- 01 Dec, 2016 8 commits
-
-
Kim Phillips authored
This is a regex converted version from the original: https://lkml.org/lkml/2016/5/19/461 Add basic support to recognise AArch64 assembly. This allows perf to identify AArch64 instructions that branch to other parts within the same function, thereby properly annotating them. Rebased onto new cross-arch annotation bits: https://lkml.org/lkml/2016/11/25/546 Sample output: security_file_permission vmlinux 5.80 │ ← ret ▒ │70: ldr w0, [x21,#68] ▒ 4.44 │ ↓ tbnz d0 ▒ │ mov w0, #0x24 // #36 ▒ 1.37 │ ands w0, w22, w0 ▒ │ ↑ b.eq 60 ▒ 1.37 │ ↓ tbnz e4 ▒ │ mov w19, #0x20000 // #131072 ▒ 1.02 │ ↓ tbz ec ▒ │90:┌─→ldr x3, [x21,#24] ▒ 1.37 │ │ add x21, x21, #0x10 ▒ │ │ mov w2, w19 ▒ 1.02 │ │ mov x0, x21 ▒ │ │ mov x1, x3 ▒ 1.71 │ │ ldr x20, [x3,#48] ▒ │ │→ bl __fsnotify_parent ▒ 0.68 │ │↑ cbnz 60 ▒ │ │ mov x2, x21 ▒ 1.37 │ │ mov w1, w19 ▒ │ │ mov x0, x20 ▒ 0.68 │ │ mov w5, #0x0 // #0 ▒ │ │ mov x4, #0x0 // #0 ▒ 1.71 │ │ mov w3, #0x1 // #1 ▒ │ │→ bl fsnotify ▒ 1.37 │ │↑ b 60 ▒ │d0:│ mov w0, #0x0 // #0 ▒ │ │ ldp x19, x20, [sp,#16] ▒ │ │ ldp x21, x22, [sp,#32] ▒ │ │ ldp x29, x30, [sp],#48 ▒ │ │← ret ▒ │e4:│ mov w19, #0x10000 // #65536 ▒ │ └──b 90 ◆ │ec: brk #0x800 ▒ Press 'h' for help on key bindings Signed-off-by:
Kim Phillips <kim.phillips@arm.com> Signed-off-by:
Chris Ryder <chris.ryder@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Link: http://lkml.kernel.org/r/20161130092344.012e18e3e623bea395162f95@arm.comSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Kim Phillips authored
Presume neglected in commit 786c1b51 "perf annotate: Start supporting cross arch annotation". This doesn't fix a bug since none of the affected arches support parsing dec/inc instructions yet. Signed-off-by:
-
David Ahern authored
Add option to allow user to control analysis window. e.g., collect data for time window and analyze a segment of interest within that window. Committer notes: Testing it: Using the perf.data file captured via 'perf kmem record': # perf report --header-only # ======== # captured on: Tue Nov 29 16:01:53 2016 # hostname : jouet # os release : 4.8.8-300.fc25.x86_64 # perf version : 4.9.rc6.g5a6aca # arch : x86_64 # nrcpus online : 4 # nrcpus avail : 4 # cpudesc : Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz # cpuid : GenuineIntel,6,61,4 # total memory : 20254660 kB # cmdline : /home/acme/bin/perf kmem record usleep 1 # event : name = kmem:kmalloc, , id = { 931980, 931981, 931982, 931983 }, type = 2, size = 112, config = 0x1b9, { sample_period, sample_freq } = 1, sample_typ # event : name = kmem:kmalloc_node, , id = { 931984, 931985, 931986, 931987 }, type = 2, size = 112, config = 0x1b7, { sample_period, sample_freq } = 1, sampl # event : name = kmem:kfree, , id = { 931988, 931989, 931990, 931991 }, type = 2, size = 112, config = 0x1b5, { sample_period, sample_freq } = 1, sample_type # event : name = kmem:kmem_cache_alloc, , id = { 931992, 931993, 931994, 931995 }, type = 2, size = 112, config = 0x1b8, { sample_period, sample_freq } = 1, s # event : name = kmem:kmem_cache_alloc_node, , id = { 931996, 931997, 931998, 931999 }, type = 2, size = 112, config = 0x1b6, { sample_period, sample_freq } = # event : name = kmem:kmem_cache_free, , id = { 932000, 932001, 932002, 932003 }, type = 2, size = 112, config = 0x1b4, { sample_period, sample_freq } = 1, sa # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu = 4, intel_pt = 7, intel_bts = 6, uncore_arb = 13, cstate_pkg = 15, breakpoint = 5, uncore_cbox_1 = 12, power = 9, software = 1, uncore_im # HEADER_CACHE info available, use -I to display # missing features: HEADER_BRANCH_STACK HEADER_GROUP_DESC HEADER_AUXTRACE HEADER_STAT # ======== # # # Looking at just the histogram entries for the first event: # # perf report | head -33 # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 40 of event 'kmem:kmalloc' # Event count (approx.): 40 # # Overhead Trace output # ........ ............................................................................................................... # 37.50% call_site=ffffffffb91ad3c7 ptr=0xffff88895fc05000 bytes_req=4096 bytes_alloc=4096 gfp_flags=GFP_KERNEL 10.00% call_site=ffffffffb9258416 ptr=0xffff888a1dc61f00 bytes_req=240 bytes_alloc=256 gfp_flags=GFP_KERNEL|__GFP_ZERO 7.50% call_site=ffffffffb9258416 ptr=0xffff888a2640ac00 bytes_req=240 bytes_alloc=256 gfp_flags=GFP_KERNEL|__GFP_ZERO 2.50% call_site=ffffffffb92759ba ptr=0xffff888a26776000 bytes_req=4096 bytes_alloc=4096 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb9276864 ptr=0xffff8886f6b82600 bytes_req=136 bytes_alloc=192 gfp_flags=GFP_KERNEL|__GFP_ZERO 2.50% call_site=ffffffffb9276903 ptr=0xffff888aefcf0460 bytes_req=32 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb92ad0ce ptr=0xffff888756c98a00 bytes_req=392 bytes_alloc=512 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb92ad0ce ptr=0xffff888756c9ba00 bytes_req=504 bytes_alloc=512 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb92ad301 ptr=0xffff888a31747600 bytes_req=128 bytes_alloc=128 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb92ad511 ptr=0xffff888a9d26a2a0 bytes_req=28 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c11a0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c12c0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c1540 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c15a0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c15e0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c16e0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff88873e8c1c20 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb936a7fb ptr=0xffff888a9d26a2a0 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL 2.50% call_site=ffffffffb9373e66 ptr=0xffff8889f1931240 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_ATOMIC|__GFP_ZERO 2.50% call_site=ffffffffb9373e66 ptr=0xffff8889f1931980 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_ATOMIC|__GFP_ZERO 2.50% call_site=ffffffffb9373e66 ptr=0xffff8889f1931a00 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_ATOMIC|__GFP_ZERO # # # And then limiting using the example for 'perf kmem stat --time' used # # in the previous changeset committer note we see that there were no # # kmem:kmalloc in that last part of the file, but there were some # # kmem:kmem_cache_alloc ones: # # perf report --time 20119.782088, --stdio # # Total Lost Samples: 0 # # Samples: 0 of event 'kmem:kmalloc' # Event count (approx.): 0 # # Overhead Trace output # ........ ............ # # Samples: 0 of event 'kmem:kmalloc_node' # Event count (approx.): 0 # # Overhead Trace output # ........ ............ # # Samples: 0 of event 'kmem:kfree' # Event count (approx.): 0 # # Overhead Trace output # ........ ............ # # Samples: 8 of event 'kmem:kmem_cache_alloc' # Event count (approx.): 8 # # Overhead Trace output # ........ .................................................................................................................. # 75.00% call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO 12.50% call_site=ffffffffb90ad33a ptr=0xffff8889f071f6e0 bytes_req=160 bytes_alloc=160 gfp_flags=GFP_ATOMIC|__GFP_NOTRACK 12.50% call_site=ffffffffb9287cc1 ptr=0xffff8889b12722d8 bytes_req=104 bytes_alloc=104 gfp_flags=GFP_NOFS|__GFP_ZERO # Signed-off-by:David Ahern <dsahern@gmail.com> Tested-by:
Namhyung Kim <namhyung@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1480439746-42695-7-git-send-email-dsahern@gmail.comSigned-off-by:
-
David Ahern authored
Add option to allow user to control analysis window. e.g., collect data for time window and analyze a segment of interest within that window. Committer notes: Testing it: # perf kmem record usleep 1 [ perf record: Woken up 0 times to write data ] [ perf record: Captured and wrote 1.540 MB perf.data (2049 samples) ] # perf evlist kmem:kmalloc kmem:kmalloc_node kmem:kfree kmem:kmem_cache_alloc kmem:kmem_cache_alloc_node kmem:kmem_cache_free # Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events # # # Use 'perf script' to get a first approach, select a chunk for then using # # with 'perf kmem stat --time' # # perf script | tail -15 usleep 9889 [0] 20119.782088: kmem:kmem_cache_free: (selinux_file_free_security+0x27) call_site=ffffffffb936aa07 ptr=0xffff888a1df49fc0 perf 9888 [3] 20119.782088: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 perf 9888 [3] 20119.782089: kmem:kmem_cache_alloc: (jbd2__journal_start+0x72) call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO perf 9888 [3] 20119.782090: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 perf 9888 [3] 20119.782090: kmem:kmem_cache_alloc: (jbd2__journal_start+0x72) call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO usleep 9889 [0] 20119.782091: kmem:kmem_cache_alloc: (__sigqueue_alloc+0x4a) call_site=ffffffffb90ad33a ptr=0xffff8889f071f6e0 bytes_req=160 bytes_alloc=160 gfp_flags=GFP_ATOMIC|__GFP_NOTRACK perf 9888 [3] 20119.782091: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 perf 9888 [3] 20119.782093: kmem:kmem_cache_free: (__sigqueue_free.part.17+0x33) call_site=ffffffffb90ad3f3 ptr=0xffff8889f071f6e0 perf 9888 [3] 20119.782098: kmem:kmem_cache_alloc: (jbd2__journal_start+0x72) call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO perf 9888 [3] 20119.782098: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 perf 9888 [3] 20119.782099: kmem:kmem_cache_alloc: (jbd2__journal_start+0x72) call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO perf 9888 [3] 20119.782100: kmem:kmem_cache_alloc: (alloc_buffer_head+0x21) call_site=ffffffffb9287cc1 ptr=0xffff8889b12722d8 bytes_req=104 bytes_alloc=104 gfp_flags=GFP_NOFS|__GFP_ZERO perf 9888 [3] 20119.782101: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 perf 9888 [3] 20119.782102: kmem:kmem_cache_alloc: (jbd2__journal_start+0x72) call_site=ffffffffb9333b42 ptr=0xffff888bdf1a39c0 bytes_req=48 bytes_alloc=48 gfp_flags=GFP_NOFS|__GFP_ZERO perf 9888 [3] 20119.782103: kmem:kmem_cache_free: (jbd2_journal_stop+0x1a1) call_site=ffffffffb9334581 ptr=0xffff888bdf1a39c0 # # # stats for the whole perf.data file, i.e. no interval specified # # perf kmem stat SUMMARY (SLAB allocator) ======================== Total bytes requested: 172,628 Total bytes allocated: 173,088 Total bytes freed: 161,280 Net total bytes allocated: 11,808 Total bytes wasted on internal fragmentation: 460 Internal fragmentation: 0.265761% Cross CPU allocations: 0/851 # # # stats for an end open interval, after a certain time: # # perf kmem stat --time 20119.782088, SUMMARY (SLAB allocator) ======================== Total bytes requested: 552 Total bytes allocated: 552 Total bytes freed: 448 Net total bytes allocated: 104 Total bytes wasted on internal fragmentation: 0 Internal fragmentation: 0.000000% Cross CPU allocations: 0/8 # Signed-off-by: -
David Ahern authored
Add option to allow user to control analysis window. e.g., collect data for time window and analyze a segment of interest within that window. Committer notes: Testing it: # perf sched record -a usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.593 MB perf.data (25 samples) ] # # perf sched timehist | head -18 Samples do not have callchains. time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) ------------- ------ --------------- --------- --------- -------- 19818.635579 [0002] <idle> 0.000 0.000 0.000 19818.635613 [0000] perf[9116] 0.000 0.000 0.000 19818.635676 [0000] <idle> 0.000 0.000 0.063 19818.635678 [0000] rcuos/2[29] 0.000 0.002 0.001 19818.635696 [0002] perf[9117] 0.000 0.004 0.116 19818.635702 [0000] <idle> 0.001 0.000 0.024 19818.635709 [0002] migration/2[25] 0.000 0.003 0.012 19818.636263 [0000] usleep[9117] 0.005 0.000 0.560 19818.636316 [0000] <idle> 0.560 0.000 0.053 19818.636358 [0002] <idle> 0.129 0.000 0.649 19818.636358 [0000] usleep[9117] 0.053 0.002 0.042 # # perf sched timehist --time 19818.635696, Samples do not have callchains. time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) ------------- ------ --------------- -------- --------- --------- 19818.635696 [0002] perf[9117] 0.000 0.120 0.000 19818.635702 [0000] <idle> 0.019 0.000 0.006 19818.635709 [0002] migration/2[25] 0.000 0.003 0.012 19818.636263 [0000] usleep[9117] 0.005 0.000 0.560 19818.636316 [0000] <idle> 0.560 0.000 0.053 19818.636358 [0002] <idle> 0.129 0.000 0.649 19818.636358 [0000] usleep[9117] 0.053 0.002 0.042 # # perf sched timehist --time 19818.635696,19818.635709 Samples do not have callchains. time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) ------------- ------ --------------- --------- --------- --------- 19818.635696 [0002] perf[9117] 0.000 0.120 0.000 19818.635702 [0000] <idle> 0.019 0.000 0.006 19818.635709 [0002] migration/2[25] 0.000 0.003 0.012 19818.635709 [0000] usleep[9117] 0.005 0.000 0.006 # Signed-off-by: -
David Ahern authored
Add option to allow user to control analysis window. e.g., collect data for some amount of time and analyze a segment of interest within that window. Committer notes: Testing it: # perf evlist -v cycles:ppp: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|CPU|PERIOD, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1 # # perf script --hide-call-graph | head -15 swapper 0 [0] 9693.370039: 1 cycles:ppp: ffffffffb90072ad x86_pmu_enable (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370044: 1 cycles:ppp: ffffffffb900ca1b intel_pmu_handle_irq (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370046: 7 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370048: 126 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370049: 2701 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370051: 58823 cycles:ppp: ffffffffb90cd2e0 idle_cpu (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370059: 1 cycles:ppp: ffffffffb91a713a ctx_resched (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370062: 1 cycles:ppp: ffffffffb900ca1b intel_pmu_handle_irq (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370064: 13 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370065: 250 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370067: 5269 cycles:ppp: ffffffffb902fe79 sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370069: 114602 cycles:ppp: ffffffffb90c1c5a atomic_notifier_call_chain (.../4.8.8-300.fc25.x86_64/vmlinux) perf 5124 [2] 9693.370076: 1 cycles:ppp: ffffffffb91a76c1 __perf_event_enable (.../4.8.8-300.fc25.x86_64/vmlinux) perf 5124 [2] 9693.370091: 1 cycles:ppp: ffffffffb900ca1b intel_pmu_handle_irq (.../4.8.8-300.fc25.x86_64/vmlinux) perf 5124 [2] 9693.370095: 3 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) # # perf script --hide-call-graph --time ,9693.370048 swapper 0 [0] 9693.370039: 1 cycles:ppp: ffffffffb90072ad x86_pmu_enable (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370044: 1 cycles:ppp: ffffffffb900ca1b intel_pmu_handle_irq (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [0] 9693.370046: 7 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) # perf script --hide-call-graph --time 9693.370064,9693.370076 swapper 0 [1] 9693.370064: 13 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370065: 250 cycles:ppp: ffffffffb902fd93 native_sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370067: 5269 cycles:ppp: ffffffffb902fe79 sched_clock (.../4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [1] 9693.370069: 114602 cycles:ppp: ffffffffb90c1c5a atomic_notifier_call_chain (.../4.8.8-300.fc25.x86_64/vmlinux) # Signed-off-by: -
David Ahern authored
Code move only; no functional change intended. Committer notes: Fix the build on Ubuntu 16.04 x86-64 cross-compiling to S/390, with this set of auto-detected features: ... dwarf: [ on ] ... dwarf_getlocations: [ on ] ... glibc: [ on ] ... gtk2: [ OFF ] ... libaudit: [ OFF ] ... libbfd: [ OFF ] ... libelf: [ on ] ... libnuma: [ OFF ] ... numa_num_possible_cpus: [ OFF ] ... libperl: [ OFF ] ... libpython: [ OFF ] ... libslang: [ OFF ] ... libcrypto: [ OFF ] ... libunwind: [ OFF ] ... libdw-dwarf-unwind: [ on ] ... zlib: [ on ] ... lzma: [ OFF ] ... get_cpuid: [ OFF ] ... bpf: [ on ] Where it was failing with: CC /tmp/build/perf/util/time-utils.o util/time-utils.c: In function 'parse_nsec_time': util/time-utils.c:17:13: error: implicit declaration of function 'strtoul' [-Werror=implicit-function-declaration] time_sec = strtoul(str, &end, 10); ^ util/time-utils.c:17:2: error: nested extern declaration of 'strtoul' [-Werror=nested-externs] time_sec = strtoul(str, &end, 10); ^ util/time-utils.c: In function 'perf_time__parse_str': util/time-utils.c:93:2: error: implicit declaration of function 'free' [-Werror=implicit-function-declaration] free(str); ^ util/time-utils.c:93:2: error: incompatible implicit declaration of built-in function 'free' [-Werror] util/time-utils.c:93:2: note: include '<stdlib.h>' or provide a declaration of 'free' Do as suggested and add a '#include <stdlib.h>' to get the free() and strtoul() declarations and fix the build. Signed-off-by: -
David Ahern authored
Add function to parse a user time string of the form <start>,<stop> where start and stop are time in sec.nsec format. Both start and stop times are optional. Add function to determine if a sample time is within a given time time window of interest. Signed-off-by:
-
- 29 Nov, 2016 7 commits
-
-
David Ahern authored
Allow user to specify list of symbols which cause the dump of callchains to stop at that symbol. Committer notes: Testing it: # perf record -ag usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.177 MB perf.data (33 samples) ] # # # Without it: # # perf script swapper 0 [000] 9693.370039: 1 cycles:ppp: 2072ad x86_pmu_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a29d7 perf_pmu_enable.part.90 (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a713a ctx_resched (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a76c1 __perf_event_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a0390 event_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a1cff remote_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 326978 flush_smp_call_function_queue (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 327413 generic_smp_call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 249b37 smp_call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) a04b2c call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 889427 cpuidle_enter (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 2e534a call_cpuidle (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 2e5730 cpu_startup_entry (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 9f5167 rest_init (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 137ffeb start_kernel ([kernel.vmlinux].init.text) 137f2ca x86_64_start_reservations ([kernel.vmlinux].init.text) 137f419 x86_64_start_kernel ([kernel.vmlinux].init.text) swapper 0 [000] 9693.370044: 1 cycles:ppp: 20ca1b intel_pmu_handle_irq (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 205b0c perf_event_nmi_handler (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a14a nmi_handle (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a6b3 default_do_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a83c do_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) a03fb1 end_repeat_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a29d7 perf_pmu_enable.part.90 (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a713a ctx_resched (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a76c1 __perf_event_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a0390 event_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a1cff remote_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 326978 flush_smp_call_function_queue (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 327413 generic_smp_call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 249b37 smp_call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) a04b2c call_function_single_interrupt (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 889427 cpuidle_enter (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 2e534a call_cpuidle (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 2e5730 cpu_startup_entry (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 9f5167 rest_init (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 137ffeb start_kernel ([kernel.vmlinux].init.text) 137f2ca x86_64_start_reservations ([kernel.vmlinux].init.text) # # # Using it to see just what are the calls from the 'remote_function' function: # # perf script --stop-bt remote_function swapper 0 [000] 9693.370039: 1 cycles:ppp: 2072ad x86_pmu_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a29d7 perf_pmu_enable.part.90 (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a713a ctx_resched (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a76c1 __perf_event_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a0390 event_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a1cff remote_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) swapper 0 [000] 9693.370044: 1 cycles:ppp: 20ca1b intel_pmu_handle_irq (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 205b0c perf_event_nmi_handler (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a14a nmi_handle (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a6b3 default_do_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 22a83c do_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) a03fb1 end_repeat_nmi (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a29d7 perf_pmu_enable.part.90 (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a713a ctx_resched (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a76c1 __perf_event_enable (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a0390 event_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) 3a1cff remote_function (/usr/lib/debug/lib/modules/4.8.8-300.fc25.x86_64/vmlinux) Signed-off-by:David Ahern <dsa@cumulusnetworks.com> Tested-by:
-
David Ahern authored
Track freed memory as well as allocations and show the net in the summary. Committer notes: Testing it: # perf kmem record usleep 1 [ perf record: Woken up 0 times to write data ] [ perf record: Captured and wrote 1.626 MB perf.data (4208 samples) ] [root@jouet ~]# perf kmem stat --slab SUMMARY (SLAB allocator) ======================== Total bytes requested: 234,011 Total bytes allocated: 234,504 Total bytes freed: 213,328 <------ Net total bytes allocated: 21,176 Total bytes wasted on internal fragmentation: 493 Internal fragmentation: 0.210231% Cross CPU allocations: 4/1,963 # Signed-off-by:
-
Arnaldo Carvalho de Melo authored
Having "test" in almost all test descriptions is redundant, simplify it removing and rewriting tests with such descriptions. End result: # perf test 1: vmlinux symtab matches kallsyms : Ok 2: Detect openat syscall event : Ok 3: Detect openat syscall event on all cpus : Ok 4: Read samples using the mmap interface : Ok 5: Parse event definition strings : Ok 6: PERF_RECORD_* events & perf_sample fields : Ok 7: Parse perf pmu format : Ok 8: DSO data read : Ok 9: DSO data cache : Ok 10: DSO data reopen : Ok 11: Roundtrip evsel->name : Ok 12: Parse sched tracepoints fields : Ok 13: syscalls:sys_enter_openat event fields : Ok 14: Setup struct perf_event_attr : Ok 15: Match and link multiple hists : Ok 16: 'import perf' in python : Ok 17: Breakpoint overflow signal handler : Ok 18: Breakpoint overflow sampling : Ok 19: Number of exit events of a simple workload : Ok 20: Software clock events period values : Ok 21: Object code reading : Ok 22: Sample parsing : Ok 23: Use a dummy software event to keep tracking: Ok 24: Parse with no sample_id_all bit set : Ok 25: Filter hist entries : Ok 26: Lookup mmap thread : Ok 27: Share thread mg : Ok 28: Sort output of hist entries : Ok 29: Cumulate child hist entries : Ok 30: Track with sched_switch : Ok 31: Filter fds with revents mask in a fdarray : Ok 32: Add fd to a fdarray, making it autogrow : Ok 33: kmod_path__parse : Ok 34: Thread map : Ok 35: LLVM search and compile : 35.1: Basic BPF llvm compile : Ok 35.2: kbuild searching : Ok 35.3: Compile source for BPF prologue generation: Ok 35.4: Compile source for BPF relocation : Ok 36: Session topology : Ok 37: BPF filter : 37.1: Basic BPF filtering : Ok 37.2: BPF prologue generation : Ok 37.3: BPF relocation checker : Ok 38: Synthesize thread map : Ok 39: Synthesize cpu map : Ok 40: Synthesize stat config : Ok 41: Synthesize stat : Ok 42: Synthesize stat round : Ok 43: Synthesize attr update : Ok 44: Event times : Ok 45: Read backward ring buffer : Ok 46: Print cpu map : Ok 47: Probe SDT events : Ok 48: is_printable_array : Ok 49: Print bitmap : Ok 50: perf hooks : Ok 51: x86 rdpmc : Ok 52: Convert perf time to TSC : Ok 53: DWARF unwind : Ok 54: x86 instruction decoder - new instructions : Ok 55: Intel cqm nmi context read : Skip # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-rx2lbfcrrio2yx1fxcljqy0e@git.kernel.orgSigned-off-by:
-
Wang Nan authored
Perf hooks allow hooking user code at perf events. They can be used for manipulation of BPF maps, taking snapshot and reporting results. In this patch two perf hook points are introduced: record_start and record_end. To avoid buggy user actions, a SIGSEGV signal handler is introduced into 'perf record'. It turns off perf hook if it causes a segfault and report an error to help debugging. A test case for perf hook is introduced. Test result: $ ./buildperf/perf test -v hook 50: Test perf hooks : --- start --- test child forked, pid 10311 SIGSEGV is observed as expected, try to recover. Fatal error (SEGFAULT) in perf hook 'test' test child finished with 0 ---- end ---- Test perf hooks: Ok Signed-off-by:
Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@fb.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Joe Stringer <joe@ovn.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/20161126070354.141764-5-wangnan0@huawei.comSigned-off-by:
-
Wang Nan authored
Add a new API to libbpf, caller is able to get bpf_map through the offset of bpf_map_def to 'maps' section. The API will be used to help jitted perf hook code find fd of a map. Signed-off-by:
Alexei Starovoitov <ast@kernel.org> Cc: He Kuang <hekuang@huawei.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Joe Stringer <joe@ovn.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/20161126070354.141764-4-wangnan0@huawei.comSigned-off-by:
-
Wang Nan authored
Similar to other classes defined in libbpf.h (map and program), allow 'object' class has its own private data. Signed-off-by:
-
Wang Nan authored
Add more BPF map operations to libbpf. Also add bpf_obj_{pin,get}(). They can be used on not only BPF maps but also BPF programs. Signed-off-by:
-
- 25 Nov, 2016 17 commits
-
-
David Ahern authored
Leverage pid/tid filtering done by symbol_conf hooks. Signed-off-by:
-
David Ahern authored
Add handlers for sched:sched_migrate_task event. Total number of migrations is added to summary display and -M/--migrations can be used to show migration events. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
To help in debugging when the wrong offset is being used, like in: │13d98: ↓ jne 13dd1 <lzma_lzma_preset@@XZ_5.0+0x28e1> That is the full line from objdump, and it seems what should be used is 13dd1, not 28e1. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-4nc0marsgst1ft6inmvqber7@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
To print some values, like in the annotation code with invalid jump offsets. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-1vk0g5twas2ioswn1mmvnvwq@git.kernel.orgSigned-off-by:
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-20161125' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: New features: - Improve ARM support in the annotation code, affecting 'perf annotate', 'perf report' and live annotation in 'perf top' (Kim Phillips) - Initial support for PowerPC in the annotation code (Ravi Bangoria) - Skip repetitive scheduler function on the top of the stack in 'perf sched timehist' (Namhyung Kim) Fixes: - Fix maps resolution in libbpf (Eric Leblond) - Get the kernel signature via /proc/version_signature, available on Ubuntu systems, to make sure BPF proggies works, as the one provided via 'uname -r' doesn't (Wang Nan) - Fix segfault in 'perf record' when running with suid and kptr_restrict is 1 (Wang Nan) Infrastructure changes: - Support per-arch instruction tables, kept via a static or dynamic table (Arnaldo Carvalho de Melo) Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
Eric Leblond authored
It is not correct to assimilate the elf data of the maps section to an array of map definition. In fact the sizes differ. The offset provided in the symbol section has to be used instead. This patch fixes a bug causing a elf with two maps not to load correctly. Wang Nan added: This patch requires a name for each BPF map, so array of BPF maps is not allowed. This restriction is reasonable, because kernel verifier forbid indexing BPF map from such array unless the index is a fixed value, but if the index is fixed why not merging it into name? For example: Program like this: ... unsigned long cpu = get_smp_processor_id(); int *pval = map_lookup_elem(&map_array[cpu], &key); ... Generates bytecode like this: 0: (b7) r1 = 0 1: (63) *(u32 *)(r10 -4) = r1 2: (b7) r1 = 680997 3: (63) *(u32 *)(r10 -8) = r1 4: (85) call 8 5: (67) r0 <<= 4 6: (18) r1 = 0x112dd000 8: (0f) r0 += r1 9: (bf) r2 = r10 10: (07) r2 += -4 11: (bf) r1 = r0 12: (85) call 1 Where instruction 8 is the computation, 8 and 11 render r1 to an invalid value for function map_lookup_elem, causes verifier report error. Signed-off-by:
Eric Leblond <eric@regit.org> Cc: Alexei Starovoitov <ast@fb.com> Cc: He Kuang <hekuang@huawei.com> Cc: Wang Nan <wangnan0@huawei.com> [ Merge bpf_object__init_maps_name into bpf_object__init_maps. Fix segfault for buggy BPF script Validate obj->maps ] Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/20161115040617.69788-5-wangnan0@huawei.comSigned-off-by:
-
Wang Nan authored
Commit 0b3c2264 ("perf symbols: Fix kallsyms perf test on ppc64le") refers struct symbol in probe_event.h, but forgets to include its definition. Gcc will complain about it when that definition is not added, by sheer luck, by some other header included before probe_event.h. Signed-off-by:
-
Wang Nan authored
Before this patch perf panics if kptr_restrict is set to 1 and perf is owned by root with suid set: $ whoami wangnan $ ls -l ./perf -rwsr-xr-x 1 root root 19781908 Sep 21 19:29 /home/wangnan/perf $ cat /proc/sys/kernel/kptr_restrict 1 $ cat /proc/sys/kernel/perf_event_paranoid -1 $ ./perf record -a Segmentation fault (core dumped) $ The reason is that perf assumes it is allowed to read kptr from /proc/kallsyms when euid is root, but in fact the kernel doesn't allow reading kptr when euid and uid do not match with each other: $ cp /bin/cat . $ sudo chown root:root ./cat $ sudo chmod u+s ./cat $ cat /proc/kallsyms | grep do_fork 0000000000000000 T _do_fork <--- kptr is hidden even euid is root $ sudo cat /proc/kallsyms | grep do_fork ffffffff81080230 T _do_fork See lib/vsprintf.c for kernel side code. This patch fixes this problem by checking both uid and euid. Signed-off-by:
-
Wang Nan authored
On ubuntu the internal kernel version code is different from what can be retrived from uname: $ uname -r 4.4.0-47-generic $ cat /lib/modules/`uname -r`/build/include/generated/uapi/linux/version.h #define LINUX_VERSION_CODE 263192 #define KERNEL_VERSION(a,b,c) (((a) << 16) + ((b) << 8) + (c)) $ cat /lib/modules/`uname -r`/build/include/generated/utsrelease.h #define UTS_RELEASE "4.4.0-47-generic" #define UTS_UBUNTU_RELEASE_ABI 47 $ cat /proc/version_signature Ubuntu 4.4.0-47.68-generic 4.4.24 The macro LINUX_VERSION_CODE is set to 4.4.24 (263192 == 0x40418), but `uname -r` reports 4.4.0. This mismatch causes LINUX_VERSION_CODE macro passed to BPF script become an incorrect value, results in magic failure in BPF loading: $ sudo ./buildperf/perf record -e ./tools/perf/tests/bpf-script-example.c ls event syntax error: './tools/perf/tests/bpf-script-example.c' \___ Failed to load program for unknown reason According to Ubuntu document (https://wiki.ubuntu.com/Kernel/FAQ), the correct kernel version can be retrived through /proc/version_signature, which is ubuntu specific. This patch checks the existance of /proc/version_signature, and returns version number through parsing this file instead of uname. Version string is untouched (value returns from uname) because `uname -r` is required to be consistence with path of kbuild directory in /lib/module. Signed-off-by: -
Namhyung Kim authored
When it records callchains, they will always have 2 scheduler functions (__schedule + schedule or __schedule + preempt_schedule) and get ignored. So it should collect 2 more functions to show the expected number of callchains to user. Committer Notes: Example of final result, using the same perf.data file as in the previous cset comment, but this time redirecting the output of 'perf sched timehist' to a file instead of copy'n'pasting from xterm: [root@jouet experimental]# perf sched timehist > /tmp/bla [root@jouet experimental]# cat /tmp/bla time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) -------- ---- -------------------- ------ ------ ----- 6.494998 [01] <idle> 0.000 0.000 0.000 6.495027 [02] perf[519] 0.000 0.000 0.000 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_timeout <- do_sys_poll <- sys_poll 6.495096 [03] <idle> 0.000 0.000 0.000 6.495100 [03] rcuos/0[9] 0.000 0.005 0.003 rcu_nocb_kthread <- kthread <- ret_from_fork 6.495113 [01] perf[520] 0.000 0.008 0.114 preempt_schedule_common <- _cond_resched <- wait_for_completion <- stop_one_cpu <- sched_exec <- do_execveat_common.isra.35 6.495121 [00] <idle> 0.000 0.000 0.000 6.495129 [01] migration/1[17] 0.000 0.003 0.016 smpboot_thread_fn <- kthread <- ret_from_fork 6.496085 [02] <idle> 0.000 0.000 1.057 6.496096 [02] kworker/u16:1[31169] 0.000 0.004 0.011 worker_thread <- kthread <- ret_from_fork 6.496096 [03] <idle> 0.003 0.000 0.996 6.496169 [02] <idle> 0.011 0.000 0.072 6.496171 [00] ls[520] 0.008 0.000 1.049 do_exit <- do_group_exit <- [unknown] <- entry_SYSCALL_64_fastpath 6.496172 [03] gnome-terminal-[4391] 0.000 0.003 0.076 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_timeout <- do_sys_poll <- sys_poll Signed-off-by: -
Namhyung Kim authored
The sched_switch event always captured from the scheduler function. So it'd be great omit them from the callchain. This patch marks the functions to be omitted by later patch. Committer notes: Testing it: Before: [root@jouet experimental]# perf sched record -g ls Dockerfile perf.data x-mips64 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.355 MB perf.data (29 samples) ] [root@jouet experimental]# perf sched timehist time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) ----------- ----- ----------------- ------ ------ ------ 6.494998 [001] <idle> 0.000 0.000 0.000 6.495027 [002] perf[519] 0.000 0.000 0.000 __schedule <- schedule <- schedule_hrtimeout_range_clock <- schedule_hrtimeou 6.495096 [003] <idle> 0.000 0.000 0.000 6.495100 [003] rcuos/0[9] 0.000 0.005 0.003 __schedule <- schedule <- rcu_nocb_kthread <- kthread <- ret_from_fork 6.495113 [001] perf[520] 0.000 0.008 0.114 __schedule <- preempt_schedule_common <- _cond_resched <- wait_for_completion 6.495121 [000] <idle> 0.000 0.000 0.000 6.495129 [001] migration/1[17] 0.000 0.003 0.016 __schedule <- schedule <- smpboot_thread_fn <- kthread <- ret_from_fork 6.496085 [002] <idle> 0.000 0.000 1.057 6.496096 [002] kworker/u16:1[31169] 0.000 0.004 0.011 __schedule <- schedule <- worker_thread <- kthread <- ret_from_fork 6.496096 [003] <idle> 0.003 0.000 0.996 6.496169 [002] <idle> 0.011 0.000 0.072 6.496171 [000] ls[520] 0.008 0.000 1.049 __schedule <- schedule <- do_exit <- do_group_exit <- [unknown] 6.496172 [003] gnome-terminal-[4391] 0.000 0.003 0.076 __schedule <- schedule <- schedule_hrtimeout_range_clock <- schedule_hrtimeo After: [root@jouet experimental]# perf sched timehist time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) ----------- ----- ----------------- ----- ----- ------ 6.494998 [001] <idle> 0.000 0.000 0.000 6.495027 [002] perf[519] 0.000 0.000 0.000 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_t 6.495096 [003] <idle> 0.000 0.000 0.000 6.495100 [003] rcuos/0[9] 0.000 0.005 0.003 rcu_nocb_kthread <- kthread <- ret_from_fork 6.495113 [001] perf[520] 0.000 0.008 0.114 preempt_schedule_common <- _cond_resched <- wait_for_completion <- stop_one_c 6.495121 [000] <idle> 0.000 0.000 0.000 6.495129 [001] migration/1[17] 0.000 0.003 0.016 smpboot_thread_fn <- kthread <- ret_from_fork 6.496085 [002] <idle> 0.000 0.000 1.057 6.496096 [002] kworker/u16:1[31169] 0.000 0.004 0.011 worker_thread <- kthread <- ret_from_fork 6.496096 [003] <idle> 0.003 0.000 0.996 6.496169 [002] <idle> 0.011 0.000 0.072 6.496171 [000] ls[520] 0.008 0.000 1.049 do_exit <- do_group_exit <- [unknown] 6.496172 [003] gnome-terminal-[4391] 0.000 0.003 0.076 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_ [root@jouet experimental]# Signed-off-by: -
Namhyung Kim authored
For tracepoint events, callchains always contain certain functions. Sometimes it'd be better to skip those functions as they have no value. Signed-off-by:
-
Ravi Bangoria authored
Support the PowerPC architecture using the ins_ops association method. Committer notes: Testing it with a perf.data file collected on a PowerPC machine and cross-annotated on a x86_64 workstation, using the associated vmlinux file: $ perf report -i perf.data.f22vm.powerdev --vmlinux vmlinux.powerpc .ktime_get vmlinux.powerpc │ clrldi r9,r28,63 8.57 │ ┌──bne e0 <- TUI cursor positioned here │54:│ lwsync 2.86 │ │ std r2,40(r1) │ │ ld r9,144(r31) │ │ ld r3,136(r31) │ │ ld r30,184(r31) │ │ ld r10,0(r9) │ │ mtctr r10 │ │ ld r2,8(r9) 8.57 │ │→ bctrl │ │ ld r2,40(r1) │ │ ld r10,160(r31) │ │ ld r5,152(r31) │ │ lwz r7,168(r31) │ │ ld r9,176(r31) 8.57 │ │ lwz r6,172(r31) │ │ lwsync 2.86 │ │ lwz r8,128(r31) │ │ cmpw cr7,r8,r28 2.86 │ │↑ bne 48 │ │ subf r10,r10,r3 │ │ mr r3,r29 │ │ and r10,r10,r5 2.86 │ │ mulld r10,r10,r7 │ │ add r9,r10,r9 │ │ srd r9,r9,r6 │ │ add r9,r9,r30 │ │ std r9,0(r29) │ │ addi r1,r1,144 │ │ ld r0,16(r1) │ │ ld r28,-32(r1) │ │ ld r29,-24(r1) │ │ ld r30,-16(r1) │ │ mtlr r0 │ │ ld r31,-8(r1) │ │← blr 5.71 │e0:└─→mr r1,r1 11.43 │ mr r2,r2 11.43 │ lwz r28,128(r31) Press 'h' for help on key bindings $ perf report -i perf.data.f22vm.powerdev --header-only # ======== # captured on: Thu Nov 24 12:40:38 2016 # hostname : pdev-f22-qemu # os release : 4.4.10-200.fc22.ppc64 # perf version : 4.9.rc1.g6298ce # arch : ppc64 # nrcpus online : 48 # nrcpus avail : 48 # cpudesc : POWER7 (architected), altivec supported # cpuid : 74,513 # total memory : 4158976 kB # cmdline : /home/ravi/Workspace/linux/tools/perf/perf record -a # event : name = cycles:ppp, , size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1 # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu = 4, software = 1, tracepoint = 2, breakpoint = 5 # missing features: HEADER_TRACING_DATA HEADER_BRANCH_STACK HEADER_GROUP_DESC HEADER_AUXTRACE HEADER_STAT HEADER_CACHE # ======== # $ Signed-off-by:Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Signed-off-by:
Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Tested-by:
-
Arnaldo Carvalho de Melo authored
By using arch->init() to set up some regular expressions to associate ins_ops to ARM instructions, ditching that old table that has instructions not present on ARM. Take advantage of having an arch->init() to hide more arm specific stuff from the common code, like the objdump details. The regular expressions comes from a patch written by Kim Phillips. Reviewed-by:
-
Arnaldo Carvalho de Melo authored
Arches like ARM will want to use regular expressions when deciding what instructions to associate with what ins_ops, provide infrastructure for that. Reviewed-by:
-
Arnaldo Carvalho de Melo authored
Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by:
-
Arnaldo Carvalho de Melo authored
The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by:
-
- 24 Nov, 2016 2 commits
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-20161123' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: New tool: - 'perf sched timehist' provides an analysis of scheduling events. Example usage: perf sched record -- sleep 1 perf sched timehist By default it shows the individual schedule events, including the wait time (time between sched-out and next sched-in events for the task), the task scheduling delay (time between wakeup and actually running) and run time for the task: time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) -------- ------ ---------------- --------- --------- -------- 1.874569 [0011] gcc[31949] 0.014 0.000 1.148 1.874591 [0010] gcc[31951] 0.000 0.000 0.024 1.874603 [0010] migration/10[59] 3.350 0.004 0.011 1.874604 [0011] <idle> 1.148 0.000 0.035 1.874723 [0005] <idle> 0.016 0.000 1.383 1.874746 [0005] gcc[31949] 0.153 0.078 0.022 ... Times are in msec.usec. (David Ahern, Namhyung Kim) Improvements: - Make 'perf c2c report' support -f/--force, to allow skipping the ownership check for root users, for instance, just like the other tools (Jiri Olsa) - Allow sorting cachelines by total number of HITMs, in addition to local and remote numbers (Jiri Olsa) Fixes: - Make sure errors aren't suppressed by the TUI reset at the end of a 'perf c2c report' session (Jiri Olsa) Infrastructure changes: - Initial work on having the annotate code better support multiple architectures, including the ability to cross-annotate, i.e. to annotate perf.data files collected on an ARM system on a x86_64 workstation (Arnaldo Carvalho de Melo, Ravi Bangoria, Kim Phillips) - Use USECS_PER_SEC instead of hard coded number in libtraceevent (Steven Rostedt) - Add retrieval of preempt count and latency flags in libtraceevent (Steven Rostedt) Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
- 23 Nov, 2016 6 commits
-
-
git://git.linux-nfs.org/projects/anna/linux-nfsLinus Torvalds authored
Pull NFS client bugfixes from Anna Schumaker: "Most of these fix regressions or races, but there is one patch for stable that Arnd sent me Stable bugfix: - Hide array-bounds warning Bugfixes: - Keep a reference on lock states while checking - Handle NFS4ERR_OLD_STATEID in nfs4_reclaim_open_state - Don't call close if the open stateid has already been cleared - Fix CLOSE rases with OPEN - Fix a regression in DELEGRETURN" * tag 'nfs-for-4.9-4' of git://git.linux-nfs.org/projects/anna/linux-nfs: NFSv4.x: hide array-bounds warning NFSv4.1: Keep a reference on lock states while checking NFSv4.1: Handle NFS4ERR_OLD_STATEID in nfs4_reclaim_open_state NFSv4: Don't call close if the open stateid has already been cleared NFSv4: Fix CLOSE races with OPEN NFSv4.1: Fix a regression in DELEGRETURN
-
git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tileLinus Torvalds authored
Pull arch/tile bugfix from Chris Metcalf: "This fixes a bug that causes reboots after 208 days of uptime :-)" * 'stable' of git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile: tile: avoid using clocksource_cyc2ns with absolute cycle count
-
Chris Metcalf authored
For large values of "mult" and long uptimes, the intermediate result of "cycles * mult" can overflow 64 bits. For example, the tile platform calls clocksource_cyc2ns with a 1.2 GHz clock; we have mult = 853, and after 208.5 days, we overflow 64 bits. Since clocksource_cyc2ns() is intended to be used for relative cycle counts, not absolute cycle counts, performance is more importance than accepting a wider range of cycle values. So, just use mult_frac() directly in tile's sched_clock(). Commit 4cecf6d4 ("sched, x86: Avoid unnecessary overflow in sched_clock") by Salman Qazi results in essentially the same generated code for x86 as this change does for tile. In fact, a follow-on change by Salman introduced mult_frac() and switched to using it, so the C code was largely identical at that point too. Peter Zijlstra then added mul_u64_u32_shr() and switched x86 to use it. This is, in principle, better; by optimizing the 64x64->64 multiplies to be 32x32->64 multiplies we can potentially save some time. However, the compiler piplines the 64x64->64 multiplies pretty well, and the conditional branch in the generic mul_u64_u32_shr() causes some bubbles in execution, with the result that it's pretty much a wash. If tilegx provided its own implementation of mul_u64_u32_shr() without the conditional branch, we could potentially save 3 cycles, but that seems like small gain for a fair amount of additional build scaffolding; no other platform currently provides a mul_u64_u32_shr() override, and tile doesn't currently have an <asm/div64.h> header to put the override in. Additionally, gcc currently has an optimization bug that prevents it from recognizing the opportunity to use a 32x32->64 multiply, and so the result would be no better than the existing mult_frac() until such time as the compiler is fixed. For now, just using mult_frac() seems like the right answer. Cc: stable@kernel.org [v3.4+] Signed-off-by:
Chris Metcalf <cmetcalf@mellanox.com>
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull perf fixes from Ingo Molnar: "Six fixes for bugs that were found via fuzzing, and a trivial hw-enablement patch for AMD Family-17h CPU PMUs" * 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: perf/x86/intel/uncore: Allow only a single PMU/box within an events group perf/x86/intel: Cure bogus unwind from PEBS entries perf/x86: Restore TASK_SIZE check on frame pointer perf/core: Fix address filter parser perf/x86: Add perf support for AMD family-17h processors perf/x86/uncore: Fix crash by removing bogus event_list[] handling for SNB client uncore IMC perf/core: Do not set cpuctx->cgrp for unscheduled cgroups
-
David Ahern authored
The -V option provides a visual aid for sched switches by cpu: $ perf sched timehist -V time cpu 0123456789abc task name b/n time sch delay run time [tid/pid] (msec) (msec) (msec) --------------- ------ ------------- -------------------- --------- --------- --------- ... 2412598.429696 [0009] i <idle> 0.000 0.000 0.000 2412598.429767 [0002] s perf[7219] 0.000 0.000 0.000 2412598.429783 [0009] s perf[7220] 0.000 0.006 0.087 2412598.429794 [0010] i <idle> 0.000 0.000 0.000 2412598.429795 [0009] s migration/9[53] 0.000 0.003 0.011 2412598.430370 [0010] s sleep[7220] 0.011 0.000 0.576 2412598.432584 [0003] i <idle> 0.000 0.000 0.000 ... Committer notes: 'i' marks idle time, 's' are scheduler events. Signed-off-by:Jiri Olsa <jolsa@kernel.org> Tested-by:
-
David Ahern authored
If callchains were recorded they are appended to the line with a default stack depth of 5: 1.874569 [0011] gcc[31949] 0.014 0.000 1.148 wait_for_completion_killable <- do_fork <- sys_vfork <- stub_vfork <- __vfork 1.874591 [0010] gcc[31951] 0.000 0.000 0.024 __cond_resched <- _cond_resched <- wait_for_completion <- stop_one_cpu <- sched_exec 1.874603 [0010] migration/10[59] 3.350 0.004 0.011 smpboot_thread_fn <- kthread <- ret_from_fork 1.874604 [0011] <idle> 1.148 0.000 0.035 cpu_startup_entry <- start_secondary 1.874723 [0005] <idle> 0.016 0.000 1.383 cpu_startup_entry <- start_secondary 1.874746 [0005] gcc[31949] 0.153 0.078 0.022 do_wait sys_wait4 <- system_call_fastpath <- __GI___waitpid --no-call-graph can be used to not show the callchains. --max-stack is used to control the number of frames shown (default of 5). -x/--excl options can be used to collapse redundant callchains to get more relevant data on screen. Signed-off-by:
-
