the OS.
Enhancements:
1. Modifications were made to IP and ARP to maintain the interface
name to device name mapping specified in the
/etc/confnet.d/inet/interface file. The interface name will now
be set to the concatenation of the "module" and "unit" portion
of the device file name (ie device name). The config.boot.sh
script was changed to report an error if the "prefix" and "unit"
fields of the interface file entry do not match the "module"
and "unit" portion of the device name.
Previously, IP and ARP created the interface name using the
device's module name and a sequential unit number
(starting at 0). The unit number was incremented by 1 for
each interface successfully configured without regard for
the interface's device name. An interface's name could vary
from one boot to the next depending on which other devices of
that type were present in the interface file or on which
devices were physically present on the system during boot.
Certain network related commands did not work properly if the
interface name did not equal the device name.
The customer that requested this change has multiple pg FDDI
boards on his system. They also want the pg device names to
remain the same even when one or more of these boards are not
physically present in the system. Since pg devices are NOT
autoconfigured, their device names are defined by entries in
the sadapters/kernel file. The pg driver maintains the correct
pg device names as long as there are entries in the
sadapters.d/kernel file for all of the pg boards that could be
configured on the system. Thus, pg1 will stay pg1, even if pg0
is not physically present on the system, IF there are
sadapter.d/kernel entries for both pg0 AND pg1.
Autoconfigured device names can vary from boot to boot if the
system hardware is reconfigured. PCI devices such as dec and
rmxf that are configured automatically at boot time are examples
of this type of device. While the IP and ARP changes take care
of the interface to device name mapping for these types of
devices, the changes made in this task do not provide any method
of overriding the autoconfiguration of their device names.
Object(s) To Be Replaced:
/etc/conf/pack.d/arpm/Driver.o
/etc/conf/pack.d/ipm/Driver.o
/etc/confnet.d/inet/config.boot.sh
/etc/confnet.d/inet/configure
/etc/confnet.d/inet/interface
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-012
Date Issued: 05/08/2003 10:37:59
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 inet package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1473: The trpt(1M) utility has not functioned properly
since the introduction of the Mentat TCP/IP stack. It
cannot find required kernel symbols from the new TCP/IP
stack that were available in the old stack.
R1. A total re-write of the utility would be needed to make it
work with the new TCP/IP stack so it was decided to remove
this utility and it's associated man page from the OS.
Enhancements:
1. Modifications were made to IP and ARP to maintain the interface
name to device name mapping specified in the
/etc/confnet.d/inet/interface file. The interface name will now
be set to the concatenation of the "module" and "unit" portion
of the device file name (ie device name). The config.boot.sh
script was changed to report an error if the "prefix" and "unit"
fields of the interface file entry do not match the "module"
and "unit" portion of the device name.
Previously, IP and ARP created the interface name using the
device's module name and a sequential unit number
(starting at 0). The unit number was incremented by 1 for
each interface successfully configured without regard for
the interface's device name. An interface's name could vary
from one boot to the next depending on which other devices of
that type were present in the interface file or on which
devices were physically present on the system during boot.
Certain network related commands did not work properly if the
interface name did not equal the device name.
The customer that requested this change has multiple pg FDDI
boards on his system. They also want the pg device names to
remain the same even when one or more of these boards are not
physically present in the system. Since pg devices are NOT
autoconfigured, their device names are defined by entries in
the sadapters/kernel file. The pg driver maintains the correct
pg device names as long as there are entries in the
sadapters.d/kernel file for all of the pg boards that could be
configured on the system. Thus, pg1 will stay pg1, even if pg0
is not physically present on the system, IF there are
sadapter.d/kernel entries for both pg0 AND pg1.
Autoconfigured device names can vary from boot to boot if the
system hardware is reconfigured. PCI devices such as dec and
rmxf that are configured automatically at boot time are examples
of this type of device. While the IP and ARP changes take care
of the interface to device name mapping for these types of
devices, the changes made in this task do not provide any method
of overriding the autoconfiguration of their device names.
Object(s) To Be Replaced:
/etc/conf/pack.d/arpm/Driver.o
/etc/conf/pack.d/ipm/Driver.o
/etc/confnet.d/inet/config.boot.sh
/etc/confnet.d/inet/configure
/etc/confnet.d/inet/interface
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-013
Date Issued: 07/08/2004 10:53:17
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 inet package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R1: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libresolv.so
/usr/lib/tcpip.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: inet-013
Date Issued: 07/26/2004 11:50:55
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 inet package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1591: A program containing strncasecmp() and socket() cannot
be statically linked. The libucbc.a and libsocket.a both
contained strcasecmp() routines which caused a conflict but
only when strncasecmp() and socket were both in the source file.
R1: The resolution was to rebuild libsocket.a with the
strcasecmp() routine renamed to Strcasecmp() as it was only
being used locally anyway.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/libresolv.so
/usr/lib/tcpip.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: inet-014
Date Issued: 03/13/2006 14:47:14
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 inet package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3156: The mpc7457 chip on newer PH920 boards caused problems in
the netsetup script which uses "uname -mp" output to check if a
particular Ethernet driver type should be configured. The gte
driver entry was restricted to mpc7455.
R1: The netsetup script was changed to use the 'uname -A' output.
The entries that require restrictions were changed to use the
model name/number instead of the PPC chip type.
P2: The RIQ10192 work overlooked how LOGFAILURES was handled if
MAXTRYS was less than LOGFAILURES. The passwd utility was not
reinitializing the counters in the restrict case on a failed
attempt so it allowed success on a second failed attempt.
R2: This has been corrected in this task.
P3: There were architecture-specific versions of the netsetup script.
R3: Replaced the architecture-specific versions with a common version
with additional enhancements.
Enhancements:
1. RIQ10192: Enhanced in.login for restrictions on remote login.
Object(s) To Be Replaced:
/usr/bin/netsetup
/usr/lib/iaf/in.login/scheme
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: inet-014
Date Issued: 03/13/2006 11:39:42
Software Package: inet pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 inet package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3156: The mpc7457 chip on newer PH920 boards caused problems in
the netsetup script which uses "uname -mp" output to check if a
particular Ethernet driver type should be configured. The gte
driver entry was restricted to mpc7455.
R1: The netsetup script was changed to use the 'uname -A' output.
The entries that require restrictions were changed to use the
model name/number instead of the PPC chip type.
P2: The RIQ10192 work overlooked how LOGFAILURES was handled if
MAXTRYS was less than LOGFAILURES. The passwd utility was not
reinitializing the counters in the restrict case on a failed
attempt so it allowed success on a second failed attempt.
R2: This has been corrected in this task.
P3: There were architecture-specific versions of the netsetup script.
R3: Replaced the architecture-specific versions with a common version
with additional enhancements.
Enhancements:
1. RIQ10192: Enhanced in.login for restrictions on remote login.
Object(s) To Be Replaced:
/usr/bin/netsetup
/usr/lib/iaf/in.login/scheme
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ip-001
Date Issued: 02/02/2000 14:05:02
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620
Related Patches: base-002, cnd-002, dec-001, egl-001, pg-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 ip package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ip/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ip-002
Date Issued: 05/30/2000 13:08:28
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620
Related Patches: cnd-003, dec-002, egl-002, nsu-002, pg-002
Related SARs: 201
Brief Description:
PowerMAX OS 4.3 ip package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ip/Driver.o
/etc/conf/pack.d/ip/space.c
/usr/include/sys/dlpi_fddi.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ip-003
Date Issued: 10/05/2000 13:54:02
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, man-003, pg-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 ip package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The ip driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ip/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: ip-004
Date Issued: 03/18/2002 11:23:10
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 ip package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/ip/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: ip-005
Date Issued: 05/08/2003 10:38:40
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 ip package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/ip
/etc/conf/pack.d/ip/Driver.o
/usr/include/sys/ip.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: ip-006
Date Issued: 01/07/2004 16:00:59
Software Package: ip pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 ip package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R3: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ip/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: is-001
Date Issued: 06/11/2001 08:35:57
Software Package: is pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 is package release updates
##############################################################################
Problem Description:
1. Driver uses async start unit command to spin up disk drives and
polls once a second until it is ready or times out.
2. Driver needed the ability to handle lists of Physcal DMA addresses.
(support new partial sector capability).
Problem Resolution:
1. Modified Driver to poll once a second until it is ready or
times out.
2. Modified Driver to support new partial sector capability.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/is/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: ise-001
Date Issued: 11/20/2000 16:06:48
Software Package: ise pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 ise package release updates
#############################################################################
Problem Description:
1. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
Problem Resolution:
1. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/sys/ise.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: kdb-001
Date Issued: 09/15/1999 11:06:55
Software Package: kdb pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 kdb package release updates
##############################################################################
Problem Description:
None.
Problem Resolution:
None.
Enhancements:
1. The number of breakpoints which can be set by the kernel debugger was
increased from 20 to 30.
Object(s) To Be Replaced:
/etc/conf/pack.d/kdb/Driver.o
/etc/conf/pack.d/kdb_util/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: kdb-001
Date Issued: 09/15/1999 13:12:52
Software Package: kdb pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 kdb package release updates
##############################################################################
Problem Description:
None.
Problem Resolution:
None.
Enhancements:
1. The number of breakpoints which can be set by the kernel debugger was
increased from 20 to 30.
Object(s) To Be Replaced:
/etc/conf/pack.d/kdb/Driver.o
/etc/conf/pack.d/kdb_util/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: librt-001
Date Issued: 02/02/2000 14:05:23
Software Package: librt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001, diskless-002, fbs-002,
man-002">fbsman-002, inet-002
Related SARs: none
Brief Description:
PowerMAX OS 4.3 librt package release updates
##############################################################################
Problem Description:
1. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
Problem Resolution:
1. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/fbslib.h
/usr/lib/libF77rt.a
/usr/lib/librt.a
/usr/lib/librt.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: librt-001
Date Issued: 02/02/2000 12:57:09
Software Package: librt pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crosslibs-001, fbs-002, man-002">fbsman-002, inet-002
Related SARs: none
Brief Description:
PowerMAX OS 4.3 librt package release updates
##############################################################################
Problem Description:
1. RCIM devices configured as distributed interrupts could only be used
as Closely Coupled FBS timing devices within the same Closely Coupled
cluster.
Problem Resolution:
1. RCIM Coupled FBS support is added to remove this restriction. Any
set of SBCs that are:
- connected to the same RCIM cable, and
- can communicate with each other over a TCP/IP connection,
may now use the same distributed interrupt RCIM device as a FBS
timing device. A new pair of device registration function calls, and
a rdevfs(4) timing device information call:
- fbs_register_rdev(3rt | 3F77rt)
- fbs_unregister_rdev(3rt | 3F77rt)
- fbsinfo_rdev(3rt | 3F77rt)
and new rtcp(1) device registration and information commands:
- rd (register Coupled FBS timing device)
- urd (unregister Coupled FBS timing device)
- vr (view rdevfs file configuration information)
are provided for support of these new RCIM Coupled timing devices.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/fbslib.h
/usr/lib/libF77rt.a
/usr/lib/librt.a
/usr/lib/librt.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-001
Date Issued: 09/15/1999 11:07:25
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001, man-001">fbsman-001, inet-001
Related SARs: #29
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. Several man pages incorrectly refers to "Closely-Coupled" systems as
"Loosely-Coupled".
Problem Resolution:
1. Corrected "Loosely-Coupled" references.
Enhancements:
1. #29: inetd(1M) updated to include enhancements to inetd. See
"inet-001" for further details.
Object(s) To Be Replaced:
/usr/share/man/cat1/inetd.1m.z
/usr/share/man/cat1/lockd.1m.z
/usr/share/man/cat1/passwd.1.z
/usr/share/man/cat2/aclipc.2.z
/usr/share/man/cat2/lvlipc.2.z
/usr/share/man/cat3/mq_open.3.z
/usr/share/man/cat3/mq_unlink.3.z
/usr/share/man/cat3/sem_open.3.z
/usr/share/man/cat3/sem_unlink.3.z
/usr/share/man/cat9/manual.9.z
/usr/share/man/man1/inetd.1m
/usr/share/man/man1/lockd.1m
/usr/share/man/man1/passwd.1
/usr/share/man/man2/aclipc.2
/usr/share/man/man2/lvlipc.2
/usr/share/man/man3/mq_open.3
/usr/share/man/man3/mq_unlink.3
/usr/share/man/man3/sem_open.3
/usr/share/man/man3/sem_unlink.3
/usr/share/man/man9/manual.9
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-001
Date Issued: 09/15/1999 13:13:21
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001, man-001">fbsman-001, inet-001
Related SARs: #29
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. Several man pages had incorrect references.
Problem Resolution:
1. Corrected man pages.
Enhancements:
1. #29: inetd(1M) updated to include enhancements to inetd. See
"inet-001" for further details.
Object(s) To Be Replaced:
/usr/share/man/cat1/inetd.1m.z
/usr/share/man/cat1/lockd.1m.z
/usr/share/man/cat1/passwd.1.z
/usr/share/man/cat2/aclipc.2.z
/usr/share/man/cat2/lvlipc.2.z
/usr/share/man/cat3/mq_open.3.z
/usr/share/man/cat3/mq_unlink.3.z
/usr/share/man/cat3/sem_open.3.z
/usr/share/man/cat3/sem_unlink.3.z
/usr/share/man/cat9/manual.9.z
/usr/share/man/man1/inetd.1m
/usr/share/man/man1/lockd.1m
/usr/share/man/man1/passwd.1
/usr/share/man/man2/aclipc.2
/usr/share/man/man2/lvlipc.2
/usr/share/man/man3/mq_open.3
/usr/share/man/man3/mq_unlink.3
/usr/share/man/man3/sem_open.3
/usr/share/man/man3/sem_unlink.3
/usr/share/man/man9/manual.9
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-002
Date Issued: 02/02/2000 14:06:09
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-002, crosslibs-001
Related SARs: #47
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The crash(1M) man page had table formatting problems. Descriptions
for several functions are inaccurate.
2. The posix_timers(2) man pages needed to be updated to reflect changes
made to POSIX timers in libthread in patch "base-002".
3. #47: Man page updates to mpadvise(3C) required to reflect changes
for added VME TurboHawk support in patch "base-002".
Problem Resolution:
1. Formatting problems were resolved. The man page descriptions of the
`page', `pte', and `vtop' functions were adjusted for accuracy.
2. Man page was updated.
3. mpadvise(3C) man page was updated. A new man page for
vme_address_expan(3C) was added.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat2/posix_timers.2.z
/usr/share/man/cat3/mpadvise.3c.z
/usr/share/man/cat3/vme_address_expan.3c.z
/usr/share/man/man1/crash.1m
/usr/share/man/man2/posix_timers.2
/usr/share/man/man3/mpadvise.3c
/usr/share/man/man3/vme_address_expan.3c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-002
Date Issued: 02/02/2000 12:58:34
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, crosslibs-001
Related SARs: #47
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The crash(1M) man page had table formatting problems. Descriptions
for several functions are inaccurate.
2. The posix_timers(2) man pages needed to be updated to reflect changes
made to POSIX timers in libthread in patch "base-002".
3. #47: Man page updates to mpadvise(3C) required to reflect changes
for added VME TurboHawk support in patch "base-002".
Problem Resolution:
1. Formatting problems were resolved. The man page descriptions of the
`page', `pte', and `vtop' functions were adjusted for accuracy.
2. Man page was updated.
3. mpadvise(3C) man page was updated. A new man page for
vme_address_expan(3C) was added.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat2/posix_timers.2.z
/usr/share/man/cat3/mpadvise.3c.z
/usr/share/man/cat3/vme_address_expan.3c.z
/usr/share/man/man1/crash.1m
/usr/share/man/man2/posix_timers.2
/usr/share/man/man3/mpadvise.3c
/usr/share/man/man3/vme_address_expan.3c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-003
Date Issued: 10/05/2000 13:54:21
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, pg-003
Related SARs: 270
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. SAR 270: A new option was added to in.bootpd. The bootpd(1M)
manual page needed to be updated to reflect the new support. The new
option was added in patch "inet-004".
2. Manual pages needed to be added for support of new clock resolution
functionality added to several C library routines in patch
"base-004".
3. A new command was added to syscx(2) for the two new C library
routines. The manual pages needed to be updated as well.
Problem Resolution:
1. Update bootpd(1M) manual page.
2. Added new manual pages for [get|set]_timeout_resolution(3C).
3. Update syscx(2) manual page.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/bootpd.1m.z
/usr/share/man/cat2/syscx.2.z
/usr/share/man/cat3/get_timeout_resolution.3c.z
/usr/share/man/cat3/set_timeout_resolution.3c.z
/usr/share/man/man1/bootpd.1m
/usr/share/man/man2/syscx.2
/usr/share/man/man3/get_timeout_resolution.3c
/usr/share/man/man3/set_timeout_resolution.3c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-003
Date Issued: 10/05/2000 14:08:28
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, inet-004, pg-003
Related SARs: 270
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. SAR 270: A new option was added to in.bootpd. The bootpd(1M)
manual page needed to be updated to reflect the new support. The new
option was added in patch "inet-004".
2. Manual pages needed to be added for support of new clock resolution
functionality added to several C library routines in patch
"base-004".
3. A new command was added to syscx(2) for the two new C library
routines. The manual pages needed to be updated as well.
Problem Resolution:
1. Update bootpd(1M) manual page.
2. Added new manual pages for [get|set]_timeout_resolution(3C).
3. Update syscx(2) manual page.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/bootpd.1m.z
/usr/share/man/cat2/syscx.2.z
/usr/share/man/cat3/get_timeout_resolution.3c.z
/usr/share/man/cat3/set_timeout_resolution.3c.z
/usr/share/man/man1/bootpd.1m
/usr/share/man/man2/syscx.2
/usr/share/man/man3/get_timeout_resolution.3c
/usr/share/man/man3/set_timeout_resolution.3c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: man-004
Date Issued: 03/29/2001 12:02:15
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 man package release updates
#############################################################################
Problem Description:
1. iobus_err(2) man page missing information for vme bus fault handing
on the TurboHawk.
2. iobus_err(2) man page missing information for vme bus fault support
on the PowerMAXION.
Problem Resolution:
1. Documented in iobus_err(2) the operation of the newly added vme
bus fault handling on the TurboHawk, for both the primary and
secondary vme busses.
2. Documented in iobus_err(2) how the newly added secondary vme bus
fault support works on the PowerMAXION.
Object(s) To Be Replaced:
/usr/share/man/man2/iobus_err.2
/usr/share/man/cat2/iobus_err.2.z
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: man-004
Date Issued: 03/29/2001 10:19:13
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-006
Related SARS: none
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. iobus_err(2) man page missing information for vme bus fault handing
on the TurboHawk.
2. iobus_err(2) man page missing information for vme bus fault support
on the PowerMAXION.
Problem Resolution:
1. Documented in iobus_err(2) the operation of the newly added vme
bus fault handling on the TurboHawk, for both the primary and
secondary vme busses.
2. Documented in iobus_err(2) how the newly added secondary vme bus
fault support works on the PowerMAXION.
Object(s) To Be Replaced:
/usr/share/man/man2/iobus_err.2
/usr/share/man/cat2/iobus_err.2.z
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: man-005
Date Issued: 08/22/2001 14:52:46
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: #507
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 man package release updates
#############################################################################
Problem Description:
1. SAR #507:
The errno data contained in the intro(2) man page is inconsistent
with the data contained in /usr/include/sys/errno.h.
2. The header file referenced in the SYNOPSIS section of
set_timeout_resolution(3C) is incorrect. The data provided
in the WARNING section of this man page is not consistent with
the functionality of the routine.
Problem Resolution:
1. The intro(2) man page has been updated to be consistent with the
data in /usr/include/sys/errno.h.
2. The set_timeout_resolution(2) man page has been corrected.
Enhancements:
None
Object(s) To Be Replaced:
/usr/share/man/man2/intro.2
/usr/share/man/cat2/intro.2.z
/usr/share/man/man3/set_timeout_resolution.3c
/usr/share/man/cat3/set_timeout_resolution.3c.z
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: man-005
Date Issued: 09/11/2001 16:24:02
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 man package release updates
#############################################################################
Problem Description:
1. SAR #507:
The errno data contained in the intro(2) man page is inconsistent
with the data contained in /usr/include/sys/errno.h.
2. The header file referenced in the SYNOPSIS section of
set_timeout_resolution(3C) is incorrect. The data provided
in the WARNING section of this man page is not consistent with
the functionality of the routine.
Problem Resolution:
1. The intro(2) man page has been updated to be consistent with the
data in /usr/include/sys/errno.h.
2. The set_timeout_resolution(2) man page has been corrected.
Enhancements:
None
Object(s) To Be Replaced:
/usr/share/man/man2/intro.2
/usr/share/man/cat2/intro.2.z
/usr/share/man/man3/set_timeout_resolution.3c
/usr/share/man/cat3/set_timeout_resolution.3c.z
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: man-006
Date Issued: 01/02/2002 14:25:01
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #579
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
2. The user of the newfs(1M) utility was never able to modify the
value of the maximum allocated inodes for an XFS filesystem.
However, this could be performed with mkfs(1M) which
is called by newfs(1M). With larger and larger XFS partitions
this has produced a serious problem since all XFS partitions
are established with the same maximum inode value regardless
of the partition size.
Problem Resolution:
1. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
2. The newfs(1M) utility has been enhanced to allow the user to
specify a maximum inode allocation value on the command line
with a new -x <max-inode> argument for XFS filesystems only.
If this argument is not used by the user, then a default value
will be applied which is based on the size of the partition
rather than using the same default value for all partitions.
Additionally, this new argument has been documented in the
newfs(1M) manual page and the C locale file for I18N has been
updated.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/newfs.1m.z
/usr/share/man/cat2/sysinfo.2.z
/usr/share/man/cat3/gethostid.3b.z
/usr/share/man/man1/newfs.1m
/usr/share/man/man2/sysinfo.2
/usr/share/man/man3/gethostid.3b
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: man-006
Date Issued: 01/02/2002 14:25:01
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #579
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The gethostid(3B) routine as implemented in PowerMAX OS did
not return any usable data. It relied upon the setting of the
kernel variable 'HW_SERIAL' in /etc/conf/pack.d/name/space.c
which is generally not used in our implementation. HW_SERIAL
is set to an empty string normally and comments in the file
/etc/conf/pack.d/name/space.c advise that it be left that way.
2. The user of the newfs(1M) utility was never able to modify the
value of the maximum allocated inodes for an XFS filesystem.
However, this could be performed with mkfs(1M) which
is called by newfs(1M). With larger and larger XFS partitions
this has produced a serious problem since all XFS partitions
are established with the same maximum inode value regardless
of the partition size.
Problem Resolution:
1. The gethostid(3B) routine has been modified such that it will
either return the value of the string at HW_SERIAL (if it
happens to be set to something other than a null terminated
empty string) or it will return the value of the string
derived from the nodename element of struct utsname which will
be converted to a long using base 36 conversion. The
gethostid(3B) and sysinfo(2) manual pages have been modified
to document this change.
2. The newfs(1M) utility has been enhanced to allow the user to
specify a maximum inode allocation value on the command line
with a new -x <max-inode> argument for XFS filesystems only.
If this argument is not used by the user, then a default value
will be applied which is based on the size of the partition
rather than using the same default value for all partitions.
Additionally, this new argument has been documented in the
newfs(1M) manual page and the C locale file for I18N has been
updated.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat1/newfs.1m.z
/usr/share/man/cat2/sysinfo.2.z
/usr/share/man/cat3/gethostid.3b.z
/usr/share/man/man1/newfs.1m
/usr/share/man/man2/sysinfo.2
/usr/share/man/man3/gethostid.3b
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: man-007
Date Issued: 03/18/2002 11:23:36
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #541, #596, #655
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The ifconfig(1M) man page is not consistent with the
command usage information.
2. SAR #541: ps(1) man page for the SZ field of ps -el was incorrect.
3. The crontab(1) man page incorrectly indicates that crontab will set
the PATH environment variable to: PATH=:/bin:/usr/bin:/usr/lbin
4. Realpath(3C) only returns fully resolved pathnames upon success.
The man page claimed it could return an absolute pathname
in certain cases.
Problem Resolution:
1. The ifconfig(1M) man page is now consistent with the
command usage information.
2. Modified ps(1) man page to match what the command displays.
Also modified the F, Flags, table of ps(1) man page.
3. The crontab(1) man page has been changed to show that crontab sets
the PATH environment variable to:
PATH=/sbin:/usr/bin:/usr/sbin:/usr/lbin:
4. Changed "relative" to "absolute" in the line:
For the other relative path names, it returns the resolved
relative name.
Enhancements:
1. Adds a description of the ui function to crash(1m).
2. Adds a description of the new -s option to the calc function
of crash(1m).
3. snmpstat(1m), Added the description of the new "-d" option that
displays the dot3 transmission group MIB statistics for Ethernet
interfaces.
4. SAR #596: Attempt to clarify the fact that for multiplexing threads,
the scheduling class and priority set or returned on
the pthread_[get|set]scheduler(), thr_[get|set]scheduler()
and thr_[get|set]prio() function calls is only the thread
library's notion of that thread's scheduling class and
priority. The kernel's scheduling class and priority of the
Lightweight Process (LWP) that this multiplexing thread is
currently executing on is NOT modified or obtained via these
pthread_xxx or thr_xxx function calls.
5. SAR #596: Attempt to clarify the fact that the main/primordial
thread is a multiplexing thread by default.
6. SAR #596: Add a description to thread(3thread) and
pthread(3pthread) on how to make the initial/primordial thread
be a bound thread instead of a multiplexing thread (the default).
7. SAR #596: Make it clearer that the sched_[set|get]scheduler() and
sched_[set/get]param() function calls should NOT be used
by multithreaded applications.
8. SAR #655: termio(7) man page was modified to give additional
information on the use of "TIOCMGET" and "TIOCMSET" ioctl commands.
9. Added new exit codes to reflect changes in the generic mount
and umount commands:
33 mount() error - misc error
19 umount() - mount point is not an absolute pathname
20 umount() - special device cannot be determined
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/crontab.1.z
/usr/share/man/cat1/ifconfig.1m.z
/usr/share/man/cat1/mount.1m.z
/usr/share/man/cat1/ps.1.z
/usr/share/man/cat1/snmpstat.1m.z
/usr/share/man/cat3/pthr_getschedparam.3x.z
/usr/share/man/cat3/pthr_setschedparam.3x.z
/usr/share/man/cat3/pthread.3x.z
/usr/share/man/cat3/realpath.3c.z
/usr/share/man/cat3/sched_getparam.3c.z
/usr/share/man/cat3/sched_getscheduler.3c.z
/usr/share/man/cat3/sched_setparam.3c.z
/usr/share/man/cat3/sched_setscheduler.3c.z
/usr/share/man/cat3/thr_getprio.3x.z
/usr/share/man/cat3/thr_getsched.3x.z
/usr/share/man/cat3/thr_setprio.3x.z
/usr/share/man/cat3/thr_setsched.3x.z
/usr/share/man/cat3/thread.3x.z
/usr/share/man/cat7/termio.7.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/crontab.1
/usr/share/man/man1/ifconfig.1m
/usr/share/man/man1/mount.1m
/usr/share/man/man1/ps.1
/usr/share/man/man1/snmpstat.1m
/usr/share/man/man3/pthr_getschedparam.3x
/usr/share/man/man3/pthr_setschedparam.3x
/usr/share/man/man3/pthread.3x
/usr/share/man/man3/realpath.3c
/usr/share/man/man3/sched_getparam.3c
/usr/share/man/man3/sched_getscheduler.3c
/usr/share/man/man3/sched_setparam.3c
/usr/share/man/man3/sched_setscheduler.3c
/usr/share/man/man3/thr_getprio.3x
/usr/share/man/man3/thr_getsched.3x
/usr/share/man/man3/thr_setprio.3x
/usr/share/man/man3/thr_setsched.3x
/usr/share/man/man3/thread.3x
/usr/share/man/man7/termio.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: man-007
Date Issued: 03/18/2002 11:34:58
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #541, #596, #655
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
Problem Description:
1. The ifconfig(1M) man page is not consistent with the
command usage information.
2. SAR #541: ps(1) man page for the SZ field of ps -el was incorrect.
3. The crontab(1) man page incorrectly indicates that crontab will set
the PATH environment variable to: PATH=:/bin:/usr/bin:/usr/lbin
4. Realpath(3C) only returns fully resolved pathnames upon success.
The man page indicated that it was possible to return an
absolute pathname in certain cases.
Problem Resolution:
1. The ifconfig(1M) man page is now consistent with the
command usage information.
2. Modified ps(1) man page to match what the command displays.
Also modified the F, Flags, table of ps(1) man page.
3. The crontab(1) man page has been changed to show that crontab sets
the PATH environment variable to:
PATH=/sbin:/usr/bin:/usr/sbin:/usr/lbin:
4. Changed "relative" to "absolute" in the line:
For the other relative path names, it returns the resolved
relative name.
Enhancements:
1. Adds a description of the ui function to crash(1m).
2. Adds a description of the new -s option to the calc function
of crash(1m).
3. snmpstat(1m), Added the description of the new "-d" option that
displays the dot3 transmission group MIB statistics for Ethernet
interfaces.
4. SAR #596: Attempt to clarify the fact that for multiplexing threads,
the scheduling class and priority set or returned on
the pthread_[get|set]scheduler(), thr_[get|set]scheduler()
and thr_[get|set]prio() function calls is only the thread
library's notion of that thread's scheduling class and
priority. The kernel's scheduling class and priority of the
Lightweight Process (LWP) that this multiplexing thread is
currently executing on is NOT modified or obtained via these
pthread_xxx or thr_xxx function calls.
5. SAR #596: Attempt to clarify the fact that the main/primordial
thread is a multiplexing thread by default.
6. SAR #596: Add a description to thread(3thread) and
pthread(3pthread) on how to make the initial/primordial thread
be a bound thread instead of a multiplexing thread (the default).
7. SAR #596: Make it clearer that the sched_[set|get]scheduler() and
sched_[set/get]param() function calls should NOT be used
by multithreaded applications.
8. SAR #655: termio(7) man page was modified to give additional
information on the use of "TIOCMGET" and "TIOCMSET" ioctl commands.
9. Added one new exit code to reflect changes in the generic mount
and umount commands:
33 mount() error - misc error
19 umount() - mount point is not an absolute pathname
20 umount() - special device cannot be determined
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/crontab.1.z
/usr/share/man/cat1/ifconfig.1m.z
/usr/share/man/cat1/mount.1m.z
/usr/share/man/cat1/ps.1.z
/usr/share/man/cat1/snmpstat.1m.z
/usr/share/man/cat3/pthr_getschedparam.3x.z
/usr/share/man/cat3/pthr_setschedparam.3x.z
/usr/share/man/cat3/pthread.3x.z
/usr/share/man/cat3/realpath.3c.z
/usr/share/man/cat3/sched_getparam.3c.z
/usr/share/man/cat3/sched_getscheduler.3c.z
/usr/share/man/cat3/sched_setparam.3c.z
/usr/share/man/cat3/sched_setscheduler.3c.z
/usr/share/man/cat3/thr_getprio.3x.z
/usr/share/man/cat3/thr_getsched.3x.z
/usr/share/man/cat3/thr_setprio.3x.z
/usr/share/man/cat3/thr_setsched.3x.z
/usr/share/man/cat3/thread.3x.z
/usr/share/man/cat7/termio.7.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/crontab.1
/usr/share/man/man1/ifconfig.1m
/usr/share/man/man1/mount.1m
/usr/share/man/man1/ps.1
/usr/share/man/man1/snmpstat.1m
/usr/share/man/man3/pthr_getschedparam.3x
/usr/share/man/man3/pthr_setschedparam.3x
/usr/share/man/man3/pthread.3x
/usr/share/man/man3/realpath.3c
/usr/share/man/man3/sched_getparam.3c
/usr/share/man/man3/sched_getscheduler.3c
/usr/share/man/man3/sched_setparam.3c
/usr/share/man/man3/sched_setscheduler.3c
/usr/share/man/man3/thr_getprio.3x
/usr/share/man/man3/thr_getsched.3x
/usr/share/man/man3/thr_setprio.3x
/usr/share/man/man3/thr_setsched.3x
/usr/share/man/man3/thread.3x
/usr/share/man/man7/termio.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: man-008
Date Issued: 07/10/2002 14:53:26
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #719, #691
Brief Description:
PowerMAX OS 4.3 man package release updates
##############################################################################
P1: The pkgadd(1M) and pkginfo(1) manpages made no mention of the TMPDIR
environment variable for use in the event that the /var filesystem
becomes too full for successful command completion.
R1: The pkgadd(1M) and pkginfo(1) manpages are modified to include a note
detailing the use of the TMPDIR environment variable.
Enhancements:
1. Added description of the lwp -e option to crash(1m).
2. Updated the setsockopt(3N) man page.
3. The pkginfo(1) man page has been modified to document
the addition of the -R option to the pkginfo(1) utility.
This enhancement to the pkginfo(1) utility allows a user
to use pkginfo to display Patch README files.
4. SAR #691: Added the following paragraph to shmget(2), msgget(2),
and semget(2) man pages:
key must be a unique value which is used by all processes
sharing the same [shared memory region|message queue|semaphore
set]. It may be generated with a call to ftok()
(see stdipc(3c)).
5. Modified the stdipc(3c) man page short NAME description
from: stdipc: ftok - standard interprocess communication package
to ftok - key generator for standard IPC system calls This will
allow "apropos key" to return ftok. Also added PowerMAX OS
Programming Guide to the REFERENCES sections.
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/pkgadd.1m.z
/usr/share/man/cat1/pkginfo.1.z
/usr/share/man/cat2/msgget.2.z
/usr/share/man/cat2/semget.2.z
/usr/share/man/cat2/shmget.2.z
/usr/share/man/cat3/getsockopt.3n.z
/usr/share/man/cat3/stdipc.3c.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/pkgadd.1m
/usr/share/man/man1/pkginfo.1
/usr/share/man/man2/msgget.2
/usr/share/man/man2/semget.2
/usr/share/man/man2/shmget.2
/usr/share/man/man3/getsockopt.3n
/usr/share/man/man3/stdipc.3c
Man pages removed in this patch:
SAR #719: /usr/share/man/cat4/issue.4.z
/usr/share/man/man4/issue.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-008
Date Issued: 07/24/2002 14:45:52
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #719, #691
Brief Description:
PowerMAX OS 4.3 man package release updates
###############################################################################
P1: The pkgadd(1M) and pkginfo(1) manpages made no mention of the TMPDIR
environment variable for use in the event that the /var filesystem
becomes too full for successful command completion.
R1: The pkgadd(1M) and pkginfo(1) manpages are modified to include a note
detailing the use of the TMPDIR environment variable.
Enhancements:
1. Added description of the lwp -e option to crash(1m).
2. Updated the setsockopt(3N) man page.
3. The pkginfo(1) man page has been modified to document
the addition of the -R option to the pkginfo(1) utility.
This enhancement to the pkginfo(1) utility allows a user
to use pkginfo to display Patch README files.
4. SAR #691: Added the following paragraph to shmget(2), msgget(2),
and semget(2) man pages:
key must be a unique value which is used by all processes
sharing the same [shared memory region|message queue|semaphore
set]. It may be generated with a call to ftok()
(see stdipc(3c)).
5. Modified the stdipc(3c) man page short NAME description
from: stdipc: ftok - standard interprocess communication package
to ftok - key generator for standard IPC system calls This will
allow "apropos key" to return ftok. Also added PowerMAX OS
Programming Guide to the REFERENCES sections.
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/pkgadd.1m.z
/usr/share/man/cat1/pkginfo.1.z
/usr/share/man/cat2/msgget.2.z
/usr/share/man/cat2/semget.2.z
/usr/share/man/cat2/shmget.2.z
/usr/share/man/cat3/getsockopt.3n.z
/usr/share/man/cat3/stdipc.3c.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/pkgadd.1m
/usr/share/man/man1/pkginfo.1
/usr/share/man/man2/msgget.2
/usr/share/man/man2/semget.2
/usr/share/man/man2/shmget.2
/usr/share/man/man3/getsockopt.3n
/usr/share/man/man3/stdipc.3c
Man pages removed in this patch:
SAR #719: /usr/share/man/cat4/issue.4.z
/usr/share/man/man4/issue.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-009
Date Issued: 12/11/2002 14:34:45
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 man package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
Enhancements:
1: Option modifications for shmdefine(1)
2: Removed references to other obsolete man pages.
3: SAR #470: xfsdump and xfsrestore man pages were modified to
include a description of the "-H" option (remote dump/restore).
4: Added descriptions of the new kmaleak and strleak commands to
the crash(1M) manpage.
5: SAR #697: Modified the hp.ps.net(5) manual page to document the
new landscape capability.
6: SAR #815: New manual page for the Operating System Messages",
osm(4).
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/shmdefine.1.z
/usr/share/man/cat1/xfsdump.1m.z
/usr/share/man/cat1/xfsrestore.1m.z
/usr/share/man/cat4/osm.4.z
/usr/share/man/cat5/hp.ps.net.5.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/shmdefine.1
/usr/share/man/man1/xfsdump.1m
/usr/share/man/man1/xfsrestore.1m
/usr/share/man/man4/osm.4
/usr/share/man/man5/hp.ps.net.5
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-009
Date Issued: 12/18/2002 17:44:42
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 man package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
Enhancements:
1: Option modifications for shmdefine(1)
2: Removed references to other obsolete man pages.
3: SAR #470: xfsdump and xfsrestore man pages were modified to
include a description of the "-H" option (remote dump/restore).
4: Added descriptions of the new kmaleak and strleak commands to
the crash(1M) manpage.
5: SAR #697: Modified the hp.ps.net(5) manual page to document the
new landscape capability.
6: SAR #815: New manual page for the Operating System Messages",
osm(4).
Object(s) To Be Replaced:
/usr/share/man/cat1/crash.1m.z
/usr/share/man/cat1/shmdefine.1.z
/usr/share/man/cat1/xfsdump.1m.z
/usr/share/man/cat1/xfsrestore.1m.z
/usr/share/man/cat5/hp.ps.net.5.z
/usr/share/man/man1/crash.1m
/usr/share/man/man1/shmdefine.1
/usr/share/man/man1/xfsdump.1m
/usr/share/man/man1/xfsrestore.1m
/usr/share/man/man5/hp.ps.net.5
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-010
Date Issued: 05/08/2003 10:39:16
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 man package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1451: The paddr_to_pci(3X) manpage describes a non-existent
library routine in /usr/lib/libud.a.
R1. The /usr/lib/libud.a routine is actually called phys_to_pci()
and the manpage has been renamed (phys_to_pci(3X)) and edited
to reflect this.
P2. SAR 1473: The trpt(1M) utility has not functioned properly
since the introduction of the Mentat TCP/IP stack. It
cannot find required kernel symbols from the new TCP/IP
stack that were available in the old stack.
R2. A total re-write of the utility would be needed to make it
work with the new TCP/IP stack so it was decided to remove
this utility and it's associated man page from the OS.
Enhancements:
1. Clarified that vme_address(3c) may also be used to obtain
correct corresponding CPU physical address for an A32 VME
address, in addition to A16 and A24 addresses.
2. Changed the description of the "prefix" and "unit" fields in
the interface(4) man page. "prefix" and "unit" must match
the device "module name" and "unit".
Object(s) To Be Replaced:
/usr/share/man/cat1/config_inet.1m.z
/usr/share/man/cat1/mount_memfs.1m.z
/usr/share/man/cat3/phys_to_pci.3x.z
/usr/share/man/cat3/vme_address.3c.z
/usr/share/man/cat4/interface.4.z
/usr/share/man/man1/config_inet.1m
/usr/share/man/man1/mount_memfs.1m
/usr/share/man/man3/phys_to_pci.3x
/usr/share/man/man3/vme_address.3c
/usr/share/man/man4/interface.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
Man pages removed by this patch:
/usr/share/man/man3/paddr_to_pci.3x
/usr/share/man/cat3/paddr_to_pci.3x.z
/usr/share/man/cat1/trpt.1m.z
/usr/share/man/man1/trpt.1m
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-010
Date Issued: 05/08/2003 10:39:16
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-103
Brief Description:
PowerMAX OS 4.3 man package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1451: The paddr_to_pci(3X) manpage describes a non-existent
library routine in /usr/lib/libud.a.
R1. The /usr/lib/libud.a routine is actually called phys_to_pci()
and the manpage has been renamed (phys_to_pci(3X)) and edited
to reflect this.
P2. SAR 1473: The trpt(1M) utility has not functioned properly
since the introduction of the Mentat TCP/IP stack. It
cannot find required kernel symbols from the new TCP/IP
stack that were available in the old stack.
R2. A total re-write of the utility would be needed to make it
work with the new TCP/IP stack so it was decided to remove
this utility and it's associated man page from the OS.
Enhancements:
1. Clarified that vme_address(3c) may also be used to obtain
correct corresponding CPU physical address for an A32 VME
address, in addition to A16 and A24 addresses.
2. Changed the description of the "prefix" and "unit" fields in
the interface(4) man page. "prefix" and "unit" must match
the device "module name" and "unit".
Object(s) To Be Replaced:
/usr/share/man/cat1/config_inet.1m.z
/usr/share/man/cat1/mount_memfs.1m.z
/usr/share/man/cat3/phys_to_pci.3x.z
/usr/share/man/cat3/vme_address.3c.z
/usr/share/man/cat4/interface.4.z
/usr/share/man/man1/config_inet.1m
/usr/share/man/man1/mount_memfs.1m
/usr/share/man/man3/phys_to_pci.3x
/usr/share/man/man3/vme_address.3c
/usr/share/man/man4/interface.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-011
Date Issued: 01/07/2004 16:01:10
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 man package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: A lwp may switch cpus during a spl_request(3X) operation,
which will result in indeterminate results.
R1: The syscx(2) man page has been updated with the new
SET_FCPU_UADDR command, which is used for fixing this problem.
Enhancements:
E1: SAR 1526: Man pages were modified to be consistent with the
source code.
- pci_cfg_read(3x) returns an unsigned long defined by PCI_DWORD, but
the man page said it returned an int.
- pci_cfg_write(3x) returns void, but the man page indicated it
returned an int.
- pci_ivec_info(3x) returns an int, but the man page indicated it
returns void.
Object(s) To Be Replaced:
/usr/share/man/cat2/syscx.2.z
/usr/share/man/man2/syscx.2
/usr/share/man/cat3/pci_cfg_read.3x.z
/usr/share/man/cat3/pci_cfg_write.3x.z
/usr/share/man/cat3/pci_ivec_info.3x.z
/usr/share/man/man3/pci_cfg_read.3x
/usr/share/man/man3/pci_cfg_write.3x
/usr/share/man/man3/pci_ivec_info.3x
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-011
Date Issued: 01/07/2004 16:01:36
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 man package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: A lwp may switch cpus during a spl_request(3X) operation,
which will result in indeterminate results.
R1: The syscx(2) man page has been updated with the new
SET_FCPU_UADDR command, which is used for fixing this problem.
Enhancements:
E1: SAR 1526: Man pages were modified to be consistent with the
source code.
- pci_cfg_read(3x) returns an unsigned long defined by PCI_DWORD, but
the man page said it returned an int.
- pci_cfg_write(3x) returns void, but the man page indicated it
returned an int.
- pci_ivec_info(3x) returns an int, but the man page indicated it
returns void.
Object(s) To Be Replaced:
/usr/share/man/cat2/syscx.2.z
/usr/share/man/man2/syscx.2
/usr/share/man/cat3/pci_cfg_read.3x.z
/usr/share/man/cat3/pci_cfg_write.3x.z
/usr/share/man/cat3/pci_ivec_info.3x.z
/usr/share/man/man3/pci_cfg_read.3x
/usr/share/man/man3/pci_cfg_write.3x
/usr/share/man/man3/pci_ivec_info.3x
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-012
Date Issued: 07/08/2004 10:53:31
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 man package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1564: Linking with the Threads Library and BSD Compatible
Library results in core dumps when the resultant program is
executed.
R1: The following note has been added to the intro(3), intro(3B),
pthread(3pthread), and thread(3thread) manpages.
"Applications linking with the Threads Libraries (-lthread)
should NOT link with the BSD-compatibility Libraries (-lucbc).
The Threads Library signal stack will get corrupted resulting
in a core dump when the program is run."
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat3/intro.3.z
/usr/share/man/cat3/intro.3b.z
/usr/share/man/cat3/pthread.3x.z
/usr/share/man/cat3/thread.3x.z
/usr/share/man/man3/intro.3
/usr/share/man/man3/intro.3b
/usr/share/man/man3/pthread.3x
/usr/share/man/man3/thread.3x
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: man-012
Date Issued: 07/26/2004 11:51:04
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 man package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1564: Linking with the Threads Library and BSD Compatible
Library results in core dumps when the resultant program is
executed.
R1: The following note has been added to the intro(3), intro(3B),
pthread(3pthread), and thread(3thread) manpages.
"Applications linking with the Threads Libraries (-lthread)
should NOT link with the BSD-compatibility Libraries (-lucbc).
The Threads Library signal stack will get corrupted resulting
in a core dump when the program is run."
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat3/intro.3.z
/usr/share/man/cat3/intro.3b.z
/usr/share/man/cat3/pthread.3x.z
/usr/share/man/cat3/thread.3x.z
/usr/share/man/man3/intro.3
/usr/share/man/man3/intro.3b
/usr/share/man/man3/pthread.3x
/usr/share/man/man3/thread.3x
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: man-013
Date Issued: 03/13/2006 14:47:18
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 man package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT2630: shmdefine(1) was missing the descriptione of the -G option.
R1: Added documentation of the -G option.
P2: DT3307: catman was changed in DT2704 to add tbl(1) filtering. A
side effect was that all manpages, with or without tbl directives,
are now filtered with col(1). col(1) has a limitation of 800
characters for lines processed, and 6 man pages produce a single
line that is greater than 800 characters.
R2: Changed the 6 man page troff formatting commands around to make
all lines be less than 800 characters. Also, 10 manual pages
with tbl directives were reformatted for readability.
P3: Description of the -A option to uname(1) was missing.
R3: Added -A description on uname(1).
Enhancements:
1. RIQ010192: The login(1), login(4), loginlog(4), passwd(1),
passwd(4), and yppasswd(1) man pages did not denote the
RESTRICTPWD and RESTRICT keyword changes for passwd(1) and
login(1), respectively. This information was added to the
above listed man pages. Also included was additional
clarification to the login(4) man page for the RIQ.
Object(s) To Be Replaced:
/usr/share/man/cat1/evgainit.1m.z
/usr/share/man/cat1/login.1.z
/usr/share/man/cat1/passwd.1.z
/usr/share/man/cat1/setcolor.1.z
/usr/share/man/cat1/setkey.1.z
/usr/share/man/cat1/shmdefine.1.z
/usr/share/man/cat1/uname.1.z
/usr/share/man/cat1/vidi.1.z
/usr/share/man/cat1/vtlmgr.1.z
/usr/share/man/cat1/yppasswd.1.z
/usr/share/man/cat3/cmn_err.3d.z
/usr/share/man/cat3/dwarf_init.3x.z
/usr/share/man/cat3/dwarf_pclines.3x.z
/usr/share/man/cat3/privname.3h.z
/usr/share/man/cat3/pthr_cancel.3x.z
/usr/share/man/cat3/pthr_creat.3x.z
/usr/share/man/cat3/rmalloc.3d.z
/usr/share/man/cat3/rmalloc_aligned.3d.z
/usr/share/man/cat3/rmalloc_lastaligned.3d.z
/usr/share/man/cat4/login.4.z
/usr/share/man/cat4/loginlog.4.z
/usr/share/man/cat7/display.7.z
/usr/share/man/cat7/keyboard.7.z
/usr/share/man/man1/evgainit.1m
/usr/share/man/man1/login.1
/usr/share/man/man1/passwd.1
/usr/share/man/man1/setcolor.1
/usr/share/man/man1/setkey.1
/usr/share/man/man1/shmdefine.1
/usr/share/man/man1/uname.1
/usr/share/man/man1/vidi.1
/usr/share/man/man1/vtlmgr.1
/usr/share/man/man1/yppasswd.1
/usr/share/man/man3/cmn_err.3d
/usr/share/man/man3/dwarf_init.3x
/usr/share/man/man3/dwarf_pclines.3x
/usr/share/man/man3/privname.3h
/usr/share/man/man3/pthr_cancel.3x
/usr/share/man/man3/pthr_creat.3x
/usr/share/man/man3/rmalloc.3d
/usr/share/man/man3/rmalloc_aligned.3d
/usr/share/man/man3/rmalloc_lastaligned.3d
/usr/share/man/man4/login.4
/usr/share/man/man4/loginlog.4
/usr/share/man/man7/display.7
/usr/share/man/man7/keyboard.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: man-013
Date Issued: 03/13/2006 11:39:51
Software Package: man pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 man package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT2630: shmdefine(1) was missing the descriptione of the -G option.
R1: Added documentation of the -G option.
P2: DT3307: catman was changed in DT2704 to add tbl(1) filtering. A
side effect was that all manpages, with or without tbl directives,
are now filtered with col(1). col(1) has a limitation of 800
characters for lines processed, and 6 man pages produce a single
line that is greater than 800 characters.
R2: Changed the 6 man page troff formatting commands around to make
all lines be less than 800 characters.
P3: Description of the -A option to uname(1) was missing.
R3: Added -A description on uname(1).
Enhancements:
1. RIQ010192: The login(1), login(4), loginlog(4), passwd(1),
passwd(4), and yppasswd(1) man pages did not denote the
RESTRICTPWD and RESTRICT keyword changes for passwd(1) and
login(1), respectively. This information was added to the
above listed man pages. Also included was additional
clarification to the login(4) man page for the RIQ.
Object(s) To Be Replaced:
/usr/share/man/cat1/login.1.z
/usr/share/man/cat1/passwd.1.z
/usr/share/man/cat1/shmdefine.1.z
/usr/share/man/cat1/uname.1.z
/usr/share/man/cat1/vtlmgr.1.z
/usr/share/man/cat1/yppasswd.1.z
/usr/share/man/cat3/cmn_err.3d.z
/usr/share/man/cat3/dwarf_init.3x.z
/usr/share/man/cat3/dwarf_pclines.3x.z
/usr/share/man/cat3/privname.3h.z
/usr/share/man/cat3/pthr_cancel.3x.z
/usr/share/man/cat3/pthr_creat.3x.z
/usr/share/man/cat3/rmalloc.3d.z
/usr/share/man/cat3/rmalloc_aligned.3d.z
/usr/share/man/cat3/rmalloc_lastaligned.3d.z
/usr/share/man/cat4/login.4.z
/usr/share/man/cat4/loginlog.4.z
/usr/share/man/man1/login.1
/usr/share/man/man1/passwd.1
/usr/share/man/man1/shmdefine.1
/usr/share/man/man1/uname.1
/usr/share/man/man1/vtlmgr.1
/usr/share/man/man1/yppasswd.1
/usr/share/man/man3/cmn_err.3d
/usr/share/man/man3/dwarf_init.3x
/usr/share/man/man3/dwarf_pclines.3x
/usr/share/man/man3/privname.3h
/usr/share/man/man3/pthr_cancel.3x
/usr/share/man/man3/pthr_creat.3x
/usr/share/man/man3/rmalloc.3d
/usr/share/man/man3/rmalloc_aligned.3d
/usr/share/man/man3/rmalloc_lastaligned.3d
/usr/share/man/man4/login.4
/usr/share/man/man4/loginlog.4
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: mvc-001
Date Issued: 02/02/2000 14:07:25
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-002
Related SARs: #116
Brief Description:
PowerMAX OS 4.3 mvc package release updates
##############################################################################
Problem Description:
1. #116: Using standard mvc ports, reads return immediately if
conditions satisfy the read (i.e., adequate data are available to be
returned). Using mvcrt ports, the initial read would stall when the
read should have been satisfied.
Additionally, if VMIN=0 and VTIME>0, a read would always wait for the
timeout before returning, even when data were available to satisfy
the read. To be correct, VTIME is the time for completion of the
read, rather than an intercharacter timer, if the read cannot
otherwise be satisfied. The read should complete if either a single
character is read or if the timer expires (no characters returned).
Problem Resolution:
1. mvcrt ports, when opened for reading, now immediately issue a read
command, permitting the first read issued by the user to be
immediately satisfied when data is available.
The VMIN/VTIME code now correctly checks for available data before
waiting for the timeout when VMIN=0 and VTIME>0.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/mvcrt/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: mvc-001
Date Issued: 02/02/2000 13:00:44
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002
Related SARs: #116
Brief Description:
PowerMAX OS 4.3 mvc package release updates
##############################################################################
Problem Description:
1. #116: Using standard mvc ports, reads return immediately if
conditions satisfy the read (i.e., adequate data are available to be
returned). Using mvcrt ports, the initial read would stall when the
read should have been satisfied.
Additionally, if VMIN=0 and VTIME>0, a read would always wait for the
timeout before returning, even when data were available to satisfy
the read. To be correct, VTIME is the time for completion of the
read, rather than an intercharacter timer, if the read cannot
otherwise be satisfied. The read should complete if either a single
character is read or if the timer expires (no characters returned).
Problem Resolution:
1. mvcrt ports, when opened for reading, now immediately issue a read
command, permitting the first read issued by the user to be
immediately satisfied when data is available.
The VMIN/VTIME code now correctly checks for available data before
waiting for the timeout when VMIN=0 and VTIME>0.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/mvcrt/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: mvc-002
Date Issued: 01/02/2002 14:25:53
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 mvc package release updates
##############################################################################
Problem Description:
1. The mvc(7) man page was vague regarding the mvc tuneables and how
they should be accessed.
Problem Resolution:
2. The new version of mvc(7) man page has been clarified.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat7/mvc.7.z
/usr/share/man/man7/mvc.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: mvc-002
Date Issued: 01/02/2002 14:25:53
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 mvc package release updates
##############################################################################
Problem Description:
1. The mvc(7) man page was vague regarding the mvc tuneables and how
they should be accessed.
Problem Resolution:
2. The new version of mvc(7) man page has been clarified.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/share/man/cat7/mvc.7.z
/usr/share/man/man7/mvc.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: mvc-003
Date Issued: 01/07/2004 16:02:00
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 mvc package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: When more than one mvc is present in a system, the configuration
of the mvc only permits all boards to be configured similarly --
it is not possible to configure one mvc to use (say) RS232 for
all ports, and have a second mvc configured to use (say) RS485
for any of that boards ports.
R1: Decompose configuration information to delineate configuration
on a per board basis, rather than using 'global' settings that
apply to all boards.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/mvc
/etc/conf/pack.d/mvc/Driver.o
/etc/conf/pack.d/mvc/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: mvc-003
Date Issued: 01/07/2004 16:02:27
Software Package: mvc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 mvc package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: When more than one mvc is present in a system, the configuration
of the mvc only permits all boards to be configured similarly --
it is not possible to configure one mvc to use (say) RS232 for
all ports, and have a second mvc configured to use (say) RS485
for any of that boards ports.
R1: Decompose configuration information to delineate configuration
on a per board basis, rather than using 'global' settings that
apply to all boards.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/mvc
/etc/conf/pack.d/mvc/Driver.o
/etc/conf/pack.d/mvc/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: ncr-001
Date Issued: 03/29/2001 12:02:47
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #492
Brief Description:
PowerMAX OS 4.3 ncr package release updates.
#############################################################################
Problem Description:
1. The warning message "NCR: ncr_intr (ncr1): Interupt routine entered
with interrupts disabled" is output on the console.
Problem Resolution:
1. The warning message "NCR: ncr_intr (ncr1): Interupt routine entered
with interrupts disabled" is output on the console whenever:
- an ncr disk shared an interrupt vector with another device,
AND
- interrupt activity is occurring for the other device,
AND
- the ncr disk was being used as a shared xfsd disk.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ncr-001
Date Issued: 03/28/2001 19:12:53
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #492
Brief Description:
PowerMAX OS 4.3 ncr package release updates.
##############################################################################
Problem Description:
1. The warning message "NCR: ncr_intr (ncr1): Interupt routine entered
with interrupts disabled" is output on the console.
Problem Resolution:
1. The warning message "NCR: ncr_intr (ncr1): Interupt routine entered
with interrupts disabled" is output on the console whenever:
- an ncr disk shared an interrupt vector with another device,
AND
- interrupt activity is occurring for the other device,
AND
- the ncr disk was being used as a shared xfsd disk.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ncr-002
Date Issued: 06/11/2001 08:09:45
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II/III, Motorola MCP750
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 ncr package release updates
##############################################################################
Problem Description:
1. Driver uses async start unit command to spin up disk drives and
polls once a second until it is ready or times out.
2. Driver needed the ability to handle lists of Physical DMA addresses.
(support new partial sector capability).
3. Driver/SCSI tag queuing can re-arrange order of I/O processing which
needs to be disabled(ordered) for RMW partial sector writes.
Problem Resolution:
1. Modified Driver to spin up disks correctly.
2. Modified Driver to support new partial sector capability.
3. Modified Driver for proper ordering of SCSI tag queuing.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: ncr-002
Date Issued: 06/11/2001 08:37:57
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 ncr package release updates
##############################################################################
Problem Description:
1. Driver uses async start unit command to spin up disk drives and
polls once a second until it is ready or times out.
2. Driver needed the ability to handle lists of Physical DMA addresses.
(support new partial sector capability).
3. Driver/SCSI tag queuing can re-arrange order of I/O processing which
needs to be disabled(ordered) for RMW partial sector writes.
Problem Resolution:
1. Updated driver to support async startup.
2. Updated driver for ability to handle lists of Physical DMA addresses.
3. Disabled tag queuing for RMW partial sector writes.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: ncr-003
Date Issued: 07/08/2004 10:54:17
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 ncr package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1584/1680: "PANIC: NCR: ncr_dstat_fatal (ncr0) NCR:
ncr_dstat_fatal: Unsolicited DMA interrupt (dstat: 0x00000080)"
during a file copy.
R1: If we get a DMA interrupt from the NCR chip but the DSTAT reg
does not identify a DMA interrupt source, then we ignore the
interrupt and return forthwith. Previously, the code would
just panic... ignoring this condition seems to have no effect
except to avoid what is apparently a bogus panic.
While this problem can occure on PowerMAX OS 4.3 and 5.1
revisions, it was originally seen during the testing for
PowerMAX OS 6.1 of the 4-port VQPM video card mounted on
the PBX3 on a Power Hawk 920. It may or may not happen every
time X is started on the graphics port. Sometimes it took as
many as 3 attempts before the PANIC would occur. It was also
seen at a customer's site on a 5.1SR6 POwer Hawk 740 system
without the graphics card.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: ncr-003
Date Issued: 07/26/2004 11:51:37
Software Package: ncr pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 ncr package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1584/1680: "PANIC: NCR: ncr_dstat_fatal (ncr0) NCR:
ncr_dstat_fatal: Unsolicited DMA interrupt (dstat: 0x00000080)"
during a file copy.
R1: If we get a DMA interrupt from the NCR chip but the DSTAT reg
does not identify a DMA interrupt source, then we ignore the
interrupt and return forthwith. Previously, the code would
just panic... ignoring this condition seems to have no effect
except to avoid what is apparently a bogus panic.
While this problem can occure on PowerMAX OS 4.3 and 5.1
revisions, it was originally seen during the testing for
PowerMAX OS 6.1 of the 4-port VQPM video card mounted on
the PBX3 on a Power Hawk 920. It may or may not happen every
time X is started on the graphics port. Sometimes it took as
many as 3 attempts before the PANIC would occur. It was also
seen at a customer's site on a 5.1SR6 POwer Hawk 740 system
without the graphics card.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ncr/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: netcmds-001
Date Issued: 08/22/2001 14:54:01
Software Package: netcmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 netcmds package release updates
#############################################################################
Problem Description:
1. The wrong type of "stat" call was causing sendmail to think
that there was no space available in /var/spool/mqueue.
Problem Resolution:
1. The #defines in sendmail where fixed for freespace() so
statvfs() is called instead of statfs().
Enhancements:
None
Object(s) To Be Replaced:
/usr/ucblib/sendmail
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: netcmds-001
Date Issued: 09/11/2001 16:25:03
Software Package: netcmds pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 netcmds package release updates
#############################################################################
Problem Description:
1. The wrong type of "stat" call was causing sendmail to think
that there was no space available in /var/spool/mqueue.
Problem Resolution:
1. The #defines in sendmail where fixed for freespace() so
statvfs() is called instead of statfs().
Enhancements:
None
Object(s) To Be Replaced:
/usr/ucblib/sendmail
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nfs-001
Date Issued: 05/30/2000 13:10:02
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 nfs package release updates
##############################################################################
Problem Description:
1. NFS problems would lock up a disk partition (i.e. it would become
inaccessible locally).
Problem Resolution:
1. The problem was due to a r/w sleep lock that would become locked more
than once by the same lwp (write and read). This would cause the
lock to remain locked forever.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfssrv/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nfs-001
Date Issued: 05/30/2000 13:50:30
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 nfs package release updates
##############################################################################
Problem Description:
1. NFS problems would lock up a disk partition (i.e. it would become
inaccessible locally).
Problem Resolution:
1. The problem was due to a r/w sleep lock that would become locked more
than once by the same lwp (write and read). This would cause the
lock to remain locked forever.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfssrv/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nfs-002
Date Issued: 03/18/2002 11:32:13
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #627
Brief Description:
PowerMAX OS 4.3 nfs package release updates
##############################################################################
Problem Description:
1. SAR #627: When one process is creating a file that is also being
accessed at the same time as a NFSed file with mlockall() and
mmap() calls, then a deadlock can occur between the segvn driver
and the nfs driver.
This particular scenario occurs especially (but not only) on a
Closely Coupled system, where Posix shared memory files for a
client SBC actually exist on the server SBC across a NFS/network
connection. One process can be creating the shared memory file
while another process is trying to open this shared memory file
to mmap() it and to lock it down in memory.
The problem occurs in segvn_lockop()-segvn_fault()-nfs_getpage()
when nfs_getpage() decides to either:
- purge the attr cache
or
- abort all mappings due to stale filehandle,
where these operations attempt to remove all page objects
that belong to the vnode via nfs_purge_caches().
However, there are pages currently locked by segvn_fault() that
are outside the current range of the nfs_getpage() parameters, and
the nfs code waits forever on a page lock in pvn_getdirty_range() or
pvn_abort_range() when it tries to remove all the page mappings
to the current vnode. The page happens to already be locked
by this same LWP from above in the segvn_fault() routine.
Problem Resolution:
1. The fix is to limit the range of pages that segvn_lockop()
passes to segvn_fault() so that segvn_fault() will not be
holding onto any pages outside of the current range of pages
that nfs_getpage() is operating on. This limitation is only
enforced for NFS-based segments.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nfs-002
Date Issued: 03/18/2002 11:40:10
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #627
Brief Description:
PowerMAX OS 4.3 nfs package release updates
##############################################################################
Problem Description:
1. SAR #627: When one process is creating a file that is also being
accessed at the same time as a NFSed file with mlockall() and
mmap() calls, then a deadlock can occur between the segvn driver
and the nfs driver.
This particular scenario occurs especially (but not only) on a
Closely Coupled system, where Posix shared memory files for a
client SBC actually exist on the server SBC across a NFS/network
connection. One process can be creating the shared memory file
while another process is trying to open this shared memory file
to mmap() it and to lock it down in memory.
The problem occurs in segvn_lockop()-segvn_fault()-nfs_getpage()
when nfs_getpage() decides to either:
- purge the attr cache or
- abort all mappings due to stale filehandle,
where these operations attempt to remove all page objects
that belong to the vnode via nfs_purge_caches().
However, there are pages currently locked by segvn_fault() that
are outside the current range of the nfs_getpage() parameters, and
the nfs code waits forever on a page lock in pvn_getdirty_range() or
pvn_abort_range() when it tries to remove all the page mappings
to the current vnode. The page happens to already be locked
by this same LWP from above in the segvn_fault() routine.
Problem Resolution:
1. The fix is to limit the range of pages that segvn_lockop()
passes to segvn_fault() so that segvn_fault() will not be
holding onto any pages outside of the current range of pages
that nfs_getpage() is operating on. This limitation is only
enforced for NFS-based segments.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nfs-003
Date Issued: 07/10/2002 14:55:32
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #731
Brief Description:
PowerMAX OS 4.3 nfs package release updates
##############################################################################
P1: SAR #731: These are fixes for using a NFS swapfile on a CCS or netboot
client.
The system would panic or hang, or the user application would abort with
SIGBUS, SIGSEGV or SIGILL terminations.
R1: Four changes were made to the nfs code in areas related to nfs swapping
support:
- the pageout daemons auth structure allocations/deallocations
- nfs attr stale cache for swap files
- nfs_putpagelist
- nfs_getpaglist
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
/usr/include/nfs/nfs_clnt.h
/usr/include/nfs/rnode.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: nfs-003
Date Issued: 07/24/2002 14:48:17
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #731
Brief Description:
PowerMAX OS 4.3 nfs package release updates
###############################################################################
P1: SAR #731: These are fixes for using a NFS swapfile on a CCS or netboot
client.
The system would panic or hang, or the user application would abort with
SIGBUS, SIGSEGV or SIGILL terminations.
R1: Four changes were made to the nfs code in areas related to nfs swapping
support:
- the pageout daemons auth structure allocations/deallocations
- nfs attr stale cache for swap files
- nfs_putpagelist
- nfs_getpaglist
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
/usr/include/nfs/nfs_clnt.h
/usr/include/nfs/rnode.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: nfs-004
Date Issued: 07/08/2004 10:54:25
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 nfs package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1553: Random nulls were being written to an nfs mounted file.
R1: Write data from user space was being copied to kernel space
by an nfs call to uiomove() when a page boundary was crossed
and a page fault occurred. A portion of the write data had
already been placed in the first page's buffer, but the write
pointer had not yet been updated. This caused a partial write
of that first page instead of a full page write. After nfs
processed of the page fault, it also determined that it should
purge its' caches. The first page was read back into the same
buffer over the top of the new data that had just been placed
in the buffer but not written out yet.
The nfs write code that calls uiomove was changed to break up
the one uiomove call into multiple calls so any one call does
not cross a page boundary. The write pointer is updated after
each subsection is moved. This keeps the nfs page fault handling
code from overwriting any partially completed write data.
P2: DT 1573: Some console messages from NFS were not printing
timestamps because they were CE_CONT type logging.
R2: Change these messages from CE_CONT to CE_NOTE type logging in
order to get timestamps.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
/etc/conf/pack.d/nfscmn/Driver.o
/etc/conf/pack.d/nfss/Driver.o
/etc/conf/pack.d/nfssrv/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: nfs-004
Date Issued: 07/26/2004 11:51:44
Software Package: nfs pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 nfs package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1553: Random nulls were being written to an nfs mounted file.
R1: Write data from user space was being copied to kernel space
by an nfs call to uiomove() when a page boundary was crossed
and a page fault occurred. A portion of the write data had
already been placed in the first page's buffer, but the write
pointer had not yet been updated. This caused a partial write
of that first page instead of a full page write. After nfs
processed of the page fault, it also determined that it should
purge its' caches. The first page was read back into the same
buffer over the top of the new data that had just been placed
in the buffer but not written out yet.
The nfs write code that calls uiomove was changed to break up
the one uiomove call into multiple calls so any one call does
not cross a page boundary. The write pointer is updated after
each subsection is moved. This keeps the nfs page fault handling
code from overwriting any partially completed write data.
P2: DT 1573: Some console messages from NFS were not printing
timestamps because they were CE_CONT type logging.
R2: Change these messages from CE_CONT to CE_NOTE type logging in
order to get timestamps.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/nfs/Driver.o
/etc/conf/pack.d/nfscmn/Driver.o
/etc/conf/pack.d/nfss/Driver.o
/etc/conf/pack.d/nfssrv/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: ngpib-001
Date Issued: 03/13/2006 11:45:01
Software Package: ngpib pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 ngpib package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3305: Earlier task removed files ngpib.mk, config/config.mk,
and lib/lib.mk which broke PowerMAX OS builds. Also, during
testing it appeared that the Makefile in the test directory had
the wrong pathname for the ngpib library. These issues were
corrected.
R1: Added the 3 files back in, and changed LIBDIR=../lib/ngpib to
LIBDIR=/usr/lib in the test/Makefile.
Enhancements:
1. Additional tests were added, ngpib_iotech.c and
ngpib_iotech_two.c, along with their header files.
Object(s) To Be Replaced:
/usr/etc/ngpibconfig
/usr/lib/libngpib.a
/usr/src/drivers/ngpib/Makefile
/usr/src/drivers/ngpib/Makefile.test
/usr/src/drivers/ngpib/ngpib_iotech.c
/usr/src/drivers/ngpib/ngpib_iotech.h
/usr/src/drivers/ngpib/ngpib_iotech_two.c
/usr/src/drivers/ngpib/ngpib_iotech_two.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nsu-001
Date Issued: 09/15/1999 11:09:52
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: HM12682
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. HM12682: Running a lot of socket processes causes a slow memory leak
that eventually hangs the system. An mblk_t structure that was
allocated by the sockmod read service routine to handle an error
message when a disconnect condition occurred was not being
deallocated.
Problem Resolution:
1. The sockmod read service routine sockmodrsrv() allocates a set of
four data structures in case it needs to queue the M_ERROR message
behind any messages already on the queue. If the queue is empty, the
M_ERROR message can be sent immediately and these data structures are
not needed. so_toss() was only deallocating three of the four
structures that had been allocated. so_toss() was modified to
deallocate all four mblk_t structures.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nsu-001
Date Issued: 09/15/1999 13:16:10
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: HM12682
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. HM12682: Running a lot of socket processes causes a slow memory leak
that eventually hangs the system. An mblk_t structure that was
allocated by the sockmod read service routine to handle an error
message when a disconnect condition occurred was not being
deallocated.
Problem Resolution:
1. The sockmod read service routine sockmodrsrv() allocates a set of
four data structures in case it needs to queue the M_ERROR message
behind any messages already on the queue. If the queue is empty, the
M_ERROR message can be sent immediately and these data structures are
not needed. so_toss() was only deallocating three of the four
structures that had been allocated. so_toss() was modified to
deallocate all four mblk_t structures.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nsu-002
Date Issued: 05/30/2000 13:10:26
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: cnd-003, dec-002, egl-002, ip-002, pg-002
Related SARs: 201, 206, 248
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
2. SAR 206: A slow memory leak can occur when using sockets.
Occasionally a call to sockmodclose() can occur before all of the
kernel users of the data structures associated with the socket have
finished using them. In certain situations, 3 of the structures
associated with a socket would not be de-allocated and a memory leak
would occur.
3. SAR 248: The maximum UDP transfer size of 16 KB was too small. The
maximum transfer size of the UDP protocol is 64KB.
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
2. The 3 socket structures that would not be de-allocated were
originally allocated during the sockmodopen(). sockmodclose() was
detecting that there was at least one outstanding user of the
structures and would exit without de-allocating them, expecting the
last structure user to de-allocate them when it was done.
free_zero() was modified to de-allocate the 3 socket structures if it
is the last user of them and the sockmodclose() for the socket has
already occurred.
3. Connectionless UDP socket transfer requests can now be up to 64 KB in
length. The amount of user data that can now be transferred is
slightly less than 64 KB however, because 28 bytes must be used for
the header. Thus, the maximum user data transfer with connectionless
UDP is now 65507 bytes ((64 KB - 1) - 28).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
/usr/include/sys/dlpi_common.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: nsu-002
Date Issued: 05/30/2000 13:50:51
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: cnd-003, dec-002, egl-002, ie-002, pg-002
Related SARs: 201, 206, 248
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
2. SAR 206: A slow memory leak can occur when using sockets.
Occasionally a call to sockmodclose() can occur before all of the
kernel users of the data structures associated with the socket have
finished using them. In certain situations, 3 of the structures
associated with a socket would not be de-allocated and a memory leak
would occur.
3. SAR 248: The maximum UDP transfer size of 16 KB was too small. The
maximum transfer size of the UDP protocol is 64KB.
Problem Resolution:
1. The N_SAPS value was increased to a more appropriate 64 channels.
2. The 3 socket structures that would not be de-allocated were
originally allocated during the sockmodopen(). sockmodclose() was
detecting that there was at least one outstanding user of the
structures and would exit without de-allocating them, expecting the
last structure user to de-allocate them when it was done.
free_zero() was modified to de-allocate the 3 socket structures if it
is the last user of them and the sockmodclose() for the socket has
already occurred.
3. Connectionless UDP socket transfer requests can now be up to 64 KB in
length. The amount of user data that can now be transferred is
slightly less than 64 KB however, because 28 bytes must be used for
the header. Thus, the maximum user data transfer with connectionless
UDP is now 65507 bytes ((64 KB - 1) - 28).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
/usr/include/sys/dlpi_common.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: nsu-003
Date Issued: 01/02/2002 14:26:13
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #566
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. SAR #566: The kernel occasionally PANICs at lock_nodbg+0x60 when
called by sockmodwsrv+0x1040 or sockmodwsrv+0x9c8. This occurs
when using a "no-delay" socket connect.
Problem Resolution:
1. A race condition caused sockmodwsrv to misinterrupt an M_PROTO
large packet on its queue. A socket "bigmsg" was being left on
sockmod's write queue after it was supposed to have been removed
along with clearing it's associated pointer and flag. The
bigmsg is used for flow control to stop write requests from
continuing down the queue until a connection response (good or bad)
is received. This mechanism is only used for "no delay" connects.
A race condition was occurring between sockmodwsrv and one of the
set of functions that should be removing the bigmsg. These
removers include sockmodwsrv, sockmodclose, sockmodrput and
sockmodwput.
The race condition occurred because sockmodwsrv did not lock the
so_lock for the entire duration of the getq call and subsequent
check for S_WRDISABLE. This allowed sockmodwsrv to take the
bigmsg off the queue at the same time one of the "removers" held
the so_lock and was attemping to remove the bigmsg. The bigmsg
contained random data that sockmodwsrv was trying to interpret as
a structure with pointers to kernel data structures. When
sockmodwsrv would use the random data as a pointer the system
would panic.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
/usr/include/sys/tihdr.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: nsu-003
Date Issued: 01/02/2002 14:26:13
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #566
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Problem Description:
1. SAR #566: The kernel occasionally PANICs at lock_nodbg+0x60 when
called by sockmodwsrv+0x1040 or sockmodwsrv+0x9c8. This occurs
when using a "no-delay" socket connect.
Problem Resolution:
1. A race condition caused sockmodwsrv to misinterrupt an M_PROTO
large packet on its queue. A socket "bigmsg" was being left on
sockmod's write queue after it was supposed to have been removed
along with clearing it's associated pointer and flag. The
bigmsg is used for flow control to stop write requests from
continuing down the queue until a connection response (good or bad)
is received. This mechanism is only used for "no delay" connects.
A race condition was occurring between sockmodwsrv and one of the
set of functions that should be removing the bigmsg. These
removers include sockmodwsrv, sockmodclose, sockmodrput and
sockmodwput.
The race condition occurred because sockmodwsrv did not lock the
so_lock for the entire duration of the getq call and subsequent
check for S_WRDISABLE. This allowed sockmodwsrv to take the
bigmsg off the queue at the same time one of the "removers" held
the so_lock and was attemping to remove the bigmsg. The bigmsg
contained random data that sockmodwsrv was trying to interpret as
a structure with pointers to kernel data structures. When
sockmodwsrv would use the random data as a pointer the system
would panic.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
/usr/include/sys/tihdr.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nsu-004
Date Issued: 03/18/2002 11:32:35
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
Object(s) To Be Replaced:
/usr/include/sys/dlpi_common.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nsu-004
Date Issued: 03/18/2002 11:40:25
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
Object(s) To Be Replaced:
/usr/include/sys/dlpi_common.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: nsu-005
Date Issued: 07/10/2002 14:56:00
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: #679
Brief Description:
PowerMAX OS 4.3 nsu package release updates
##############################################################################
P1: SAR #679: System panics in mi_timer() during a tcp_xclose().
R1: The user application was calling accept with the no-delay option set on
the socket fd. The system was "panicing" during a tcp_xclose() because
the address of the queue's tcp pointer was changed while tcp_xclose()
was waiting for a lock. When the wait returned, the old tcp was no
longer valid. This caused a series of functions to be called with bad
pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: nsu-005
Date Issued: 07/24/2002 14:48:56
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: #679
Brief Description:
PowerMAX OS 4.3 nsu package release updates
###############################################################################
P1: SAR #679: System panics in mi_timer() during a tcp_xclose().
R1: The user application was calling accept with the no-delay option set on
the socket fd. The system was panic'ing during a tcp_xclose() because
the address of the queue's tcp pointer was changed while tcp_xclose()
was waiting for a lock. When the wait returned, the old tcp was no
longer valid. This caused a series of functions to be called with bad
pointer values.
Code was added to the I_FDINSERT processing in strfdinsert() to leave
the file descriptor reference count > 0 if the connect response message
is sent downstream. This will cause all close calls on the stream to be
delayed until the connect response message has been completely processed
by sockmod and tcp. When the message is finally freed fd reference count
will be decremented which will allow close calls for that stream to
proceed.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/sockmod/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: nsu-006
Date Issued: 03/13/2006 14:48:14
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 nsu package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT4378: A streams buffer is freed twice causing system panic.
(Other related Bugzillas: DT1826, DT1868, DT1870, DT1888)
R1: ptmclose() performed the first free of the buffer, but a race
condition existed between ptmclose() and ptsclose(). ptsclose()
releases the lock, sends the buffer upstream, retakes the lock,
and clears the pointer. During the time ptsclose() does not
hold the lock, ptmclose() is able to take the lock and free the
buffer. ptsclose() was modified to clear the pointer before
releasing the lock, and ptmclose() was modified to check for a
null pointer before calling freemsg().
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ptm/Driver.o
/etc/conf/pack.d/pts/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: nsu-006
Date Issued: 03/13/2006 11:45:18
Software Package: nsu pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: base-016
Brief Description: PowerMAX_OS 4.3 nsu package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT4378: A streams buffer is freed twice causing system panic.
(Other related Bugzillas: DT1826, DT1868, DT1870, DT1888)
R1: ptmclose() performed the first free of the buffer, but a race
condition existed between ptmclose() and ptsclose(). ptsclose()
releases the lock, sends the buffer upstream, retakes the lock,
and clears the pointer. During the time ptsclose() does not
hold the lock, ptmclose() is able to take the lock and free the
buffer. ptsclose() was modified to clear the pointer before
releasing the lock, and ptmclose() was modified to check for a
null pointer before calling freemsg().
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ptm/Driver.o
/etc/conf/pack.d/pts/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: oam-001
Date Issued: 07/18/2002 14:53:09
Software Package: oam pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 oam package release updates
##############################################################################
Enhancements:
1. Modified header file so the building of the end objects
requiring devmgmt.h will complete without WARNINGS.
Object(s) To Be Replaced:
/usr/include/devmgmt.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: oam-001
Date Issued: 07/24/2002 14:49:15
Software Package: oam pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 oam package release updates
###############################################################################
Enhancements:
1. Modified header file so the building of the end objects
requiring devmgmt.h will complete without WARNINGS.
Object(s) To Be Replaced:
/usr/include/devmgmt.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-001
Date Issued: 02/02/2000 14:07:46
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-002, cnd-002, dec-001, egl-001, ip-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. config(1M) does not display pg (Peregrine) description correctly.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. The %%Extended%% fields were incorrectly labeled as %%Keyword%%
fields in the pg Mtune(4) file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/dlpi_pg.h
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-001
Date Issued: 02/02/2000 13:01:20
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: base-002, cnd-002, dec-001, egl-001, ie-001
Related SARs: #39
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. #39: One byte SAPs were not processed correctly on FDDI causing the
destination SAP to be based on random user data. This caused the
following problems:
- data was thrown away even though a properly bound SAP existed
- data was sent to the wrong SAP
- the amount of data returned to the user was incorrect
2. #39: The destination SAP specified in the DL_UNITDATA_REQ was being
ignored for two-byte SAP messages. For the one-byte SAP case, it
always assumed that a destination SAP was specified in the
dl_unitdata_req_t even when it had not been.
3. #39: Because both 802.2 and Ethernet II format messages are
supported by the PowerMAX OS Ethernet drivers, it is impossible to
support SAPs in the range 0x100 to 0x5dc correctly. PowerMAX OS did
not report an error if SAPs in this range were used for DL_ETHER.
4. #39: DL_TEST_REQ messages do not work correctly.
5. #39: The mac_type field being returned by DL_INFO_REQ was
incorrectly set to DL_CSMACD instead of the board type (DL_ETHER or
DL_FDDI) if the DL_INFO_REQ was issued before it had done the bind to
the desired SAP.
6. #39: Data packets were being dropped when using DLPI with network
buffers to achieve maximum throughput over 100BaseT dec Ethernet and
pg FDDI.
7. #39: When a VME device board (i.e pg FDDI) was configured at a
VMEIRQ level other than level 5 on a PowerHawk 620 or 640, the system
would either:
- panic in the dmac interrupt routine dmac_intr()
- hang CPU 0 and the console
when a large amount of data was transferred over the pg FDDI
interface at high rates. These hangs and panics could also occur
when using other VME devices (i.e cnd).
8. config(1M) does not display pg (Peregrine) description correctly.
Problem Resolution:
1. The code was using the wrong message offsets to determine the
destination SAP for one-byte SAPs for both FDDI and Ethernet.
The format of the header for a one-byte SAP message is different than
a two-byte SAP message header. This caused the destination SAP to be
based on random user data, which may or may not match one of the
bound SAPs. The code now uses the actual message type (one-byte or
two-byte format) to process the message, instead of assuming the
message type based on the type of SAP that ends up as the primary
receiver (it may not be the same type as the message). This can
occur if the promiscuous SAP is the only receiver.
2. The destination SAP specified in the unitdata_req was being ignored
for two-byte SAP messages. For the one-byte SAP case, it always
assumed that a destination SAP was specified in the dl_unitdata_req_t
even when it had not been. If a two-byte destination SAP was
incorrectly specified when the source was a one-byte SAP, it would
use the first byte of the two-byte SAP specified as the destination
instead of producing an error. If a destination is not specified,
the destination SAP will now default to be the source SAP number.
Both cases now check to make sure that the source and destination
SAPs are the same type (one-byte vs two-byte).
3. For SAP numbers in the range 0x100 to DL_MAX_PACKET (= 0x5dc):
- The sending side was formatting the message in the two-byte
SAP format.
- The receiving side was interpreting it as a one-byte SAP
format.
When both 802.2 and Ethernet II format messages are supported, it is
impossible to support SAPs in the range 0x100 to 0x5dc correctly.
DLunitdata_req() was changed so it will not allow bind requests to
SAPS in the range 0x100 to 0x5dc. It will now return an error,
rather than sending a message that cannot be interpreted correctly by
the receiving side. This problem occurred only with the Ethernet
drivers. FDDI uses different header format standards that allow a
message to be properly identified for all possible one and two-byte
SAP values.
4. A PowerMAX OS 4.3 change modified the order of the promiscous SAP on
the board's valid sap list (SAPs that have been bound to). Prior to
PowerMAX OS 4.3, the NULL SAP (SAP 0) would be in the valid SAP list
immediately before the promiscous SAP. With the change to 4.3, the
NULL SAP was placed in the list after the promiscous SAP. This can
cause a difference in the SAP that is selected by DLrecv() to receive
the message.
DL_TEST_REQ messages do not work correctly on PowerMAX OS 4.3 with
this SAP ordering if there is a user bound to the promiscuous SAP.
The promiscuous SAP would be chosen as the primary recepient of the
message by DLrecv() instead of the proper destination one-byte SAP.
DLrsrv() was incorrectly processing the message as a two-byte SAP
header. DLproc_llc() that handles the special processing for test
messages was never called.
The ordering of the SAPs done by DLinsert_sap() was changed back to
the way it was prior to 4.3. (i.e NULL SAP then promiscuous SAP).
DLrsrv() was changed to allocate an indication message for the
TEST/XID cases if there is a SAP that wants to receive it in addition
to the primary destination SAP. A user bound to the promiscuous SAP
will now receive all messages, including all test messages.
5. The mac_type field being returned by DL_INFO_REQ was incorrectly set
to DL_CSMACD instead of DL_FDDI if the DL_INFO_REQ was issued before
it had done the bind to the desired SAP. The board type in the
DL_sap_t structure being created was not being initialized during the
DLopen(). If a SAP structure had originally been bound to a one byte
SAP and then closed, the next open that used the DL_sap_t structure
would inherit the first bind's type (DL_CSMACD) instead of the board
type (DL_FDDI or DL_ETHER). Bind will change this later to DL_CSMACD
if it gets bound to a one byte SAP.
6. Data packets were being dropped when using DLPI with network buffers
to achieve maximum throughput over 100BaseT dec Ethernet and pg FDDI.
- The DLrecv() routine could not queue some of the incoming data
packets to the pg device queue because it was full. The flow
control q_hiwat value on that queue was too small. If these
packets were not processed by the read service routine DLrsrv()
quick enough, the interrupt routine would discard new packets that
arrived while the queue was full. The q_hiwat value for pg was
changed from (16 * DL_MAX_PACKET) to (96 * DL_MAX_PACKET).
- The DLrsrv() routine was not able to queue some of the packets to
the stream head because its queue was also full. The flow control
q_hiwat value on the stream head was too small. If the receiving
user process did not read the packets fast enough, the next packet
received by DLrsrv() from the device queue would be discarded.
DLopen() was changed to set the stream head's q_hiwat to be the
same value as the q_hiwat for the device's queue. This change
effects all Ethernet and FDDI devices.
- The Receive Load Adjustment constant that is downloaded to the pg
FDDI board was too large. This value defines the maximum number of
packets that will be queued by the board to the host on one
interrupt. The original value for PG_RX_LOADK was 8 which meant up
to 8 packets would be queued to the pg interrupt routine at one
time for an interrupt. At a size of 8, packets were being dropped
by the board because it was running out of available receive
buffers. With the new value of 4, the kernel is able to start
processing the received packets sooner and replenish the supply of
receive buffers fast enough so the pg FDDI board does not drop any
packets when the transfer rate is high.
7. The cause of these problems was not actually in the VME device
drivers but the dmac_intr() and the Xintr() routines. A problem in
dmac_intr() caused the panic to occur. It assumed that its queue of
work to do would never be empty (i.e point to itself) when it was
called. If this occurred, dmac_intr() would branch to location 0
thereby causing the system to panic. Checks for this condition were
added to dmac_intr() that increment a counter and return to its
caller. The system hang was caused by a problem in Xintr() where it
processes a VME interrupt using the Universe II Tundra registers.
The MPIC chip was not being cleared properly in certain situations
before the exit from the Xintr() routine. This caused CPU 0 and the
console to hang.
8. The %%Extended%% fields were incorrectly labeled as %%Keyword%%
fields in the pg Mtune(4) file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/dlpi_pg.h
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-002
Date Issued: 05/30/2000 13:11:27
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, Power Hawk 640
Related Patches: cnd-003, dec-002, egl-002, ip-002, nsu-002
Related SARs: 137, 201
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. SAR 137: Allow configurability of up to four FDDI boards per system.
Previously only three Peregrine FDDI boards could be configured per
system.
2. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The pg driver was enhanced to allow a fourth FDDI board on any
system.
2. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/pg
/etc/conf/mtune.d/pg
/etc/conf/node.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/dlpi_fddi.h
/usr/include/sys/dlpi_pg.h
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
1. There is a possible address conflict in /etc/conf/sadapters.d/kernel
between the fourth Peregrine FDDI board (pg3) and the first Eagle
Ethernet board (egl0). This conflict can be avoided by ensuring that
both devices, if configured, are not assigned the same address.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-002
Date Issued: 05/30/2000 13:51:23
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: cnd-003, dec-002, egl-002, ie-002, nsu-002
Related SARs: 137, 201
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. SAR 137: Allow configurability of up to four FDDI boards per system.
Previously only three Peregrine FDDI boards could be configured per
system.
2. SAR 201: The Ethernet drivers have N_SAPS (# Service Access Ports)
limited to 8. The hard limit of 8 for the N_SAPS value can give the
following message in some applications:
open: channel number out of range
Problem Resolution:
1. The pg driver was enhanced to allow a fourth FDDI board on any
system.
2. The N_SAPS value was increased to a more appropriate 64 channels.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/pg
/etc/conf/mtune.d/pg
/etc/conf/node.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/dlpi_fddi.h
/usr/include/sys/dlpi_pg.h
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
1. There is a possible address conflict in /etc/conf/sadapters.d/kernel
between the fourth Peregrine FDDI board (pg3) and the first Eagle
Ethernet board (egl0). This conflict can be avoided by ensuring that
both devices, if configured, are not assigned the same address.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-003
Date Issued: 10/05/2000 13:55:13
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, diskless-004, egl-003,
fibre-003, inet-004, ip-003, man-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the pg driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The pg driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
2. The flow control limit of the number of outstanding packets that can
be queued to each queue was set too small. Packets would be
discarded when this queue hiwater limit was reached. This could
occur on either the transmit or receive side. Changed the default
value for the pg driver's flow control queue hiwater level from 40 to
8192 packets. Added the capability to change the queue's hiwater
level for the pg driver in its space.c file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/dlpi_pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: pg-003
Date Issued: 10/05/2000 14:09:19
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-004, cmds-003, cnd-004, crosslibs-003, crypt-002,
crypt-int-002, dec-003, egl-003, ie-003, inet-004,
man-003
Related SARs: 39
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Problem Description:
1. SAR 39: Packets were being sent by the network drivers to the wire
out-of-order from the sequence that they were queued to the network
drivers.
2. SAR 39: Packets were being dropped by the pg driver when using
network buffers (NBUFFs) with either DLPI operations or datagram
sockets (UDP) when transmitting at high transfer rates.
Problem Resolution:
1. Multiple changes were made to fix this problem:
- A check for the TX_QUEUED flag was added to the common DLPI
code in DLunitdata_req() that checks for TX_BUSY before it
calls DLxmit_packet(). Without the check for the TX_QUEUED
flag, there was a race condition with the driver's queued
packet processing routine that is called by the driver's
transmit interrupt routine. New packets were being
transmitted to the board ahead of packets that were already
waiting on the queue to be transmitted. This situation
would occur if the transmit interrupt routine unlocked the
bd_lock after TX_BUSY was reset, but before the call to the
driver's queued packet processing routine.
- The pg driver's queued packet processing routines were
changed to set the TX_BUSY and TX_QUEUED flags if their call
to the transmit packet routine is unsuccessful. This
situation occurs if the transmit list is still full. When
these flags are set, DLunitdata_req() will place all new
packets at the end of the queue of packets already waiting
to be transmitted.
2. The flow control limit of the number of outstanding packets that can
be queued to each queue was set too small. Packets would be
discarded when this queue hiwater limit was reached. This could
occur on either the transmit or receive side. Changed the default
value for the pg driver's flow control queue hiwater level from 40 to
8192 packets. Added the capability to change the queue's hiwater
level for the pg driver in its space.c file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/dlpi_pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: pg-004
Date Issued: 03/18/2002 11:32:57
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: pg-004
Date Issued: 03/18/2002 11:40:39
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: none
Brief Description:
PowerMAX OS 4.3 pg package release updates
##############################################################################
Enhancements:
1. Changes made for RFC 1512 and RFC 1643 dot3 support.
2. Added support for the RFC 1512 fddimib and RFC 1643 dot3 trans-
mission group MIBs in the in.snmpd daemon. v_dot3.c processes the
RFC 1643 dot3 MIB attributes. v_fddi.c processes the RFC 1512
fddimib MIB attributes.
3. Added new '-d' option to the snmpstat(1m) command to display
some of the dot3 statistics information for ethernet type devices.
4. Fixed the existing '-i' option display. Incorrect values were
being returned for several fields.
5. Added the new DL_dot3stats_t structure for RFC 1643 dot3
transmission group support for use by Ethernet type devices.
Modified the board control structure DL_bdconfig_t to add the
DL_dot3stats_t structure. All dlpi devices will have the modified
board control structure, but at the present time only dec and rmxf
fully support it. The drivers for ie, cnd, egl provide minimal
support (EtherChipSet) only.
6. Changed the interface to DLbdspecioctl in DLioctl. The board
specific DLbdspecioctl routine that is called is now responsible
for issuing the qreply for the request instead of DLioctl. This
was needed to support delayed responses for the fddi get MIB
attribute commands.
7. Fixed a problem in DLopen that would not allow pg1, pg2, etc to
be opened properly.
8. Changed the name of the output queue length attribute from
"ifOutQlen" to "ifOutQLen" in order to match the mib2 RFC
attribute name.
9. Additional fixes in the dlpi common code for all networking
drivers. The MIB In/Out counters weren't being updated when
XID Test packet types (used by PowerIO) were transmitted or
received. This caused a discrepancy in the values returned
by netstat -i and snmpstat -i.
Object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-005
Date Issued: 12/11/2002 14:36:25
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #736, #789
Brief Description:
PowerMAX OS 4.3 pg package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #736: pg driver panics system with "pg%d: unsupported
command id" message.
R1: Several changes were made to the pg Driver.o to enhance the
logic surrounding the manipulation of the read and write
segment offset registers for the B2H (board to host) channel.
This area of the pgintr_cmn() interrupt routine was suspected
as being the cause of these panics, even though the values of
these hardware registers were not avaliable in the crashfile.
Because values written to these hardware locations are not
immediately reflected in the register values when read by
the CPU, the driver must stall itself until the values written
to these registers "stick" in the registers. Additional
"eieio" instructions were also added in these code sequences
to insure the proper order of writes to these registers.
Two arrays were added to provide debug information about the
read and write segments and segment offsets in the pgintr_cmn
routine. The pg_debug[PGBOARDS][12] array contains the prev,
current and new values of the read and write B2H segments and
segment offsets. The pg_spins[PGBOARDS][8] array contains
spin counters that are incremented whenever the driver must
stall itself to wait for either the value of the read or write
segment or segment offset to stabilize. While these changes
did not solve this customer's panic problem, it does close
some potential holes in the logic that could have caused
similar problems for other customers. It also provides
additional information should this panic occur at some other
customer site that will help us provide a solution to the
problem.
P2: SAR #789: When a pg device is installed in the system and not
connected to a token ring the message:
pgN: ring connect requested
will continuously display to the console every 10 seconds.
R2: A tunable has been added to the pg/Driver.o to allow
enabling/disabling of this message.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-005
Date Issued: 12/18/2002 17:45:35
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #736, #789
Brief Description:
PowerMAX OS 4.3 pg package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #736: pg driver panics system with "pg%d: unsupported
command id" message.
R1: Several changes were made to the pg Driver.o to enhance the
logic surrounding the manipulation of the read and write
segment offset registers for the B2H (board to host) channel.
This area of the pgintr_cmn() interrupt routine was suspected
as being the cause of these panics, even though the values of
these hardware registers were not avaliable in the crashfile.
Because values written to these hardware locations are not
immediately reflected in the register values when read by
the CPU, the driver must stall itself until the values written
to these registers "stick" in the registers. Additional
"eieio" instructions were also added in these code sequences
to insure the proper order of writes to these registers.
Two arrays were added to provide debug information about the
read and write segments and segment offsets in the pgintr_cmn
routine. The pg_debug[PGBOARDS][12] array contains the prev,
current and new values of the read and write B2H segments and
segment offsets. The pg_spins[PGBOARDS][8] array contains
spin counters that are incremented whenever the driver must
stall itself to wait for either the value of the read or write
segment or segment offset to stabilize. While these changes
did not solve this customer's panic problem, it does close
some potential holes in the logic that could have caused
similar problems for other customers. It also provides
additional information should this panic occur at some other
customer site that will help us provide a solution to the
problem.
P2: SAR #789: When a pg device is installed in the system and not
connected to a token ring the message:
pgN: ring connect requested
will continuously display to the console every 10 seconds.
R2: A tunable has been added to the pg/Driver.o to allow
enabling/disabling of this message.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-006
Date Issued: 05/08/2003 10:40:44
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 pg package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
Enhancements:
1. Developed a new utility pgstat(1M) to read and display the
firmware version and RC Statistics information for one or more
pg devices.
2. Added a man page for the new pgstat(1M) command.
3. Added a reference for the new pgstat(1M) command
to the pg(7) man page.
4. Changed the description of the "prefix" and "unit" fields in
the interface(4) and config_inet(1M) man pages. "prefix" and
"unit" must match the device "module name" and "unit".
5. Modifications to the kernel driver to support an interface to the
pg 5211 FDDI device's Report Command/Request Statistics
(RC statistics) command.
6. Modified the kernel driver to read and display during
system boot the firmware version of the 5211 device
using the RC Statistics command.
object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
/usr/sbin/pgstat
/usr/share/man/cat1/pgstat.1m.z
/usr/share/man/cat7/pg.7.z
/usr/share/man/man1/pgstat.1m
/usr/share/man/man7/pg.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-006
Date Issued: 05/08/2003 10:40:42
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 pg package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
Enhancements:
1. Developed a new utility pgstat(1M) to read and display the
firmware version and RC Statistics information for one or more
pg devices.
2. Added a man page for the new pgstat(1M) command.
3. Added a reference for the new pgstat(1M) command
to the pg(7) man page.
4. Changed the description of the "prefix" and "unit" fields in
the interface(4) and config_inet(1M) man pages. "prefix" and
"unit" must match the device "module name" and "unit".
5. Modifications to the kernel driver to support an interface to the
pg 5211 FDDI device's Report Command/Request Statistics
(RC statistics) command.
6. Modified the kernel driver to read and display during
system boot the firmware version of the 5211 device
using the RC Statistics command.
Object(s) To Be Replaced:
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
/usr/sbin/pgstat
/usr/share/man/cat1/pgstat.1m.z
/usr/share/man/cat7/pg.7.z
/usr/share/man/man1/pgstat.1m
/usr/share/man/man7/pg.7
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-007
Date Issued: 01/07/2004 16:02:21
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 pg package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: The IP layer would switch from true hardware based multicast address support to broadcast based multicast support if an attempt to exceed the maximum number of multicast addresses occurred.
R3: Changed the DLPI common code to pass to the IP layer the actual error
code returned by the driver specific code. The drivers that support
true hardware based multicast addressing (dec, gte, pg, sym) were
changed to return different error codes based on the error type. The
IP layer is now able to distinguish between problems with the hardware
based multicast address support or whether other software based errors
have occurred. When a problem occurs with the hardware based multicast
address support, the IP layer will change the interface to use
broadcasts for multicasting.
P4: SAR 1545: The FDDI pg board sometimes wrote over lowcore during the
early startup phase of the next system boot. This caused a SIGSEGV
abort in ypbind when using NIS with the fddi pg driver.
R4: The FDDI pg board was not being reset during system shutdown.
Added a shutdown/halt routine to the pg driver to perform the
reset. This stops the FDDI pg board from sending receive packets
to the cpu before the pg driver is able to initialize the software
tables that are needed by the board.
P5: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R5: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
1: SAR 261: Added true hardware based multicast address support into the
pg VME FDDI driver. In previous revisions, multicast addresses were
handled on the pg FDDI using broadcast addresses. The pg hardware
supports a maximum number of 32 multicast addresses.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/pg
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: pg-007
Date Issued: 01/07/2004 16:02:48
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 pg package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: The networking common code used for all drivers would print a
warning message on the console if it could not allocate memory
making the low memory situation even worse and possibly flooding
the console with messages.
R1: The interface ifSpecific.etherRcvResources counter will be
incremented instead of printing the warning message.
P2: The networking common code used for all drivers was not identifing
the reason why some packets were being discarded.
R2: The DLrsrv read servcie routine common code will now increment the
ifSpecific.etherReadqFull counter when it cannot pass a packet to
the next level streams read queue. This usually occurs when the
hiwater mark for that level has been exceeded.
P3: The IP layer would switch from true hardware based multicast address
support to broadcast based multicast support if an attempt to exceed
the maximum number of multicast addresses occurred.
R3: Changed the DLPI common code to pass to the IP layer the actual error
code returned by the driver specific code. The drivers that support
true hardware based multicast addressing (dec, gte, pg, sym) were
changed to return different error codes based on the error type. The
IP layer is now able to distinguish between problems with the hardware
based multicast address support or whether other software based errors
have occurred. When a problem occurs with the hardware based multicast
address support, the IP layer will change the interface to use
broadcasts for multicasting.
P4: SAR 1545: The FDDI pg board sometimes wrote over lowcore during the
early startup phase of the next system boot. This caused a SIGSEGV
abort in ypbind when using NIS with the fddi pg driver.
R4: The FDDI pg board was not being reset during system shutdown.
Added a shutdown/halt routine to the pg driver to perform the
reset. This stops the FDDI pg board from sending receive packets
to the cpu before the pg driver is able to initialize the software
tables that are needed by the board.
P5: A PANIC occurred in freemsg() from a call by the transmit queued
packet processing routine if an error occurred in the putbq() call.
R5: putbq() always consumes the packet, even if an error occurs. The
putbq() caller does not need to free the packet on an error. All
of the ethernet drivers on all platforms had this unlikely but
potential problem with an extra freemsg() call in the error path.
Enhancements:
1: SAR 261: Added true hardware based multicast address support into the
pg VME FDDI driver. In previous revisions, multicast addresses were
handled on the pg FDDI using broadcast addresses. The pg hardware
supports a maximum number of 32 multicast addresses.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/pg
/etc/conf/pack.d/pg/Driver.o
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: pg-008
Date Issued: 03/13/2006 14:48:19
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 pg package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
Enhancements:
1. ER3866: Added a new set of pg driver tunables, PG_BURST, PG0_BURST,
PG1_BURST, PG2_BURST, and PG3_BURST, to the pg driver's mtune file.
These are used to set the VME DMA Burst size. Originally, the
board defaulted this value to 256.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: pg-008
Date Issued: 03/13/2006 11:45:29
Software Package: pg pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 pg package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
Enhancements:
1. ER3866: Added a new set of pg driver tunables, PG_BURST, PG0_BURST,
PG1_BURST, PG2_BURST, and PG3_BURST, to the pg driver's mtune file.
These are used to set the VME DMA Burst size. Originally, the
board defaulted this value to 256.
Object(s) To Be Replaced:
/etc/conf/mtune.d/pg
/etc/conf/pack.d/pg/Driver.o
/etc/conf/pack.d/pg/space.c
/usr/include/sys/pg.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rmxf-001
Date Issued: 05/08/2003 10:41:54
Software Package: rmxf pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 rmxf package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Build warning messages in the rmxf driver.
R1. Fixed warning messages in the rmxf driver.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/rmxf/Driver.o
/usr/include/sys/rmxf.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rmxf-001
Date Issued: 05/08/2003 10:41:54
Software Package: rmxf pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 rmxf package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Build warning messages in the rmxf driver.
R1. Fixed warning messages in the rmxf driver.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/rmxf/Driver.o
/usr/include/sys/rmxf.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: rpc-001
Date Issued: 11/20/2000 15:58:43
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 rpc package release updates
#############################################################################
Problem Description:
1. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
Problem Resolution:
1. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/rpcsvc/yp_prot.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: rpc-001
Date Issued: 11/20/2000 16:07:06
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 rpc package release updates
#############################################################################
Problem Description:
1. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
Problem Resolution:
1. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/rpcsvc/yp_prot.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rpc-002
Date Issued: 12/11/2002 14:37:49
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: #780, #806
Brief Description:
PowerMAX OS 4.3 rpc package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
P2: SAR #806: NIS is not functional on a Closely Coupled Client
System. The ypinit(1M) utility fails.
R2: The vroot.uniqlist file has been changed to add
/etc/init.d/nis and /etc/rc2.d/S51domain to the unique list to
allow the diskless software to build a vroot that will support
the running of ypinit(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/nis.so
/usr/lib/tcpip_nis.so
Special Conditions for Installation:
None.
Possible Side Effects:
During the install of rpc-002 patch you will see pkginstall
WARNING messages similar to the 1 shown below. The WARNING
is generated because the file type is changing from a hard
link to a symbolic link.
"UX:pkginstall: WARNING: /etc/rc0.d/K80nis <no longer a linked file>"
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rpc-002
Date Issued: 12/18/2002 17:46:28
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: #780, #806
Brief Description:
PowerMAX OS 4.3 rpc package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: SAR #780: nfsping(1M) will randomly fail with: RPC: Timed out
nfsping: rpcbind is not running
The problem is due to the use by rpcbind(1M) of
netdir_options(3N) with the ND_MERGEADDR option.
There was a previous change to netdir_options() that would
return an error when it was called and there was no default
gateway or route configured. Usually in this case the
loopback address would be returned which was not correct.
However, when the specified server address on the
netdir_options() call was the loopback address, then it would
be expected to return the loopback address instead of an
error. Since nfsping ultimately called netdir_options() with
the loopback address it would receive the error return and
believe that rpcbind was not running.
R1: The netdir_options() routines in resolv.so and tcpip.so
have been corrected to return the loopback address if that
is the specified address to be merged. This keeps the
previous change to netdir_options() intact and allows nfsping
to function correctly.
P2: SAR #806: NIS is not functional on a Closely Coupled Client
System. The ypinit(1M) utility fails.
R2: The vroot.uniqlist file has been changed to add
/etc/init.d/nis and /etc/rc2.d/S51domain to the unique list to
allow the diskless software to build a vroot that will support
the running of ypinit(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/usr/lib/nis.so
/usr/lib/tcpip_nis.so
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rpc-003
Date Issued: 07/01/2003 11:18:39
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 rpc package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Kernel panic'd at t_ksndudata+0x130 with a kernel-mode address
fault on kernel address type panic.
R1. t_ksndudata() was referencing a queue structure that had already
been kmem_free'd. A race was occurring between the creation of
the stream and a user of the stream. The creator of the stream
popped the timod module off the stream just as t_ksndudata()
tried to reference the queue structure for the timod module
on the stream. The problem occurred because t_ksndudata()
released the lock for the stream's queue too early before it
had finished referencing the queue structure. The unlock
was moved to be after the final references to the queue.
Enhancement:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ktli/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: rpc-003
Date Issued: 07/03/2003 08:19:33
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 rpc package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. Kernel panic'd at t_ksndudata+0x130 with a kernel-mode address
fault on kernel address type panic.
R1. t_ksndudata() was referencing a queue structure that had already
been kmem_free'd. A race was occurring between the creation of
the stream and a user of the stream. The creator of the stream
popped the timod module off the stream just as t_ksndudata()
tried to reference the queue structure for the timod module
on the stream. The problem occurred because t_ksndudata()
released the lock for the stream's queue too early before it
had finished referencing the queue structure. The unlock
was moved to be after the final references to the queue.
Enhancement:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/ktli/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: rpc-005
Date Issued: 03/13/2006 14:48:38
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 rpc package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3209: Previous changes to the make utility to enhance VPATH
caused problem.
R1: Removed VPATH enhancements.
Enhancements:
1. RIQ10192: Enhanced passwd and yppasswd for restrictions on
mixed-case requirements. Enhanced in.login for restrictions
on remote logins.
Object(s) To Be Replaced:
/usr/bin/yppasswd
/var/yp/ypbuild
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: rpc-005
Date Issued: 03/13/2006 11:46:07
Software Package: rpc pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 rpc package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT3209: Previous changes to the make utility to enhance VPATH
caused problem.
R1: Removed VPATH enhancements.
Enhancements:
1. RIQ10192: Enhanced passwd and yppasswd for restrictions on
mixed-case requirements. Enhanced in.login for restrictions
on remote logins.
Object(s) To Be Replaced:
/usr/bin/yppasswd
/var/yp/ypbuild
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: rvsrv-001
Date Issued: 05/30/2000 13:13:29
Software Package: rvsrv pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: PowerStack II
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 rvsrv package release updates
##############################################################################
Problem Description:
1. Health Monitor support is needed to communicate with the fibre driver
and backplane for drive hot swapping.
2. SNMP MIBs used by the Residential VOD system were not ASN-1 syntax
compliant.
3. Minor issues with 'vemon' resilience needed correction.
Problem Resolution:
1. New functions for vemon to support fibre drive hot swap.
2. Redo vemon SNMP data numbering.
3. Corrected various vemon bugs.
Enhancements:
1. An automatic tool was needed to allow users to configure various
scripts for VOD system generation. Added new RSinstall script and
support files to allow in-field system generation and configuration.
Object(s) To Be Replaced:
/etc/init.d/vs_inetcfg
/etc/init.d/vs_init
/etc/init.d/vs_mcftp
/etc/init.d/vs_media
/etc/init.d/vs_reset
/etc/init.d/vs_route
/sbin/fumount
/sbin/killexports
/usr/etc/mhvs/install/README.first
/usr/etc/mhvs/install/RSinstall
/usr/etc/mhvs/install/S10savecore
/usr/etc/mhvs/install/S25client
/usr/etc/mhvs/install/S35syslogd
/usr/etc/mhvs/install/S40CCURvod.gs
/usr/etc/mhvs/install/S40CCURvod.vp
/usr/etc/mhvs/install/dskconfig
/usr/etc/mhvs/install/fib_devmap0.new
/usr/etc/mhvs/install/fib_devmap0.old
/usr/etc/mhvs/install/gd.24
/usr/etc/mhvs/install/gd.36
/usr/etc/mhvs/install/gd.client
/usr/etc/mhvs/install/gd.host
/usr/etc/mhvs/install/kernel.modlist.add
/usr/etc/mhvs/install/nodes.netboot
/usr/etc/mhvs/install/params
/usr/etc/mhvs/install/profile
/usr/etc/mhvs/install/resource.cfg
/usr/etc/mhvs/install/rsprofile.1
/usr/etc/mhvs/install/rsprofile.2
/usr/etc/mhvs/install/stune.1
/usr/etc/mhvs/install/stune.2
/usr/etc/mhvs/install/stune.3
/usr/etc/mhvs/install/stune.4
/usr/etc/mhvs/install/stune.5
/usr/etc/mhvs/install/stune.6
/usr/etc/mhvs/install/stune.7
/usr/etc/mhvs/install/stune.8
/usr/etc/mhvs/install/sysconfig.rs
/usr/etc/mhvs/install/vfstab.1
/usr/etc/mhvs/install/vfstab.2
/usr/etc/mhvs/install/vfstab.client
/usr/etc/mhvs/install/vptab.1
/usr/etc/mhvs/sbin/enable_tftp
/usr/etc/mhvs/sbin/eng_reset
/usr/etc/mhvs/sbin/getsvcmap
/usr/etc/mhvs/sbin/mk_bootptab
/usr/etc/mhvs/sbin/netreconfig
/usr/etc/mhvs/sbin/param_config
/usr/etc/mhvs/sbin/reset_all
/usr/etc/mhvs/sbin/restart_mhgs
/usr/etc/mhvs/sbin/rsrename
/usr/etc/mhvs/sbin/vemon
/usr/etc/mhvs/sbin/vpmon
Special Conditions for Installation:
The following object(s) will also be replaced by rvsrv-001:
* /etc/rc0.d/K12vs
* /etc/rc0.d/K25media
* /etc/rc1.d/K89vs
* /etc/rc2.d/S79inetcfg
* /etc/rc2.d/K89vs
* /etc/rc2.d/S45reset
* /etc/rc3.d/S37media
* /etc/rc3.d/S45mcftp
* /etc/rc3.d/S50route
* /etc/rc3.d/S89vs
* /opt/sbin/restart_mhgs
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: softint-001
Date Issued: 07/10/2002 14:56:14
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 softint package release updates
##############################################################################
Enhancements:
1. The /usr/bin/pkginfo utility is being enhanced to allow
it to display Patch README files. A -R option has been added
which is documented in the pkginfo(1) manual page.
Object(s) To Be Replaced:
/usr/bin/pkgtrans
/usr/bin/pkgmk
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: softint-001
Date Issued: 07/24/2002 14:49:33
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-011
Related SARs: none
Brief Description:
PowerMAX OS 4.3 softint package release updates
###############################################################################
Enhancements:
1. The /usr/bin/pkginfo utility is being enhanced to allow
it to display Patch README files. A -R option has been added
which is documented in the pkginfo(1) manual page.
Object(s) To Be Replaced:
/usr/bin/pkgtrans
/usr/bin/pkgmk
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: softint-002
Date Issued: 12/11/2002 14:38:15
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 softint package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: This task merely backfits recent changes to libpkg in 5.1SR5
which handle large pkg names. The associated tasks were:
pkgnmlen.mew and pkgnmlen2.mew
The problem was that pkgtrans would not properly handle the
pkg name (xdevc51synergy-001). This was because of a few
remaining hardcoded array lengths where the pkg name was
involved.
R1: The fix was to change these hardcoded array lengths to use
the value of PKGSIZ which is defined in /usr/include/pkgstrct.h.
P1: Performing a pkgtrans(1M) on a large number of input packages
to an output datastream specified as "stdout" (-) where stdout
is redirected to the actual datastream can result in a corrupted
datastream. This is due to transference messages normally
destined for stdout being interspersed with the packages being
transferred which are also now destined for stdout.
R1: The fix for this problem is to turn on the PT_SILENT option
during the transference of the packages so that the
transferrence messages will not be displayed.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgmk
/usr/bin/pkgproto
/usr/bin/pkgtrans
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: softint-002
Date Issued: 12/18/2002 17:46:45
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-012
Related SARs: none
Brief Description:
PowerMAX OS 4.3 softint package release updates
------------------------------------------------------------------------------
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: This task merely backfits recent changes to libpkg in 5.1SR5
which handle large pkg names. The associated tasks were:
pkgnmlen.mew and pkgnmlen2.mew
The problem was that pkgtrans would not properly handle the
pkg name (xdevc51synergy-001). This was because of a few
remaining hardcoded array lengths where the pkg name was
involved.
R1: The fix was to change these hardcoded array lengths to use
the value of PKGSIZ which is defined in /usr/include/pkgstrct.h.
P1: Performing a pkgtrans(1M) on a large number of input packages
to an output datastream specified as "stdout" (-) where stdout
is redirected to the actual datastream can result in a corrupted
datastream. This is due to transference messages normally
destined for stdout being interspersed with the packages being
transferred which are also now destined for stdout.
R1: The fix for this problem is to turn on the PT_SILENT option
during the transference of the packages so that the
transference messages will not be displayed.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgmk
/usr/bin/pkgproto
/usr/bin/pkgtrans
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: softint-003
Date Issued: 07/08/2004 10:54:58
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 softint package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1602: Due to the size of some package names now, there is
the possibility that menu displays can scroll off the top of
the screen. This is a problem on dumb terminals where one
cannot scroll backwards in that the data is lost before it
can be seen.
R1: The printmenu() routine in libadm.a has been modified to
ensure that no more than 24 lines will be displayed when
displaying the menu selections.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgmk
/usr/bin/pkgproto
/usr/bin/pkgtrans
/usr/lib/libadm.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: softint-003
Date Issued: 07/26/2004 11:52:09
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 softint package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: DT 1602: Due to the size of some package names now, there is
the possibility that menu displays can scroll off the top of
the screen. This is a problem on dumb terminals where one
cannot scroll backwards in that the data is lost before it
can be seen.
R1: The printmenu() routine in libadm.a has been modified to
ensure that no more than 24 lines will be displayed when
displaying the menu selections.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgmk
/usr/bin/pkgproto
/usr/bin/pkgtrans
/usr/lib/libadm.a
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: softint-004
Date Issued: 03/13/2006 14:48:43
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 softint package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT2128: The libpkg library did not allow greater than 128 packages
to be selected in gpkglist().
R1: The library routine has been changed to allow up to 256 packages.
The change to the libpkg library affects the pkgtrans utility.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgtrans
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
################################################################################
Patch Name: softint-004
Date Issued: 03/13/2006 11:46:17
Software Package: softint pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: @<platforms>
Related Patches: none
Brief Description: PowerMAX_OS 4.3 softint package release updates
################################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
################################################################################
P1: DT2128: The libpkg library did not allow greater than 128 packages
to be selected in gpkglist().
R1: The library routine has been changed to allow up to 256 packages.
The change to the libpkg library affects the pkgtrans utility.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/pkgtrans
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-001
Date Issued: 09/15/1999 11:10:17
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. It is possible to trace the kernel using the RCIM synchronized tick
clock as the event time stamp source even though the clock is not
enabled. The resulting trace data is meaningless.
2. The interface for selecting the RCIM synchronized tick clock as the
event time stamp source was obscure.
Problem Resolution:
1. Enhance ktrace to read the RCIM synchronized tick clock status
register during initialization if the tick clock was selected.
ktrace will abort with an error message if the clock is not enabled.
2. The interface for selecting the RCIM synchronized tick clock as the
event time stamp source was changed. Selection of the clock is now
effected by "-clock <source>", where "<source>" can be either
"default" or "rcim_tick". The previous option, "-altclock sync_tick",
is no longer supported.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-001
Date Issued: 09/29/1999 13:08:22
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, TurboHawk, PowerMAXION
Related Patches: base-001
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. The interface for selecting the RCIM synchronized tick clock as the
event time stamp source was obscure.
Problem Resolution:
1. The interface for selecting the RCIM synchronized tick clock as the
event time stamp source was changed. This change does not affect use
of ktrace on Night Hawk systems.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-002
Date Issued: 02/02/2000 14:09:41
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. When NightTrace encounters an "int on no int" interrupt event, it
displays the incorrect interrupt name for the event. This is caused
by ktrace(1) not setting up the proper interrupt vector as the
"int on no int" interrupt.
2. When initializing the name of the 'Rescheduling with Streams'
pseudo-interrupt, ntfilter(1) uses an invalid array index and stores
a pointer to the interrupt name beyond the dimensioned size of the
interrupt name array. This could cause either memory corruption or a
segmentation fault.
3. As ntfilter(1) is processing trace events, a rescheduling interrupt
may be assigned an incorrect vector number, and thus an incorrect
vector name, if the rescheduling interrupt occurs near the end of a
trace buffer.
4. When processing a ktrace file generated with the -bufferwrap
ktrace(1) option, ntfilter(1) does not copy all trace data to its
output file. Upon subsequent analysis with NightTrace(1), event
times may be wrong after a certain point.
Problem Resolution:
1. Change ktrace to first determine if the "int on no int" interrupt is
supported and, if it is, to set up "int on no int" as the interrupt
handler name for the correct interrupt vector.
2. Change ntfilter so that the proper array index is used when storing
a pointer to the 'Rescheduling with Streams' pseudo-interrupt name.
3. Change ntfilter to correctly assign vector numbers to rescheduling
interrupts.
4. Change ntfilter to correctly position itself within the kernel trace
file so that trace event processing begins with the trace buffer
containing the oldest trace data.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-002
Date Issued: 02/02/2000 13:03:06
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. When NightTrace encounters an "int on no int" interrupt event, it
displays the incorrect interrupt name for the event. This is caused
by ktrace(1) not setting up the proper interrupt vector as the
"int on no int" interrupt.
2. When initializing the name of the 'Rescheduling with Streams'
pseudo-interrupt, ntfilter(1) uses an invalid array index and stores
a pointer to the interrupt name beyond the dimensioned size of the
interrupt name array. This could cause either memory corruption or a
segmentation fault.
3. As ntfilter(1) is processing trace events, a rescheduling interrupt
may be assigned an incorrect vector number, and thus an incorrect
vector name, if the rescheduling interrupt occurs near the end of a
trace buffer.
4. When processing a ktrace file generated with the -bufferwrap
ktrace(1) option, ntfilter(1) does not copy all trace data to its
output file. Upon subsequent analysis with NightTrace(1), event
times may be wrong after a certain point.
Problem Resolution:
1. Change ktrace to first determine if the "int on no int" interrupt is
supported and, if it is, to set up "int on no int" as the interrupt
handler name for the correct interrupt vector.
2. Change ntfilter so that the proper array index is used when storing
a pointer to the 'Rescheduling with Streams' pseudo-interrupt name.
3. Change ntfilter to correctly assign vector numbers to rescheduling
interrupts.
4. Change ntfilter to correctly position itself within the kernel trace
file so that trace event processing begins with the trace buffer
containing the oldest trace data.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-003
Date Issued: 05/30/2000 13:21:06
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. ktrace(1) gets confused when creating a verbose report from a kernel
trace run in which the trace event time stamps wrapped around.
Starting with the trace buffer following the one in which the
wraparound occurred, ktrace prints an incorrect elapsed time for all
events and also prints an incorrect total run time in its summary
output. The total run time in a raw report summary is also printed
incorrectly.
Problem Resolution:
1. Change ktrace to:
a. internally view time stamps as 64 bits.
b. generate a pseudo value for the upper 32 bits of each
event's time stamp.
c. perform time stamp arithmetic using unsigned 64 bit values.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: trace-003
Date Issued: 05/30/2000 13:54:45
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. ktrace(1) gets confused when creating a verbose report from a kernel
trace run in which the trace event time stamps wrapped around.
Starting with the trace buffer following the one in which the
wraparound occurred, ktrace prints an incorrect elapsed time for all
events and also prints an incorrect total run time in its summary
output. The total run time in a raw report summary is also printed
incorrectly.
Problem Resolution:
1. Change ktrace to:
a. internally view time stamps as 64 bits.
b. generate a pseudo value for the upper 32 bits of each
event's time stamp.
c. perform time stamp arithmetic using unsigned 64 bit values.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: trace-004
Date Issued: 03/29/2001 12:06:51
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II, Motorola MCP750
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #499, #500
Brief Description:
PowerMAX OS 4.3 trace package release updates
############################################################################
Problem Description:
1. Combining kernel and user trace data logged from significantly
disjoint time periods caused erroneous kernel event times when
displayed with NightTrace.
2. Valid kernel trace data was discarded when actual kernel trace
records were lost due to heavy system activity. NightTrace did
not indicate the loss of data in kernel displays.
3. Erroneous filtering of kernel data by ntfilter caused erroneous
NightTrace display of kernel events related to system calls logged
by user level interrupt handlers when the corresponding interrupt
preempted an active system call.
Problem Resolution:
1. The kernel was modified to mark each kernel buffer with a full
64-bit timestamp. The ntfilter command was modified to consume
the timestamp and attribute the header of the filtered trace data
file accordingly, thus indicating the absolute time frame for the
kernel data.
2. The ktrace utility no longer discards kernel buffers associated
with lost kernel events. The ntfilter command annotates the
filtered data file with appropriate events (a single occurrence
of NT_LOST_DATA and one TR_EVENT_LOST event per cpu with arg2
indicating the number of events lost for that CPU) where actual
event loss occurred. The NightTrace utility was modified to
terminate all active Interrupt, Exception, and System call
states/graphs when a TR_EVENT_LOST event is seen. Thus, users
who have customized kernel display pages may wish to update their
State Graph and Data configurations by examining the new config-
urations for default kernel pages. The NightTrace modifications
are available with a patch to NightTrace 4.2 (ntrace-001).
3. ntfilter(1) now supports nested syscalls which are displayed
correctly by ntrace(1) via two new trace events:
TR_SYSCALL_NESTED_ENTRY and TR_SYSCALL_NESTED_EXIT.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
############################################################################
Patch Name: trace-004
Date Issued: 03/28/2001 19:16:27
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Patch Set Number: 4.3P6
Related Patches: base-006
Related SARS: #499, #500
Brief Description:
PowerMAX OS 4.3 trace package release updates
############################################################################
Problem Description:
1. Combining kernel and user trace data logged from significantly
disjoint time periods caused erroneous kernel event times when
displayed with NightTrace.
2. Valid kernel trace data was discarded when actual kernel trace
records were lost due to heavy system activity. NightTrace did
not indicate the loss of data in kernel displays.
3. Erroneous filtering of kernel data by ntfilter caused erroneous
NightTrace display of kernel events related to system calls logged
by user level interrupt handlers when the corresponding interrupt
preempted an active system call.
Problem Resolution:
1. The kernel was modified to mark each kernel buffer with a full
64-bit timestamp. The ntfilter command was modified to consume
the timestamp and attribute the header of the filtered trace data
file accordingly, thus indicating the absolute time frame for the
kernel data.
2. The ktrace utility no longer discards kernel buffers associated
with lost kernel events. The ntfilter command annotates the
filtered data file with appropriate events (a single occurrence
of NT_LOST_DATA and one TR_EVENT_LOST event per cpu with arg2
indicating the number of events lost for that CPU) where actual
event loss occurred. The NightTrace utility was modified to
terminate all active Interrupt, Exception, and System call
states/graphs when a TR_EVENT_LOST event is seen. Thus, users
who have customized kernel display pages may wish to update their
State Graph and Data configurations by examining the new config-
urations for default kernel pages. The NightTrace modifications
are available with a patch to NightTrace 4.2 (ntrace-001).
3. ntfilter(1) now supports nested syscalls which are displayed
correctly by ntrace(1) via two new trace events:
TR_SYSCALL_NESTED_ENTRY and TR_SYSCALL_NESTED_EXIT.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: trace-005
Date Issued: 08/22/2001 14:54:41
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
#############################################################################
Problem Description:
1. On single CPU moto-based platforms, the allocation and
use of the next available kernel trace record would
sometimes allocate the same trace record structure twice,
thus corrupting the information stored into this same trace
record for two trace events.
This corruption would occur when the tracing of a spurious
interrupt event takes place just as the CPU is currently
allocating a kernel trace record for another event. The method
used to serialize the kernel trace record allocation was unable
to successfully hold out spurious interrupts.
Problem Resolution:
1. The kernel trace record allocation code now disables interrupts
on the Power PC processor chip, rather than attempting to raise the
Interrupt Priority Level (IPL). By disabling interrupts inside
the processor, we are now able to successfully hold out spurious
interrupts while we allocate the next available ktrace record
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/trace/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: trace-005
Date Issued: 09/11/2001 16:25:39
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-008
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
#############################################################################
Problem Description:
1. On single CPU moto-based platforms, the allocation and
use of the next available kernel trace record would
sometimes allocate the same trace record structure twice,
thus corrupting the information stored into this same trace
record for two trace events.
This corruption would occur when the tracing of a spurious
interrupt event took place just as the CPU was currently
allocating a kernel trace record for another event. The method
used to serialize the kernel trace record allocation was unable
to successfully hold out spurious interrupts.
Problem Resolution:
1. The kernel trace record allocation code now disables interrupts
on the Power PC processor chip, rather than attempting to raise the
Interrupt Priority Level (IPL). By disabling interrupts inside
the processor, we are now able to successfully hold out spurious
interrupts while we allocate the next available ktrace record.
Enhancements:
None
Object(s) To Be Replaced:
/etc/conf/pack.d/trace/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: trace-006
Date Issued: 01/02/2002 14:26:34
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #416
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. SAR #416: NightTrace would not display kernel pages properly for
files generated by systems whose node name is hyphenated.
Problem Resolution:
1. Ntfilter now makes an internal coversion to all table names
going into a filtered kernel trace file so that NightTrace
can properly display the file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/trace/Driver.o
/etc/conf/pack.d/trace/space.c
/usr/bin/ntfilter
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: trace-006
Date Issued: 01/02/2002 14:26:34
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #416
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. SAR #416: NightTrace would not display kernel pages properly for
files generated by systems whose node name is hyphenated.
Problem Resolution:
1. Ntfilter now makes an internal coversion to all table names
going into a filtered kernel trace file so that NightTrace
can properly display the file.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/trace/Driver.o
/etc/conf/pack.d/trace/space.c
/usr/bin/ntfilter
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: trace-007
Date Issued: 03/18/2002 11:35:26
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Packages: base-010
Related SARs: #671
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. The ktrace(1) manpage was improperly being supplied as
part of the ntracelog package when it should have
been supplied as part of the trace package.
2. SAR #671: TR_HO_TIME event can be timestamped one tick behind the
next kernel event. If that next kernel event happens to be a
context switch, ntfilter may add a pseudo event at time -2 ticks
to "suspend" a system call or exception. This will cause the
events to appear out of chronological order to ntrace.
Problem Resolution:
1. The trace package will now supply the ktrace(1) man page.
2. Changed pseudo events to be timestamped currenttime - 1 tick and
hope to never see 2 events read by ntfilter with the same timestamp.
Enhancements:
1. Modified the format of a raw kernel trace file to include
an addition field, the clock frequency related to the
trace points in the trace file.
2. Utilized the newest version of the NightTrace header for
the filtered data to also include the clock frequency
in the header. A future version of ntrace will utilize
this information.
Changes 1 and 2 are backward compatible; i.e., if you were
to somehow get an old ntfilter (not possible if patch procedures
are correct), it would cope. Likewise, the existing NightTrace
can read the new headers without trouble.
3. Modified various files used by ntfilter so that "ntfilter" can
be built as part of the PowerWorks Linux Development Environment;
i.e. it runs on Linux.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
/usr/share/man/cat1/ktrace.1.z
/usr/share/man/man1/ktrace.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Service Release Report
##############################################################################
Software Update Name: trace-007
Date Issued: 03/18/2002 11:42:14
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Packages: base-010
Related SARs: #671
Brief Description:
PowerMAX OS 4.3 trace package release updates
##############################################################################
Problem Description:
1. The ktrace(1) manpage was improperly being supplied as
part of the ntracelog package when it should have
been supplied as part of the trace package.
2. SAR #671: TR_HO_TIME event can be timestamped one tick behind the
next kernel event. If that next kernel event happens to be a
context switch, ntfilter may add a pseudo event at time -2 ticks
to "suspend" a system call or exception. This will cause the
events to appear out of chronological order to ntrace.
Problem Resolution:
1. The trace package will now supply the ktrace(1) man page.
2. Changed pseudo events to be timestamped currenttime - 1 tick and
hope to never see 2 events read by ntfilter with the same timestamp.
Enhancements:
1. Modified the format of a raw kernel trace file to include
an addition field, the clock frequency related to the
trace points in the trace file.
2. Utilized the newest version of the NightTrace header for
the filtered data to also include the clock frequency
in the header. A future version of ntrace will utilize
this information.
Changes 1 and 2 are backward compatible; i.e., if you were
to somehow get an old ntfilter (not possible if patch procedures
are correct), it would cope. Likewise, the existing NightTrace
can read the new headers without trouble.
3. Modified various files used by ntfilter so that "ntfilter" can
be built as part of the PowerWorks Linux Development Environment;
i.e. it runs on Linux.
Object(s) To Be Replaced:
/usr/bin/ktrace
/usr/bin/ntfilter
/usr/share/man/cat1/ktrace.1.z
/usr/share/man/man1/ktrace.1
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: trace-008
Date Issued: 01/21/2004 13:26:46
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 trace package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
###############################################################################
P1: When in input mode (-i <file>), ktrace can print out
"vector does not match" over and over and over again.
R1: It is possible to have a rescheduling interrupt exit event without
a rescheduling interrupt entry event. This can happen when we enter
the rescheduling code not via a rescheduling interrupt. This
"unmatched" state confuses ktrace and thus the message. Resolve by
not printing out the message and ignoring the event, just as ntfilter
and the newer ntrace/autofiltering do.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: trace-008
Date Issued: 01/21/2004 13:42:10
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX_OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX_OS 4.3 trace package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
###############################################################################
P1: When in input mode (-i <file>), ktrace can print out
"vector does not match" over and over and over again.
R1: It is possible to have a rescheduling interrupt exit event without
a rescheduling interrupt entry event. This can happen when we enter
the rescheduling code not via a rescheduling interrupt. This
"unmatched" state confuses ktrace and thus the message. Resolve by
not printing out the message and ignoring the event, just as ntfilter
and the newer ntrace/autofiltering do.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: trace-009
Date Issued: 07/08/2004 10:55:22
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Architecture: moto
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 trace package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: When ktrace determines the input file is in wrapped mode, it
unwraps and stores the events as unwrapped. There are 2
problems after ktrace unwraps the events:
#1 - header marked as still in wrapped mode causing future
ktrace invokations to get confused.
#2 - new unwrapped file nukes post-run process table
R1: Change post_trace_process_table(), before it closes the new
unwrapped header, to :
1) Zero wrapped start flag if input file present, and
2) Corretly read&store the post-run table and header
pointers to it.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: trace-009
Date Issued: 07/26/2004 11:52:27
Software Package: trace pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Architecture: nh
Related Patches: base-015
Brief Description: PowerMAX OS 4.3 trace package release updates
###############################################################################
README format has changed. There will no longer be separate subsections
for "Problem Description:" or "Problem Resolution:". The new format will
be P# followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1: When ktrace determines the input file is in wrapped mode, it
unwraps and stores the events as unwrapped. There are 2
problems after ktrace unwraps the events:
#1 - header marked as still in wrapped mode causing future
ktrace invokations to get confused.
#2 - new unwrapped file nukes post-run process table
R1: Change post_trace_process_table(), before it closes the new
unwrapped header, to :
1) Zero wrapped start flag if input file present, and
2) Corretly read&store the post-run table and header
pointers to it.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/bin/ktrace
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: via-001
Date Issued: 11/20/2000 15:59:05
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 via package release updates
#############################################################################
Problem Description:
1. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
Problem Resolution:
1. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/sys/via.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
#############################################################################
Patch Name: via-001
Date Issued: 11/20/2000 16:07:23
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 via package release updates
#############################################################################
Problem Description:
1. There were various inconsistencies and bugs in the headers that
previously were fixed within the c++ 5.1 compiler release itself.
Problem Resolution:
1. The headers have now been updated so that the c++ 5.2 compiler
release will not need to fix them separately. These header changes
are needed for this 5.2 compiler as well as the upcoming PowerWorks
Linux Development Environment (PLDE) release.
Enhancements:
None.
Object(s) To Be Replaced:
/usr/include/sys/via.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: via-002
Date Issued: 06/11/2001 08:10:23
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 via package release updates
##############################################################################
Problem Description:
1. Driver uses async start unit command to spin up disk drives and
polls once a second until it is ready or times out.
2. Driver needed the ability to handle lists of Physical DMA addresses.
(support new partial sector capability).
3. Driver/SCSI tag queuing can re-arrange order of I/O processing which
needs to be disabled(ordered) for RMW partial sector writes.
Problem Resolution:
1. Modified Driver to poll disks once a second until it is ready or
times out.
2. Modofied Driver to support new partial sector capability.
3. Modified Driver to do correct SCSI tag queuing.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: via-002
Date Issued: 06/11/2001 08:38:36
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6200/6800, PowerMAXION-4/8, TurboHawk
Related Patches: base-007
Related SARs: none
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 via package release updates
##############################################################################
Problem Description:
1. Driver uses async start unit command to spin up disk drives and
polls once a second until it is ready or times out.
2. Driver needed the ability to handle lists of Physical DMA addresses.
(support new partial sector capability).
3. Driver/SCSI tag queuing can re-arrange order of I/O processing which
needs to be disabled(ordered) for RMW partial sector writes.
Problem Resolution:
1. Added support for async unit startup.
2. Added support to handle lists of Physical DMA addresses.
3. Added support to allow disabled for RMW partial sector writes.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: via-003
Date Issued: 01/02/2002 14:27:04
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #583
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 via package release updates
##############################################################################
Problem Description:
1. If the kernel tunable SCSISCAN is turned on, then any disk
device connected to a VIA which had an ID greater than 6
would fail to be opened by disk utilities even though it was
probed at boot time. The value set for VIA_MAXDRIVE was set to
15 in /usr/include/sys/via.h which meant that not enough structures
were initialized prior to probing for attached devices.
Subsequently, when the probe for device 15 was done a 0 was
inadvertently written into the daughter card id element of
some other via structure. This incorrectly informed the
system that the VIA was an HSA (which does not support device IDs
greater than 7).
Problem Resolution:
1. In the header /usr/include/sys/via.h the value of VIA_MAXDRIVE
was changed to the more appropriate 16 (0 - 15) which allows
enough structures to be allocated for all possible devices
attached to the VIA.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
/usr/include/sys/via.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: via-003
Date Issued: 01/02/2002 14:27:04
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #583
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 via package release updates
##############################################################################
Problem Description:
1. If the kernel tunable SCSISCAN is turned on, then any disk
device connected to a VIA which had an ID greater than 6
would fail to be opened by disk utilities even though it was
probed at boot time. The value set for VIA_MAXDRIVE was set to
15 in /usr/include/sys/via.h which meant that not enough structures
were initialized prior to probing for attached devices.
Subsequently, when the probe for device 15 was done a 0 was
inadvertently written into the daughter card id element of
some other via structure. This incorrectly informed the
system that the VIA was an HSA (which does not support device IDs
greater than 7).
Problem Resolution:
1. In the header /usr/include/sys/via.h the value of VIA_MAXDRIVE
was changed to the more appropriate 16 (0 - 15) which allows
enough structures to be allocated for all possible devices
attached to the VIA.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
/usr/include/sys/via.h
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: via-004
Date Issued: 01/07/2004 16:03:23
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 via package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: Setting the tape drive type to "archivepythonv" leads to tape
read errors.
R1: The tape device is initialized only once in "via" driver. This default behaviour was changed with support for removable hard disks.
Enhancements:
None,
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: via-004
Date Issued: 01/07/2004 16:03:47
Software Package: via pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-014
Brief Description: PowerMAX OS 4.3 via package release updates
###############################################################################
README format has changed. There will no longer be separate subsections for
"Problem Description:" or "Problem Resolution:". The new format will be
P#(Problem) followed by R#(Resolution) one after the other in 1 subsection.
###############################################################################
P1: Setting the tape drive type to "archivepythonv" leads to tape
read errors.
R1: The tape device is initialized only once in "via" driver. This default
behaviour was changed with support for removable hard disks.
Enhancements:
None,
Object(s) To Be Replaced:
/etc/conf/pack.d/via/Driver.o
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Date Issued: 05/08/2003 10:42:23
Software Package: vmet pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 vmet package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mdevice.d/vmet
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: vp-001
Date Issued: 01/02/2002 14:27:27
Software Package: vp pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-009
Related SARs: #569
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 vp package release updates
##############################################################################
Problem Description:
1. SAR #569: The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
Problem Resolution:
1. Problem was due to the count of pages attached to an IO
operation not being correct.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/vp/Driver.o
Special Conditions for Installation:
1. The size of the b_numpages field of the internal kernel-only data
structure, buf_t, increased from a byte to a short. This change
was done in a way that the size of the buf_t structure did not
change nor did any offset of that structure, therefore, only
users of b_numpages itself are effected. All kernel Driver.o
modules that use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays the contents
of the field b_numpages.
Possible Side Effects:
1. crash(1m) from earlier patch-releases of PowerMAX OS will no
longer correctly display the b_numpages field.
return to index
================================================================================
##############################################################################
Concurrent Computer Corporation Software Development
Software Patch Report
Patch Name: vp-001
Date Issued: 01/02/2002 14:27:27
Software Package: vp pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-009
Related SARs: #569
Defect Tickets: none
Brief Description:
PowerMAX OS 4.3 vp package release updates
##############################################################################
Problem Description:
1. SAR #569: The kernel panics in the kpageflt() with the message,
`kernel-mode address fault on kernel address 0x0000003C'.
Problem Resolution:
1. Problem was due to the count of pages attached to an IO
operation not being correct.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/vp/Driver.o
Special Conditions for Installation:
1. The size of the b_numpages field of the internal kernel-only data
structure, buf_t, increased from a byte to a short. This change
was done in a way that the size of the buf_t structure did not
change nor did any offset of that structure, therefore, only
users of b_numpages itself are effected. All kernel Driver.o
modules that use buf_t are replaced by this patch, as well as
/usr/sbin/crash, as it has a command that displays the contents
of the field b_numpages.
Possible Side Effects:
1. crash(1m) from earlier patch-releases of PowerMAX OS will no
longer correctly display the b_numpages field.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: xfsd-001
Date Issued: 05/30/2000 13:21:26
Software Package: xfsd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620, 640, PowerStack II, Motorola MCP750
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 xfsd package release updates
##############################################################################
Problem Description:
1. An xfsd kernel thread may panic if a file is opened on a remote
system immediately after it has been written using buffered writes.
The remote file open flushes all modified buffer pages to disk so
that the remote system will read the correct data from the shared
disk. A missing header file declaration caused the compiler to assume
an incorrect function parameter size in the flush path resulting in an
uninitialized parameter value. This may cause the called function to
panic.
Problem Resolution:
1. The xfsd system has been enhanced to allow a file to be written on
one system and concurrently read on a remote system. Three new
configuration fields in xfsd enable and control this enhancement.
The options are documented in the configuration file. The
enhancement works by periodic update of the remote file size and
flush of any buffered data on the file update system. Truncate of
file open on remote system is still not permitted. File space must
be preallocated before remote open to use this new feature.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/xfsd/Driver.o
/etc/conf/pack.d/xfsd/space.c
/etc/conf/mtune.d/xfsd
Special Conditions for Installation:
1. The xfsd enhancement is disabled by default. Users of this feature
should be aware of the extra overhead caused by the periodic file
flush if using buffered writes to files while they are read on remote
systems. This overhead can be minimized by increasing the
synchronization delay and flushing to a page boundary. Writing with
direct I/O eliminates the flush overhead.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
##############################################################################
Patch Name: xfsd-001
Date Issued: 05/30/2000 13:55:11
Software Package: xfsd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION-4, PowerMAXION, TurboHawk
Related Patches: none
Related SARs: none
Brief Description:
PowerMAX OS 4.3 xfsd package release updates
##############################################################################
Problem Description:
1. An xfsd kernel thread may panic if a file is opened on a remote
system immediately after it has been written using buffered writes.
The remote file open flushes all modified buffer pages to disk so
that the remote system will read the correct data from the shared
disk. A missing header file declaration caused the compiler assume an
incorrect function parameter size in the flush path resulting in an
uninitialized parameter value. This may cause the called function to
panic.
Problem Resolution:
1. The xfsd system has been enhanced to allow a file to be written on
one system and concurrently read on a remote system. Three new
configuration fields in xfsd enable and control this enhancement.
The options are documented in the configuration file. The
enhancement works by periodic update of the remote file size and
flush of any buffered data on the file update system. Truncate of
file open on remote system is still not permitted. File space must
be preallocated before remote open to use this new feature.
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/pack.d/xfsd/Driver.o
/etc/conf/pack.d/xfsd/space.c
/etc/conf/mtune.d/xfsd
Special Conditions for Installation:
1. The xfsd enhancement is disabled by default. Users of this feature
should be aware of the extra overhead caused by the periodic file
flush if using buffered writes to files while they are read on remote
systems. This overhead can be minimized by increasing the
synchronization delay and flushing to a page boundary. Writing with
direct I/O eliminates the flush overhead.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: xfsd-002
Date Issued: 05/08/2003 10:42:33
Software Package: xfsd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: moto
Platforms: Power Hawk 620/640, PowerStack II/III, Motorola MCP750
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 xfsd package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/xfsd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================
Concurrent Computer Corporation Software Development
Software Patch Report
###############################################################################
Patch Name: xfsd-002
Date Issued: 05/08/2003 10:42:33
Software Package: xfsd pkg (Version 4.3)
OS Release: PowerMAX OS 4.3
Architecture: nh
Platforms: Night Hawk 6800, PowerMAXION, TurboHawk
Related Patches: base-013
Brief Description:
PowerMAX OS 4.3 xfsd package release updates
With the release of PowerMAX OS 4.3 Patch Set 11 the README format has
changed. There will no longer be separate subsections for "Problem
Description:" or "Problem Resolution:". The new format will be P#
followed by R# one after the other in 1 subsection.
Problem Description:=P#, Problem Resolution:=R#
###############################################################################
P1. SAR 1446: Some tunables have improper MIN/MAX values and
config(1M) will fail when attempting to modify the tunable
from the current value. In some cases idtune(1M) will fail
as well.
R1. The tunable MIN/MAX values have been set to proper values to
resolve the problems with config(1M) and idtune(1M).
Enhancements:
None.
Object(s) To Be Replaced:
/etc/conf/mtune.d/xfsd
Special Conditions for Installation:
None.
Possible Side Effects:
None.
return to index
================================================================================