c_cpp/papi
SHA256
8
0
Fork 0
forked from pool/papi
Code Pull Requests Activity
Files
f30042993cef3b3ca356cf9fbd5cd5bba2aece9eefbcd2df2fac7df0ed81d482
papi/papi.changes

353 lines
15 KiB
Plaintext
Raw Blame History

-------------------------------------------------------------------
Sun Oct 28 08:54:58 UTC 2012 - coolo@suse.com
- do not require libquadmath by old name but by provides, so it
works in factory too
-------------------------------------------------------------------
Thu Sep 27 11:45:28 UTC 2012 - idonmez@suse.com
- Update baselibs.conf for libpfm4
-------------------------------------------------------------------
Mon Sep 3 17:19:13 CEST 2012 - pth@suse.de
- Modify spec so that build succeeds (requiring libquadmath versions).
-------------------------------------------------------------------
Mon Sep 3 11:31:43 UTC 2012 - idonmez@suse.com
- Add papi-glibc-siginfo.patch: newer glibc no longer defines
siginfo, we need to use siginfo_t
-------------------------------------------------------------------
Wed Jul 4 11:45:05 CEST 2012 - pth@suse.de
- libquadmath is only needed/available from 12.1 up.
- Remove patches and files not needed anymore.
-------------------------------------------------------------------
Mon Apr 23 14:21:06 CEST 2012 - pth@suse.de
- Update to 4.4.0:
New Platforms:
--------------
Added PAPI support for Blue Gene/Q.
Bug Fixes:
----------
* Fix buffer overrun in lmsensors component
* libpfm4: Update to current git libpfm4 snapshot
* Fix broken Pentium 4 Prescott support we were missing the netburst_p
declaration in papi_events.csv
* Fix various locking issues in the threaded code.
* Fix multiplexing of large eventsets on perf_events systems.
This presented when using more than 31 multiplexed events on perf_event
- Add a devel-static package that contains the static libraries.
- For other changes since 4.2.1 see RELEASENOTES.txt in the packages
documentation.
-------------------------------------------------------------------
Wed Feb 22 16:21:25 CET 2012 - pth@suse.de
- Update to 4.2.1:
Bug Fixes:
----------
* multiplexing math errors were fixed in perf_events.c
* more multiplexing read path errors were identified and fixed
* Fix SMP ARM issues reported by Harald Servat.
* src/papi_events.csv: Update the coreduo (not core2) events. Most
notably the FP events were wrong.
* Modify Intel Sandybridge PAPI_FP_OPS and PAPI_FP_INS events to
not count x87 fp instructions. The problem is that the current
predefines were made by adding 5 events. With the NMI watchdog
stealing an event and/or hyperthreading reducing the number of
available counters by half, we just couldn't fit.
This now raises the potential for people using x87-compiled
floating point on Sandybridge and getting 0 FP_OPS. This is only
likely if running a 32-bit kernel and *not* compiling your code
with -msse.
A long-term solution might be trying to find a better set of FP
predefines for sandybridge.
* Apply patch suggested by Will Cohen to check for system return
values.
* Patch to cleanup dependencies, allowing for parallel makes.
* Fix two race conditions that are probably the cause of the
pthrtough double-free error.
When freeing a thread, we remove and free all eventsets belonging
to that thread. This could race with the thread itself removing
the evenset, causing some ESI fields to be freed twice.
The problem was found by using the Valgrind 3.8 Helgrind tool
valgrind --tool=helgrind --free-is-write=yes ctests/pthrtough
In order for Helgrind to work, I had to temporarily modify PAPI
to use POSIX pthread mutexes for locking.
Enhancements:
-------------
* general doxygen cleanups
* cleanup output of overflow_allcounters for clarity in debugging
* updates to most recent (as of Feb 1) libpfm4
* remove now-opaque event codes from papi_native_avail
and papi_xml_event_info
* Update the component initialization code so that it can handle a
PAPI ERROR return gracefully. Previously there was no way to
indicate initialization failure besides just setting
num_native_events to 0.
New Platforms:
-------------
* Support for AMD Family 11.
* Support for AMD Family 12.
Deprecated Platforms:
---------------------
* remove obsolete ACPI component
New / Improved Components:
---------------
* PAPI CUDA component updated for CUDA / CUPTI 4.1.
* SetCudaDevice() now works with the latest CUDA 4.1 version.
* Auto-detection of CUDA version for backward compatibility.
* PAPI_read() now accumulates event values. This fixes a bug
in earlier versions.
* extensive updates and cleanups to the example and coretemp components.
* significant updates of lustre, and mx components
* The linux net component underwent extensive updates and cleanups.
In particular, it nows dynamically detects the network
interface names [1] and export 16 counters for each interface
(see also src/components/net/{CHANGES,README}).
Experimental:
-------------
* a new vmware component has been added to report a variety of soft events
when running as a guest in a VMWare environment
- Add upstream patch for libpfm4 to fix signed/unsigned issues.
- Readd ExcludeArch for s390/s390x as it isn't supported by PAPI.
-------------------------------------------------------------------
Sun Jan 29 16:02:44 UTC 2012 - jengelh@medozas.de
- Remove redundant tags/sections per specfile guideline suggestions
- Remove unexplained s390 exclusion
-------------------------------------------------------------------
Fri Apr 15 20:00:00 CEST 2011 - pth@suse.de
- Adapt patch from IBM to fix the memory info for POWER7 CPUs.
-------------------------------------------------------------------
Fri Apr 15 15:12:37 CEST 2011 - pth@suse.de
- Clean up the patches and remove those that aren't needed anymore.
-------------------------------------------------------------------
Sat Jan 29 18:16:17 UTC 2011 - alexandre@exatati.com.br
- Update to 4.1.2.1:
* fixed a long-standing subtle bug identified by Richard Strong that caused
segfaults when multiplexing
* fixed several bugs that were causing test failures on POWER6/AIX
* properly detect Pentium M in configure
* fixed a problem with perf_events not properly handling overflows; first
identified by Mark Krentel
* fixed a problem where perfctr was silently adding uncountable events
* fixed a lock bug identified by Martin Schindewolf
* fixed forking order for {multi|zero}_attach.c
* updated support for freeBSD submitted by Harald Servat
* a plethora of code cleanups submitted by Robert Richter
* addressed compatibility issues in run_tests.sh to make it posix comliant
* refreshed PAPI_Matlab support
* reimplemented SUBDBG print capabilities to address an issue first
identified by Maynard Johnson
* refreshed preset event definitions for Nehalem, including implementations
for PAPI_HW_INT; submitted by Michel Brown
* added 3 new error codes: PAPI_EATTR, PAPI_ECOUNT, and PAPI_ECOMBO.
These provide more detail on why an event add fails
* implement cpuid leaf4 mtrics required by Intel Westmere
* Intel Westmere on perfctr and perf_events
* libpfm4 support is experimentally available and subject to change
- Added "export SUSE_ASNEEDED=0" to build section.
-------------------------------------------------------------------
Tue Jun 22 19:02:33 UTC 2010 - alexandre@exatati.com.br
- Update to 4.0.1:
- configure was mis-identifying some Pentium 4 processors
- the ctests/shlib test now tests against the shared math library, libm.so,
instead of libpapi.so, which works more predictably with library renaming.
- multiplexing was silently returning without setting multiplex TRUE in
cases where no event had been assigned to an eventset. An event must be
added to an eventset or PAPI_assign_eventset_component() must be called
before multiplexing can be enabled. This silent error has been removed.
- the perfmon and perf_events counter interfaces were not properly handling
event unit masks. This has been fixed.
- PAPI_name_to_code() was not exiting properly in certain circumstances,
failing on events where there should have been a match. This is corrected.
- a serious but insidious bug in the overflow logic was corrected. This bug
would only show up when PAPI_overflow was called between calls to
PAPI_add_event. Overflow would only be set for the last call of
PAPI_overflow. This has been corrected.
- IBM Blue Gene P systems were corrupting stack frames and crashing when the
papi_get_event_info call was executed. This has been fixed.
- The PAPI cycles event was not working for IBM Blue Gene P. This is fixed.
papi_native_avail was exiting improperly when using the -e option.
This caused problems with batch execution systems (like Blue Gene P).
This has been fixed.
- a significant number of memory leaks have been purged.
- compiler warning flags have been tightened and a range of warnings have
been eliminated.
- removed implicit type conversions in prototypes.
- the utils/papi_version utility now reports four digits where the last digit
matches the patch number.
- Pentium II and Athlon now use libpfm for event decoding like all other x86
platforms.
- Doxygen documentation has been added to the API and components.
- Component compilation has been completely restructured. See
/papi/src/components/README for details.
- PAPI can now be compiled with a no-cpu-counters option
- the ultrasparc architecture has been resurrected
- freebsd support was migrated from PAPI 3.7
- Intel Nehalem EX and Westmere support has been added
- IBM BG/L has been deprecated.
- POWER 3 and POWER4 have been deprecate
- Infiniband: Experimental
- Lustre: Experimental
- example: provides simple test case and template code.
-------------------------------------------------------------------
Mon Jun 14 19:01:14 UTC 2010 - alexandre@exatati.com.br
- Add Upstream patch3:
* the utils/papi_version utility reports PAPI 4.0.0.3, where the last
digit matches the patch number.
* the ultrasparc architecture has been resurrected and supported in this
patch.
* IBM Blue Gene P systems were corrupting stack frames and crashing when
the papi_get_event_info call was executed. This has been fixed.
* papi_native_avail was exiting improperly when using the -e option.
This caused problems with batch execution systems (like Blue Gene P).
This has been fixed.
- Spec file cleaned with spec-cleaner.
-------------------------------------------------------------------
Wed Mar 24 19:04:12 UTC 2010 - pth@novell.com
- Don't run testsuite as there are too many reasons why it won't
work on the build machine.
-------------------------------------------------------------------
Tue Mar 23 15:10:21 CET 2010 - pth@suse.de
- Update to 4.0.0:
- fixed L3 cache size reporting for AMD Family 10h processors
- fixed std deviation underflow in sdsc2 and sdsc4 tests
- fixed bug in counter assignment for FreeBSD Atom implementation
- updated cache tables for Intel Nehalem i7 processors
- configure provides better autodetection of 2.6.31 or 2.6.32 kernels and
perf_counter interface (in most cases)
- configure provides better detection and autoselection of perfctr or perfmon
drivers for linux
- configure and sources have been modified to support perf_counter on
kernel 2.6.31 and perf_event on kernel 2.6.32
- a papi.spec file has been added to simplify creation of rpms
- Add Upstream patch2:
* the utils/papi_version utility now reports PAPI 4.0.0.2, where the
last digit matches the patch number.
* the papi.spec file has a correction for a minor typo in the
ExcludeArch line.
* a serious but insidious bug in the overflow logic was corrected. This
bug would only show up when PAPI_overflow was called between calls to
PAPI_add_event. Overflow would only be set for the last call of
PAPI_overflow. This has been corrected.
* configure was mis-identifying some Pentium 4 processors; this has been
corrected.
* the ctests/shlib test now tests against the shared math library,
libm.so, instead of libpapi.so
* the PAPI_set_multiplex() call was silently returning without setting
multiplex TRUE in cases where no event had been assigned to an eventset.
This silent error has been removed.
* the perfmon and perf_events counter interfaces were not properly
handling event unit masks. This has been fixed.
* PAPI_name_to_code() was not exiting properly in certain circumstances,
failing on events where there should have been a match. This has been
corrected.
- Add RPM_OPT_FLAG to CFLAGS in configure.
- Initialize variables where the compiler can't determine that they're
used initialized.
- Add full prototypes to src/ctests/mendes-alt.c
- Delete the test for pfm_msg_t.type in configure for external libpfm instead
of disabling the test all together.
- Fix all compiler warnings in libpfm.
-------------------------------------------------------------------
Tue Mar 2 15:45:53 CET 2010 - rw@suse.de
- build against libpfm3 on ia64 (bnc#578864)
-------------------------------------------------------------------
Tue Feb 16 00:27:29 CET 2010 - ro@suse.de
- fix filelist for ia64 again, perfmon file is back
-------------------------------------------------------------------
Tue Feb 2 14:45:58 CET 2010 - pth@suse.de
- Fix dependencies in baselibs.conf (papi-devel-32bit requires
libpapi-32bit not papi-32bit which doesn't exist) (#bnc565236)
-------------------------------------------------------------------
Thu Jan 21 13:18:11 CET 2010 - sf@suse.de
- add baselibs.conf including libpapi and papi-devel (bnc #565236)
-------------------------------------------------------------------
Thu Jan 21 11:50:50 CET 2010 - ro@suse.de
- fix typo in last change
-------------------------------------------------------------------
Wed Jan 13 14:20:26 CET 2010 - sf@suse.de
- fix filelist for ia64 (remove /usr/share/papi), as the only
file there does not exist on ia64
-------------------------------------------------------------------
Fri Dec 18 15:13:30 CET 2009 - ro@suse.de
- fix filelist for ia64
- remove "," from excludearch
-------------------------------------------------------------------
Tue Dec 8 17:57:16 CET 2009 - pth@suse.de
- Update to 3.7.2:
Bug Fixes:
----------
* fixed L3 cache size reporting for AMD Family 10h processors
* fixed std deviation underflow in sdsc2 and sdsc4 tests
Enhancements:
-------------
* updated cache tables for Intel Nehalem i7 processors
* configure provides better autodetection of 2.6.31 or 2.6.32 kernels and
perf_counter interface (in most cases)
* configure provides better detection and autoselection of perfctr or perfmon
drivers for linux
* configure and sources have been modified to support perf_counter on
kernel 2.6.31 and perf_event on kernel 2.6.32
* a papi.spec file has been added to simplify creation of rpms
-------------------------------------------------------------------
Fri Nov 20 14:36:26 CET 2009 - pth@suse.de
- Initial package for SLES (fate#307192)
- Include perf_counter.h from kernel because kernel-headers
doesn't include the file.
View Git Blame Copy Permalink