aboutsummaryrefslogtreecommitdiff
path: root/tools/perf
AgeCommit message (Collapse)AuthorFilesLines
2021-06-18perf probe: Support probes on init functions for offline kernelMasami Hiramatsu1-2/+7
'perf probe' internally checks the probe target is in the text area in post-process (after analyzing debuginfo). But it fails if the probe target is in the "inittext". This is a good limitation for the online kernel because such functions have gone after booting. However, for using it for boot-time tracing, user may want to put a probe on init functions. This skips the post checking process if the target is offline kenrel so that user can get the probe definition on the init functions. Without this patch: $ perf probe -k ./build-x86_64/vmlinux -D do_mount_root:10 Probe point 'do_mount_root:10' not found. Error: Failed to add events. With this patch: $ perf probe -k ./build-x86_64/vmlinux -D do_mount_root:10 p:probe/do_mount_root_L10 mount_block_root+300 Signed-off-by: Masami Hiramatsu <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/162282410293.452340.13347006295826431632.stgit@devnote2 Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-18perf test: Make stat bpf counters test more robustIan Rogers1-0/+8
If the test is run on a hypervisor then the cycles event may not be counted, skip the test in this situation. Fail the test if cycles are not counted in the subsequent bpf counter run. Signed-off-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Song Liu <[email protected]> Cc: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-18perf test: Add verbose skip output for bpf countersIan Rogers1-1/+7
Provide additional context for when the stat bpf counters test skips. Signed-off-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Song Liu <[email protected]> Cc: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-16perf annotate: Add itrace options supportYang Jihong2-0/+18
The "auxtrace_info" and "auxtrace" functions are not set in "tool" member of "annotate". As a result, perf annotate does not support parsing itrace data. Before: # perf record -e arm_spe_0/branch_filter=1/ -a sleep 1 [ perf record: Woken up 9 times to write data ] [ perf record: Captured and wrote 20.874 MB perf.data ] # perf annotate --stdio Error: The perf.data data has no samples! Solution: 1. Add itrace options in help, 2. Set hook functions of "id_index", "auxtrace_info" and "auxtrace" in perf_tool. After: # perf record --all-user -e arm_spe_0/branch_filter=1/ ls Couldn't synthesize bpf events. perf.data [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.010 MB perf.data ] # perf annotate --stdio Percent | Source code & Disassembly of libc-2.28.so for branch-miss (1 samples, percent: local period) ------------------------------------------------------------------------------------------------------------ : : : : Disassembly of section .text: : : 0000000000066180 <__getdelim@@GLIBC_2.17>: 0.00 : 66180: stp x29, x30, [sp, #-96]! 0.00 : 66184: cmp x0, #0x0 0.00 : 66188: ccmp x1, #0x0, #0x4, ne // ne = any 0.00 : 6618c: mov x29, sp 0.00 : 66190: stp x24, x25, [sp, #56] 0.00 : 66194: stp x26, x27, [sp, #72] 0.00 : 66198: str x28, [sp, #88] 0.00 : 6619c: b.eq 66450 <__getdelim@@GLIBC_2.17+0x2d0> // b.none 0.00 : 661a0: stp x22, x23, [x29, #40] 0.00 : 661a4: mov x22, x1 0.00 : 661a8: ldr w1, [x3] 0.00 : 661ac: mov w23, w2 0.00 : 661b0: stp x20, x21, [x29, #24] 0.00 : 661b4: mov x20, x3 0.00 : 661b8: mov x21, x0 0.00 : 661bc: tbnz w1, #15, 66360 <__getdelim@@GLIBC_2.17+0x1e0> 0.00 : 661c0: ldr x0, [x3, #136] 0.00 : 661c4: ldr x2, [x0, #8] 0.00 : 661c8: str x19, [x29, #16] 0.00 : 661cc: mrs x19, tpidr_el0 0.00 : 661d0: sub x19, x19, #0x700 0.00 : 661d4: cmp x2, x19 0.00 : 661d8: b.eq 663f0 <__getdelim@@GLIBC_2.17+0x270> // b.none 0.00 : 661dc: mov w1, #0x1 // #1 0.00 : 661e0: ldaxr w2, [x0] 0.00 : 661e4: cmp w2, #0x0 0.00 : 661e8: b.ne 661f4 <__getdelim@@GLIBC_2.17+0x74> // b.any 0.00 : 661ec: stxr w3, w1, [x0] 0.00 : 661f0: cbnz w3, 661e0 <__getdelim@@GLIBC_2.17+0x60> 0.00 : 661f4: b.ne 66448 <__getdelim@@GLIBC_2.17+0x2c8> // b.any 0.00 : 661f8: ldr x0, [x20, #136] 0.00 : 661fc: ldr w1, [x20] 0.00 : 66200: ldr w2, [x0, #4] 0.00 : 66204: str x19, [x0, #8] 0.00 : 66208: add w2, w2, #0x1 0.00 : 6620c: str w2, [x0, #4] 0.00 : 66210: tbnz w1, #5, 66388 <__getdelim@@GLIBC_2.17+0x208> 0.00 : 66214: ldr x19, [x29, #16] 0.00 : 66218: ldr x0, [x21] 0.00 : 6621c: cbz x0, 66228 <__getdelim@@GLIBC_2.17+0xa8> 0.00 : 66220: ldr x0, [x22] 0.00 : 66224: cbnz x0, 6623c <__getdelim@@GLIBC_2.17+0xbc> 0.00 : 66228: mov x0, #0x78 // #120 0.00 : 6622c: str x0, [x22] 0.00 : 66230: bl 20710 <malloc@plt> 0.00 : 66234: str x0, [x21] 0.00 : 66238: cbz x0, 66428 <__getdelim@@GLIBC_2.17+0x2a8> 0.00 : 6623c: ldr x27, [x20, #8] 0.00 : 66240: str x19, [x29, #16] 0.00 : 66244: ldr x19, [x20, #16] 0.00 : 66248: sub x19, x19, x27 0.00 : 6624c: cmp x19, #0x0 0.00 : 66250: b.le 66398 <__getdelim@@GLIBC_2.17+0x218> 0.00 : 66254: mov x25, #0x0 // #0 0.00 : 66258: b 662d8 <__getdelim@@GLIBC_2.17+0x158> 0.00 : 6625c: nop 0.00 : 66260: add x24, x19, x25 0.00 : 66264: ldr x3, [x22] 0.00 : 66268: add x26, x24, #0x1 0.00 : 6626c: ldr x0, [x21] 0.00 : 66270: cmp x3, x26 0.00 : 66274: b.cs 6629c <__getdelim@@GLIBC_2.17+0x11c> // b.hs, b.nlast 0.00 : 66278: lsl x3, x3, #1 0.00 : 6627c: cmp x3, x26 0.00 : 66280: csel x26, x3, x26, cs // cs = hs, nlast 0.00 : 66284: mov x1, x26 0.00 : 66288: bl 206f0 <realloc@plt> 0.00 : 6628c: cbz x0, 66438 <__getdelim@@GLIBC_2.17+0x2b8> 0.00 : 66290: str x0, [x21] 0.00 : 66294: ldr x27, [x20, #8] 0.00 : 66298: str x26, [x22] 0.00 : 6629c: mov x2, x19 0.00 : 662a0: mov x1, x27 0.00 : 662a4: add x0, x0, x25 0.00 : 662a8: bl 87390 <explicit_bzero@@GLIBC_2.25+0x50> 0.00 : 662ac: ldr x0, [x20, #8] 0.00 : 662b0: add x19, x0, x19 0.00 : 662b4: str x19, [x20, #8] 0.00 : 662b8: cbnz x28, 66410 <__getdelim@@GLIBC_2.17+0x290> 0.00 : 662bc: mov x0, x20 0.00 : 662c0: bl 73b80 <__underflow@@GLIBC_2.17> 0.00 : 662c4: cmn w0, #0x1 0.00 : 662c8: b.eq 66410 <__getdelim@@GLIBC_2.17+0x290> // b.none 0.00 : 662cc: ldp x27, x19, [x20, #8] 0.00 : 662d0: mov x25, x24 0.00 : 662d4: sub x19, x19, x27 0.00 : 662d8: mov x2, x19 0.00 : 662dc: mov w1, w23 0.00 : 662e0: mov x0, x27 0.00 : 662e4: bl 807b0 <memchr@@GLIBC_2.17> 0.00 : 662e8: cmp x0, #0x0 0.00 : 662ec: mov x28, x0 0.00 : 662f0: sub x0, x0, x27 0.00 : 662f4: csinc x19, x19, x0, eq // eq = none 0.00 : 662f8: mov x0, #0x7fffffffffffffff // #9223372036854775807 0.00 : 662fc: sub x0, x0, x25 0.00 : 66300: cmp x19, x0 0.00 : 66304: b.lt 66260 <__getdelim@@GLIBC_2.17+0xe0> // b.tstop 0.00 : 66308: adrp x0, 17f000 <sys_sigabbrev@@GLIBC_2.17+0x320> 0.00 : 6630c: ldr x0, [x0, #3624] 0.00 : 66310: mrs x2, tpidr_el0 0.00 : 66314: ldr x19, [x29, #16] 0.00 : 66318: mov w3, #0x4b // #75 0.00 : 6631c: ldr w1, [x20] 0.00 : 66320: mov x24, #0xffffffffffffffff // #-1 0.00 : 66324: str w3, [x2, x0] 0.00 : 66328: tbnz w1, #15, 66340 <__getdelim@@GLIBC_2.17+0x1c0> 0.00 : 6632c: ldr x0, [x20, #136] 0.00 : 66330: ldr w1, [x0, #4] 0.00 : 66334: sub w1, w1, #0x1 0.00 : 66338: str w1, [x0, #4] 0.00 : 6633c: cbz w1, 663b8 <__getdelim@@GLIBC_2.17+0x238> 0.00 : 66340: mov x0, x24 0.00 : 66344: ldr x28, [sp, #88] 0.00 : 66348: ldp x20, x21, [x29, #24] 0.00 : 6634c: ldp x22, x23, [x29, #40] 0.00 : 66350: ldp x24, x25, [sp, #56] 0.00 : 66354: ldp x26, x27, [sp, #72] 0.00 : 66358: ldp x29, x30, [sp], #96 0.00 : 6635c: ret 100.00 : 66360: tbz w1, #5, 66218 <__getdelim@@GLIBC_2.17+0x98> 0.00 : 66364: ldp x20, x21, [x29, #24] 0.00 : 66368: mov x24, #0xffffffffffffffff // #-1 0.00 : 6636c: ldp x22, x23, [x29, #40] 0.00 : 66370: mov x0, x24 0.00 : 66374: ldp x24, x25, [sp, #56] 0.00 : 66378: ldp x26, x27, [sp, #72] 0.00 : 6637c: ldr x28, [sp, #88] 0.00 : 66380: ldp x29, x30, [sp], #96 0.00 : 66384: ret 0.00 : 66388: mov x24, #0xffffffffffffffff // #-1 0.00 : 6638c: ldr x19, [x29, #16] 0.00 : 66390: b 66328 <__getdelim@@GLIBC_2.17+0x1a8> 0.00 : 66394: nop 0.00 : 66398: mov x0, x20 0.00 : 6639c: bl 73b80 <__underflow@@GLIBC_2.17> 0.00 : 663a0: cmn w0, #0x1 0.00 : 663a4: b.eq 66438 <__getdelim@@GLIBC_2.17+0x2b8> // b.none 0.00 : 663a8: ldp x27, x19, [x20, #8] 0.00 : 663ac: sub x19, x19, x27 0.00 : 663b0: b 66254 <__getdelim@@GLIBC_2.17+0xd4> 0.00 : 663b4: nop 0.00 : 663b8: str xzr, [x0, #8] 0.00 : 663bc: ldxr w2, [x0] 0.00 : 663c0: stlxr w3, w1, [x0] 0.00 : 663c4: cbnz w3, 663bc <__getdelim@@GLIBC_2.17+0x23c> 0.00 : 663c8: cmp w2, #0x1 0.00 : 663cc: b.le 66340 <__getdelim@@GLIBC_2.17+0x1c0> 0.00 : 663d0: mov x1, #0x81 // #129 0.00 : 663d4: mov x2, #0x1 // #1 0.00 : 663d8: mov x3, #0x0 // #0 0.00 : 663dc: mov x8, #0x62 // #98 0.00 : 663e0: svc #0x0 0.00 : 663e4: ldp x20, x21, [x29, #24] 0.00 : 663e8: ldp x22, x23, [x29, #40] 0.00 : 663ec: b 66370 <__getdelim@@GLIBC_2.17+0x1f0> 0.00 : 663f0: ldr w2, [x0, #4] 0.00 : 663f4: add w2, w2, #0x1 0.00 : 663f8: str w2, [x0, #4] 0.00 : 663fc: tbz w1, #5, 66214 <__getdelim@@GLIBC_2.17+0x94> 0.00 : 66400: mov x24, #0xffffffffffffffff // #-1 0.00 : 66404: ldr x19, [x29, #16] 0.00 : 66408: b 66330 <__getdelim@@GLIBC_2.17+0x1b0> 0.00 : 6640c: nop 0.00 : 66410: ldr x0, [x21] 0.00 : 66414: strb wzr, [x0, x24] 0.00 : 66418: ldr w1, [x20] 0.00 : 6641c: ldr x19, [x29, #16] 0.00 : 66420: b 66328 <__getdelim@@GLIBC_2.17+0x1a8> 0.00 : 66424: nop 0.00 : 66428: mov x24, #0xffffffffffffffff // #-1 0.00 : 6642c: ldr w1, [x20] 0.00 : 66430: b 66328 <__getdelim@@GLIBC_2.17+0x1a8> 0.00 : 66434: nop 0.00 : 66438: mov x24, #0xffffffffffffffff // #-1 0.00 : 6643c: ldr w1, [x20] 0.00 : 66440: ldr x19, [x29, #16] 0.00 : 66444: b 66328 <__getdelim@@GLIBC_2.17+0x1a8> 0.00 : 66448: bl e3ba0 <pthread_setcanceltype@@GLIBC_2.17+0x30> 0.00 : 6644c: b 661f8 <__getdelim@@GLIBC_2.17+0x78> 0.00 : 66450: adrp x0, 17f000 <sys_sigabbrev@@GLIBC_2.17+0x320> 0.00 : 66454: ldr x0, [x0, #3624] 0.00 : 66458: mrs x1, tpidr_el0 0.00 : 6645c: mov w2, #0x16 // #22 0.00 : 66460: mov x24, #0xffffffffffffffff // #-1 0.00 : 66464: str w2, [x1, x0] 0.00 : 66468: b 66370 <__getdelim@@GLIBC_2.17+0x1f0> 0.00 : 6646c: ldr w1, [x20] 0.00 : 66470: mov x4, x0 0.00 : 66474: tbnz w1, #15, 6648c <__getdelim@@GLIBC_2.17+0x30c> 0.00 : 66478: ldr x0, [x20, #136] 0.00 : 6647c: ldr w1, [x0, #4] 0.00 : 66480: sub w1, w1, #0x1 0.00 : 66484: str w1, [x0, #4] 0.00 : 66488: cbz w1, 66494 <__getdelim@@GLIBC_2.17+0x314> 0.00 : 6648c: mov x0, x4 0.00 : 66490: bl 20e40 <gnu_get_libc_version@@GLIBC_2.17+0x130> 0.00 : 66494: str xzr, [x0, #8] 0.00 : 66498: ldxr w2, [x0] 0.00 : 6649c: stlxr w3, w1, [x0] 0.00 : 664a0: cbnz w3, 66498 <__getdelim@@GLIBC_2.17+0x318> 0.00 : 664a4: cmp w2, #0x1 0.00 : 664a8: b.le 6648c <__getdelim@@GLIBC_2.17+0x30c> 0.00 : 664ac: mov x1, #0x81 // #129 0.00 : 664b0: mov x2, #0x1 // #1 0.00 : 664b4: mov x3, #0x0 // #0 0.00 : 664b8: mov x8, #0x62 // #98 0.00 : 664bc: svc #0x0 0.00 : 664c0: b 6648c <__getdelim@@GLIBC_2.17+0x30c> Signed-off-by: Yang Jihong <[email protected]> Tested-by: Leo Yan <[email protected]> Acked-by: Adrian Hunter <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-16perf mem-events: Remove duplicate #undefLi Huafei1-2/+0
Remove duplicate '#undef E'. Signed-off-by: Li Huafei <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Kan Liang <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Zhang Jinhao <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-11perf session: Correct buffer copying when peeking eventsLeo Yan1-0/+1
When peeking an event, it has a short path and a long path. The short path uses the session pointer "one_mmap_addr" to directly fetch the event; and the long path needs to read out the event header and the following event data from file and fill into the buffer pointer passed through the argument "buf". The issue is in the long path that it copies the event header and event data into the same destination address which pointer "buf", this means the event header is overwritten. We are just lucky to run into the short path in most cases, so we don't hit the issue in the long path. This patch adds the offset "hdr_sz" to the pointer "buf" when copying the event data, so that it can reserve the event header which can be used properly by its caller. Fixes: 5a52f33adf02 ("perf session: Add perf_session__peek_event()") Signed-off-by: Leo Yan <[email protected]> Acked-by: Adrian Hunter <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Kan Liang <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-10perf evsel: Adjust hybrid event and global event mixed groupJin Yao1-0/+25
A group mixed with hybrid event and global event is allowed. For example, group leader is 'intel_pt//' and the group member is 'cpu_atom/cycles/'. e.g.: # perf record --aux-sample -e '{intel_pt//,cpu_atom/cycles/}:u' The challenge is that their available cpus are not fully matched. For example, 'intel_pt//' is available on CPU0-CPU23, but 'cpu_atom/cycles/' is available on CPU16-CPU23. When getting the group id for group member, we must be very careful. Because the cpu for 'intel_pt//' is not equal to the cpu for 'cpu_atom/cycles/'. Actually the cpu here is the index of evsel->core.cpus, not the real CPU ID. e.g. cpu0 for 'intel_pt//' is CPU0, but cpu0 for 'cpu_atom/cycles/' is CPU16. Before: # perf record --aux-sample -e '{intel_pt//,cpu_atom/cycles/}:u' -vv uname ... ------------------------------------------------------------ perf_event_attr: type 10 size 128 config 0xe601 { sample_period, sample_freq } 1 sample_type IP|TID|TIME|CPU|IDENTIFIER read_format ID disabled 1 inherit 1 exclude_kernel 1 exclude_hv 1 enable_on_exec 1 sample_id_all 1 exclude_guest 1 ------------------------------------------------------------ sys_perf_event_open: pid 4084 cpu 0 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 4084 cpu 1 group_fd -1 flags 0x8 = 6 sys_perf_event_open: pid 4084 cpu 2 group_fd -1 flags 0x8 = 7 sys_perf_event_open: pid 4084 cpu 3 group_fd -1 flags 0x8 = 9 sys_perf_event_open: pid 4084 cpu 4 group_fd -1 flags 0x8 = 10 sys_perf_event_open: pid 4084 cpu 5 group_fd -1 flags 0x8 = 11 sys_perf_event_open: pid 4084 cpu 6 group_fd -1 flags 0x8 = 12 sys_perf_event_open: pid 4084 cpu 7 group_fd -1 flags 0x8 = 13 sys_perf_event_open: pid 4084 cpu 8 group_fd -1 flags 0x8 = 14 sys_perf_event_open: pid 4084 cpu 9 group_fd -1 flags 0x8 = 15 sys_perf_event_open: pid 4084 cpu 10 group_fd -1 flags 0x8 = 16 sys_perf_event_open: pid 4084 cpu 11 group_fd -1 flags 0x8 = 17 sys_perf_event_open: pid 4084 cpu 12 group_fd -1 flags 0x8 = 18 sys_perf_event_open: pid 4084 cpu 13 group_fd -1 flags 0x8 = 19 sys_perf_event_open: pid 4084 cpu 14 group_fd -1 flags 0x8 = 20 sys_perf_event_open: pid 4084 cpu 15 group_fd -1 flags 0x8 = 21 sys_perf_event_open: pid 4084 cpu 16 group_fd -1 flags 0x8 = 22 sys_perf_event_open: pid 4084 cpu 17 group_fd -1 flags 0x8 = 23 sys_perf_event_open: pid 4084 cpu 18 group_fd -1 flags 0x8 = 24 sys_perf_event_open: pid 4084 cpu 19 group_fd -1 flags 0x8 = 25 sys_perf_event_open: pid 4084 cpu 20 group_fd -1 flags 0x8 = 26 sys_perf_event_open: pid 4084 cpu 21 group_fd -1 flags 0x8 = 27 sys_perf_event_open: pid 4084 cpu 22 group_fd -1 flags 0x8 = 28 sys_perf_event_open: pid 4084 cpu 23 group_fd -1 flags 0x8 = 29 ------------------------------------------------------------ perf_event_attr: size 128 config 0x800000000 { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|PERIOD|IDENTIFIER|AUX read_format ID inherit 1 exclude_kernel 1 exclude_hv 1 freq 1 sample_id_all 1 exclude_guest 1 aux_sample_size 4096 ------------------------------------------------------------ sys_perf_event_open: pid 4084 cpu 16 group_fd 5 flags 0x8 sys_perf_event_open failed, error -22 The group_fd 5 is not correct. It should be 22 (the fd of 'intel_pt' on CPU16). After: # perf record --aux-sample -e '{intel_pt//,cpu_atom/cycles/}:u' -vv uname ... ------------------------------------------------------------ perf_event_attr: type 10 size 128 config 0xe601 { sample_period, sample_freq } 1 sample_type IP|TID|TIME|CPU|IDENTIFIER read_format ID disabled 1 inherit 1 exclude_kernel 1 exclude_hv 1 enable_on_exec 1 sample_id_all 1 exclude_guest 1 ------------------------------------------------------------ sys_perf_event_open: pid 5162 cpu 0 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 5162 cpu 1 group_fd -1 flags 0x8 = 6 sys_perf_event_open: pid 5162 cpu 2 group_fd -1 flags 0x8 = 7 sys_perf_event_open: pid 5162 cpu 3 group_fd -1 flags 0x8 = 9 sys_perf_event_open: pid 5162 cpu 4 group_fd -1 flags 0x8 = 10 sys_perf_event_open: pid 5162 cpu 5 group_fd -1 flags 0x8 = 11 sys_perf_event_open: pid 5162 cpu 6 group_fd -1 flags 0x8 = 12 sys_perf_event_open: pid 5162 cpu 7 group_fd -1 flags 0x8 = 13 sys_perf_event_open: pid 5162 cpu 8 group_fd -1 flags 0x8 = 14 sys_perf_event_open: pid 5162 cpu 9 group_fd -1 flags 0x8 = 15 sys_perf_event_open: pid 5162 cpu 10 group_fd -1 flags 0x8 = 16 sys_perf_event_open: pid 5162 cpu 11 group_fd -1 flags 0x8 = 17 sys_perf_event_open: pid 5162 cpu 12 group_fd -1 flags 0x8 = 18 sys_perf_event_open: pid 5162 cpu 13 group_fd -1 flags 0x8 = 19 sys_perf_event_open: pid 5162 cpu 14 group_fd -1 flags 0x8 = 20 sys_perf_event_open: pid 5162 cpu 15 group_fd -1 flags 0x8 = 21 sys_perf_event_open: pid 5162 cpu 16 group_fd -1 flags 0x8 = 22 sys_perf_event_open: pid 5162 cpu 17 group_fd -1 flags 0x8 = 23 sys_perf_event_open: pid 5162 cpu 18 group_fd -1 flags 0x8 = 24 sys_perf_event_open: pid 5162 cpu 19 group_fd -1 flags 0x8 = 25 sys_perf_event_open: pid 5162 cpu 20 group_fd -1 flags 0x8 = 26 sys_perf_event_open: pid 5162 cpu 21 group_fd -1 flags 0x8 = 27 sys_perf_event_open: pid 5162 cpu 22 group_fd -1 flags 0x8 = 28 sys_perf_event_open: pid 5162 cpu 23 group_fd -1 flags 0x8 = 29 ------------------------------------------------------------ perf_event_attr: size 128 config 0x800000000 { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|PERIOD|IDENTIFIER|AUX read_format ID inherit 1 exclude_kernel 1 exclude_hv 1 freq 1 sample_id_all 1 exclude_guest 1 aux_sample_size 4096 ------------------------------------------------------------ sys_perf_event_open: pid 5162 cpu 16 group_fd 22 flags 0x8 = 30 sys_perf_event_open: pid 5162 cpu 17 group_fd 23 flags 0x8 = 31 sys_perf_event_open: pid 5162 cpu 18 group_fd 24 flags 0x8 = 32 sys_perf_event_open: pid 5162 cpu 19 group_fd 25 flags 0x8 = 33 sys_perf_event_open: pid 5162 cpu 20 group_fd 26 flags 0x8 = 34 sys_perf_event_open: pid 5162 cpu 21 group_fd 27 flags 0x8 = 35 sys_perf_event_open: pid 5162 cpu 22 group_fd 28 flags 0x8 = 36 sys_perf_event_open: pid 5162 cpu 23 group_fd 29 flags 0x8 = 37 ------------------------------------------------------------ ... Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jin Yao <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-08perf probe: Provide clearer message permission error for tracefs accessMasami Hiramatsu1-30/+65
Report permission error for the tracefs open and rewrite whole the error message code around it. You'll see a hint according to what you want to do with perf probe as below. $ perf probe -l No permission to read tracefs. Please try 'sudo mount -o remount,mode=755 /sys/kernel/tracing/' Error: Failed to show event list. $ perf probe -d \* No permission to write tracefs. Please run this command again with sudo. Error: Failed to delete events. This also fixes -ENOTSUP checking for mounting tracefs/debugfs. Actually open returns -ENOENT in that case and we have to check it with current mount point list. If we unmount debugfs and tracefs perf probe shows correct message as below. $ perf probe -l Debugfs or tracefs is not mounted Please try 'sudo mount -t tracefs nodev /sys/kernel/tracing/' Error: Failed to show event list. Signed-off-by: Masami Hiramatsu <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Aneesh Kumar K.V <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Ravi Bangoria <[email protected]> Link: http://lore.kernel.org/lkml/162299456839.503471.13863002017089255222.stgit@devnote2 Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-08perf auxtrace: Change to use SMP memory barriersLeo Yan1-3/+3
The kernel and the userspace tool can access the AUX ring buffer head and tail from different CPUs, thus SMP class of barriers are required on SMP system. This patch changes to use SMP barriers to replace mb() and rmb() barriers. Signed-off-by: Leo Yan <[email protected]> Acked-by: Adrian Hunter <[email protected]> Acked-by: Peter Zijlstra <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: H. Peter Anvin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Mathieu Poirier <[email protected]> Cc: Mike Leach <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Suzuki Poulouse <[email protected]> Cc: Thomas Gleixner <[email protected]> Cc: [email protected] Cc: [email protected] Cc: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-08perf srccode: Use list_move() instead of equivalent list_del() + list_add() ↵Zou Wei1-2/+1
sequence Using list_move() instead of list_del() + list_add(), shorter, equivalent. Reported-by: Hulk Robot <[email protected]> Signed-off-by: Zou Wei <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf probe: Report possible permission error for map__load() failureMasami Hiramatsu1-3/+22
Report possible permission error including kptr_restrict setting for map__load() failure. This can happen when non-superuser runs perf probe. With this patch, perf probe shows the following message. $ perf probe vfs_read Failed to load symbols from /proc/kallsyms Please ensure you can read the /proc/kallsyms symbol addresses. If the /proc/sys/kernel/kptr_restrict is '2', you can not read kernel symbol address even if you are a superuser. Please change it to '1'. If kptr_restrict is '1', the superuser can read the symbol addresses. In that case, please run this command again with sudo. Error: Failed to add events. Signed-off-by: Masami Hiramatsu <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Aneesh Kumar K.V <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Ravi Bangoria <[email protected]> Link: http://lore.kernel.org/lkml/162282065877.448336.10047912688119745151.stgit@devnote2 Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf env: Fix memory leak of bpf_prog_info_linear memberRiccardo Mancini1-0/+1
ASan reported a memory leak caused by info_linear not being deallocated. The info_linear was allocated during in perf_event__synthesize_one_bpf_prog(). This patch adds the corresponding free() when bpf_prog_info_node is freed in perf_env__purge_bpf(). $ sudo ./perf record -- sleep 5 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.025 MB perf.data (8 samples) ] ================================================================= ==297735==ERROR: LeakSanitizer: detected memory leaks Direct leak of 7688 byte(s) in 19 object(s) allocated from: #0 0x4f420f in malloc (/home/user/linux/tools/perf/perf+0x4f420f) #1 0xc06a74 in bpf_program__get_prog_info_linear /home/user/linux/tools/lib/bpf/libbpf.c:11113:16 #2 0xb426fe in perf_event__synthesize_one_bpf_prog /home/user/linux/tools/perf/util/bpf-event.c:191:16 #3 0xb42008 in perf_event__synthesize_bpf_events /home/user/linux/tools/perf/util/bpf-event.c:410:9 #4 0x594596 in record__synthesize /home/user/linux/tools/perf/builtin-record.c:1490:8 #5 0x58c9ac in __cmd_record /home/user/linux/tools/perf/builtin-record.c:1798:8 #6 0x58990b in cmd_record /home/user/linux/tools/perf/builtin-record.c:2901:8 #7 0x7b2a20 in run_builtin /home/user/linux/tools/perf/perf.c:313:11 #8 0x7b12ff in handle_internal_command /home/user/linux/tools/perf/perf.c:365:8 #9 0x7b2583 in run_argv /home/user/linux/tools/perf/perf.c:409:2 #10 0x7b0d79 in main /home/user/linux/tools/perf/perf.c:539:3 #11 0x7fa357ef6b74 in __libc_start_main /usr/src/debug/glibc-2.33-8.fc34.x86_64/csu/../csu/libc-start.c:332:16 Signed-off-by: Riccardo Mancini <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Alexei Starovoitov <[email protected]> Cc: Andrii Nakryiko <[email protected]> Cc: Daniel Borkmann <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: John Fastabend <[email protected]> Cc: KP Singh <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Martin KaFai Lau <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Song Liu <[email protected]> Cc: Yonghong Song <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf probe: Add permission and sysctl notice to man pageMasami Hiramatsu1-3/+16
Add a section to notify the permission and sysctl setting for perf probe. And fix some indentations. Reported-by: Ravi Bangoria <[email protected]> Signed-off-by: Masami Hiramatsu <[email protected]> Reviewed-by: Ravi Bangoria <[email protected]> Cc: Aneesh Kumar K.V <[email protected]> Cc: Jiri Olsa <[email protected]> Link: http://lore.kernel.org/lkml/162204068898.388434.16842705842611255787.stgit@devnote2 Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf symbol-elf: Fix memory leak by freeing sdt_note.argsRiccardo Mancini1-0/+1
Reported by ASan. Signed-off-by: Riccardo Mancini <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Fabian Hemmer <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Remi Bernon <[email protected]> Cc: Jiri Slaby <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf stat: Honor event config name on --no-mergeNamhyung Kim1-5/+3
If user gave an event name explicitly, it should be displayed in the output as is. But with --no-merge option it adds a pmu name at the end so might confuse users. Actually this is true for hybrid pmus, I think we should do the same for others. Signed-off-by: Namhyung Kim <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jin Yao <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-04perf evsel: Add missing cloning of evsel->use_config_nameNamhyung Kim2-2/+3
The evsel__clone() should copy all fields in the evsel which are set during the event parsing. But it missed the use_config_name field. Fixes: 12279429d862 ("perf stat: Uniquify hybrid event name") Signed-off-by: Namhyung Kim <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jin Yao <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-02Revert "perf vendor events intel: Add metrics for Icelake Server"Arnaldo Carvalho de Melo1-327/+0
It is making 'perf test 10' fail: ⬢[acme@toolbox perf]$ perf test 10 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : FAILED! ⬢[acme@toolbox perf] This reverts commit d89bf9cab1f613e4496f929d89477b2baaad1ea9.
2021-06-01Merge remote-tracking branch 'torvalds/master' into perf/coreArnaldo Carvalho de Melo13-304/+304
To pick up fixes from perf/urgent to allow perf/core to be used for new development. Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf c2c: Support record for hybrid platformJin Yao1-17/+23
Support 'perf c2c record' for hybrid platform. On hybrid platform, such as Alderlake, when executing 'perf c2c record', it actually calls: record -W -d --phys-data --sample-cpu -e {cpu_core/mem-loads-aux/,cpu_core/mem-loads,ldlat=30/}:P -e cpu_atom/mem-loads,ldlat=30/P -e cpu_core/mem-stores/P -e cpu_atom/mem-stores/P Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf mem: Disable 'mem-loads-aux' group before reportingJin Yao3-0/+28
For some platforms, such as Alderlake, the 'mem-loads' event is required to use together with 'mem-loads-aux' within a group and 'mem-loads-aux' must be the group leader. Now we disable this group before reporting because 'mem-loads-aux' is just an auxiliary event. It doesn't carry any valid memory load result. If we show the 'mem-loads-aux' + 'mem-loads' as a group in report, it needs many of changes but they are totally unnecessary. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf mem: Fix wrong verbose output for recording eventsJin Yao1-3/+5
Current code: for (j = 0; j < argc; j++, i++) rec_argv[i] = argv[j]; if (verbose > 0) { pr_debug("calling: record "); while (rec_argv[j]) { pr_debug("%s ", rec_argv[j]); j++; } pr_debug("\n"); } The entries of argv[] are copied to the end of rec_argv[], not copied to the beginning of rec_argv[]. So the index j at rec_argv[] doesn't point to the first event. Now we record the start index and end index for events in rec_argv[], and print them if verbose is enabled. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf mem: Support record for hybrid platformJin Yao3-17/+93
Support 'perf mem record' for hybrid platform. On hybrid platform, such as Alderlake, when executing 'perf mem record', it actually calls: record -e {cpu_core/mem-loads-aux/,cpu_core/mem-loads,ldlat=30/}:P -e cpu_atom/mem-loads,ldlat=30/P -e cpu_core/mem-stores/P -e cpu_atom/mem-stores/P Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf tools: Check if mem_events is supported for hybrid platformJin Yao1-6/+26
Check if the mem_events ('mem-loads' and 'mem-stores') exist in the sysfs path. For Alderlake, the hybrid cpu pmu are "cpu_core" and "cpu_atom". Check the existing of following paths: /sys/devices/cpu_atom/events/mem-loads /sys/devices/cpu_atom/events/mem-stores /sys/devices/cpu_core/events/mem-loads /sys/devices/cpu_core/events/mem-stores If the patch exists, the mem_event is supported. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf tools: Support pmu prefix for mem-store eventJin Yao1-1/+11
For enabling mem-store event, it doesn't need an auxiliary event. So just build an event name string with the pmu prefix. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf tools: Support pmu prefix for mem-load eventJin Yao7-16/+37
The perf_mem_events__name() can generate the mem-load event name. It uses a variable 'mem_loads_name__init' to avoid generating the event name every time (because perf_pmu__scan takes some time). The perf_mem_events__name() assumes the pmu is "cpu" but it's not correct for hybrid platform. For Alderlake, the pmu is "cpu_core" or "cpu_atom" Introduce a new parameter 'pmu_name' in perf_mem_events__name to let the caller specify a pmu name. Considering such event name is x86 specific, so move perf_mem_events[] to arch/x86/util/mem-events.c. We still keep the variable 'mem_loads_name__init' but it's only used when pmu_name is NULL (compatible for original behavior). When pmu_name is not NULL (e.g. "cpu_core"), this patch doesn't have optimization. That can be implemented in follow up patch. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf tools: Check mem-loads auxiliary eventJin Yao1-2/+7
For some platforms, an auxiliary event has to be enabled simultaneously with the load latency event. For Alderlake, the auxiliary event is created in "cpu_core" pmu. So first we need to check the existing of "cpu_core" pmu and then check if this pmu has auxiliary event. Signed-off-by: Jin Yao <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Kan Liang <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf test: Test 17 fails with make LIBPFM4=1 on s390 z/VMThomas Richter1-1/+1
This test case fails on s390 virtual machine z/VM which has no PMU support when the perf tool is built with LIBPFM4=1. Using make LIBPFM4=1 builds the perf tool with support for libpfm event notation. The command line flag --pfm-events is valid: # ./perf record --pfm-events cycles -- true [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.001 MB perf.data (2 samples) ] # However the command 'perf test -Fv 17' fails on s390 z/VM virtual machine with LIBPFM4=1: # perf test -Fv 17 17: Setup struct perf_event_attr : --- start --- ..... running './tests/attr/test-record-group2' unsupp './tests/attr/test-record-group2' running './tests/attr/test-record-pfm-period' expected exclude_hv=0, got 1 FAILED './tests/attr/test-record-pfm-period' - match failure ---- end ---- Setup struct perf_event_attr: FAILED! When --pfm-event system is not supported, the test returns unsupported and continues. Here is an example using a virtual machine on x86 and Fedora 34: [root@f33 perf]# perf test -Fv 17 17: Setup struct perf_event_attr : --- start --- ..... running './tests/attr/test-record-group2' unsupp './tests/attr/test-record-group2' running './tests/attr/test-record-pfm-period' unsupp './tests/attr/test-record-pfm-period' .... The issue is file ./tests/attr/test-record-pfm-period which requires perf event attribute member exclude_hv to be zero. This is not the case on s390 where the value of exclude_hv is one when executing on a z/VM virtual machine without PMU hardware support. Fix this by allowing value exlucde_hv to be zero or one. Output before: # /usr/bin/python ./tests/attr.py -d ./tests/attr/ -t \ test-record-pfm-period -p ./perf -vvv 2>&1| fgrep match matching [event:base-record] match: [event:base-record] matches [] FAILED './tests/attr//test-record-pfm-period' - match failure # Output after: # /usr/bin/python ./tests/attr.py -d ./tests/attr/ -t \ test-record-pfm-period -p ./perf -vvv 2>&1| fgrep match matching [event:base-record] match: [event:base-record] matches ['event-1-0-6', 'event-1-0-5'] matched Background: Using libpfm library ends up in this function call sequence pfm_get_perf_event_encoding() +-- pfm_get_os_event_encoding() +-- pfmlib_perf_event_encode() is called when no hardware specific PMU unit can be detected as in the s390 z/VM virtual machine case. This uses the "perf_events generic PMU" data structure which sets exclude_hv to 1 per default. Using this PMU that test case always fails. That is the reason why exclude_hv attribute setting varies. Version 2: As suggested by Ian Rogers make perf_event_attribute member exclude_hv more robust and accept value 0 or 1 to handle more test cases which might fail on s390 virtual machine z/VM. Suggested-by: Ian Rogers <[email protected]> Signed-off-by: Thomas Richter <[email protected]> Reviewed-by: Ian Rogers <[email protected]> Cc: Heiko Carstens <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Sumanth Korikkar <[email protected]> Cc: Sven Schnelle <[email protected]> Cc: Vasily Gorbik <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf stat: Fix error return code in bperf__load()Yu Kuai1-2/+4
Fix to return a negative error code from the error handling case instead of 0, as done elsewhere in this function. Committer notes: Added the missing {} for the now multiline 'if' block, fixing this error: CC /tmp/build/perf/util/bpf_counter.o util/bpf_counter.c: In function ‘bperf__load’: util/bpf_counter.c:523:9: error: this ‘if’ clause does not guard... [-Werror=misleading-indentation] 523 | if (evsel->bperf_leader_link_fd < 0 && | ^~ util/bpf_counter.c:526:17: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’ 526 | goto out; | ^~~~ cc1: all warnings being treated as errors Fixes: 7fac83aaf2eecc9e ("perf stat: Introduce 'bperf' to share hardware PMCs with BPF") Reported-by: Hulk Robot <[email protected]> Signed-off-by: Yu Kuai <[email protected]> Cc: Alexei Starovoitov <[email protected]> Cc: Andrii Nakryiko <[email protected]> Cc: Daniel Borkmann <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Song Liu <[email protected]> Cc: Yu Kuai <[email protected]> Cc: Zhang Yi <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf record: Move probing cgroup sampling supportNamhyung Kim3-0/+17
I found that checking cgroup sampling support using the missing features doesn't work on old kernels. Because it added both attr.cgroup bit and PERF_SAMPLE_CGROUP bit, it needs to check whichever comes first (usually the actual event, not dummy). But it only checks the attr.cgroup bit which is set only in the dummy event so cannot detect failtures due the sample bits. Also we don't ignore the missing feature and retry, it'd be better checking it with the API probing logic. Committer notes: Extracted the minimal part to check using the new cgroup API probe routine, the part that removes the cgroup member can be left for further discussion. Signed-off-by: Namhyung Kim <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf probe: Fix NULL pointer dereference in convert_variable_location()Li Huafei2-2/+9
If we just check whether the variable can be converted, 'tvar' should be a null pointer. However, the null pointer check is missing in the 'Constant value' execution path. The following cases can trigger this problem: $ cat test.c #include <stdio.h> void main(void) { int a; const int b = 1; asm volatile("mov %1, %0" : "=r"(a): "i"(b)); printf("a: %d\n", a); } $ gcc test.c -o test -O -g $ sudo ./perf probe -x ./test -L "main" <main@/home/lhf/test.c:0> 0 void main(void) { 2 int a; const int b = 1; asm volatile("mov %1, %0" : "=r"(a): "i"(b)); 6 printf("a: %d\n", a); } $ sudo ./perf probe -x ./test -V "main:6" Segmentation fault The check on 'tvar' is added. If 'tavr' is a null pointer, we return 0 to indicate that the variable can be converted. Now, we can successfully show the variables that can be accessed. $ sudo ./perf probe -x ./test -V "main:6" Available variables at main:6 @<main+13> char* __fmt int a int b However, the variable 'b' cannot be tracked. $ sudo ./perf probe -x ./test -D "main:6 b" Failed to find the location of the 'b' variable at this address. Perhaps it has been optimized out. Use -V with the --range option to show 'b' location range. Error: Failed to add events. This is because __die_find_variable_cb() did not successfully match variable 'b', which has the DW_AT_const_value attribute instead of DW_AT_location. We added support for DW_AT_const_value in __die_find_variable_cb(). With this modification, we can successfully track the variable 'b'. $ sudo ./perf probe -x ./test -D "main:6 b" p:probe_test/main_L6 /home/lhf/test:0x1156 b=\1:s32 Fixes: 66f69b219716 ("perf probe: Support DW_AT_const_value constant value") Signed-off-by: Li Huafei <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Frank Ch. Eigler <[email protected]> Cc: Jianlin Lv <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Srikar Dronamraju <[email protected]> Cc: Yang Jihong <[email protected]> Cc: Zhang Jinhao <[email protected]> http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf tools: Copy uapi/asm/perf_regs.h from the kernel for MIPSTiezhu Yang2-1/+1
To allow the build to complete on older systems, where those files are either not uptodate, lacking some recent additions or not present at all. And check if the copy drifts from the kernel. This commit is similar with commit 12f020338a2c ("tools: Copy uapi/asm/perf_regs.h from the kernel") With this commit, we can avoid the following build error in any case: tools/perf/arch/mips/include/perf_regs.h:7:10: fatal error: asm/perf_regs.h: No such file or directory #include <asm/perf_regs.h> ^~~~~~~~~~~~~~~~~ compilation terminated. Signed-off-by: Tiezhu Yang <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Xuefeng Li <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: intel-pt-events.py: Add --insn-trace and --src-traceAdrian Hunter2-19/+163
Add an instruction trace and a source trace to the intel-pt-events.py script. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: exported-sql-viewer.py: Factor out libxed.pyAdrian Hunter2-88/+108
Factor out libxed.py so it can be reused. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Update documentation for srcline etcAdrian Hunter1-4/+42
Add new fields and functions to the perf-script-python documentation. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Add perf_sample_srcline() and perf_sample_srccode()Adrian Hunter1-0/+56
Add perf_sample_srcline() and perf_sample_srccode() to the perf_trace_context module so that a script can get the srcline or srccode information. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Add perf_set_itrace_options()Adrian Hunter1-2/+42
Add perf_set_itrace_options() to the perf_trace_context module so that a script can set the itrace options for a session if they have not been set already. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf auxtrace: Factor out itrace_do_parse_synth_opts()Adrian Hunter2-3/+17
Factor out itrace_do_parse_synth_opts() so that it can be reused. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Add perf_sample_insn()Adrian Hunter1-0/+27
Add perf_sample_insn() to the perf_trace_context module so that a script can get the instruction bytes. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf script: Factor out script_fetch_insn()Adrian Hunter2-2/+11
Factor out script_fetch_insn() so it can be reused. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Assign perf_script_contextAdrian Hunter2-1/+35
The scripting_context pointer itself does not change and nor does it need to. Put it directly into the script as a variable at the start so it does not have to be passed on each call into the script. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting: Add perf_session to scripting_contextAdrian Hunter5-6/+15
This is preparation for allowing a script to set the itrace options for the session if they have not already been set. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting: Add scripting_context__update()Adrian Hunter4-7/+38
Move scripting_context update to a separate function and add the arguments of ->process_event() to it. This prepares the way for adding more methods to the perf_trace_context module, by providing the context information that they will need. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Simplify perf-trace-context module functionsAdrian Hunter1-22/+17
Simplify perf-trace-context module functions by factoring out some common code. Signed-off-by: Adrian Hunter <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-06-01perf scripting python: Remove unnecessary 'static'Adrian Hunter1-3/+3
The variables are always assigned before use, making the 'static' storage class unnecessary. Signed-off-by: Adrian Hunter <[email protected]> Cc: Arnaldo Carvalho de Melo <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-28perf vendor events powerpc: Fix eventcode of power10 JSON eventsKajol Jain10-299/+299
Fixed the eventcode values in the power10 JSON event files to prepend "0x" since these are hexadecimal values. The patch also changes the event description of the PM_EXEC_STALL_LOAD_FINISH and PM_EXEC_STALL_NTC_FLUSH event and move some events to correct files. Fixes: 32daa5d7899e ("perf vendor events: Initial JSON/events list for power10 platform") Signed-off-by: Kajol Jain <[email protected]> Reviewed-by: Paul A. Clarke <[email protected]> Tested-by: Nageswara R Sastry <[email protected]> Cc: Athira Jajeev <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Madhavan Srinivasan <[email protected]> Cc: Michael Ellerman <[email protected]> Cc: Ravi Bangoria <[email protected]> Cc: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-27perf stat: Fix error check for bpf_program__attachNamhyung Kim1-2/+2
It seems the bpf_program__attach() returns a negative error code instead of a NULL pointer in case of error. Fixes: 7fac83aaf2ee ("perf stat: Introduce 'bperf' to share hardware PMCs with BPF") Signed-off-by: Namhyung Kim <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Song Liu <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-27perf test: Test 2 libpfm4 error casesIan Rogers1-0/+10
Proposed in: https://lore.kernel.org/lkml/[email protected]/ Signed-off-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Cc: Thomas Richter <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-27perf test: Iterate over shell tests in alphabetical orderRiccardo Mancini1-17/+21
The for_each_shell_test macro iterated over all shell tests in the directory using readdir, which does not guarantee any ordering, causing problems on certain fs. However, the order in which they are visited determines the id of the test, in case one wants to run a single test. This patch replaces readdir with scandir using alphabetical sorting. This guarantees that, given the same set of tests, all machines will see the tests in the same order, and, thus, that test ids are consistent. Signed-off-by: Riccardo Mancini <[email protected]> Reported-by: Ian Rogers <[email protected]> Acked-by: Ian Rogers <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Fabian Hemmer <[email protected]> Cc: Leo Yan <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Cc: Tommi Rantala <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-27perf probe: Provide more detail with relocation warningRavi Bangoria1-3/+8
When run as normal user with default sysctl kernel.kptr_restrict=0 and kernel.perf_event_paranoid=2, perf probe fails with: $ ./perf probe move_page_tables Relocated base symbol is not found! The warning message is not much informative. The reason perf fails is because /proc/kallsyms is restricted by perf_event_paranoid=2 for normal user and thus perf fails to read relocated address of the base symbol. Tweaking kptr_restrict and perf_event_paranoid can change the behavior of perf probe. Also, running as root or privileged user works too. Add these details in the warning message. Plus, kmap->ref_reloc_sym might not be always set even if host_machine is initialized. Above is the example of the same. Remove that comment. Reported-by: Aneesh Kumar K.V <[email protected]> Signed-off-by: Ravi Bangoria <[email protected]> Acked-by: Masami Hiramatsu <[email protected]> Cc: Jiri Olsa <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2021-05-27perf parse-events: Add bison --file-prefix-map optionDenys Zagorui2-3/+12
During a perf build with O= bison stores full paths in generated files and those paths are stored in resulting perf binary. Starting from bison v3.7.1 those paths can be remapped by using the --file-prefix-map option. Use this option if possible to make perf binary more reproducible. Signed-off-by: Denys Zagorui <[email protected]> Acked-by: Jiri Olsa <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.