Skip to content

perf/x86/amd: Miscellaneous updates

Kim Phillips lkml stable 3 versions – Applied

Status from this message: The following commit has been merged into the perf/core branch of tip:

Get series with b4
b4 am [email protected]

Versions newest first · linked by author and title

  1. v2 Latest7 patches 4 repliesApplied
  2. v1 posting 2/27 patches 1 replySuperseded
  3. v1 posting 1/21 patch 0 repliesSuperseded

Patches in v2 Applied

#Subject and filesSizeRepliesApplied
1/7perf/amd/uncore: Set all slices and threads to restore perf stat -a behaviourarch/x86/events/amd/uncore.c+8 -201 reply
2/7perf/x86/amd: Fix sampling Large Increment per Cycle eventsarch/x86/events/core.c+2 -20 replies
3/7arch/x86/amd/ibs: Fix re-arming IBS Fetcharch/x86/events/amd/ibs.c+8 -13 replies
4/7perf/x86/amd/ibs: Don't include randomized bits in get_ibs_op_count()arch/x86/events/amd/ibs.c+8 -40 replies
5/7perf/x86/amd/ibs: Fix raw sample data accumulationarch/x86/events/amd/ibs.c arch/x86/include/asm/msr-index.h+17 -100 replies
6/7perf/x86/amd/ibs: Support 27-bit extended Op/cycle counterarch/x86/events/amd/ibs.c arch/x86/include/asm/perf_event.h+32 -110 replies
7/7perf/x86/rapl: Add AMD Fam19h RAPL supportarch/x86/events/rapl.c+1 -00 replies

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v2 yet.

Read the v2 thread

Cover letter [PATCH v2 0/7] perf/x86/amd: Miscellaneous updates

Hi,

Please merge:

Patch 1/7 restores 'perf stat -a' behaviour to program the uncore PMU to count all CPU threads.

Patch 2/7 fixes setting the proper count when sampling Large Increment per Cycle events / 'paired' events.

The next 4 patches fix IBS Fetch sampling on F17h and some other IBS fine tuning, the last one greatly reducing the number of interrupts when large sample periods are specified.

Finally, patch 7/7 extends Family 17h RAPL support to also work on compatible F19h machines.

Thanks,

Kim

v2: Added this cover letter. The only patch that changed in this new submission is the fix for IBS Fetch sampling, which got a new family check after testing on some other family machines.

Link to 1st patch in original submission of this series:

https://lore.kernel.org/patchwork/patch/1289806/

Cc Stephane Eranian <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Alexander Shishkin <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc "H. Peter Anvin" <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Mark Rutland <[email protected]>
Cc Michael Petlan <[email protected]>
Cc Namhyung Kim <[email protected]>

Kim Phillips (7):
perf/amd/uncore: Set all slices and threads to restore perf stat -a
behaviour
perf/x86/amd: Fix sampling Large Increment per Cycle events arch/x86/amd/ibs: Fix re-arming IBS Fetch perf/x86/amd/ibs: Don't include randomized bits in get_ibs_op_count() perf/x86/amd/ibs: Fix raw sample data accumulation perf/x86/amd/ibs: Support 27-bit extended Op/cycle counter perf/x86/rapl: Add AMD Fam19h RAPL support

 arch/x86/events/amd/ibs.c         | 87 ++++++++++++++++++++++---------
 arch/x86/events/amd/uncore.c      | 28 +++-------
 arch/x86/events/core.c            |  4 +-
 arch/x86/events/rapl.c            |  1 +
 arch/x86/include/asm/msr-index.h  |  1 +
 arch/x86/include/asm/perf_event.h |  1 +
 6 files changed, 75 insertions(+), 47 deletions(-)
Signature · 1 line
2.27.0