Skip to content

more perf_counter stuff

Peter Zijlstra lkml 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest6 patches 51 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/6perf_counter: move the event overflow output bits to record_typeinclude/linux/perf_counter.h kernel/perf_counter.c+67 -828 replies
2/6RFC perf_counter: singleshot supportarch/powerpc/kernel/perf_counter.c arch/x86/kernel/cpu/perf_counter.c include/linux/perf_counter.h +1 more+89 -307 replies
3/6perf_counter: per event wakeupsinclude/linux/perf_counter.h kernel/perf_counter.c+11 -22 replies
4/6perf_counter: kerneltop: update to new ABIDocumentation/perf_counter/kerneltop.c+4 -510 replies
5/6perf_counter: add more context informationarch/x86/kernel/cpu/perf_counter.c include/linux/perf_counter.h kernel/perf_counter.c+12 -323 replies
6/6perf_counter: update mmap() counter readinclude/linux/perf_counter.h kernel/perf_counter.c+16 -181 reply

Reviews trailers per patch

None of the 6 patches has a Reviewed-by or Acked-by.

PatchPaul Mackerras
1/6
2/6
3/6
4/6
5/6
6/6Noticed-by

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.

Read the v1 thread

Cover letter [PATCH 0/6] more perf_counter stuff

This patch set implements some features requested by Paul/Corey. The main one being the singleshot thing, which is still RFC due to being incomplete.

It really needs a mechanism to run bits in a regular context ASAP after the event context. My faviourite is a self-IPI, because one can issue it from NMI context, or IRQ-disabled rq->lock context and have it work as expected.

I'll look at doing a x86 self-IPI.