perf fixes
Linus,
Please pull the latest perf-urgent-for-linus git tree from:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git perf-urgent-for-linus
# HEAD: 57e04eeda515ee979fec3bc3d64c408feae18acc Merge tag 'perf-urgent-for-mingo-5.5-20191216' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent
These are all perf tooling changes: most of them are fixes.
Note that the large CPU count related fixes go beyond regression fixes - but the IPI-flood symptoms are severe enough that I think justifies their inclusion. If it's too much we'll redo the branch.
Thanks,
Ingo
------------------>
Adrian Hunter (1):
perf inject: Fix processing of ID index for injected instruction tracing
Andi Kleen (10):
perf cpumap: Maintain cpumaps ordered and without dups perf evlist: Maintain evlist->all_cpus
perf evsel: Add iterator to iterate over events ordered by CPU perf evsel: Add functions to close evsel on a CPU perf stat: Use affinity for closing file descriptors perf stat: Factor out open error handling
perf stat: Use affinity for opening events
perf stat: Use affinity for reading
perf evsel: Add functions to enable/disable for a specific CPU perf stat: Use affinity for enabling/disabling events
Arnaldo Carvalho de Melo (15):
perf machine: Fill map_symbol->maps in append_inlines() to fix segfault perf bench: Update the copies of x86's mem{cpy,set}_64.S tools arch x86: Sync the msr-index.h copy with the kernel sources tools headers uapi: Sync linux/fscrypt.h with the kernel sources tools headers uapi: Sync linux/stat.h with the kernel sources tools headers kvm: Sync kvm headers with the kernel sources tools headers UAPI: Sync sched.h with the kernel
perf beauty: Add CLEAR_SIGHAND support for clone's flags arg tools arch x86: Sync asm/cpufeatures.h with the kernel sources perf kvm: Clarify the 'perf kvm' -i and -o command line options tools headers UAPI: Sync drm/i915_drm.h with the kernel sources tools headers UAPI: Update tools's copy of drm.h headers tools headers kvm: Sync linux/kvm.h with the kernel sources perf arch: Make the default get_cpuid() return compatible error perf top: Do not bail out when perf_env__read_cpuid() returns ENOSYS
Ed Maste (2):
perf vendor events s390: Fix counter long description for DTLB1_GPAGE_WRITES perf vendor events s390: Remove name from L1D_RO_EXCL_WRITES description
Ian Rogers (1):
perf jit: Move test functionality in to a test
Kajol Jain (1):
perf metricgroup: Fix printing event names of metric group with multiple events
Michael Petlan (1):
perf header: Fix false warning when there are no duplicate cache entries
Ravi Bangoria (4):
perf report/top TUI: Replace pr_err() with ui__error() perf report: Make -F more strict like -s
perf report: Bail out --mem-mode if mem info is not available perf/x86/pmu-events: Fix Kernel_Utilization metric
Sudip Mukherjee (3):
libtraceevent: Fix lib installation with O=
libtraceevent: Copy pkg-config file to output folder when using O=
libtraceevent: Allow custom libdir path
tools/arch/arm/include/uapi/asm/kvm.h | 3 +- tools/arch/arm64/include/uapi/asm/kvm.h | 5 +- tools/arch/powerpc/include/uapi/asm/kvm.h | 3 + tools/arch/x86/include/asm/cpufeatures.h | 3 + tools/arch/x86/include/asm/msr-index.h | 18 ++ tools/arch/x86/lib/memcpy_64.S | 20 +- tools/arch/x86/lib/memset_64.S | 16 +- tools/include/uapi/drm/drm.h | 3 +- tools/include/uapi/drm/i915_drm.h | 128 ++++++++- tools/include/uapi/linux/fscrypt.h | 3 +- tools/include/uapi/linux/kvm.h | 12 + tools/include/uapi/linux/sched.h | 60 +++-- tools/include/uapi/linux/stat.h | 2 +- tools/lib/traceevent/Makefile | 11 +- tools/lib/traceevent/plugins/Makefile | 5 +- tools/perf/Documentation/perf-kvm.txt | 5 +- tools/perf/arch/arm/tests/regs_load.S | 4 +- tools/perf/arch/arm64/tests/regs_load.S | 4 +- tools/perf/arch/x86/tests/regs_load.S | 8 +- tools/perf/builtin-inject.c | 13 +- tools/perf/builtin-record.c | 2 +- tools/perf/builtin-report.c | 8 + tools/perf/builtin-stat.c | 288 +++++++++++++++------ tools/perf/builtin-top.c | 10 +- tools/perf/check-headers.sh | 4 +- tools/perf/lib/cpumap.c | 73 +++++- tools/perf/lib/evlist.c | 1 + tools/perf/lib/evsel.c | 76 ++++-- tools/perf/lib/include/internal/evlist.h | 1 + tools/perf/lib/include/perf/cpumap.h | 2 + tools/perf/lib/include/perf/evsel.h | 3 + .../perf/pmu-events/arch/s390/cf_z13/extended.json | 2 +- .../perf/pmu-events/arch/s390/cf_z14/extended.json | 2 +- .../pmu-events/arch/x86/broadwell/bdw-metrics.json | 2 +- .../arch/x86/broadwellde/bdwde-metrics.json | 2 +- .../arch/x86/broadwellx/bdx-metrics.json | 2 +- .../arch/x86/cascadelakex/clx-metrics.json | 2 +- .../pmu-events/arch/x86/haswell/hsw-metrics.json | 2 +- .../pmu-events/arch/x86/haswellx/hsx-metrics.json | 2 +- .../pmu-events/arch/x86/ivybridge/ivb-metrics.json | 2 +- .../pmu-events/arch/x86/ivytown/ivt-metrics.json | 2 +- .../pmu-events/arch/x86/jaketown/jkt-metrics.json | 2 +- .../arch/x86/sandybridge/snb-metrics.json | 2 +- .../pmu-events/arch/x86/skylake/skl-metrics.json | 2 +- .../pmu-events/arch/x86/skylakex/skx-metrics.json | 2 +- tools/perf/tests/Build | 1 + tools/perf/tests/builtin-test.c | 9 + tools/perf/tests/cpumap.c | 16 ++ tools/perf/tests/event-times.c | 4 +- tools/perf/tests/genelf.c | 51 ++++ tools/perf/tests/tests.h | 2 + tools/perf/trace/beauty/clone.c | 1 + tools/perf/util/cpumap.h | 1 + tools/perf/util/evlist.c | 113 +++++++- tools/perf/util/evlist.h | 11 +- tools/perf/util/evsel.c | 35 ++- tools/perf/util/evsel.h | 9 +- tools/perf/util/genelf.c | 46 ---- tools/perf/util/header.c | 23 +- tools/perf/util/include/linux/linkage.h | 89 ++++++- tools/perf/util/machine.c | 1 + tools/perf/util/metricgroup.c | 7 +- tools/perf/util/sort.c | 16 +- tools/perf/util/stat.c | 5 +- tools/perf/util/stat.h | 3 +- 65 files changed, 976 insertions(+), 289 deletions(-) create mode 100644 tools/perf/tests/genelf.c
Quoted 4 lines · Ingo Molnar
Note that the large CPU count related fixes go beyond regression fixes - but the IPI-flood symptoms are severe enough that I think justifies their inclusion. If it's too much we'll redo the branch.
I'm much less sensitive to the perf _tooling_ changes than I am to actual kernel code changes.
Of course, if you start having a history of breaking things during the rc kernels, that may change, but at least for now, I tend to look at diffstats and go "oh, this is perf tooling", and not worry too much about it...
Linus
The pull request you sent on Tue, 17 Dec 2019 12:34:25 +0100:
The pull request you sent on Tue, 17 Dec 2019 12:34:25 +0100:
Quoted 1 line
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.gitperf-urgent-for-linus
has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/89c683cd06e03dfd3186c4cab1e2a39982c42a48
Thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.wiki.kernel.org/userdoc/prtracker
Quoted 6 lines · Linus Torvalds
Quoted 4 lines · Ingo Molnar
Note that the large CPU count related fixes go beyond regression fixes - but the IPI-flood symptoms are severe enough that I think justifies their inclusion. If it's too much we'll redo the branch.
Quoted 7 lines
I'm much less sensitive to the perf _tooling_ changes than I am to actual kernel code changes.
Of course, if you start having a history of breaking things during the rc kernels, that may change, but at least for now, I tend to look at diffstats and go "oh, this is perf tooling", and not worry too much about it...
Thanks for pulling these in any case!
Ingo