src/utils/native_avail.c | 29 20 + 9 - 0 !
1 file changed, 20 insertions(+), 9 deletions(-)

 [patch 04/44] native_avail: add a --validate flag

--validate attempts to decide if each event can be added to an eventset.

Patch by Steve Kaufmann

src/papi_events.csv | 1 1 + 0 - 0 !
1 file changed, 1 insertion(+)

 [patch 06/44] add aarch64 cortex a57 presets

The events for the ARM Cortex A57 events look to be the same as the
events for the ARM Cortex A15 events. Add an entry so the preset are
available for the ARM Cortex A57.

src/utils/native_avail.c | 41 32 + 9 - 0 !
1 file changed, 32 insertions(+), 9 deletions(-)

 [patch 10/44] native_avail.c: bug fixes and updates

Thanks to Gary Mohr

src/papi_internal.c | 64 39 + 25 - 0 !
1 file changed, 39 insertions(+), 25 deletions(-)

 [patch 11/44] papi_internal.c: change subdbg to intdbg

Thanks to Gary Mohr

src/run_tests.sh | 5 0 + 5 - 0 !
1 file changed, 5 deletions(-)

 [patch 12/44] run_tests.sh: more exclude cleanups

Thanks Gary Mohr

src/papi.c | 72 51 + 21 - 0 !
1 file changed, 51 insertions(+), 21 deletions(-)

 [patch 13/44] papi.c: add information to api entry debuging

Thanks to Gary Mohr

src/papi_events.csv | 9 7 + 2 - 0 !
1 file changed, 7 insertions(+), 2 deletions(-)

 [patch 14/44] update hsw presets

Thanks to Gary Mohr

src/components/perf_event/perf_event.c | 63 41 + 22 - 0 !
1 file changed, 41 insertions(+), 22 deletions(-)

 [patch 15/44] perf_event.c: cleanup error messages

Thanks to Gary Mohr

src/libpfm4/include/perfmon/pfmlib_perf_event.h | 2 1 + 1 - 0 !
src/libpfm4/lib/pfmlib_perf_event.c | 11 11 + 0 - 0 !
src/libpfm4/lib/pfmlib_perf_event_priv.h | 2 2 + 0 - 0 !
src/libpfm4/perf_examples/perf_util.c | 1 1 + 0 - 0 !
src/libpfm4/perf_examples/perf_util.h | 1 1 + 0 - 0 !
5 files changed, 16 insertions(+), 1 deletion(-)

 [patch 16/44] libpfm4 update: catch the cpu= modifier update

Inline with:
commit c9482351a6ce2cf39433527fdad41a78fed2b6f2

src/components/nvml/README | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

 [patch 17/44] update nvml readme

We changed the options to simplify the configure line.
Bad information is worse than no information...

src/utils/error_codes.c | 13 9 + 4 - 0 !
1 file changed, 9 insertions(+), 4 deletions(-)

 [patch 18/44] error_codes utility: remove internal bits

Remove dependency on _papi_hwi_num_errors, just keep calling
PAPI_strerror until it fails. We shouldn't be using internal symbols
anyways.

src/utils/native_avail.c | 428 223 + 205 - 0 !
1 file changed, 223 insertions(+), 205 deletions(-)

 [patch 22/44] extend 'papi_native_event --validate' to check for
 umasks.

src/utils/event_info.c | 3 1 + 2 - 0 !
1 file changed, 1 insertion(+), 2 deletions(-)

 [patch 23/44] event_info utility: address coverity defect

From Gary Mohr

src/testlib/test_utils.c | 6 5 + 1 - 0 !
1 file changed, 5 insertions(+), 1 deletion(-)

 [patch 24/44] testlib/test_util.c: check enum return value

Addresses an issue found by Coverity.
Thanks Gary Mohr,

src/ctests/branches.c | 2 1 + 1 - 0 !
src/ctests/calibrate.c | 3 2 + 1 - 0 !
src/ctests/describe.c | 12 5 + 7 - 0 !
src/ctests/get_event_component.c | 10 7 + 3 - 0 !
src/ctests/krentel_pthreads.c | 13 10 + 3 - 0 !
5 files changed, 25 insertions(+), 15 deletions(-)

 [patch 25/44] ctests/ address coverity reported defects

Thanks to Gary Mohr for the patch

src/components/perf_event/perf_event_lib.h | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

 [patch 26/44] bump num_mpx_counters for linux-perf

Uncore on SNB and newer systems has enough counters to go beyond the 64
array spaces we allocate. This needs a better long term solution.

Reported by Gary Mohr

src/utils/native_avail.c | 310 195 + 115 - 0 !
1 file changed, 195 insertions(+), 115 deletions(-)

 [patch 35/44] adding the nativeavailvalidate patch provided by gary
 Mohr.

The problem being addresed is that if there were any problems validating
event masks, then those problems would result in the entire event being
invalid.  The desired action was to test each event mask, and if any
basic event mask can make the event succeed, then the event should be
returned as valid and available.

The solution is to create a large buffer and write events and masks into
this buffer as they are processed, tracking their validity.  At the end
go back and mark the validity of the entire event.  This matches the
standard output of PAPI.

src/papi_events.csv | 23 23 + 0 - 0 !
1 file changed, 23 insertions(+)

 [patch 40/44] support for the arm x-gene processor.

Thanks to William Cohen for this patch.
The events for the Applied Micro X-Gene processor are slightly

src/papi_events.csv | 1 1 + 0 - 0 !
1 file changed, 1 insertion(+)

 [patch 41/44] update papi_events.csv to match libpfm support for
 Intel family 6 model 63 (hsw_ep)

Thanks to William Cohen for this patch and its information:
'A recent September 11, 2014 patch (98c00b) to the upstream libpfm
split out Intel family 6 model 63 into its own name of "hsw_ep".  The
papi_events.csv needs to be updated to support that new name.  This
should have no impact for older libpfms that still identify Intel
family 6 model 63 as "hswv" and "hsw_ep" map to the same papi presets.'

src/papi_events.csv | 5 0 + 5 - 0 !
1 file changed, 5 deletions(-)

'Commit 4c87d753ab56688acad5bf0cb3b95eae8aa80458 added some events
meant for Intel Haswell to the Intel Ivy bridge presets.  This patch
removes those stray events.  Without this patch on Intel Ivy Bridge
machines would see messages like the following:
PAPI Error: papi_preset: Error finding event L2_TRANS:DEMAND_DATA_RD.
PAPI Error: papi_preset: Error finding event L2_RQSTS:ALL_DEMAND_REFERENCES.'

This patch was not tested by the PAPI team (no appropriate hardware).

src/linux-memory.c | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

The variable sys_mem_info is an array of type PAPI_mh_info_t.  It is
statically declared as size 4.  The data for POWER8 is statically
declared in entry 4 of the array which is beyond the allocated array.  The
array should be declared without a size so the compiler will automatically
determine the correct size based on the number of elements being
initialized.  This patch makes the change.

src/ctests/sprofile.c | 9 7 + 2 - 0 !
1 file changed, 7 insertions(+), 2 deletions(-)

The POWER 8 platform is Little Endian.  It uses ELF
version 2 which does not use function descriptors.  This
patch adds the needed #ifdef support to correctly compile
the test case for Big Endian or Little Endian.

src/ctests/byte_profile.c | 7 6 + 1 - 0 !
1 file changed, 6 insertions(+), 1 deletion(-)

The POWER 8 platform is Little Endian.  It uses ELF
version 2 which does not use function descriptors.  This
patch adds the needed #ifdef support to correctly compile
the test case for Big Endian or Little Endian.

This patch is untested by the PAPI developers (hardware not easily accessible).

src/linux-common.c | 1 1 + 0 - 0 !
1 file changed, 1 insertion(+)

The Power 8 system can be run as a non-virtualized machine.  In
this case, the platform is "PowerNV".  This patch adds the platform
to the possible IBM platform types.

src/configure.in | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

---

src/configure.in | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

---

man/man1/papi_avail.1 | 14 7 + 7 - 0 !
man/man1/papi_command_line.1 | 6 3 + 3 - 0 !
man/man1/papi_component_avail.1 | 4 2 + 2 - 0 !
man/man1/papi_cost.1 | 12 6 + 6 - 0 !
man/man1/papi_decode.1 | 6 3 + 3 - 0 !
man/man1/papi_hybrid_native_avail.1 | 26 13 + 13 - 0 !
man/man1/papi_multiplex_cost.1 | 12 6 + 6 - 0 !
man/man1/papi_native_avail.1 | 24 12 + 12 - 0 !
man/man1/papi_xml_event_info.1 | 8 4 + 4 - 0 !
man/man3/PAPI_accum.3 | 2 1 + 1 - 0 !
man/man3/PAPI_accum_counters.3 | 2 1 + 1 - 0 !
man/man3/PAPI_attach.3 | 2 1 + 1 - 0 !
man/man3/PAPI_detach.3 | 2 1 + 1 - 0 !
man/man3/PAPI_read.3 | 2 1 + 1 - 0 !
man/man3/PAPI_read_counters.3 | 2 1 + 1 - 0 !
man/man3/PAPI_set_debug.3 | 2 1 + 1 - 0 !
man/man3/PAPI_thread_id.3 | 2 1 + 1 - 0 !
17 files changed, 64 insertions(+), 64 deletions(-)

 fix manpage hyphenation and typos

src/Rules.pfm4_pe | 3 2 + 1 - 0 !
1 file changed, 2 insertions(+), 1 deletion(-)

 we link against libpfm4.so, so don't embed a copy of libpfm4.a

src/components/Makefile_comp_tests.target.in | 7 4 + 3 - 0 !
src/ctests/Makefile.recipies | 3 3 + 0 - 0 !
src/ctests/Makefile.target.in | 4 2 + 2 - 0 !
src/ftests/Makefile.recipies | 3 3 + 0 - 0 !
src/ftests/Makefile.target.in | 4 2 + 2 - 0 !
5 files changed, 14 insertions(+), 7 deletions(-)

 support building the examples out-of-tree and link them dynamically

src/configure.in | 2 1 + 1 - 0 !
src/linux-context.h | 2 2 + 0 - 0 !
src/linux-timer.c | 2 1 + 1 - 0 !
3 files changed, 4 insertions(+), 2 deletions(-)

 fix build on hppa/linux
Bug-Debian: https://bugs.debian.org/751411