|
Packit |
577717 |
2012-08-08
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4b4f87ff ChangeLogP5000.txt: Changelog for PAPI5
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6f208c06 doc/Doxyfile-common papi.spec src/Makefile.in...: Bump version
|
|
Packit |
577717 |
numbers in prep for a 5.0 release.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c6fdbd11 release_procedure.txt: Update release_procedure.txt Change the
|
|
Packit |
577717 |
order of when we branch git, so that the main dev branch gets some of the
|
|
Packit |
577717 |
release related changes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 97d4687f ChangeLogP440.txt: Pickup the changelog from papi 4.4 This was
|
|
Packit |
577717 |
only included in the stable-4.4 branch.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 628c2b6e src/buildbot_configure_with_components.sh: Take debug out of the
|
|
Packit |
577717 |
with several components build test config. When built with PAPI's memory
|
|
Packit |
577717 |
wrapper routines, the threaded stress tests will sometimes get into poor
|
|
Packit |
577717 |
performing situations. See trac ticket 148 for discussion.
|
|
Packit |
577717 |
http://icl.utk.edu/trac/papi/ticket/148
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-22
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 46faae8e src/ctests/overflow2.c src/ctests/overflow_single_event.c
|
|
Packit |
577717 |
src/ctests/overflow_twoevents.c...: Move find_nonderived_event() from
|
|
Packit |
577717 |
overflow_twoevents to test_utils and call it from overflow2 and
|
|
Packit |
577717 |
overflow_single_event to insure that we're not trying to overflow on a
|
|
Packit |
577717 |
derived event.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3e7d8455 src/ctests/zero_smp.c: Fix a memory leak reported on the aix
|
|
Packit |
577717 |
power7 machine. zero_smp.c did not unregister at the end of its thread
|
|
Packit |
577717 |
function.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3ad5782f src/perf_events.c: perf_events: fix segfault if DEBUG is enabled
|
|
Packit |
577717 |
Was incorrectly using "i" as an index where it should be "0" in a debug
|
|
Packit |
577717 |
statement.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-21
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a3cadbdb src/ftests/accum.F src/ftests/avail.F src/ftests/case1.F...: Take
|
|
Packit |
577717 |
#2. Changing len_trim function in ftests to last_char. This time, I respect
|
|
Packit |
577717 |
72 char line limit.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c9db8fbf src/ctests/overflow_force_software.c: overflow_force_software was
|
|
Packit |
577717 |
the only test that used a different hard_tolerance value (0.25) than the
|
|
Packit |
577717 |
other overflow tests (0.75). This caused trouble on Power7/AIX. Now we are
|
|
Packit |
577717 |
using the same hard_tolerance value in all overflow tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 70515343 src/ftests/accum.F src/ftests/avail.F src/ftests/case1.F...:
|
|
Packit |
577717 |
Changed name of function len_trim to last_char.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 95168d79 src/components/cuda/linux-cuda.c: Cleanup cuda shutdown code.
|
|
Packit |
577717 |
* The
|
|
Packit |
577717 |
shutdown_thread code cleaned out the whole component's state. This has been
|
|
Packit |
577717 |
split into shutdown_global for the whole component, and shutdown_thread is
|
|
Packit |
577717 |
left to cleanup some control state info.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 56284f81 src/ctests/multiplex1_pthreads.c: Fix memory leaks in pthread
|
|
Packit |
577717 |
multiplex tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* aeead8b6 src/threads.c: Remove an outdated comment about
|
|
Packit |
577717 |
_papi_hwi_free_EventSet holding INTERNAL lock
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e598647b src/perf_events.c: perf_events: fix issue where we dereference a
|
|
Packit |
577717 |
pointer before NULL check. Fix suggested by Will Cohen, based on a coverity
|
|
Packit |
577717 |
report.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4e0ed976 src/ctests/calibrate.c: Modify warning message to eliminate the
|
|
Packit |
577717 |
word "error" Hopefully this will suppress it in buildbot outputs.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 50fbba18 src/ctests/api.c src/ftests/case2.F: Cleanup a few more warnings
|
|
Packit |
577717 |
from the PAPI_perror change.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1f06bf28 src/ftests/case2.F: Missed an instance of perror in the fortran
|
|
Packit |
577717 |
code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 93e6ae2c src/ftests/ftests_util.F: Fix warning in ftest_util.F
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 60c6029e src/perf_events.c: perf_events: Update multiplexing code It
|
|
Packit |
577717 |
* turns
|
|
Packit |
577717 |
out the PERF_EVENT_IOC_RESET ioctl resets the count but not the multiplexing
|
|
Packit |
577717 |
info. This means that when we fiddle with the events then reset them in
|
|
Packit |
577717 |
check_scheduability(), we are not really resetting things to zero. The
|
|
Packit |
577717 |
effect might be small, but since the new multiplex code by definition is
|
|
Packit |
577717 |
always scheduable, then let's skip the test if multiplexing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9079236c src/ctests/zero.c: Change error reporting so FLOPS > 100% above
|
|
Packit |
577717 |
theoretical FAIL and FLOPS > 25% above theoretical WARN.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 980558af src/papi_internal.c: papi_internal: fix memory leak When I made
|
|
Packit |
577717 |
some changes a while back I forgot to free ESI->NativeBits properly. This
|
|
Packit |
577717 |
was causing memory leak warnings on buildbot.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 83a14612 src/perf_events.c: perf_events: more cleanups and comments We
|
|
Packit |
577717 |
really need to go back and figure out in more detail what the
|
|
Packit |
577717 |
profile/sampling/overflow code is doing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7cafb941 src/perf_events.c: perf_events: more cleanups and comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e9e39a4b src/perf_events.c: perf_events: disable kernel multiplexing
|
|
Packit |
577717 |
* before
|
|
Packit |
577717 |
2.6.34 It turns out even our simple multiplexing won't work on kernels
|
|
Packit |
577717 |
before 2.6.34, so fall back to sw multiplex in that case.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 05801901 src/perf_events.c: perf_events: more cleanup and comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 268e31d7 src/perf_events.c: perf_events: more cleanup and commenting
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d62fc2bf src/perf_events.c: perf_events: more cleanup and comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fb0081bc src/perf_events.c: perf_events: more cleanups and comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a1142fc8 src/perf_events.c: perf_events: cleanup and comment the kernel
|
|
Packit |
577717 |
* bug
|
|
Packit |
577717 |
workarounds
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b8560369 src/perf_events.c: perf_events: minor cleanups and new comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6c320bb2 src/perf_events.c: perf_events: fix some debug messages I forget
|
|
Packit |
577717 |
to test with --with-debug enabled
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f7a3cccf src/perf_events.c: perf_events: enable new read_code This makes
|
|
Packit |
577717 |
the read code much simpler. It finishes the multiplexing changes. To avoid
|
|
Packit |
577717 |
complication, we no longer enable PERF_FORMAT_ID as reading that extra info
|
|
Packit |
577717 |
is unnecessary with the current implementation. This passes all the tests
|
|
Packit |
577717 |
on
|
|
Packit |
577717 |
a recent kernel, but on 2.6.32 there are still a few issues.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 15749cff src/ctests/all_events.c src/ctests/all_native_events.c: Fix
|
|
Packit |
577717 |
warning in all_events and all_native_events. In the perror semantic change,
|
|
Packit |
577717 |
several strings for use in the old interface were left.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-16
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* afdd25fa src/perf_events.c: perf_events: always enable kernel multiplexing
|
|
Packit |
577717 |
The new code should work on any kernel version.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9f5e23ae src/perf_events.c: Rewrite multiplex support. Drop support for
|
|
Packit |
577717 |
the former "partitioned" multiplexing, as we could never use it. Instead
|
|
Packit |
577717 |
use
|
|
Packit |
577717 |
the simple/braindead model. This still needs more work, as sometimes reads
|
|
Packit |
577717 |
are failing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* cdd29909 src/ftests/strtest.F: Fix strtest.F ftest It was still making
|
|
Packit |
577717 |
some assumptions about PAPI_perror() writing to a string rather than
|
|
Packit |
577717 |
directly
|
|
Packit |
577717 |
to standard error.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 565f60b3 src/papi_internal.c: Missing code to set num_error_chunks to 0
|
|
Packit |
577717 |
The new _papi_hwi_cleanup_errors() function was not resetting
|
|
Packit |
577717 |
num_error_chunks to 0, leading to a segfault in the fmultiplex1 test.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-02
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bb85bafd src/genpapifdef.c src/papi.c src/papi_common_strings.h...: Remove
|
|
Packit |
577717 |
usage of _papi_hwi_err. Move PAPI over to storing errors in a runtime list.
|
|
Packit |
577717 |
* Functions to add/lookup errors. * Generate the list of PAPI_E* errors at
|
|
Packit |
577717 |
library_init time. * genpapifdef pulled the values for the PAPI_* error
|
|
Packit |
577717 |
return codes from the _papi_hwi_err structure at configure time. Since this
|
|
Packit |
577717 |
is now built at run-time, I added the appropriate values to genpapifdef's
|
|
Packit |
577717 |
builting describe_t table. See : _papi_hwi_publish_error
|
|
Packit |
577717 |
_papi_hwi_init_errors For usage hints.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-10
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e27af085 src/perf_events.c: perf_event: rename BRAINDEAD_MULTIPLEXING It
|
|
Packit |
577717 |
is now "simple_multiplexing" and is a variable not an #ifdef This is needed
|
|
Packit |
577717 |
before perf_event multiplexing can be sorted out. It's unclear if it
|
|
Packit |
577717 |
actually works anyway.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7f8e8c58 src/perf_events.c: perf_event: remove context "cookie" field It
|
|
Packit |
577717 |
was a bit of overkill, we just need an initialized field. Also revamp how
|
|
Packit |
577717 |
context and control are initialized.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8cb8ac6d src/perf_events.c: perf_event: move all event specific info to
|
|
Packit |
577717 |
* the
|
|
Packit |
577717 |
control state previously half was in the context state and half in the
|
|
Packit |
577717 |
control state perf_event has a strange architecture with each event being
|
|
Packit |
577717 |
created having its own fd, which is context wide. In PAPI though we usually
|
|
Packit |
577717 |
only have one eventset (control state) active at once, so there's no need to
|
|
Packit |
577717 |
have the context be aware of this.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-09
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8d7782cb src/perf_events.c: perf_event: rename evt_t to perf_event_info_t
|
|
Packit |
577717 |
This just makes the code easier to follow.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 349de05c src/perf_events.c: perf_event: remove the superfluous
|
|
Packit |
577717 |
per_event_info_t structure
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-08
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* da8ad0a2 src/ctests/all_native_events.c src/ctests/get_event_component.c
|
|
Packit |
577717 |
src/utils/native_avail.c: Fix warnings about PAPI_enum_cmp_event() return
|
|
Packit |
577717 |
not
|
|
Packit |
577717 |
being checked Reported by coverity checker via Will Cohen Harmless
|
|
Packit |
577717 |
warnings, and now the checker will likely complain about the value being
|
|
Packit |
577717 |
checked but ignored.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b4719888 src/papi_user_events.c: Fix unused value in papi_user_events.c
|
|
Packit |
577717 |
Reported by Coverity checker by Will Cohen
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6a8f255c src/utils/event_chooser.c: remove unused
|
|
Packit |
577717 |
* PAPI_get_component_info()
|
|
Packit |
577717 |
call in event_chooser Reported by Will Cohen from coverity checker
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 62cda478 src/genpapifdef.c src/papi_common_strings.h
|
|
Packit |
577717 |
src/papi_internal.c...: Remove usage of _papi_hwi_err. genpapifdef pulled
|
|
Packit |
577717 |
the values for the PAPI_* error return codes from the _papi_hwi_err structure
|
|
Packit |
577717 |
at configure time. Since this is now built at run-time, I added the
|
|
Packit |
577717 |
appropriate values to genpapifdef's builting describe_t table.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-02
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d11259f3 src/papi.c src/papi_internal.c src/papi_internal.h...: Move over
|
|
Packit |
577717 |
to generating the list of PAPI errors at library_init time.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 097ffc44 src/papi_internal.c: Functions to add/lookup errors.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-07
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2530533f src/papi_events.csv: tests/zero fails on Power7 due to PAPI_FP_INS
|
|
Packit |
577717 |
Error of 50%. Preset definition has been redefined and test now passes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8e17836f src/components/appio/Rules.appio src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/appio.h...: We now intercept recv(). The support for
|
|
Packit |
577717 |
recv() requires dynamic linkage. For static linkage, recv is not intercepted.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8b1eb84c src/perf_events.c: perf_events: some whitespace cleanup and extra
|
|
Packit |
577717 |
comments
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f10edba6 src/perf_events.c: perf_events: MAX_READ was no longer being used,
|
|
Packit |
577717 |
remove it
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 08c06ed1 src/perf_events.c: perf_event event_id is actually 64-bit, so make
|
|
Packit |
577717 |
our copy match
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a33e8d9c src/perf_events.c: Rename context_t pe_context_t in perf_events.c
|
|
Packit |
577717 |
Makes the code a bit clearer and matches how other components name things.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 96ce9dcd src/perf_events.c: Rename control_state_t pe_control_state_t This
|
|
Packit |
577717 |
makes the code a bit easier to follow and matches how other components name
|
|
Packit |
577717 |
things.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-03
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4c5dce7f src/ctests/zero.c: Beef up error reporting.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 83b5d28a src/ctests/cycle_ratio.c: Have the cycle_ratio test skip if
|
|
Packit |
577717 |
PAPI_REF_CYC event is not defined.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-02
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 25b1ba41 src/ctests/cycle_ratio.c: Removed all TESTS_QUIET blocks. They
|
|
Packit |
577717 |
aren't needed because tests_quiet() overloads printf. We should probably
|
|
Packit |
577717 |
remove TEST_QUIET blocks in ALL tests at some point for code clarity…
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8777d7d4 src/ctests/zero.c: Fixed error reporting. The error computation
|
|
Packit |
577717 |
was inside a TESTS_QUIET block and wasn't getting executed when run quietly.
|
|
Packit |
577717 |
Thus this test always passed on buildbot, even when it didn't.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 006fe8e9 src/ctests/Makefile: Fix typo in cycle_ratio make line.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 88e6d6a4 src/aix.c src/aix.h: Setting number of multiplex counters back to
|
|
Packit |
577717 |
32 for AIX. Before it was set equal to number of max HW counters. This caused
|
|
Packit |
577717 |
ctests/sdsc-mpx to fail.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ab78deda src/papi_events.csv: ctests/calibrate on Power7/AIX failed with a
|
|
Packit |
577717 |
50% error all the way through. Updated the preset FP_OPS with a more
|
|
Packit |
577717 |
appropriate definition. Now the calibrate errors range from 0.0002 to 0.0011%
|
|
Packit |
577717 |
for double and single precision
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fadce32f src/ctests/calibrate.c: Modify calibrate test in two ways: 1. add
|
|
Packit |
577717 |
a -f flag to suppress failures and allow test to run to completion; 2. change
|
|
Packit |
577717 |
error detection to allow warnings above MAX_WARN and failures above MAX_FAIL.
|
|
Packit |
577717 |
Currently set to 10% and 80% respectively. This allows speculative over
|
|
Packit |
577717 |
counting to pass with warning rather than fail completely.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8a39ac9d src/papi_events.csv: LST_INS for Power7 was defined from 3 native
|
|
Packit |
577717 |
events that cannot be counted together at the same time. Caused
|
|
Packit |
577717 |
ctests/all_events to fail. Updated the preset with a more appropriate
|
|
Packit |
577717 |
definition.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* cdc16e5d src/papi_events.csv: L1_DCA for Power7 was defined from 3 native
|
|
Packit |
577717 |
events that cannot be counted together at the same time. That caused
|
|
Packit |
577717 |
ctests/tenth to fail. Updated the preset with a more appropriate definition.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-08-01
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2bf44d13 src/papi_internal.c src/perf_events.c: icc does not like
|
|
Packit |
577717 |
arithmetic on void pointers. Added cast to unsigned char* when arithmetic was
|
|
Packit |
577717 |
being performed on void pointers in papi_internal and perf_events.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7825ec14 src/ctests/api.c src/ctests/attach2.c src/ctests/attach3.c...:
|
|
Packit |
577717 |
Modify tests that FAIL if PAPI_FP_OPS or PAPI_FP_INS not implemented. Now
|
|
Packit |
577717 |
they will warn and continue. This is specifically to accommodate the
|
|
Packit |
577717 |
brain-dead IvyBridge implementation.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fd70a015 src/testlib/test_utils.c: Re-writing of test_utils introduced new
|
|
Packit |
577717 |
bugs that caused ctests/tenth to fail. test_events struct lists the same
|
|
Packit |
577717 |
event twice (MASK_L1_DCW), hence PAPI_add_event() fails because it's forced
|
|
Packit |
577717 |
to add the same preset twice.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 74ece3a0 src/run_tests.sh: run_tests.sh was clobbering $EXCLUDE variable if
|
|
Packit |
577717 |
$CUDA was defined. Changed to add entries from run_tests_exclude_cuda.txt to
|
|
Packit |
577717 |
$EXCLUDE which should already contain entries from run_tests_exclude.txt
|
|
Packit |
577717 |
instead of replacing the entries already contained.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 11ed2364 src/libpfm4/config.mk: Added check in libpfm4/config.mk to check
|
|
Packit |
577717 |
if using icc. If so, the -Wno-unused-parameter flag will no longer be used
|
|
Packit |
577717 |
because icc does not provide it and provides no alternative.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* dedf73f6 src/papi_user_events.c: fget() returns an int it should be treated
|
|
Packit |
577717 |
as an int The coverity scan flagged that the int return by fget was stored
|
|
Packit |
577717 |
in a char. The main concern with this is the EOF that fget() could return is
|
|
Packit |
577717 |
-1. Do not want to mess up that value by typecasting to char and then back to
|
|
Packit |
577717 |
int.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c4fcbe7e src/ctests/kufrin.c: Check return values of PAPI_get_cmp_opt() and
|
|
Packit |
577717 |
calloc A coverity scan showed that PAPI_get_cmp_opt() could potentially
|
|
Packit |
577717 |
return a negative number. Also it is good form to check the return value of
|
|
Packit |
577717 |
calloc to ensure it is a non-null pointer.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e89d6ffa src/testlib/test_utils.c: Clean up test_print_event_header()
|
|
Packit |
577717 |
There were a couple warnings flagged by coverity on
|
|
Packit |
577717 |
test_print_events_header(). The function now checks for error conditions
|
|
Packit |
577717 |
flagged by PAPI_get_cmp_opt() and also frees memory allocated by a calloc()
|
|
Packit |
577717 |
function.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c81d8b60 src/threads.h: Eliminate deadcode from threads.h If
|
|
Packit |
577717 |
HAVE_THREAD_LOCAL_STORAGE is defined, a portion of the
|
|
Packit |
577717 |
_papi_hwi_lookup_thread() will never be executed. This patch make either one
|
|
Packit |
577717 |
section or the other section of code be compiled. This will eliminate a
|
|
Packit |
577717 |
coverity scan warning about unreachable code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f70f3f56 src/ctests/all_native_events.c: Eliminate unused variable in
|
|
Packit |
577717 |
ctests/all_native_events.c Coverity identified a variable that was set but
|
|
Packit |
577717 |
never used in all_native_events.c. This patch removes the unused variable to
|
|
Packit |
577717 |
eliminate that warning.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a9f29840 src/components/appio/appio.c: A couple places in appio.c used the
|
|
Packit |
577717 |
FD_SET() without initializing the variable. Coverity scan pointed out this
|
|
Packit |
577717 |
issue.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9e535ae2 src/components/rapl/linux-rapl.c: A Coverity scan pointed out that
|
|
Packit |
577717 |
read_msr() could potentially use an invalid value of fd for pread(). Need to
|
|
Packit |
577717 |
check the value of fd before using it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7b55c675 src/components/rapl/linux-rapl.c: The arrays used for
|
|
Packit |
577717 |
initialization were hard coded to 1024 packages. Want to avoid hard coding
|
|
Packit |
577717 |
that so the day when machines with 1025 packages are available is a
|
|
Packit |
577717 |
non-event. Also changed the initialization code to avoid having the
|
|
Packit |
577717 |
initialization be O((number of packages)^2) in time complexity.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-27
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3703995a src/papi_internal.c: Fix the component name predending code. When
|
|
Packit |
577717 |
presented with a NULL component .short_name, the code did the wrong thing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5258db8b src/components/cuda/linux-cuda.c: Fix a warning in cuda.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-26
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ddd6f193 src/ctests/Makefile src/ctests/cycle_ratio.c: Add a test to
|
|
Packit |
577717 |
compute nominal CPU MHz from real cycles and use PAPI_TOT_CYC and
|
|
Packit |
577717 |
PAPI_REF_CYC to compute effective MHz. Warns if PAPI_REF_CYC is zero, which
|
|
Packit |
577717 |
can happen on kernels < ~3.3.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fab5e9ef src/papiStdEventDefs.h src/papi_common_strings.h
|
|
Packit |
577717 |
src/papi_events.csv: Add PAPI_REF_CYC preset event. Define it as
|
|
Packit |
577717 |
UNHALTED_REFERENCE_CYCLES for all Intel platforms on which this native event
|
|
Packit |
577717 |
is supported.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-25
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8b9b6bef src/papi_events.csv: Modify SandyBridge and IvyBridge tables:
|
|
Packit |
577717 |
SandyBridge FP_OPS only counts scalars; SP_OPS and DP_OPS now count
|
|
Packit |
577717 |
correctly, including SSE and AVX. IvyBridge can't count FP at all;
|
|
Packit |
577717 |
adjustments made to eliminate event differences with SandyBridge.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-26
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5b11c982 src/components/cuda/linux-cuda.c: Fix the cuda component. The
|
|
Packit |
577717 |
cuda component prepended CUDA. to all its event names, this is no longer the
|
|
Packit |
577717 |
case.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-25
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* db5b0857 src/papi_events.csv: Added 2 new preset definitions for BGQ. Note,
|
|
Packit |
577717 |
these presets use the new feature where a generic event mask together with an
|
|
Packit |
577717 |
ORed opcode string is used. This won't work until the new Q driver is
|
|
Packit |
577717 |
released (currently scheduled for end of August).
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-24
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* af7cd721 src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/coretemp/tests/coretemp_pretty.c
|
|
Packit |
577717 |
src/components/cuda/linux-cuda.c...: Enforce all our components to use the
|
|
Packit |
577717 |
same naming. We setteled on :'s as inter-event seperators. This also touches
|
|
Packit |
577717 |
a few of the components' tests, we changed the name field so their searches
|
|
Packit |
577717 |
needed help.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 57aeb9d4 src/papi_internal.c: Prepend component .short_name to each event
|
|
Packit |
577717 |
name. Use ::: as a sep.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-24
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 762e9584 src/ctests/multiplex2.c src/sw_multiplex.c: Fix multiplex2 test
|
|
Packit |
577717 |
It complained if it tried to add a multiplex event and PAPI properly told it
|
|
Packit |
577717 |
that it couldn't.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 531870f1 src/papi_internal.c: Add sanity check at component init time
|
|
Packit |
577717 |
Looks for num_cntrs being larger than num_mpx_cntrs which doesn't make much
|
|
Packit |
577717 |
sense.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 53ad0259 src/extras.c src/genpapifdef.c src/papi.c...: Rename
|
|
Packit |
577717 |
PAPI_MAX_HWCTRS to PAPI_EVENTS_IN_DERIVED_EVENT Hopefully this will make
|
|
Packit |
577717 |
things a little less confusing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 700af24b src/papi_internal.c: Change EventInfoArrayLength to always return
|
|
Packit |
577717 |
num_mpx_cntrs Things should be consistently using num_mpx_cntrs rather than
|
|
Packit |
577717 |
num_cntrs now. Issue reported by Steve Kaufmann
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d1570bec src/sw_multiplex.c: Fix sw_multiplex bug when max SW counters is
|
|
Packit |
577717 |
less than max HW counters this was causing kufrin and max_multiplex to fail
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f47f5d6a src/aix.c src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/linux-CNKunit.c...: Remove PAPI_MPX_DEF_DEG It
|
|
Packit |
577717 |
was not well documented and being used in confused ways all over the code.
|
|
Packit |
577717 |
Now there is a different define PAPI_MAX_SW_MPX_EVENTS used by the software
|
|
Packit |
577717 |
multiplex code. All other components have had the value replaced with just
|
|
Packit |
577717 |
the maximum number of counters. If a component can handle its own
|
|
Packit |
577717 |
(non-software) multiplexing it is up to it to set .num_mpx_cntrs to a value
|
|
Packit |
577717 |
that's different from .num_cntrs
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0d83f5db src/papi_internal.c src/papi_internal.h: Split NativeBits off of
|
|
Packit |
577717 |
NativeInfoArray in EventSet previously we were doing some crazy thing where
|
|
Packit |
577717 |
we allocated both at once and then split them afterward. The new code is
|
|
Packit |
577717 |
easier to follow.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 98f2ecbd src/papi_internal.c: Clean up EventSet creation Sort out which
|
|
Packit |
577717 |
sizes are used for allocating which structures.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e1024579 src/Makefile.inc src/multiplex.c src/multiplex.h...: Rename the
|
|
Packit |
577717 |
multiplex files to be sw_multiplex That way it's more clear the stuff
|
|
Packit |
577717 |
included only relates to software multiplexing, not generic multiplexing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a6adc7ff src/multiplex.h src/papi_internal.c src/papi_internal.h: Move some
|
|
Packit |
577717 |
sw-multiplex specific terms out of papi_internal.h and into multiplex.h
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1ddbe117 src/components/README: Added note that lmsensors component
|
|
Packit |
577717 |
requires lmsensors version >=3.0.0
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 94676869 src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_pthreads.c: proper checking of return
|
|
Packit |
577717 |
codes in response to tests using coverity
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ea958b18 src/components/appio/tests/appio_list_events.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_values_by_code.c: As component name in table
|
|
Packit |
577717 |
has been changed from appio.c to appio, we now use appio in the tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f212cc34 src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/coretemp_freebsd/coretemp_freebsd.c...: Add .short_name
|
|
Packit |
577717 |
entries to each component.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1e755836 src/papi_libpfm4_events.c src/perf_events.c: Fix use-after-free
|
|
Packit |
577717 |
bug in perf_events.c This turned up in the ctests/flops test, and Valgrind
|
|
Packit |
577717 |
made it easy to track down.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4580ed1d src/perf_events.c: Update perf_event.c rdpmc support Use the
|
|
Packit |
577717 |
libpfm4 definition for mmap rather than our custom one, now that libpfm4 has
|
|
Packit |
577717 |
been updated
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 47558b2c src/libpfm4/examples/showevtinfo.c
|
|
Packit |
577717 |
src/libpfm4/include/perfmon/perf_event.h
|
|
Packit |
577717 |
src/libpfm4/lib/pfmlib_intel_nhm_unc.c...: Import current libpfm4 from
|
|
Packit |
577717 |
libpfm4 git It has some minor uncore fixes plus the header changes needed
|
|
Packit |
577717 |
for rdpmc.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 65d4c06c src/linux-common.c: Reorder statements to ensure the fclose() are
|
|
Packit |
577717 |
performed Coverity pointed out that it was possible for resources to be
|
|
Packit |
577717 |
leaked in linux-common.c if the fscanf() encountered error. This reordering
|
|
Packit |
577717 |
of the statements ensures that the fclose() calls are done regards of the
|
|
Packit |
577717 |
results of the fscanf() functions.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7bf071ff src/papi_user_events.c: Ensure that load_user_event_table() frees
|
|
Packit |
577717 |
files and memory on error A Coverity scan showed that an error condition in
|
|
Packit |
577717 |
load_user_event_table() function would exit the the function without closing
|
|
Packit |
577717 |
the table file or freeing allocated memory. This patch addresses those
|
|
Packit |
577717 |
issues.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1ba52e35 src/components/stealtime/linux-stealtime.c: Ensure that
|
|
Packit |
577717 |
read_stealtime() closes the file in case of an error condition A Coverity
|
|
Packit |
577717 |
scan showed that an error condition could cause read_stealtime() to exit
|
|
Packit |
577717 |
without closing the file. This patch ensures that the file is closed
|
|
Packit |
577717 |
regardless of success or failure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f37f22e5 src/papi_libpfm4_events.c: Fix warning in papi_libpfm4_events.c
|
|
Packit |
577717 |
We were setting a value but never using it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8e8401bc src/testlib/test_utils.c: Remove unused variable in test_utils.c
|
|
Packit |
577717 |
Most of the machines in buildbot were complaining about this.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 133ce6a9 src/linux-timer.c: Add missing papi_vector.h include to
|
|
Packit |
577717 |
linux-timer.c This was breaking on PPC Linux
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6fd3cedd src/perf_events.c: Fix perf_events.c warnings reported by ICC
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 21c8f932 src/perfctr-x86.c: Fix perfctr-x86 build with debug enabled
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 08f76743 src/configure src/configure.in src/linux-bgq.c: Attempt to fix
|
|
Packit |
577717 |
linux-bgq compilation error. It turns out BGQ uses the standard
|
|
Packit |
577717 |
linux-context.h header
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 43457f4f src/linux-bgq.c: Made check for opcodes more robust.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d58116b4 src/perf_events.c: More cleanups of perf_events.c file
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 409438b7 src/freebsd-context.h src/freebsd.c src/freebsd.h: Fix FreeBSD
|
|
Packit |
577717 |
compile warnings Similar to the perfctr issues.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1e6dfb02 src/aix.c src/aix.h: Fix AIX build warnings They were similar in
|
|
Packit |
577717 |
cause to the perfctr issues.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3d0b5785 src/Rules.perfmon2 src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/linux-CNKunit.h...: Remove papi_vector.h include
|
|
Packit |
577717 |
from papi_internal.h There were some semi-circular dependencies that came up
|
|
Packit |
577717 |
with the context split changes. The easiest way to fix things for perfctr
|
|
Packit |
577717 |
was just move papi_vector.h out to be included explicitly. This touches a
|
|
Packit |
577717 |
lot of files because a lot of files include papi_internal.h This should also
|
|
Packit |
577717 |
fix the perfctr and perfmon2 builds that were broken yesterday.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-16
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a7a14a5b src/ctests/zero.c src/testlib/test_utils.c: Modify zero test to
|
|
Packit |
577717 |
warm up processor before measuring events, and report timing errors as signed
|
|
Packit |
577717 |
deviations. Modify test_utils add_two_events code to check for errors after
|
|
Packit |
577717 |
adding nominally valid events. This is a more rigorous test than just
|
|
Packit |
577717 |
counting available registers.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* de0860d3 src/perf_events.c: Remove perf_events.h module header It's no
|
|
Packit |
577717 |
longer needed, everything important is merged into the perf_events.c file.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 22975f14 src/perf_events.c: Remove perf_event SYNCHRONIZED_RESET code This
|
|
Packit |
577717 |
was never defined and never used, just remove the code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 48750b8c src/perf_events.c: Remove papi_pe_allocate_registers On
|
|
Packit |
577717 |
perf_event this code wasn't really doing anything useful, as
|
|
Packit |
577717 |
update_control_state would end up re-doing any possible tests we could want
|
|
Packit |
577717 |
to do here.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1775566f src/Makefile.in src/Makefile.inc src/Rules.pfm4_pe...: Remove
|
|
Packit |
577717 |
"include CPUCOMPONENT" from papi_internal.h This was the last major
|
|
Packit |
577717 |
dependency on CPU component in common PAPI code. It was mostly necessary for
|
|
Packit |
577717 |
the ucontext definitions when trying to get the instruction pointer when
|
|
Packit |
577717 |
doing sampling. This change creates new OS-specific header files that handle
|
|
Packit |
577717 |
the ucontext related code, and has papi_internal.h include that instead.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 969ce035 src/Rules.pfm4_pe src/Rules.pfm_pe src/configure...: Make
|
|
Packit |
577717 |
perf_event libpfm4 only perf_event libpfm3 support is not really needed
|
|
Packit |
577717 |
anymore and supporting it was cluttering up the perf_event component.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-13
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* adad1d2a src/perf_events.c: Add init time error messages to perf_event
|
|
Packit |
577717 |
component
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 827ccc07 src/perf_events.c: Add perf_event rdpmc / fast_real_timer
|
|
Packit |
577717 |
detection Currently we need a custom copy of struct perf_event_mmap_page
|
|
Packit |
577717 |
because the version included with libpfm4 doesn't define the fields we need
|
|
Packit |
577717 |
yet.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4f82fe94 src/perf_events.c: Read in paranoid info on perf_events This
|
|
Packit |
577717 |
indicates whether a regular user can read CPU-specific or system-wide
|
|
Packit |
577717 |
measurements.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 03080450 src/perf_events.c: Add proper perf_event detection Using the
|
|
Packit |
577717 |
official /proc/sys/kernel/perf_event_paranoid file
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6e71d3f7 src/linux-bgq.c: Updated BGQ opcode stuff; cleaned up code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-11
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3114d3dc src/multiplex.c src/papi_internal.c src/perf_events.c: Minor
|
|
Packit |
577717 |
documentation improvements Plus fixes some typos
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-09
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b60c0f0c src/perf_events.c: Minor cleanups to perf_events.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 075278a0 src/aix.c src/freebsd.c src/linux-bgp.c...: Change return value
|
|
Packit |
577717 |
for .allocate_registers For some reason it returned 1 on success and 0 on
|
|
Packit |
577717 |
failure. Change it so you return PAPI_OK on success and a PAPI error on
|
|
Packit |
577717 |
failure, to better match all of the other component vectors.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 29d9e62b src/testlib/test_utils.c: Fixed the print_header routine to report
|
|
Packit |
577717 |
an error message if counters are not found, instead of a negative counter
|
|
Packit |
577717 |
number. Tested by forcing the return value negative; not by running on a Mac,
|
|
Packit |
577717 |
where the error first appeared.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 74257334 src/ctests/Makefile src/ctests/remove_events.c: Add remove_events
|
|
Packit |
577717 |
test This just makes sure EventSets still work after an event has been
|
|
Packit |
577717 |
removed. This is probably covered by other more elaborate tests, but I
|
|
Packit |
577717 |
needed a simple test to make sure I wasn't breaking anything.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1372714f src/papi.c src/papi_internal.c src/papi_internal.h: Clean up,
|
|
Packit |
577717 |
rename, and comment _papi_hwi_remap_event_position I've found this section
|
|
Packit |
577717 |
of code to be confusing for a long time. I think I finally have it mostly
|
|
Packit |
577717 |
figured out. I've renamed it _papi_hwi_map_events_to_native() to better
|
|
Packit |
577717 |
describe what it does. The issue is that the native event list in an
|
|
Packit |
577717 |
EventSet can change at various times. At event add, event remove, and
|
|
Packit |
577717 |
somewhat unexpectedly, whenever ->update_control_state is called (a component
|
|
Packit |
577717 |
can re-arrange native events if it wants, to handle conflicts, etc.) Once
|
|
Packit |
577717 |
the native event list has been changed, _papi_hwi_map_events_to_native() has
|
|
Packit |
577717 |
to be called to make sure the events all map to the proper native_event
|
|
Packit |
577717 |
again. Currently we have the _papi_hwi_map_events_to_native() calls in odd
|
|
Packit |
577717 |
places. It seems to cover all possible needed locations, but analyzing that
|
|
Packit |
577717 |
we do takes a lot of analysis...
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f1b837d8 src/papi.c: Remove unused variable in papi.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 541bcf44 src/papi_internal.h: Update commens in papi_internal.h Some of
|
|
Packit |
577717 |
the EventSetInfo comments were out of date.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e6587847 src/papi.c src/papi_internal.c src/papi_internal.h: Remove unused
|
|
Packit |
577717 |
paramater from _papi_hwi_remap_event_position The mysterious
|
|
Packit |
577717 |
_papi_hwi_remap_event_position function had a "thisindex" field that was
|
|
Packit |
577717 |
ignored. This will possibly speed PAPI_start() time as it was running a loop
|
|
Packit |
577717 |
over num_native_events on _papi_hwi_remap_event_position even though each
|
|
Packit |
577717 |
call did the same thing since the value being passed was ignored.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3ad3d14b src/papi_internal.c: Clean up and comment add_native_events in
|
|
Packit |
577717 |
papi_internal.c I'm chasing some weird perf_events behavior with the
|
|
Packit |
577717 |
papi_event_chooser. The add_native_events code is very hard to understand,
|
|
Packit |
577717 |
working on commenting it more.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4e5e7664 src/utils/event_chooser.c: Fix coverity warning in
|
|
Packit |
577717 |
papi_event_chooser
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 666249a8 src/jni/EventSet.java src/jni/FlipInfo.java
|
|
Packit |
577717 |
src/jni/FlopInfo.java...: RIP Java. Java PAPI wrappers have not been
|
|
Packit |
577717 |
supported for years (2005?). They are being removed to declutter the source.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e18561fc src/papi_preset.c: Update cmpinfo->num_preset_events properly
|
|
Packit |
577717 |
This value wasn't being set if we were reading the presets directly from the
|
|
Packit |
577717 |
CSV file.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 290ab7c3 src/utils/component.c: Have papi_component_avail report counter
|
|
Packit |
577717 |
and event info
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7c421b9c src/testlib/test_utils.c src/utils/native_avail.c: Remove counter
|
|
Packit |
577717 |
number from the testlib header. The header was only reporting number of
|
|
Packit |
577717 |
counters for the CPU component, even though the header is printed for many
|
|
Packit |
577717 |
utils and the CPU component might not even be involved. This could be a bit
|
|
Packit |
577717 |
confusing, so remove it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 26432359 src/darwin-common.c src/darwin-memory.c: Improve OSX support This
|
|
Packit |
577717 |
properly detects CPU information now. You can get results like this:
|
|
Packit |
577717 |
Available native events and hardware information.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
- PAPI Version : 4.9.0.0 Vendor string and code : GenuineIntel
|
|
Packit |
577717 |
(1) Model string and code : Intel(R) Core(TM) i5-2435M CPU @ 2.40GHz (42)
|
|
Packit |
577717 |
CPU Revision : 7.000000 CPUID Info : Family: 6
|
|
Packit |
577717 |
Model: 42 Stepping: 7 CPU Max Megahertz : 2400 CPU Min Megahertz
|
|
Packit |
577717 |
: 2400 CPUs per Node : 0 Total CPUs : 4 Running in
|
|
Packit |
577717 |
a VM : no Number Hardware Counters : -4 Max Multiplex Counters :
|
|
Packit |
577717 |
-4
|
|
Packit |
577717 |
|
|
Packit |
577717 |
-
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-08
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 845d9ecb src/Makefile.inc src/configure src/configure.in...: Add Mac OSX
|
|
Packit |
577717 |
support This is enough that things compile and simple utilities run. No CPU
|
|
Packit |
577717 |
perf counter support.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ff6f9ab4 src/linux-bgq.c: missed to delete a debug output.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 12e9a11a RELEASENOTES.txt: Release notes for the 4.4 release.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ac2eac56 src/papi.c src/papi.h: Add a PAPI_disable_component_by_name entry
|
|
Packit |
577717 |
point.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8c490849 src/components/coretemp_freebsd/coretemp_freebsd.c src/freebsd.c:
|
|
Packit |
577717 |
Fix FreeBSD to work I'm not sure how it ever worked in the past. With these
|
|
Packit |
577717 |
changes I can at least do a papi_component_avail and a papi_native_avail and
|
|
Packit |
577717 |
get sane results
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 108b5ce6 src/freebsd.c src/freebsd.h src/freebsd/map-atom.c...: Fix FreeBSD
|
|
Packit |
577717 |
build some of the recent changes broke the FreeBSD build
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 40a60f0a src/linux-bgq.c src/linux-bgq.h: Added BGQ's opcode and generic
|
|
Packit |
577717 |
event functionality to PAPI. For BGQ there are multiple ways to define
|
|
Packit |
577717 |
presets. The naive way is to derive from multiple events. This eats up
|
|
Packit |
577717 |
multiple counters and we lose sample capability as well as overflow
|
|
Packit |
577717 |
capability. On the other side, some events come with multiple InstrGroup
|
|
Packit |
577717 |
derived in the field. If that's the case we can use a generic event and
|
|
Packit |
577717 |
opcodes to filter multiple groups in a single counter. This is not working
|
|
Packit |
577717 |
properly yet due to a known error in BGPM. Bgpm_AddEvent() does not work
|
|
Packit |
577717 |
properly when multiple generic events are added to an EventSet. The BGPM
|
|
Packit |
577717 |
folks have been made aware of this issue, they confirmed the error, and they
|
|
Packit |
577717 |
are currently working on a fix.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6f72b70f src/papi_events_table.sh: Make this script robust enough to handle
|
|
Packit |
577717 |
any line ending, including CR (Mac), CRLF (Windows), and LF (Unix). It
|
|
Packit |
577717 |
appears that google mail now automagically converts attached files to CRLF
|
|
Packit |
577717 |
format.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 765ed0d2 src/papi_internal.c: Fix a type warning in the UE code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 94bc1b15 src/MACROS: Remove the MACROS file it held out of date info and
|
|
Packit |
577717 |
hasn't been touched since 2004
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d19e73ba src/ctests/Makefile src/ctests/clockcore.c
|
|
Packit |
577717 |
src/testlib/Makefile...: Move the clockcore.c file from ctests to testlib
|
|
Packit |
577717 |
it's common code used by multiple tests, including some in the utils
|
|
Packit |
577717 |
directory also add a function definition to fix a build-time warning
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1101a6aa src/aix-lock.h src/aix.h src/configure...: Make papi_lock.h
|
|
Packit |
577717 |
changes for non Linux architectures
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-05
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3b82b03d src/Makefile.in src/Makefile.inc src/aix.c...: Make the PAPI locks
|
|
Packit |
577717 |
be tied to OS, not to CPU There is not a papi_lock.h file that when included
|
|
Packit |
577717 |
gets the proper lock include for the OS. This fixes a lot of previous build
|
|
Packit |
577717 |
hacks where a CPU component was needed in order for locks to work.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0632ef42 src/threads.c: Fix spurious init_thread call in threads.c
|
|
Packit |
577717 |
threads.c was calling init_thread() on all components, even ones that were
|
|
Packit |
577717 |
disabled Fix it to honor the disable bit, as well as for shutdown_thread().
|
|
Packit |
577717 |
This was causing perfctr disable code to not work.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 19d9de7f src/Makefile.in src/Makefile.inc src/Rules.pfm4_pe...: Replace
|
|
Packit |
577717 |
SUBSTRATE with CPUCOMPONENT in build This was mostly a configure/build
|
|
Packit |
577717 |
change but it also cleaned up some cases where we were including SUBSTRATE
|
|
Packit |
577717 |
where we didn't have to.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 829780db src/solaris-common.c src/solaris-common.h
|
|
Packit |
577717 |
src/solaris-niagara2.c...: Move some common solaris code to solaris-common
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 681ef027 src/configure src/configure.in src/solaris-memory.c...: Merge
|
|
Packit |
577717 |
solaris-memory.c and solaris-niagara2-memory.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bbd41743 src/solaris-ultra/get_tick.S src/solaris.h: Remove
|
|
Packit |
577717 |
solaris-ultra/get_tick.S Nothing was using it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* dc3b6920 src/papi_sys_headers.h src/solaris.h: Remove papi_sys_headers.h
|
|
Packit |
577717 |
Solaris was the only thing including it, and it wasn't really using it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7ccfa9df src/Makefile.in src/Makefile.inc src/configure...: Move move OS
|
|
Packit |
577717 |
specific code into the new OSFILESSRC Linux in particular was using MISC for
|
|
Packit |
577717 |
this.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6f16c0c5 src/configure: Re-run autoconf to pickup the substrate=>component
|
|
Packit |
577717 |
change.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* cfff1ede src/Makefile.in src/Makefile.inc src/configure...: Remove
|
|
Packit |
577717 |
MEMSUBSTR In reality, what we want instead of a Memory Substrate is an idea
|
|
Packit |
577717 |
of the OS-specific common code that includes the memory substrate. This
|
|
Packit |
577717 |
change adds OSFILESSRC and OSFILESOBJ to handle this case in configure
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ca4729e6 src/configure.in: Separate out MEMSUBSTR and make it per-OS
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3148cba5 src/Matlab/PAPI_Matlab.dsp src/ctests/calibrate.c
|
|
Packit |
577717 |
src/ctests/flops.c...: RIP Windows, remove the windows support code. Windows
|
|
Packit |
577717 |
has not been activly supported since the transition to Component PAPI (4.0)
|
|
Packit |
577717 |
This cleans up the code-base.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-03
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a366adf7 src/papi.c src/utils/error_codes.c: Change PAPI_strerror and
|
|
Packit |
577717 |
PAPI_perror to behave more like thir POSIX namesakes. PAPI_error_descr is
|
|
Packit |
577717 |
made redundant and removed as a result.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-05
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7df46f81 src/Rules.pfm src/aix.c
|
|
Packit |
577717 |
src/components/coretemp/linux-coretemp.c...: Move uses of PAPI_ESBSTR to
|
|
Packit |
577717 |
PAPI_ECMP I left PAPI_ESBSTR defined too for backward compatability. Also
|
|
Packit |
577717 |
some of the changes update PAPI_ESBSTR to be a more relevant error code, it
|
|
Packit |
577717 |
one is available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-03
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fdb348ad src/components/coretemp_freebsd/coretemp_freebsd.c
|
|
Packit |
577717 |
src/components/example/example.c src/components/net/linux-net.c...: A few
|
|
Packit |
577717 |
more substrate removals
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 791747c1 src/cpus.c src/papi.h src/perf_events.c...: Fix bugs introduced by
|
|
Packit |
577717 |
substrate -> component change Fix some stupid compile bugs that I missed.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 79b01a47 src/aix.c src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/linux-CNKunit.c...: More substrate -> component
|
|
Packit |
577717 |
changes This changes the vectors .init_substrate -> .init_component
|
|
Packit |
577717 |
.shutdown_substrate -> .shutdown_substrate .init -> .init_thread .shutdown ->
|
|
Packit |
577717 |
.shutdown_thread hopefully this will make the code clearer.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 02a10d71 src/Makefile.inc src/aix.c src/cpus.c...: Rename "substrate" to
|
|
Packit |
577717 |
"component" This first pass only re-names things in comments.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-07-02
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c4bbff1c src/papi.c src/papi.h: Minor documentation fixes Found when
|
|
Packit |
577717 |
writing up the PAPI 5.0 changes document
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-30
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f9cb7346 src/components/vmware/vmware.c: Fix vmware component apparently I
|
|
Packit |
577717 |
forgot to test the build with the vmguestlib support disabled.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-22
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 599040d1 src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/rapl/linux-rapl.c
|
|
Packit |
577717 |
src/components/stealtime/linux-stealtime.c...: Fix libpfm4 ntv_event_to_info
|
|
Packit |
577717 |
event_info_t on other components This was actually a widespead problem due
|
|
Packit |
577717 |
to cut-and-paste.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2b51b439 src/papi_libpfm4_events.c: Properly fix libpfm4 ntv_event_to_info
|
|
Packit |
577717 |
event_info_t event value The previous fix was subtly wrong. This is the
|
|
Packit |
577717 |
proper fix, which is to do nothing inside of papi_libpfm4_events.c because
|
|
Packit |
577717 |
papi_internal.c does the right thing for us and we were overwriting that with
|
|
Packit |
577717 |
the wrong value.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a4f576bf src/ctests/overflow_allcounters.c src/testlib/papi_test.h
|
|
Packit |
577717 |
src/testlib/test_utils.c: Clean up overflow_allcounters code While tracking
|
|
Packit |
577717 |
down a previous issue I also cleaned up the overflow_allcounters test code to
|
|
Packit |
577717 |
use some of the new interfaces.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6903e053 src/papi_libpfm4_events.c: Fix libpfm4 ntv_event_to_info
|
|
Packit |
577717 |
event_info_t event value The recently added libpfm4 ntv_event_to_info
|
|
Packit |
577717 |
function was not properly oring PAPI_NATIVE_MASK to the event value in the
|
|
Packit |
577717 |
event_info_t struct. That means if you tried to use that event value to add
|
|
Packit |
577717 |
an event it would fail. The overflow_allcounters test broke because of this.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 420c3d11 src/ctests/Makefile src/ctests/disable_component.c src/papi.c...:
|
|
Packit |
577717 |
Add PAPI_get_component_index() and PAPI_disable_component()
|
|
Packit |
577717 |
PAPI_get_component_index() will return a component index if you give it the
|
|
Packit |
577717 |
name of a component to match. This saves you having to iterate the entire
|
|
Packit |
577717 |
component list looking. PAPI_disable_component() will manually mark a
|
|
Packit |
577717 |
component as disabled. It has to be run before PAPI_library_init() is called.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 11946525 src/aix.c src/components/cuda/linux-cuda.c
|
|
Packit |
577717 |
src/components/example/example.c...: Standardize component names to not end
|
|
Packit |
577717 |
in .c We were being inconsistent; the time to make them all be the same is
|
|
Packit |
577717 |
now before 5.0 gets out.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-21
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 274e1ad8 src/components/vmware/tests/Makefile: Fix cut-and-paste error in
|
|
Packit |
577717 |
the vmware component Makefile
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 85d6438d src/utils/event_chooser.c: Update papi_event_chooser to work with
|
|
Packit |
577717 |
components Now you can specify events from components and it will tell you
|
|
Packit |
577717 |
all the other events on that component that can run with it. Previously the
|
|
Packit |
577717 |
utility was limited to the CPU component (0) only.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3c2fcc83 src/papi_libpfm3_events.c src/papi_libpfm4_events.c
|
|
Packit |
577717 |
src/perf_events.c: Hook up .ntv_code_to_info on perf_event
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 36e864b3 src/papi_libpfm4_events.c src/papi_libpfm_events.h
|
|
Packit |
577717 |
src/perf_events.c: Enable support for showing extended umasks on perf_event
|
|
Packit |
577717 |
With this change, papi_native_avail now shows event umasks such as :u, :k,
|
|
Packit |
577717 |
:c, :e, and :i. (user, kernel, cmask, edge-trigger, invert) Thes are boolean
|
|
Packit |
577717 |
or integer events. They were supported by previous PAPI but they were never
|
|
Packit |
577717 |
enumerated.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8f3e305e src/components/coretemp/linux-coretemp.c: Fix cut-and-paste error
|
|
Packit |
577717 |
in linux-coretemp.c that could lead to wrong size being copied
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0eedd562 src/libpfm4/lib/events/intel_atom_events.h
|
|
Packit |
577717 |
src/libpfm4/lib/events/intel_core_events.h
|
|
Packit |
577717 |
src/libpfm4/lib/events/intel_coreduo_events.h...: Import most recent libpfm4
|
|
Packit |
577717 |
git This fixes an issue where there can be confusion between :i and :i=1
|
|
Packit |
577717 |
type events. It also has initial support for Uncore, though you need a
|
|
Packit |
577717 |
specially patched kernel and PAPI does not support it yet.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2f86ec78 src/components/appio/tests/Makefile
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_blocking.c
|
|
Packit |
577717 |
.../appio/tests/appio_test_fread_fwrite.c...: - Fixed tests verbosity by
|
|
Packit |
577717 |
using TESTS_QUIET macro - Fixed Makefile to only include necessary tests for
|
|
Packit |
577717 |
automatic builds (skip blocking tests that read from stdin)
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6936b955 src/components/appio/README src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/appio.h...: Added polling of read/write descriptor to
|
|
Packit |
577717 |
check which operations would block.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 48cacccf src/papi.h: Add back PAPI_COMPONENT_INDEX() for backward
|
|
Packit |
577717 |
compatability It turns out some people were using this for cases other than
|
|
Packit |
577717 |
enumeration. The proper way to do things now is to use
|
|
Packit |
577717 |
PAPI_get_event_component() which is what this PAPI_COMPONENT_INDEX() now maps
|
|
Packit |
577717 |
to.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d1ed12b7 src/ctests/Makefile src/ctests/get_event_component.c
|
|
Packit |
577717 |
src/papi.c...: Add PAPI_get_event_component() This function returns the
|
|
Packit |
577717 |
component an event belongs to. It also adds a test to test this
|
|
Packit |
577717 |
functionality.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ffccf633 src/papi.h: Add component_type field to .cmp_info The idea is
|
|
Packit |
577717 |
we'll specify CPU, I/O, GPU, hardware, etc.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9998eecc src/components/lmsensors/Rules.lmsensors: Another lmsensors build
|
|
Packit |
577717 |
fix
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* caa94d64 src/components/lmsensors/linux-lmsensors.c: Update lmsensors
|
|
Packit |
577717 |
component to actually compile. I finally found a machine with lmsensors
|
|
Packit |
577717 |
installed.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fbcde325 src/components/lmsensors/linux-lmsensors.c
|
|
Packit |
577717 |
src/components/lmsensors/linux-lmsensors.h: Update lmsensor component Unlike
|
|
Packit |
577717 |
the other components it hadn't been updated to PAPI 5 standards. Also, it
|
|
Packit |
577717 |
was wrongly de-allocating all state in "_shutdown" rather than
|
|
Packit |
577717 |
"_shutdown_substrate" which was causing double-frees during tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0d3c0ae2 src/papi_internal.c: Add some extra debugging to
|
|
Packit |
577717 |
_papi_hwi_get_native_event_info
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5961c03d src/aix.c src/components/nvml/linux-nvml.c
|
|
Packit |
577717 |
src/ctests/subinfo.c...: Remove cntr_groups from .cmp_info This information
|
|
Packit |
577717 |
is better exposed by enumeration.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2b4193fd src/utils/event_chooser.c: Cleanup and comment event_chooser code
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7f9fab2b src/ctests/all_native_events.c: Cleanup and add comments to
|
|
Packit |
577717 |
all_native_events.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a245b502 src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/freebsd.c...: Remove profile_ear from .cmp_info The CPU components
|
|
Packit |
577717 |
should handle this internally.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bca07f3c src/papi.c: Add comments to the PAPI_sprofil code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b1e2090c src/papi.c: Minor papi.c cleanups Fix some minor cosmetic things,
|
|
Packit |
577717 |
including a typo in a comment.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8f3aef4a src/ctests/subinfo.c src/papi.h: Remove opcode_match_width field
|
|
Packit |
577717 |
from .cmp_info This should be exposed via enumeration and not by a field in
|
|
Packit |
577717 |
the generic cmp_info structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 047af629 src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/papi.h...: Remove cntr_OPCM_events field from .cmp_info This should be
|
|
Packit |
577717 |
exposed via enumeration and not by a field in the generic cmp_info structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3f1f9e10 src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/papi.h...: Remove cntr_DEAR_events field from .cmp_info This should be
|
|
Packit |
577717 |
exposed via enumeration and not by a field in the generic cmp_info structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 962c642a src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/papi.h...: Remove cntr_IEAR_events field from .cmp_info This should be
|
|
Packit |
577717 |
exposed via enumeration and not by a field in the generic cmp_info structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5aa7eac1 src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/papi.h...: Remove instr_address_range from .cmp_info This feature should
|
|
Packit |
577717 |
be deteced via enumeration, not via a flag in the generic .cmp_info
|
|
Packit |
577717 |
structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1bf68d5d src/components/nvml/linux-nvml.c src/ctests/subinfo.c
|
|
Packit |
577717 |
src/papi.h...: Remove data_address_range field from .cmp_info The proper way
|
|
Packit |
577717 |
to detect this feature is via enumeration.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-19
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 90037307 src/linux-context.h: Change Linux from using "struct siginfo" to
|
|
Packit |
577717 |
"siginfo_t" This conforms to POSIX, and fixes newer Fedora where struct
|
|
Packit |
577717 |
siginfo is no longer supported. This can in theory break on older setups
|
|
Packit |
577717 |
(possibly kernel 2.4). If that happens, we need to somehow detect this using
|
|
Packit |
577717 |
autoconf.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ad48b4fa src/Rules.perfctr-pfm: Fix the perfctr-pfm build; for buildbot,
|
|
Packit |
577717 |
mostly. Have the perfctr-pfm build only build libpfm, like the perfevents
|
|
Packit |
577717 |
builds. The icc build was choking on warnings (-Werror => errors) in the
|
|
Packit |
577717 |
example programs with libpfm, this is not something we depend upon.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 358b14f9 src/papi_events.csv: Update BGQ presets
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* cf26fc87 src/components/bgpm/CNKunit/linux-CNKunit.c
|
|
Packit |
577717 |
src/components/bgpm/IOunit/linux-IOunit.c
|
|
Packit |
577717 |
src/components/bgpm/L2unit/linux-L2unit.c...: Update bgpm components
|
|
Packit |
577717 |
according to the papi5 changes
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a7b08a91 src/configure src/configure.in src/linux-bgq.c: Merging the BG/Q
|
|
Packit |
577717 |
stuff from stable_4.2 to PAPI 5 did break it. It's corrected now; also
|
|
Packit |
577717 |
predefined events are now working.)
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-15
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2d5a4205 src/components/bgpm/IOunit/linux-IOunit.c
|
|
Packit |
577717 |
src/components/bgpm/L2unit/linux-L2unit.c src/configure...: Merging the BG/Q
|
|
Packit |
577717 |
stuff from stable_4.2 to PAPI 5 did break it. It's corrected now (almost);
|
|
Packit |
577717 |
predefined events are not working yet.)
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1b034920 src/papi.c: Re-enable PAPI_event_name_to_code() optimization In
|
|
Packit |
577717 |
PAPI_event_name_to_code() there was a commented-out optimization where we
|
|
Packit |
577717 |
would check if an event name begins with "PAPI_" before searching the entire
|
|
Packit |
577717 |
preset list for an event name. The comment says we had to disable this due
|
|
Packit |
577717 |
to "user events", but a check shows that was introduced in e7bd768850ecf90
|
|
Packit |
577717 |
and that the "user events" it means is not the current support, but the
|
|
Packit |
577717 |
now-removed PAPI_set_event_info() function where you could change the names
|
|
Packit |
577717 |
of presets on the fly (even to something not starting with PAPI_). Since we
|
|
Packit |
577717 |
don't support that anymore, we can re-enable the optimization.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-14
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9a26b43d src/papi_internal.c src/papi_internal.h src/papi_preset.c: Remove
|
|
Packit |
577717 |
the 16-component limit This turned out to be easier than I thought it would
|
|
Packit |
577717 |
be. Now determining which component an event is in is a two step process.
|
|
Packit |
577717 |
Before, the code shifted and masked to find the component from bits 26-30.
|
|
Packit |
577717 |
Now, _papi_hwi_component_index() is used. There's a new native event table
|
|
Packit |
577717 |
which maps all native events (which are allocated incrementally at first use
|
|
Packit |
577717 |
starting with 0x4000000) to two values, a component number and an "internal"
|
|
Packit |
577717 |
event number.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-13
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d5c50353 src/papi_internal.c: Fix for the PAPI_COMPONENT_MASK change I
|
|
Packit |
577717 |
missed two cases in papi_internal.c This was causing the overflow_allcounters
|
|
Packit |
577717 |
test to fail
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 46fd84ce src/components/bgpm/CNKunit/linux-CNKunit.c
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/linux-CNKunit.h
|
|
Packit |
577717 |
src/components/bgpm/IOunit/linux-IOunit.c...: Updating the Q substrate
|
|
Packit |
577717 |
according to the PAPI 5 changes
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 05a8dcbf src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_list_events.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_values_by_code.c...: First steps of removing
|
|
Packit |
577717 |
16-component limit This change removes PAPI_COMPONENT_INDEX(),
|
|
Packit |
577717 |
PAPI_COMPONENT_MASK and PAPI_COMPONENT_AND_MASK. It adds the new functions
|
|
Packit |
577717 |
_papi_hwi_component_index() _papi_hwi_native_to_eventcode()
|
|
Packit |
577717 |
_papi_hwi_eventcode_to_native() By replacing all of the former macros with
|
|
Packit |
577717 |
the equivelant of the latter functions, it allows all of the future
|
|
Packit |
577717 |
16-component limit changes to be made in the functions. Components now
|
|
Packit |
577717 |
receive as events a plain 32-bit value as their internal native event; the
|
|
Packit |
577717 |
high bits are not set. This may break some external components. This change
|
|
Packit |
577717 |
should not break things, but a lot of testing is needed.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* af4cbb86 src/run_tests_exclude.txt: Exclude iozone helper scripts from
|
|
Packit |
577717 |
run_tests. run_tests.sh looks for executible files under components/*/tests
|
|
Packit |
577717 |
Some of the plotting scripts in appio/iozone were getting picked up.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-12
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c10c7ccb src/configure src/configure.in: Configure does not work on BGQ due
|
|
Packit |
577717 |
to missing subcomp feature. It worked for stable-4.2 but got lost in current
|
|
Packit |
577717 |
git origin.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d9a58148 src/aix.c src/ctests/hwinfo.c src/ctests/overflow.c...: Update
|
|
Packit |
577717 |
hw_info_t CPU frequency reporting. Previously PAPI reported "float mhz" and
|
|
Packit |
577717 |
"int clock_mhz". In theory the first was the current CPU speed, and the
|
|
Packit |
577717 |
latter was the resolution of the cycle counter. In practice they were both
|
|
Packit |
577717 |
set to the same value (on Linux, read from /proc/cpuinfo) and not very useful
|
|
Packit |
577717 |
when DVFS was enabled, as the value reported was usually lower than the
|
|
Packit |
577717 |
actual frequency running once CPU started being used. This change adds two
|
|
Packit |
577717 |
new values "cpu_max_mhz" and "cpu_min_mhz" which are read from
|
|
Packit |
577717 |
/sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq and
|
|
Packit |
577717 |
/sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_min_freq if they are available,
|
|
Packit |
577717 |
and falls back to /proc/cpuinfo otherwise. All of the tests were updated to
|
|
Packit |
577717 |
use cpu_max_mhz. The old mhz and clock_mhz values are left for compatability
|
|
Packit |
577717 |
reasons (and set to cpu_max_mhz) but are currently unused otherwise.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-11
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0f124891 src/papi_events.csv: Initial PAPI Ivy Bridge Support for now try
|
|
Packit |
577717 |
to re-use the sandy bridge event presets
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a1f46077 src/libpfm4/docs/man3/libpfm_intel_ivb.3
|
|
Packit |
577717 |
src/libpfm4/include/perfmon/err.h
|
|
Packit |
577717 |
src/libpfm4/lib/events/intel_ivb_events.h...: Import libpfm4 git snapshot
|
|
Packit |
577717 |
This adds IvyBridge support
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3bb983cc src/libpfm-3.y/examples_v2.x/self_smpl_multi.c: Fix a libpfm3
|
|
Packit |
577717 |
example program for icc, local fix because libpfm3 is deprecated. icc does
|
|
Packit |
577717 |
have more enjoyable warnings than gcc, : error 186: pointless comparison of
|
|
Packit |
577717 |
unsigned integer with zero on this: unsigned int foo; ... if ( foo < 0 )
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d28adccf src/papi_user_events.c: The user events code had a call to exit,
|
|
Packit |
577717 |
this was bad...
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-04
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6bf43022 src/testlib/test_utils.c: Further the hack for testing for perf SW
|
|
Packit |
577717 |
events. Events like
|
|
Packit |
577717 |
|
|
Packit |
577717 |
- | perf::CPU-CLOCK
|
|
Packit |
577717 |
| | PERF_COUNT_SW_CPU_CLOCK
|
|
Packit |
577717 |
|
|
|
Packit |
577717 |
|
|
Packit |
577717 |
- were passing the check, now we also check the event_info_struct.long_descr
|
|
Packit |
577717 |
field for PERF_COUNT_SW....
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fa4b1a28 src/components/nvml/linux-nvml.c: Cleanup nvml code a little. A
|
|
Packit |
577717 |
few print statements were left over from debugging. Also check errors from
|
|
Packit |
577717 |
nvml and cuda pciinfo functions, disabling the component in a few more cases.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-06-01
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* da144a94 src/components/nvml/Makefile.nvml.in src/components/nvml/README
|
|
Packit |
577717 |
src/components/nvml/Rules.nvml...: Rewrite and merge of the nVidia Management
|
|
Packit |
577717 |
library component. This component attempts to expose all supported
|
|
Packit |
577717 |
'performance counters' on each card cuda knows about at runtime. Much like
|
|
Packit |
577717 |
the cuda component reads happen on the card you're executing on at PAPI_read
|
|
Packit |
577717 |
time. The test included is a copy of the cuda helloworld test, but it
|
|
Packit |
577717 |
attempts to start/stop the event on each gpgpu. If you select an event that
|
|
Packit |
577717 |
is not supported on the card you're running on we should fail gracefully but
|
|
Packit |
577717 |
this has not been tested.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b2d414dc src/components/stealtime/linux-stealtime.c: At units to stealtime
|
|
Packit |
577717 |
component Added the function but forgot to add a function vector for it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ce9d4500 src/components/stealtime/linux-stealtime.c: Add units to stealtime
|
|
Packit |
577717 |
Properly report that the units are in micro seconds.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 149948c8 src/components/rapl/linux-rapl.c: Minor cleanup of RAPL code
|
|
Packit |
577717 |
missing "void" paramter in init_substrate function
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6a7e22fa src/components/vmware/vmware.c: More vmware component fixes. This
|
|
Packit |
577717 |
makes the component thread-safe. Also makes it fail more gracefully if the
|
|
Packit |
577717 |
guestlib SDK is installed but does not support our hypervisor (for example,
|
|
Packit |
577717 |
if we are running under VM Workstation). Still need to test on ESX.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 072d6473 src/components/appio/tests/appio_test_select.c: added code to
|
|
Packit |
577717 |
intercept and time select() calls.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-22
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 12b6d0d7 src/components/vmware/vmware.c: Some more minor fixes to VMware
|
|
Packit |
577717 |
component Try to handle things better if VMguest SDK not working
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6e015bc5 src/components/vmware/Rules.vmware src/components/vmware/vmware.c:
|
|
Packit |
577717 |
More vmware component fixups Now works with the events from the VMguest SDK
|
|
Packit |
577717 |
library
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5fc0f646 src/components/vmware/vmware.c: More cleanup of vmware component
|
|
Packit |
577717 |
The pseudo-performance counters work again. Now they behave in accumulate
|
|
Packit |
577717 |
mode, like all other PAPI counters.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f72b0967 src/components/vmware/tests/vmware_basic.c: Make vmware test a bit
|
|
Packit |
577717 |
more complete
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 070e5481 src/components/vmware/tests/Makefile
|
|
Packit |
577717 |
src/components/vmware/tests/vmware_basic.c: Add a test for the vmware
|
|
Packit |
577717 |
component
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7cf62498 src/components/vmware/Makefile.vmware.in
|
|
Packit |
577717 |
src/components/vmware/Rules.vmware src/components/vmware/configure...: Clean
|
|
Packit |
577717 |
up the vmware component. bring it up to date with other components. make it
|
|
Packit |
577717 |
possible to build it without the vmguest library being installed
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b32ae1ae src/components/stealtime/Rules.stealtime
|
|
Packit |
577717 |
src/components/stealtime/linux-stealtime.c
|
|
Packit |
577717 |
src/components/stealtime/tests/Makefile...: Add a stealtime component When
|
|
Packit |
577717 |
running in a VM, this provides information on how much time was "stolen" by
|
|
Packit |
577717 |
the hypervisor due to the VM being disabled. This info is gathered from
|
|
Packit |
577717 |
column 8 of /proc/stat This currently only works on KVM.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9e95b480 src/components/appio/tests/appio_test_blocking.c: Use a
|
|
Packit |
577717 |
non-blocking select to determine which reads and writes would block
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-19
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f60d991f src/components/appio/README src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_read_write.c...: Interception of
|
|
Packit |
577717 |
close() implemented. This allows us to correctly determine the number of
|
|
Packit |
577717 |
currently open descriptors.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7cd8b5a3 src/libpfm4/.gitignore src/libpfm4/config.mk
|
|
Packit |
577717 |
src/libpfm4/lib/Makefile...: Update libpfm4 to current git tree
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ebffdb7e src/components/rapl/tests/rapl_overflow.c: Skip rapl_overflow test
|
|
Packit |
577717 |
if RAPL not available
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 98d21ef3 src/components/example/example.c src/components/rapl/linux-rapl.c:
|
|
Packit |
577717 |
Fix some component warnings.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0447f373 src/configure src/configure.in src/linux-generic.h: Make build not
|
|
Packit |
577717 |
stall if PAPI_EVENTS_CSV not set This is some fallout from the FreeBSD
|
|
Packit |
577717 |
changes. PAPI_EVENTS_CSV could not be set, which would make the event
|
|
Packit |
577717 |
creation script hang forever. Also catch various fallthroughs in the code
|
|
Packit |
577717 |
where SUBSTR wasn't being set, which is how the above problem can happen.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ef484c00 src/linux-timer.h: Fix typo in linux-timer.h
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-14
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7c3385f4 src/components/bgpm/CNKunit/CVS/Entries
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/CVS/Repository
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/CVS/Root...: Removed CVS stuff from Q code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2cf4aeb2 src/configure src/configure.in src/linux-bgq.c...: Removed
|
|
Packit |
577717 |
papi_events.csv parsing from Q code. (CVS stuff still needs to be taken care
|
|
Packit |
577717 |
of.)
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-12
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 153c2bb1 INSTALL.txt: Updated INSTALL notes for Q
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ff6a43fb src/Makefile.in src/Makefile.inc src/components/README...: Added
|
|
Packit |
577717 |
missing files for Q merge. Conflicts: src/configure src/configure.in
|
|
Packit |
577717 |
src/freq.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-12
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0e142630 src/Rules.bgpm src/components/bgpm/CNKunit/CVS/Entries
|
|
Packit |
577717 |
src/components/bgpm/CNKunit/CVS/Repository...: Added PAPI support for Blue
|
|
Packit |
577717 |
Gene/Q.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-14
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ad7e3fa0 src/components/rapl/linux-rapl.c: Properly accumulate RAPL results
|
|
Packit |
577717 |
Previously it was resetting the counts on read, instead of continuing to
|
|
Packit |
577717 |
count as per other PAPI events.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c79e3018 src/components/rapl/tests/rapl_overflow.c: Fix some warnings in
|
|
Packit |
577717 |
rapl_overflow test
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 731afd1a src/components/rapl/tests/Makefile
|
|
Packit |
577717 |
src/components/rapl/tests/rapl_overflow.c: Add rapl_overflow test This test
|
|
Packit |
577717 |
sees if we can measure RAPL in conjunction with overflow CPU performance
|
|
Packit |
577717 |
events.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b0e201bb src/components/rapl/utils/Makefile
|
|
Packit |
577717 |
src/components/rapl/utils/rapl_plot.c: Remove derived "uncore" values from
|
|
Packit |
577717 |
rapl tool They weren't really measuring uncore, but were just TOTAL - PP0.
|
|
Packit |
577717 |
It was causing some confusion.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-09
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 547e9379 doc/Doxyfile-common papi.spec src/Makefile.in...: Bump the version
|
|
Packit |
577717 |
number to 4.9.0.0 Read 4.9 as pre-5.0 master was at version number 4.2.1,
|
|
Packit |
577717 |
this was archaic... Sorry for the confusion Tushar, master is the correct
|
|
Packit |
577717 |
branch for the latest development code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 133e3d67 src/configure src/configure.in: Fix perfctr build In the FreeBSD
|
|
Packit |
577717 |
changes I removed the CPU determination by reading /proc/cpuinfo as that was
|
|
Packit |
577717 |
prone to failure and non-portable. This broke perfctr as it was doing a huge
|
|
Packit |
577717 |
CPU name lookup to determine if it was on an x86 system or not. This change
|
|
Packit |
577717 |
fixes that.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-08
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 42b21d67 src/papi_libpfm4_events.c: Fix PAPI event enumeration inside of
|
|
Packit |
577717 |
VMware VMware disables the availability of architectural events when
|
|
Packit |
577717 |
virtualized PMU is not available. libpfm4 was checkign this when enumerating
|
|
Packit |
577717 |
events, and we would end up in the situation where ix86arch was marked active
|
|
Packit |
577717 |
but 0 events were available. We didn't check for this error condition and
|
|
Packit |
577717 |
thus end up thoroughly confused.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-07
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fd79a584 src/freebsd.c: Fix event enumeration on FreeBSD It was passing
|
|
Packit |
577717 |
PAPI_OK in all cases, causing papi_native_avail to try to do things like
|
|
Packit |
577717 |
report groups even when groups weren't available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 53732c2e src/freebsd.c: Add Virtual Machine detection support to FreeBSD
|
|
Packit |
577717 |
again, support for this on x86 is OS Neutral
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7b4d7c96 src/configure src/configure.in src/freebsd-memory.c...: Add
|
|
Packit |
577717 |
x86_cacheinfo support to FreeBSD The x86 cache and memory info is
|
|
Packit |
577717 |
OS-independent, so add support for it to FreeBSD.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 91033df6 src/Makefile.in src/Makefile.inc src/configure...: Re-enable
|
|
Packit |
577717 |
predefined events on FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 36f6dc1b src/freebsd.c src/freebsd/map.c src/freebsd/map.h: Modify FreeBSD
|
|
Packit |
577717 |
to use _papi_load_preset_table
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 45651746 src/freebsd.c: Cleanup the freebsd code a bit.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e1554ed8 src/configure: re-run autoconf for updated configure
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1deb2f5d src/Makefile.inc: Make sure a proper dependency for
|
|
Packit |
577717 |
papi_events_table.h exists Our Makefile code that builds a shared library is
|
|
Packit |
577717 |
way broken; it will fail to rebuild in many cases where the static library
|
|
Packit |
577717 |
properly detects thing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 28e28006 src/configure.in: Make papi_events_table.h build normally, not by
|
|
Packit |
577717 |
configure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9a66dfa5 src/configure.in: Another place papi_events_table.sh is called
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 12e4a934 src/Makefile.inc src/papi_events_table.sh: Make
|
|
Packit |
577717 |
papi_events_table.sh take a command line argument This way we can use it on
|
|
Packit |
577717 |
any .csv file, not just papi_events.csv
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7018528f src/freebsd/memory.c: Remove unused freebsd/memory.c file
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 819e5826 src/freebsd_events.csv: Make freebsd_events.csv a valid PAPI event
|
|
Packit |
577717 |
file
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9cc4a468 src/freebsd.c src/freebsd/map-atom.c src/freebsd/map-core.c...:
|
|
Packit |
577717 |
Fix FreeBSD build on head. This temporarily disables preset events. There
|
|
Packit |
577717 |
are also a few other minor fixes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-05-01
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ab36c0a2 src/Makefile.inc src/configure src/configure.in: Update build
|
|
Packit |
577717 |
system for FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2b61d8b7 src/freebsd.c src/freebsd.h: Fix various compiler warnings on
|
|
Packit |
577717 |
FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2c0bcc84 src/freebsd.c: Enable new Westmere events on FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b0499663 src/freebsd/map-i7.c src/freebsd/map-i7.h
|
|
Packit |
577717 |
src/freebsd/map-westmere.c...: Add Westmere event support for FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e54cabc6 src/ctests/inherit.c: Fix the inherit ctest to compile on FreeBSD
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d9dbdd31 src/components/appio/appio.c: - change in appio component
|
|
Packit |
577717 |
(appio.c): removed reference to .ntv_bits_to_info as it doesn't exist in the
|
|
Packit |
577717 |
PAPI component interface.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-27
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5d661b2d src/Rules.pfm src/Rules.pfm_pe: Add the libpfm -Wno-override-init
|
|
Packit |
577717 |
bandaid to the other rules files. In
|
|
Packit |
577717 |
b33331b66137668155c02e52c98a7e389fad402e we test if gcc -Wextra complains
|
|
Packit |
577717 |
about some structure initialization that libpfm does. This was incoperated
|
|
Packit |
577717 |
into Rules.pfm4_pe only. Jim Galarowicz noticed the other Rules files didn't
|
|
Packit |
577717 |
have it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4349b6fd src/Rules.pfm4_pe src/Rules.pfm_pe: Cleanup the perf events Rules
|
|
Packit |
577717 |
files. Steve Kaufmann reported that CONFIG_PFMLIB_OLD_PFMV2 is only used for
|
|
Packit |
577717 |
libpfm3 builds targeting old versions of perfmon2.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-26
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8a7fef68 src/mb.h: Add memory barries for ia64
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-24
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9af4dd4a src/libpfm4/README src/libpfm4/config.mk
|
|
Packit |
577717 |
src/libpfm4/include/perfmon/perf_event.h...: Import libpfm4 git snapshot
|
|
Packit |
577717 |
This brings libpfm4 up to 9ffc45e048661a29c2a8ba4bfede78d3feb828f4 The
|
|
Packit |
577717 |
important change is support for Intel Atom Cedarview.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fac6aec0 src/linux-bgp-memory.c src/linux-bgp.c: Some BG/P cleanups.
|
|
Packit |
577717 |
Removed a lot of dead code, noticed when looking for any potential BG/P
|
|
Packit |
577717 |
issues.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 977709f6 src/linux-bgp-preset-events.c src/linux-bgp.c: Fix PAPI compile on
|
|
Packit |
577717 |
BG/P Thanks to Harald Servat <harald.servat@bsc.es>
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-19
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5207799e release_procedure.txt: Modified release_procedure.txt to push
|
|
Packit |
577717 |
tags.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b248ae80 doc/Makefile: Have clean remove the doxygen error file.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1d4f75a3 doc/Doxyfile-man1 doc/Doxyfile-man3: Fix an error in the Doxygen
|
|
Packit |
577717 |
config files. Doxygen includes things with @INCLUDE not @include. The html
|
|
Packit |
577717 |
file had this, the man page files did not...
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 979cda20 cvs2cl.pl delete_before_release.sh gitlog2changelog.py...: Update
|
|
Packit |
577717 |
the release machinery for git. gitlog2changelog.py takes the output of git
|
|
Packit |
577717 |
log and parses it to something like a changelog.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 67bdd45f doc/Doxyfile-html: Cover up an instance of doxygen using full
|
|
Packit |
577717 |
paths. Doxygen ( up to 1.8.0, the most recent at this writing ) would use
|
|
Packit |
577717 |
full paths in directory dependencies ignoring the use relative paths config
|
|
Packit |
577717 |
option.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-13
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c38eb0b7 src/libpfm-3.y/lib/intel_corei7_events.h
|
|
Packit |
577717 |
src/libpfm-3.y/lib/intel_wsm_events.h src/libpfm-3.y/lib/pfmlib_intel_nhm.c:
|
|
Packit |
577717 |
Add missing update to libpfm3 Somehow during all of the troubles we had with
|
|
Packit |
577717 |
importing libpfm3 into CVS, we lost some Nehalem/Westmere updates. Tested on
|
|
Packit |
577717 |
a Nehalem machine to make sure this doesn't break anything.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 193d8d06 src/papi_libpfm3_events.c: Fix max_multiplex case on
|
|
Packit |
577717 |
perf_event/libpfm3 num_mpx_cntrs was being set to 512 even though the real
|
|
Packit |
577717 |
maximum is 32, causing a buffer overflow and segfault.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-12
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f1f7fb5b src/threads.h: Fix minor typo in a comment
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0373957d src/linux-timer.c: Fix potential fd leak Noticed by coverity
|
|
Packit |
577717 |
checker.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 71727e38 src/ctests/max_multiplex.c: Improve max_multiplex ctest on
|
|
Packit |
577717 |
perfmon2, this test was failing because the maximum number of multiplexed
|
|
Packit |
577717 |
counters was much more than the available counters we could test with. This
|
|
Packit |
577717 |
change modifies the test to not fail in this case.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-11
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* fdbdac9f src/perfmon.c: Fix the perfmon substrate. It was missing a
|
|
Packit |
577717 |
_papi_libpfm_init() call, which meant the number of events was being left at
|
|
Packit |
577717 |
0.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-09
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2a44df97 src/libpfm-3.y/examples_v2.x/multiplex.c
|
|
Packit |
577717 |
src/libpfm-3.y/examples_v2.x/pfmsetup.c
|
|
Packit |
577717 |
src/libpfm-3.y/examples_v2.x/rtop.c...: Catch a few libpfm-3.y files up to
|
|
Packit |
577717 |
libpfm-3.10. More skeletons keep falling out of the cvs closet. This is just
|
|
Packit |
577717 |
what diff -q -r catches.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-04
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0e05da68 src/components/rapl/utils/Makefile
|
|
Packit |
577717 |
src/components/rapl/utils/README src/components/rapl/utils/rapl_plot.c: Add
|
|
Packit |
577717 |
the rapl_plot utility to the RAPL component. This utility uses PAPI to
|
|
Packit |
577717 |
periodicly poll the RAPL counters and generate average power results suitable
|
|
Packit |
577717 |
for plotting. There's been a lot of interest in this utility so it's
|
|
Packit |
577717 |
probably useful to include it with the RAPL component.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2daa03ac src/papi_internal.c: Check if a component is disabled at init
|
|
Packit |
577717 |
time. This change modifies the code so that at PAPI_library_init() time we
|
|
Packit |
577717 |
check the component disable field, and we don't call the init routines for
|
|
Packit |
577717 |
components the user has disabled. This allows code like the following to
|
|
Packit |
577717 |
happen _before_ PAPI_library_init(): numcmp = PAPI_num_components();
|
|
Packit |
577717 |
for(cid=0; cid
|
|
Packit |
577717 |
(strstr(cmpinfo->name,"cuda")) { cmpinfo->disabled=1;
|
|
Packit |
577717 |
strncpy(cmpinfo->disabled_reason,"Disabled by user",PAPI_MAX_STR_LEN); } }
|
|
Packit |
577717 |
We might want to add a specific PAPI_disable_component(int cid) call of maybe
|
|
Packit |
577717 |
even a PAPI_disable_component(char *name) as the above code causes compiler
|
|
Packit |
577717 |
warnings since cmpinfo is returned as a const pointer. This all works
|
|
Packit |
577717 |
because currently PAPI currently statically allocates all of the components
|
|
Packit |
577717 |
at compile time, so we can view and modify the cmp_info structure before
|
|
Packit |
577717 |
PAPI_library_init() is called.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3fd2b21e src/components/appio/README src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/appio.h...: Added support to count reads that are
|
|
Packit |
577717 |
interrupted or would block.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-04-03
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* dd3a192f release_procedure.txt: Change chmod flags for doxygen stuff from
|
|
Packit |
577717 |
755 to 775 to allow group write permissions.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-30
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* deac54cc src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/coretemp/tests/coretemp_basic.c
|
|
Packit |
577717 |
src/components/coretemp/tests/coretemp_pretty.c...: Add new
|
|
Packit |
577717 |
PAPI_enum_cmp_event() function This will be needed when we remove the
|
|
Packit |
577717 |
16-component limit. Currently in PAPI_enum_event() the component number is
|
|
Packit |
577717 |
gathered from bits 29-26 of the eventcode. This won't work anymore once we
|
|
Packit |
577717 |
remove those bits. Also update the various components to not use
|
|
Packit |
577717 |
PAPI_COMPONENT_MASK() as this too will go away in the transition.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 48331cc9 src/configure src/configure.in src/papi.c...: Place all
|
|
Packit |
577717 |
compiled-in components in the _papi_hwd[] array. Previously we had separate
|
|
Packit |
577717 |
compiled_in[] and _papi_hwd[] arrays. At init time a pointer to the
|
|
Packit |
577717 |
compiled_in[] was copied to _papi_hwd[] if initialization passed. This kind
|
|
Packit |
577717 |
of code setup makes enumerating components hard, and finding info from
|
|
Packit |
577717 |
non-available components would require additional function entry points.
|
|
Packit |
577717 |
This change leaves all compiled in components to _papi_hwd[]. Availability of
|
|
Packit |
577717 |
the component can be checked with the new "disabled" field. This will make
|
|
Packit |
577717 |
enumeration support a lot easier to add. It can possibly cause user confusion
|
|
Packit |
577717 |
if they try to access component structures directly without checking the
|
|
Packit |
577717 |
"disabled" field first. This change should also make any eventual support
|
|
Packit |
577717 |
for run-time component enabling/disabling a lot easier.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 66a72f44 src/papi.c: Documentation was referring to nonexistent
|
|
Packit |
577717 |
"PAPI_enum_events()" The actual function we have is PAPI_enum_event()
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0f2c2593 src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/lustre/linux-lustre.c src/components/mx/linux-mx.c...: Add
|
|
Packit |
577717 |
support for reporting reason for failed component initialization. This
|
|
Packit |
577717 |
change adds the fields "disabled" and "disabled_reason" to the
|
|
Packit |
577717 |
component_info_t structure. At initialization time, PAPI will set the
|
|
Packit |
577717 |
"disabled" field to the value returned by component init (that is PAPI_OK if
|
|
Packit |
577717 |
OK or an error otherwise). This can be checked later to find why component
|
|
Packit |
577717 |
init failed. Also provided is the "disabled_reason" string. The components
|
|
Packit |
577717 |
can set this at failure time, and this can be printed later. For example,
|
|
Packit |
577717 |
this is sample output of the updated papi_component_avail routine:
|
|
Packit |
577717 |
|
|
Packit |
577717 |
- Compiled-in components: Name: perf_events.c Linux perf_event
|
|
Packit |
577717 |
CPU counters Name: linux-rapl Linux SandyBridge RAPL energy
|
|
Packit |
577717 |
measurements \-> Disabled: Not a SandyBridge processor Name: example.c
|
|
Packit |
577717 |
A simple example component Name: linux-coretemp Linux
|
|
Packit |
577717 |
hwmon temperature and other info \-> Disabled: No coretemp events found Name:
|
|
Packit |
577717 |
linux-net.c Linux network driver statistics Name: linux-mx.c
|
|
Packit |
577717 |
Myricom MX (Myrinet Express) statistics \-> Disabled: No MX
|
|
Packit |
577717 |
utilities found Name: linux-lustre.c Lustre filesystem statistics
|
|
Packit |
577717 |
\-> Disabled: No lustre filesystems found Active components: Name:
|
|
Packit |
577717 |
perf_events.c Linux perf_event CPU counters Name: example.c
|
|
Packit |
577717 |
A simple example component Name: linux-net.c Linux
|
|
Packit |
577717 |
network driver statistics
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-29
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d84b144e src/components/rapl/Rules.rapl src/components/rapl/linux-rapl.c
|
|
Packit |
577717 |
src/components/rapl/tests/Makefile...: Add a SandyBridge RAPL (Running
|
|
Packit |
577717 |
Average Power Level) Component This component allows energy measurement at
|
|
Packit |
577717 |
the package-level on Sandybridge machines. To run, you need the Linux
|
|
Packit |
577717 |
x86-msr kernel module installed and read permissions to /dev/cpu/*/msr The
|
|
Packit |
577717 |
output from the rapl_busy test looks like this on a SandyBridge-EP machine:
|
|
Packit |
577717 |
Trying all RAPL events Found rapl component at cid 2 Starting
|
|
Packit |
577717 |
measurements... Doing a naive 1024x1024 MMM... Matrix multiply sum:
|
|
Packit |
577717 |
s=1016404871450364.375000 Stopping measurements, took 3.979s, gathering
|
|
Packit |
577717 |
results... Energy measurements: PACKAGE_ENERGY:PACKAGE0 175.786011J (Average
|
|
Packit |
577717 |
Power 44.2W) PACKAGE_ENERGY:PACKAGE1 73.451096J (Average Power 18.5W)
|
|
Packit |
577717 |
DRAM_ENERGY:PACKAGE0 11.663467J (Average Power 2.9W) DRAM_ENERGY:PACKAGE1
|
|
Packit |
577717 |
8.055389J (Average Power 2.0W) PP0_ENERGY:PACKAGE0 119.215500J (Average
|
|
Packit |
577717 |
Power 30.0W) PP0_ENERGY:PACKAGE1 16.315216J (Average Power 4.1W) Fixed
|
|
Packit |
577717 |
values: THERMAL_SPEC:PACKAGE0 135.000W THERMAL_SPEC:PACKAGE1 135.000W
|
|
Packit |
577717 |
MINIMUM_POWER:PACKAGE0 51.000W MINIMUM_POWER:PACKAGE1 51.000W
|
|
Packit |
577717 |
MAXIMUM_POWER:PACKAGE0 215.000W MAXIMUM_POWER:PACKAGE1 215.000W
|
|
Packit |
577717 |
MAXIMUM_TIME_WINDOW:PACKAGE0 0.046s MAXIMUM_TIME_WINDOW:PACKAGE1 0.046s
|
|
Packit |
577717 |
rapl_basic.c PASSED
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-26
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b44d60ca src/components/appio/appio.c src/components/appio/appio.h
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_read_write.c: Added support for
|
|
Packit |
577717 |
intercepting open calls.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9e9fac4b src/Makefile.in src/Rules.pfm4_pe src/configure...: Fix the test
|
|
Packit |
577717 |
case in configure at 0cea1848 Make use of the structure we're using for the
|
|
Packit |
577717 |
override-init test case.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0cea1848 src/configure src/configure.in: Doctor CFLAGS when testing for a
|
|
Packit |
577717 |
gcc warning. -Wextra was not in CFLAGS when I attempted to check for the
|
|
Packit |
577717 |
initialized field overwritten warning. So we set -Wall -Wextra -Werror when
|
|
Packit |
577717 |
running the test code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-22
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b33331b6 src/Makefile.in src/Rules.pfm4_pe src/configure...: Fix
|
|
Packit |
577717 |
initialized field overwritten warning when building libpfm4 on some gcc
|
|
Packit |
577717 |
versions. In gcc 4.2 or so, -Woverride-init was added to -Wextra causing
|
|
Packit |
577717 |
issues with code like struct foo { int a; int b;}; struct foo bar = { .a=0,
|
|
Packit |
577717 |
.b=0, .b=5; }; --Wno-override-init allows us to keep -Werror for libpfm4
|
|
Packit |
577717 |
compiles.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-21
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ae149766 src/papi_internal.h: Delete an old comment. Yes, Dan in 2003, we
|
|
Packit |
577717 |
should and do use MAX_COUNTER_TERMS as the size of the event position array.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b937cdd8 src/papi_user_events.c: Move the user events code over to using
|
|
Packit |
577717 |
the new preset event data structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-14
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6ca599e2 src/papi_internal.c: Fix a small memory leak. We weren't freeing
|
|
Packit |
577717 |
_papi_hwd, causing a lot of MEM_LEAK warnings in buildbot.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-13
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 473b8203 src/aix.h src/configure src/configure.in...: Remove last MY_VECTOR
|
|
Packit |
577717 |
usage. Have configure explicitly set the name of the perf counter substrate
|
|
Packit |
577717 |
vector in the components_config.h file This removes one more special case,
|
|
Packit |
577717 |
and gets us slightly closer to being able to have multiple CPU substrates
|
|
Packit |
577717 |
compiled in at once.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 360c3003 src/papi.c src/papi_libpfm3_events.c src/papi_libpfm_events.h...:
|
|
Packit |
577717 |
Clean up the papi_libpfm3_events.c code. Move code that was perfctr specific
|
|
Packit |
577717 |
into perfctr-x86.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 03de65e3 src/libpfm-3.y/examples_v2.x/multiplex.c
|
|
Packit |
577717 |
src/libpfm-3.y/examples_v2.x/pfmsetup.c
|
|
Packit |
577717 |
src/libpfm-3.y/examples_v2.x/rtop.c...: Fix some libpfm3 warnings. libpfm3
|
|
Packit |
577717 |
is not maintained anymore, so applied these changes locally. libpfm3 is
|
|
Packit |
577717 |
compiled with -Werror so they broke the build with newer gcc even though they
|
|
Packit |
577717 |
are just warnings in example programs.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ad490353 src/ctests/zero_named.c src/utils/multiplex_cost.c: Fix a few
|
|
Packit |
577717 |
compiler warnings in the tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a0fec783 src/linux-timer.c: Fix another linux-timer.c compile problem. I
|
|
Packit |
577717 |
hadn't tested with debug enabled, so all of buildbot failed last night.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-12
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a3733ecd src/linux-timer.h: Fix typo in the linux-timer.h header
|
|
Packit |
577717 |
_linux_get_virt_usec_timess should have been _linux_get_virt_usec_times
|
|
Packit |
577717 |
Thanks to Steve Kaufmann for noticing this.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 785db5ae src/linux-common.c src/linux-timer.c: Fix timer compile on Power
|
|
Packit |
577717 |
machines Power, ARM, and MIPS have no get_cycles() call so provide a dummy
|
|
Packit |
577717 |
function on these architectures.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 708090ee src/linux-common.c src/linux-timer.h: Another fix for non-POSIX
|
|
Packit |
577717 |
timers The recent changes had the name of the fallback usec method wrong.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 88e8d355 src/papi_libpfm3_events.c: Fix a warning in the libpfm3 code.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8ca63705 src/configure src/configure.in src/linux-common.c...: Fix build
|
|
Packit |
577717 |
when not using POSIX timers The PAPI build system was being overly clever
|
|
Packit |
577717 |
with how it defined what kind of wall clock timers were to be used, so of
|
|
Packit |
577717 |
course I broke things when breaking the timer code out to make it a bit more
|
|
Packit |
577717 |
understandable. This patch breaks out the timer define into two pieces; one
|
|
Packit |
577717 |
saying it's a POSIX timer and one saying whether to use HR timers or not.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-09
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b69ad727 src/linux-common.c src/linux-timer.c src/linux-timer.h: Add Linux
|
|
Packit |
577717 |
posix gettime() nanosecond functions
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* af2c9a49 src/papi.c src/papi_vector.c src/papi_vector.h: Add
|
|
Packit |
577717 |
->get_virt_nsec() and ->get_real_nsec() OS vectors Currently PAPI was just
|
|
Packit |
577717 |
cheating and running the usec functions and multiplying by 1000. Make this
|
|
Packit |
577717 |
the default, but allow the OS code to override if they have timers capable of
|
|
Packit |
577717 |
returning nsec percision.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 24c68dbe src/aix.c src/freebsd.c src/linux-bgp.c...: Clean up
|
|
Packit |
577717 |
->get_virt_usec() It no longer needs to be passed a context, so remove that
|
|
Packit |
577717 |
from all callers. Also, ->get_virt_cycles() was just a get_virt_usec()*MHz
|
|
Packit |
577717 |
on most platforms. While this is a bit dubious (especially as MHz can't be
|
|
Packit |
577717 |
relied on) make this a common routine that will be added at innoculate time
|
|
Packit |
577717 |
if ->get_virt_cycles() is set to NULL.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a3ef7cef src/linux-common.c src/linux-timer.c src/linux-timer.h: Cleanup
|
|
Packit |
577717 |
the Linux timer code. Split things up a bit to make the code more readable.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 50ce8ea0 src/papi_internal.c: Change a strcpy() to strncpy() just to be a
|
|
Packit |
577717 |
bit safer.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0526b125 src/components/lmsensors/linux-lmsensors.c: Fix buffer overrun in
|
|
Packit |
577717 |
lmsensors component
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b088db70 src/libpfm4/config.mk
|
|
Packit |
577717 |
src/libpfm4/docs/man3/pfm_get_os_event_encoding.3
|
|
Packit |
577717 |
src/libpfm4/examples/showevtinfo.c...: Update to current git libpfm4 snapshot
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ccb45f61 src/aix.c src/extras.c: Fix segfault on AIX During some of the
|
|
Packit |
577717 |
cleanups, the extras.h header was not added to aix.c This made some of the
|
|
Packit |
577717 |
functions (silently) use default data types for the function parameters,
|
|
Packit |
577717 |
leading to segfaults in some of the tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-08
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1cb22d0b src/components/coretemp/linux-coretemp.c src/utils/native_avail.c:
|
|
Packit |
577717 |
Make "native_avail -d" report units if available Add units support to the
|
|
Packit |
577717 |
coretemp component, have native_avail -d (detailed mode) print it to make
|
|
Packit |
577717 |
sure it works.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9c54840e src/extras.c src/extras.h src/papi_internal.c...: Add new
|
|
Packit |
577717 |
ntv_code_to_info vector This will allow components to return the extended
|
|
Packit |
577717 |
event_info data for native events. If a component doesn't implement
|
|
Packit |
577717 |
ntv_code_to_info then get_event_info falls back to the old way of just
|
|
Packit |
577717 |
reporting symbol name and long description.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c4579559 src/papi.h: Add new event_info fields New fields are added to
|
|
Packit |
577717 |
event_info that allow passing on extended information. This includes things
|
|
Packit |
577717 |
such as measurement units, data type, location, timescope, etc.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 17533e4e src/ctests/all_events.c src/ctests/derived.c
|
|
Packit |
577717 |
src/ctests/kufrin.c...: Restore fields to event_info structure The changes
|
|
Packit |
577717 |
made were probably too ambitious, even for a 5.0 release. In the end it
|
|
Packit |
577717 |
looks like we can remain API compatible while just using up a little more
|
|
Packit |
577717 |
memory. We can still save space by shrinking preset_t behind the scenes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6f13a5f6 src/aix.c src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/coretemp_freebsd/coretemp_freebsd.c...: Remove
|
|
Packit |
577717 |
->ntv_bits_to_info vector from component interface We weren't using it
|
|
Packit |
577717 |
anymore, and many of the components were just setting it to NULL
|
|
Packit |
577717 |
unncessarily. We'll be replacing the functionality soon with
|
|
Packit |
577717 |
ntv_code_to_info
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 401f37bc src/components/example/example.c src/ctests/subinfo.c src/papi.h:
|
|
Packit |
577717 |
Remove invert and edge_detect fields from component info These fields were
|
|
Packit |
577717 |
there to indicate if a CPU component supported these attributes (for Intel
|
|
Packit |
577717 |
processors) but in the end we never used these. The proper way to export
|
|
Packit |
577717 |
this info is during event enumeration.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f32fe481 src/papi_events.csv: We had the PAPI_VEC_INS preset wrong on amd
|
|
Packit |
577717 |
fam12h llano
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 38a8d8a7 src/ctests/multiplex2.c src/papi_preset.c: Fix preset adding code
|
|
Packit |
577717 |
to be more robust. If an invalid event is in a preset definition, we'd
|
|
Packit |
577717 |
currently add it with an eventcode of 0 to the preset, which would break if
|
|
Packit |
577717 |
you tried to use the event. This change properly prints a warning in this
|
|
Packit |
577717 |
case, and sets the preset to be unavailable.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2591a546 src/ctests/val_omp.c src/ctests/zero_omp.c: Remove the hw_info
|
|
Packit |
577717 |
field from add_two_events calls. Two ctests missed the bus when Vince
|
|
Packit |
577717 |
reworked the add_two_events call.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 358a2e32 src/papi_internal.c src/papi_preset.c: Fix segfault seen on an AMD
|
|
Packit |
577717 |
fusion machine With the recent preset and component hanges, we were not
|
|
Packit |
577717 |
properly resetting papi_num_components if PAPI_library_init()/PAPI_shutdown()
|
|
Packit |
577717 |
was called multiple times.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-07
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7751f5d8 src/ftests/zeronamed.F: Fix a compile error on aix. Dan ran over
|
|
Packit |
577717 |
72 characters on a single line. xlf actually enforced that part the Fortran
|
|
Packit |
577717 |
spec.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-06
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1c87d89c src/ftests/Makefile src/ftests/zeronamed.F src/papi_fwrappers.c:
|
|
Packit |
577717 |
Add support for {add, remove, query}_named to Fortran interface; add zero
|
|
Packit |
577717 |
named.F test case; modify ftests Makefile to support "all" tag.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 71bd4fdd src/configure src/configure.in: Modify configure to define the
|
|
Packit |
577717 |
default FTEST_TARGETS as "all"
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 54e39855 src/components/vmware/vmware.c: Changed tri8ggering environment
|
|
Packit |
577717 |
variable to PAPI_VMWARE_PSEUDOPERFORMANCE per Vince's earlier email. This
|
|
Packit |
577717 |
should complete all the VMware component changes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-05
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 845503fb src/Makefile.inc: Add missing MISCSRCS line to Makefile.inc This
|
|
Packit |
577717 |
was breaking the shared library build
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-01
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 11be8e4b .../appio/tests/appio_test_fread_fwrite.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_pthreads.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_test_read_write.c: updated these tests to
|
|
Packit |
577717 |
print timing information
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9ad62ab1 src/components/appio/README src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/appio.h...: Added support for timing I/O calls. Updated
|
|
Packit |
577717 |
tests and README.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-01-31
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* beaa5ff0 src/components/appio/tests/iozone/Changes.txt
|
|
Packit |
577717 |
src/components/appio/tests/iozone/Generate_Graphs
|
|
Packit |
577717 |
src/components/appio/tests/iozone/Gnuplot.txt...: added the latest stable
|
|
Packit |
577717 |
iozone to the appio tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 4af58174 src/components/appio/README src/components/appio/tests/Makefile
|
|
Packit |
577717 |
src/components/appio/tests/init_fini.c: added a hook to run the appio test
|
|
Packit |
577717 |
for iozone.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-01-21
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 15c733cf src/components/appio/CHANGES src/components/appio/README
|
|
Packit |
577717 |
src/components/appio/appio.c...: Removed stray 'net' references. All
|
|
Packit |
577717 |
remaining references are only for the purpose of giving credit. Updated
|
|
Packit |
577717 |
change log.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-01-20
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ca4b6785 src/components/appio/README src/components/appio/appio.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_list_events.c...: - general cleanup -
|
|
Packit |
577717 |
improved tests to be quiet and be conform to other PAPI tests - replaced
|
|
Packit |
577717 |
hardwire constants in appio.c with symbolic ones - tests will now write to
|
|
Packit |
577717 |
/dev/null to avoid filling the terminal screen with useless text - more
|
|
Packit |
577717 |
comments added - @author added to files - updated README
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-01-18
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bb22ed9f src/components/appio/README src/components/appio/Rules.appio
|
|
Packit |
577717 |
src/components/appio/appio.c...: - Added support to measure
|
|
Packit |
577717 |
bytes/calls/eof/short calls for read/write calls. - Interception of
|
|
Packit |
577717 |
read/write and fread/fwrite calls. - Works for static and dynamic linkage
|
|
Packit |
577717 |
(without need for LD_PRELOAD) - Tested OK on 32-bit i686 Linux 2.6.38.
|
|
Packit |
577717 |
Tushar
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2011-12-03
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d58b34b6 src/components/appio/tests/Makefile
|
|
Packit |
577717 |
src/components/appio/tests/appio_list_events.c
|
|
Packit |
577717 |
src/components/appio/tests/appio_values_by_code.c...: *** empty log message
|
|
Packit |
577717 |
***
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* cd7d7acc src/components/appio/tests/appio_values_by_name.c: file
|
|
Packit |
577717 |
appio_values_by_name.c was added on branch appio on 2011-12-03 05:22:06 +0000
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 425e4d09 src/components/appio/tests/appio_values_by_code.c: file
|
|
Packit |
577717 |
appio_values_by_code.c was added on branch appio on 2011-12-03 05:22:06 +0000
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 596ad9bb src/components/appio/tests/appio_list_events.c: file
|
|
Packit |
577717 |
appio_list_events.c was added on branch appio on 2011-12-03 05:22:06 +0000
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 119543dc src/components/appio/tests/Makefile: file Makefile was added on
|
|
Packit |
577717 |
branch appio on 2011-12-03 05:22:06 +0000
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-05
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ba748a41 src/components/vmware/configure: Remove old configuration
|
|
Packit |
577717 |
parameters from vmware/configure
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-02
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2b7e2abb src/ctests/Makefile src/ctests/max_multiplex.c: Add a new
|
|
Packit |
577717 |
max_multiplex test This tries to use the maximum number of multiplexed
|
|
Packit |
577717 |
events. This was written in response to the 32/64 perf_event multiplexed
|
|
Packit |
577717 |
event limit reported by Mohammad j. Ranji
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a0985ff5 src/multiplex.c src/papi_internal.c src/papi_libpfm4_events.c...:
|
|
Packit |
577717 |
Fix issue when using more than 32 multiplexed events on perf_event On
|
|
Packit |
577717 |
perf_event we were setting num_mpx_cntrs to 64. This broke, as the
|
|
Packit |
577717 |
MPX_EventSet struct only allocates room for PAPI_MPX_DEF_DEG events, which is
|
|
Packit |
577717 |
32. This patch makes perf_event use a value of 32 for num_mpx_cntrs,
|
|
Packit |
577717 |
especially as 64 was arbitrarily chosen at some point (the actual value
|
|
Packit |
577717 |
perf_event can support is static, but I'm pretty sure it is higher than 64).
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 331c516c src/ctests/acpi.c: Remove the acpi.c file from ctests It wasn't
|
|
Packit |
577717 |
being built, and we removed the ACPI component a while ago.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 73e7d191 src/components/vmware/vmware.c: Removed all old references to
|
|
Packit |
577717 |
#define VMWARE_PSEUDO_PPERF and switched over to getenv
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-03-01
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 969b8aa9 src/ctests/Makefile src/ctests/zero_named.c src/papi.c: Three new
|
|
Packit |
577717 |
APIs: PAPI_query_named_event PAPI_add_named_event PAPI_remove_named_event and
|
|
Packit |
577717 |
a new test: zero_named Still to do: maybe test named native events and
|
|
Packit |
577717 |
support Fortran
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 97bf9bf8 src/papi.c src/papi.h: First pass implementation of {add, remove,
|
|
Packit |
577717 |
query}_named_event
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2416af88 src/components/vmware/vmware.c: Add functionality to getenv
|
|
Packit |
577717 |
selectors
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 297f9cd6 src/papi.c: Fix possible race in _papi_hwi_gather_all_thrspec_data
|
|
Packit |
577717 |
The valgrind helgrind tool noticed this with the thrspecific test
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* be599976 src/papi_internal.c: Add some locking in
|
|
Packit |
577717 |
_papi_hwi_shutdown_global_internal This caused a glibc double-free warning,
|
|
Packit |
577717 |
and was caught by the Valgrind helgrind tool in krentel_pthreads There are
|
|
Packit |
577717 |
some other potential locking issues in PAPI_shutdown, especially when debug
|
|
Packit |
577717 |
is enabled.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8444d577 src/utils/clockres.c src/utils/command_line.c: Cleanup the oxygen
|
|
Packit |
577717 |
markup for the utilities.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7144394f doc/Doxyfile-html: Missed a recursive tag for the html config
|
|
Packit |
577717 |
file.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 63b2efc4 src/papi_preset.c: Fix segfaults in tests on AMD machines The
|
|
Packit |
577717 |
papi_preset code was wrongly calling papi_free() on some code that was
|
|
Packit |
577717 |
allocated with strdup() (not with papi_malloc). We were only noticing this
|
|
Packit |
577717 |
on AMD machines because it was the code for freeing developer notes in
|
|
Packit |
577717 |
presets, and currently only AMD events have developer notes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0b1350df src/linux-common.c: Touch 'virtual_vendor_name' to cleanup a
|
|
Packit |
577717 |
warning on bluegrass.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-29
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1f17b571 src/Makefile.inc src/Rules.perfctr-pfm src/Rules.pfm4_pe...: Merge
|
|
Packit |
577717 |
the contents of papi_libpfm_presets.c into papi_preset.c The code isn't
|
|
Packit |
577717 |
libpfm specific at all anymore, it's the generic "read presets from a file"
|
|
Packit |
577717 |
code. It makes more sense to find it in papi_presets.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d087d49f src/papi_fwrappers.c: Fix Fortran breakage after the preset event
|
|
Packit |
577717 |
changes
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 156141ec src/papi_libpfm_presets.c src/papi_preset.c src/papi_preset.h:
|
|
Packit |
577717 |
Simplify papi_libpfm_presets.c Previously adding presets from
|
|
Packit |
577717 |
papi_events.csv was a three step process. 1. Load the presets from the
|
|
Packit |
577717 |
file, put in temporary structure. 2. Convert this temporary structure to a
|
|
Packit |
577717 |
"findem" dense structure 3. Pass this dense structure to
|
|
Packit |
577717 |
_papi_hwi_setup_all_presets for final assignment. This change creates the
|
|
Packit |
577717 |
final assignment directly without the intermediate two steps.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8bc2bafd src/papi.c src/papi.h src/papi_common_strings.h...: Make the
|
|
Packit |
577717 |
internal preset_info match the one exported by papi.h There were a lot of
|
|
Packit |
577717 |
cases where the same structure fields were available, just with different
|
|
Packit |
577717 |
names. That was confusing. Also, this allows using a pointer to the preset
|
|
Packit |
577717 |
info instead of having to copy values out of the structure when gathering
|
|
Packit |
577717 |
event info for presets.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8fda68cb src/genpapifdef.c src/papi.c src/papi_common_strings.h...: Merge
|
|
Packit |
577717 |
the 4 separate preset structs into one. _papi_hwi_presets was a structure
|
|
Packit |
577717 |
containing pointers to 4 other arrays of structures which held the actual
|
|
Packit |
577717 |
preset data. This change merges all of these into one big structure.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-28
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e69815d7 src/linux-bgp.c src/papi_internal.c src/papi_internal.h...:
|
|
Packit |
577717 |
Removing remaining vestiges of references to bipartite routines. Now the only
|
|
Packit |
577717 |
references are in papi_bipartite.h, perfctr-x86.c and winpmc-p3.c.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5766b641 src/papi_bipartite.h src/perfctr-x86.c
|
|
Packit |
577717 |
src/win2k/substrate/winpmc-p3.c: These changes implement the bipartite
|
|
Packit |
577717 |
allocation routine as a header file to be included in whatever cpu component
|
|
Packit |
577717 |
needs it. Right now, that's just perfecter-x86 and windows. Both components
|
|
Packit |
577717 |
have been modified and perfecter-86 compiles cleanly. Neither has been tested
|
|
Packit |
577717 |
since I don't have access to a test bed.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7f444b76 src/papi_libpfm_presets.c src/papi_preset.c src/papi_preset.h:
|
|
Packit |
577717 |
Merge the hwi_dev_notes structure into hwi_preset_data
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 21a1d197 src/components/vmware/vmware.c: add getenv
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 08c1b474 src/perfctr-x86.c: Merge bipartite routine into perfecter-x86
|
|
Packit |
577717 |
component, since this is effectively the only place it is used.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9ed9b1f5 src/papi.c: Remove a reference to PAPI_set_event_info() which was
|
|
Packit |
577717 |
removed for PAPI 4
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c626f064 src/ctests/all_events.c src/ctests/derived.c
|
|
Packit |
577717 |
src/ctests/kufrin.c...: Convert PAPI_event_info_t to separate preset event
|
|
Packit |
577717 |
info This moves the preset event info to its own separate structure, which
|
|
Packit |
577717 |
reduces greatly the large string overhead that is not used by the native
|
|
Packit |
577717 |
events.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 787d6822 src/perfctr-x86.c: Move bipartite stuff to perfctr_x86 since
|
|
Packit |
577717 |
that's really the only place it's currently used.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 229c8b41 src/components/vmware/vmware.h: Add env_var definition to vmware.h
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 46aaf6ca src/components/vmware/vmware.c: Remove all unneeded cases
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 874a5718 src/freebsd.c src/perfctr-ppc64.c: Remove more unused references
|
|
Packit |
577717 |
to .bpt_ routines in preparation for refactoring.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 74e5a5fd src/components/vmware/vmware.h: Remove uneeded defines from
|
|
Packit |
577717 |
vmware.h header
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 58b51367 src/components/coretemp_freebsd/coretemp_freebsd.c
|
|
Packit |
577717 |
src/components/vmware/vmware.c src/solaris-niagara2.c...: Remove unused
|
|
Packit |
577717 |
references to .bpt_ routines in preparation for refactoring.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-27
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6b184158 src/Makefile.inc src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/configure...: Have separate concept of "compiled-in" versus "active"
|
|
Packit |
577717 |
components With this change, the _papi_hwd[] component info array only
|
|
Packit |
577717 |
contains a null-terminated list of _active_ components. The
|
|
Packit |
577717 |
_papi_compiled_components[] array has the original full list. At
|
|
Packit |
577717 |
init_substrate[] time a pointer to a component is only put in the _papi_hwd[]
|
|
Packit |
577717 |
list if it is successfully initialized. In addition the
|
|
Packit |
577717 |
PAPI_num_compiled_components() and PAPI_get_compiled_component_info() calls
|
|
Packit |
577717 |
have been added, but this is probably a confusing interface so they might
|
|
Packit |
577717 |
only be temporary additions.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 042bfd5b src/Makefile.inc src/papi.c src/papi_data.c...: Split the contents
|
|
Packit |
577717 |
of papi_data.c to various other files. The data declarations in papi_data.c
|
|
Packit |
577717 |
were mostly used in other files. Move these into more relevant locations.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1877862c src/papiStdEventDefs.h src/papi_common_strings.h: Remove the BGL
|
|
Packit |
577717 |
and BGP specific pre-defined events. They can be better replaced by
|
|
Packit |
577717 |
user-events, and we also had already removed BGL support completely a while
|
|
Packit |
577717 |
back. This removes some ifdefs from the pre-defined event list and keeps
|
|
Packit |
577717 |
future pre-defined events from having different eventcodes on different
|
|
Packit |
577717 |
platforms.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* c3986b79 src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/cuda/linux-cuda.c
|
|
Packit |
577717 |
src/components/infiniband/linux-infiniband.c...: Add names and descriptions
|
|
Packit |
577717 |
for components. Also fixes cuda and lmsensors build issues introduced by
|
|
Packit |
577717 |
vector.h cleanup
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2c84f920 src/aix.c src/freebsd.c src/perf_events.c...: Add names and
|
|
Packit |
577717 |
descriptions to all of the CPU substrates.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9f3e634a src/components/example/example.c src/papi.h src/utils/component.c:
|
|
Packit |
577717 |
Add new "description" and "short_name" fields to .cmp_info structure This
|
|
Packit |
577717 |
description field allows components to provide extra information on what they
|
|
Packit |
577717 |
do. The short_name field will eventually be used to pre-pend event names.
|
|
Packit |
577717 |
The papi_component_avail utility has been updated to print the description.
|
|
Packit |
577717 |
The example component was updated to fill in these values.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ab61c9a7 src/Makefile.inc src/genpapifdef.c src/papi_common_strings.h...:
|
|
Packit |
577717 |
Split papi_data.c into two parts papi_data.c was half data structure
|
|
Packit |
577717 |
definitions for all of PAPI and half string definitions used by both PAPI
|
|
Packit |
577717 |
*and* genpapifdef This splits the common string definitions to
|
|
Packit |
577717 |
papi_common_strings.h so that genpapifdef can still be built w/o linking
|
|
Packit |
577717 |
libpapi.a while making the code a lot easier to follow.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b8e6294c src/solaris-ultra.c: Remove unncessary extern declarations from
|
|
Packit |
577717 |
solaris-ultra.c.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 5ddaff91 src/sys_perf_event_open.c: Remove unncessary extern declarations
|
|
Packit |
577717 |
from sys_perf_event_open.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a6c463b7 doc/Doxyfile-common.config: Create a common config file for
|
|
Packit |
577717 |
doxygen. As part of streamlining the doxygen process, this is a new template
|
|
Packit |
577717 |
doxygen config files. This is a blank template file generated by doxygen
|
|
Packit |
577717 |
1.7.4 (the version currently mandated by the release procedure )
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* dc2c11fa src/aix.c src/aix.h src/perfmon.c...: The vector pre-definition
|
|
Packit |
577717 |
should be in the .c file, not the .h file
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0b3c83c3 src/perf_events.c: Remove unnecessary extern declarations in
|
|
Packit |
577717 |
perf_events.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b93efca0 src/perfmon.c src/perfmon.h: Remove unnecessary extern
|
|
Packit |
577717 |
declarations in perfmon.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7f7a2359 src/papi_preset.c: Remove unnecessary extern declarations from
|
|
Packit |
577717 |
papi_preset.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ecec03ad src/papi_libpfm_presets.c: Remove extraneous extern declarations
|
|
Packit |
577717 |
from papi_libpfm_presets.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7b5f3991 src/extras.c: remove extraneous extern declarations from extras.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f6470e4d src/aix-memory.c src/aix.c src/aix.h: Remove unncessary extern
|
|
Packit |
577717 |
declarations from aix.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f197d4ab src/papi_data.h src/papi_internal.c: Remove unncessary extern
|
|
Packit |
577717 |
declarations in papi_internal.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e7b39d48 src/papi.c src/papi_data.c src/papi_data.h...: remove unnecessary
|
|
Packit |
577717 |
extern definitions from papi.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-24
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 92689f62 src/configure src/configure.in src/linux-common.c...: Add a
|
|
Packit |
577717 |
--with-pthread-mutexes option to enable using pthread mutexes rather than
|
|
Packit |
577717 |
PAPI custom locks This is useful when running on new architectures that
|
|
Packit |
577717 |
don't have a custom PAPI lock coded yet, and also for running valgrind
|
|
Packit |
577717 |
deadlock detection utilities that only work with pthread based locking.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ca51ae67 src/papi_events.csv: Fix broken Pentium 4 Prescott support We
|
|
Packit |
577717 |
were missing the netbusrt_p declaration in papi_events.csv
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f6460736 src/linux-common.c: Fix build on POWER, broken by the
|
|
Packit |
577717 |
virtualization change.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 91d32585 src/perfctr-x86.c src/perfmon.c: Fix some warnings that have
|
|
Packit |
577717 |
appeared due to recent changes.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ae0cf00f src/linux-common.c src/papi_libpfm3_events.c
|
|
Packit |
577717 |
src/papi_libpfm4_events.c...: Clean up the Linux lock files The locking
|
|
Packit |
577717 |
primitives for some reason were spread among the libpfm code and the
|
|
Packit |
577717 |
substrate codes. This change moves them into linux-common and has them part
|
|
Packit |
577717 |
of the OS code. This way they will get properly initialized even if the perf
|
|
Packit |
577717 |
counter or libpfm code isn't being used.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-23
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 88847e52 src/papi.c src/papi_memory.h: Remove _papi_cleanup_all_memory
|
|
Packit |
577717 |
define from papi_memory.h The code in papi_memory.h said: /* define an
|
|
Packit |
577717 |
alternate entry point for compatibility with papi.c for 3.1.x*/ /* this line
|
|
Packit |
577717 |
should be deleted for the papi 4.0 head */ Since we are post papi-4.0 I
|
|
Packit |
577717 |
thought it was time to act on this. Of course papi.c was still using the old
|
|
Packit |
577717 |
name in one place.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1d29dfc6 src/papi_libpfm_presets.c src/perfctr.c src/perfmon.c: Fix some
|
|
Packit |
577717 |
missing includes found after the header cleanup.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* b425a9f4 src/Makefile.inc src/extras.c src/extras.h...: Header file cleanup
|
|
Packit |
577717 |
The papi_protos.h file contained a lot of no-longer in use exports. I split
|
|
Packit |
577717 |
up the ones that are still relevant to header files corresponding to the C
|
|
Packit |
577717 |
file that the functions are defined in.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 07199b41 src/extras.c src/papi_vector.c src/papi_vector.h: Clean up the
|
|
Packit |
577717 |
papi_vector code. Remove things no longer being used, mark static functions
|
|
Packit |
577717 |
as static.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d7496311 src/linux-common.c src/x86_cpuid_info.c src/x86_cpuid_info.h: Fix
|
|
Packit |
577717 |
a missing "return 1" which meant that the virtualization flag wasn't being
|
|
Packit |
577717 |
set right. With this fix, on saturn-vm1 we now get: Running in a VM
|
|
Packit |
577717 |
: yes VM Vendor: : VMwareVMware in the papi_native_avail
|
|
Packit |
577717 |
header
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8da36222 src/freebsd.c src/linux-bgp.c src/papi.c...: Remove the
|
|
Packit |
577717 |
->add_prog_event function vector As far as I can tell this is a PAPI 2.0
|
|
Packit |
577717 |
remnant that was never properly removed. This also removes
|
|
Packit |
577717 |
PAPI_add_pevent(), PAPI_save(), and PAPI_restore(), none of which were
|
|
Packit |
577717 |
exported in papi.h so in theory no one could have been using them. Also
|
|
Packit |
577717 |
removes _papi_hwi_add_pevent()
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a5f3c8b5 src/aix.c src/freebsd.c src/linux-timer.c...: Reduce the usage of
|
|
Packit |
577717 |
MY_VECTOR whenever possible. This is an attempt to make the cpu-counter
|
|
Packit |
577717 |
components to be as similar as possible to external components.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* abbcbf29 src/any-null.h: Missed removing any-null.h during the any-null
|
|
Packit |
577717 |
removal.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 665d4c5c src/linux-common.c: Somehow missed an include during the
|
|
Packit |
577717 |
virtualization addition.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0c06147b src/perfctr-2.6.x/usr.lib/event_set_centaur.os
|
|
Packit |
577717 |
src/perfctr-2.6.x/usr.lib/event_set_p5.os
|
|
Packit |
577717 |
src/perfctr-2.6.x/usr.lib/event_set_p6.os: Removes the last of the binary
|
|
Packit |
577717 |
files from perfctr2.6.x Some binary files were left out in the cold after a
|
|
Packit |
577717 |
mishap trying to configure perfctr for the build test.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3acb7d57 src/Makefile.inc src/configure src/configure.in...: Add support
|
|
Packit |
577717 |
for reporting if we are running in a virtualized environment to the
|
|
Packit |
577717 |
PAPI_hw_info_t structure. This currently only works on x86. it works by
|
|
Packit |
577717 |
looking at bit 31 of ecx after a cpuid (the "in a VM" bit) and then using
|
|
Packit |
577717 |
leaf 0x40000000 to get the name of the VM software (this works for VMware and
|
|
Packit |
577717 |
Xen at least) x86_cache_info.c was renamed to x86_cpuid_info.c to better
|
|
Packit |
577717 |
reflect what goes on in that file (it does various things based on the cpuid
|
|
Packit |
577717 |
instruction). the testlib header was updated to report virtualization status
|
|
Packit |
577717 |
in the papi header (printed for things like papi_native_avail).
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-22
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9c7659b5 src/Makefile.inc src/freq.c: Remove the freq.c file as nothing
|
|
Packit |
577717 |
seemed to be using it.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d205e2d3 src/perfctr-x86.c: Made a stupid typo when converting perfctr to
|
|
Packit |
577717 |
call libpfm functions with the component id.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 25b41779 src/papi_libpfm3_events.c src/papi_libpfm4_events.c
|
|
Packit |
577717 |
src/papi_libpfm_events.h...: When updating the preset code to take a
|
|
Packit |
577717 |
component index I missed a few callers.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a713ffb1 src/papi_internal.c src/papi_vector.c: Remove any-null component
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 27e1c2c5 src/any-null-memory.c src/any-null.c src/any-proc-null.c...:
|
|
Packit |
577717 |
Remove the any-null component.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 25779ae0 PAPI_FAQ.html: Saving another version of the FAQ after adding a
|
|
Packit |
577717 |
git section, and removing several obsolete sections. These questions still
|
|
Packit |
577717 |
need detailed review for relevance and timeliness.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 449a1a61 src/ctests/overflow_allcounters.c: Fix overflow_allcounters which
|
|
Packit |
577717 |
was making assumptions about component 0 existing.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f21be742 src/ctests/hwinfo.c: Make the hwinfo test not bail out if no
|
|
Packit |
577717 |
counters are available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ebc675e6 src/ctests/memory.c: Make sure the memory ctest runs even if no
|
|
Packit |
577717 |
components are available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 9b3de551 src/linux-common.c src/perf_events.c src/perfmon-ia64.c...: Make
|
|
Packit |
577717 |
sure the system info init happens at os init time. Otherwise the system info
|
|
Packit |
577717 |
never gets set if a perfcounter component isn't available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 59e47e12 src/papi_internal.c: Make sure that _papi_hwi_assign_eventset()
|
|
Packit |
577717 |
does the right thing if no components are available.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* dd51e5d6 src/ctests/api.c: The api test would fail in the no cpu component
|
|
Packit |
577717 |
case. Fix it to properly check for errors before attempting to run
|
|
Packit |
577717 |
high-level PAPI tests.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 069e9d2f src/aix.c src/papi.c src/papi_internal.h...: Fix code that was
|
|
Packit |
577717 |
depending on _papi_hwd[0] existing. Most of this was in the presets code.
|
|
Packit |
577717 |
The preset code had many assumptions so that you can only code presets with
|
|
Packit |
577717 |
component[0]. This fixes some of them by passing the component index around.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 7259eaec src/papi_vector.c: Fix up papi_vector to get rid of some warnings
|
|
Packit |
577717 |
introduced on AIX.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 16fe0a61 src/aix.c src/solaris-ultra.c: Fix two last substrates where I
|
|
Packit |
577717 |
missed some fields in the OS structure conversion.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 625871ec src/perfmon.c: Missed a cmp_info field in perfmon.c
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 680919d9 PAPI_FAQ.html: Saving the latest version of the FAQ before
|
|
Packit |
577717 |
undertaking major revisions.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3d4fa2e5 src/linux-timer.c src/perfctr-x86.h: Fix the perfctr code to
|
|
Packit |
577717 |
compile if configured with --with-virtualtimer=perfctr
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bbd7871f src/perfctr.c: Missed two OS vector calls in the perfctr code
|
|
Packit |
577717 |
during the conversion.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* bc6d1713 src/Makefile.inc: Removed one of the two instances of MISCOBJS
|
|
Packit |
577717 |
listed in Makefile.inc.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-21
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 40bc4c57 src/papi_vector.c src/papi_vector.h: Remove now-unused OS vectors
|
|
Packit |
577717 |
from the main papi vector table.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 3c6a0f7b src/aix.c src/freebsd.c src/linux-bgp.c...: Convert PAPI to use
|
|
Packit |
577717 |
the _papi_os_vector for the operating-system specific function vectors.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 568abad5 src/papi_vector.h: Add new _papi_os_vector structure to hold
|
|
Packit |
577717 |
operating-system specific function vectors.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* a39d2373 src/ctests/subinfo.c: Missed removing a field from the subinfo
|
|
Packit |
577717 |
ctest.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 1d930868 src/papi.h: Remove fields now in PAPI_os_info_t from the
|
|
Packit |
577717 |
component_info_t struct.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* d397d74a src/components/example/example.c: Remove fields now in
|
|
Packit |
577717 |
PAPI_os_info_t from the example component.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 8cd5c8e0 src/aix.c src/freebsd.c src/linux-bgp.c...: Modify all the
|
|
Packit |
577717 |
substrates to use _papi_os_info. instead of _papi_hwd[0]->cmp_info for the
|
|
Packit |
577717 |
values moved to the OS struct
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 58855d3a src/papi_internal.h: Add padding for future expansion to
|
|
Packit |
577717 |
PAPI_os_info_t Add _papi_hwi_init_os(void); definition
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* ea1930e1 src/papi_internal.h: Add new PAPI_os_info_t structure to
|
|
Packit |
577717 |
papi_internal.h
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 0eac1b29 src/utils/multiplex_cost.c: Modify multiplex_cost to properly use
|
|
Packit |
577717 |
the API_get_opt() interface to get itimer data, rather than directly
|
|
Packit |
577717 |
accessing the fields from the cmp_info structure. This would have broken
|
|
Packit |
577717 |
after the OS split.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 87c2aa2f src/ctests/subinfo.c: subinfo was printing itimer data from the
|
|
Packit |
577717 |
cmpinfo structure. These values will not be in cmpinfo once the OS split
|
|
Packit |
577717 |
happens.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f2c62d50 src/components/vmware/vmware.h: Clean up the VMware Header a bit
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-17
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 6f0c1230 src/aix.c src/components/coretemp/linux-coretemp.c
|
|
Packit |
577717 |
src/components/coretemp_freebsd/coretemp_freebsd.c...: The git conversion
|
|
Packit |
577717 |
reset all of the CVS $Id$ lines to just $Id$ Since we depend on the $Id$
|
|
Packit |
577717 |
lines for the component names, I had to go back and fix all of them to be the
|
|
Packit |
577717 |
component names again.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 2d208d0e src/perfctr-2.6.x/usr.lib/event_set_centaur.o
|
|
Packit |
577717 |
src/perfctr-2.6.x/usr.lib/event_set_p5.o
|
|
Packit |
577717 |
src/perfctr-2.6.x/usr.lib/event_set_p6.o: Remove a few binary files in
|
|
Packit |
577717 |
perfctr-2.6.x
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* f78bf1af src/libpfm-3.y/Makefile src/libpfm-3.y/README
|
|
Packit |
577717 |
src/libpfm-3.y/docs/Makefile...: More cleanups from the migration, latest
|
|
Packit |
577717 |
version of libpfm-3.y perfctr-2.[6,7] Version numbers got really confused in
|
|
Packit |
577717 |
cvs and the git cvsimport didn't know that eg 1.1.1.28 > 1.1 ( see
|
|
Packit |
577717 |
perfctr-2.6.x/CHANGES revision 1.1.1.28.6.1 :~)
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* e8aa2e61 INSTALL.txt: Explicitly state that 3.7 was the last version of
|
|
Packit |
577717 |
PAPI with good windows support.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 546901fa src/components/cuda/linux-cuda.c: Modified CUDA component so that
|
|
Packit |
577717 |
a PAPI version - that was configured with CUDA - will successfully build on a
|
|
Packit |
577717 |
machine that does not have GPUs.
|
|
Packit |
577717 |
|
|
Packit |
577717 |
2012-02-16
|
|
Packit |
577717 |
|
|
Packit |
577717 |
* 49d9f71c src/.gitignore: Add a .gitignore file with the files that PAPI
|
|
Packit |
577717 |
autogenerates. This way they won't clutter up "git status" messages
|
|
Packit |
577717 |
|