aboutsummaryrefslogtreecommitdiff
path: root/tools/perf
AgeCommit message (Collapse)AuthorFilesLines
2020-09-03perf jevents: Fix suspicious code in fixregex()Namhyung Kim1-1/+1
The new string should have enough space for the original string and the back slashes IMHO. Fixes: fbc2844e84038ce3 ("perf vendor events: Use more flexible pattern matching for CPU identification for mapfile.csv") Signed-off-by: Namhyung Kim <[email protected]> Reviewed-by: Ian Rogers <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: John Garry <[email protected]> Cc: Kajol Jain <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Cc: William Cohen <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-03perf parse-events: Use uintptr_t when casting numbers to pointersArnaldo Carvalho de Melo1-4/+4
To address these errors found when cross building from x86_64 to MIPS little endian 32-bit: CC /tmp/build/perf/util/parse-events-bison.o util/parse-events.y: In function 'parse_events_parse': util/parse-events.y:514:6: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast] 514 | (void *) $2, $6, $4); | ^ util/parse-events.y:531:7: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast] 531 | (void *) $2, NULL, $4)) { | ^ util/parse-events.y:547:6: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast] 547 | (void *) $2, $4, 0); | ^ util/parse-events.y:564:7: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast] 564 | (void *) $2, NULL, 0)) { | ^ Fixes: cabbf26821aa210f ("perf parse: Before yyabort-ing free components") Cc: Adrian Hunter <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Alexei Starovoitov <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Daniel Borkmann <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jin Yao <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: John Garry <[email protected]> Cc: Kan Liang <[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: Stephane Eranian <[email protected]> Cc: Yonghong Song <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf report: Disable ordered_events for raw dumpJiri Olsa1-0/+3
Disable ordered_events for report raw dump, because for raw dump we want to see events as they are stored in the perf.data file, not sorted by time. Signed-off-by: Jiri Olsa <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Michael Petlan <[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]>
2020-09-01perf intel-pt: Fix corrupt data after perf inject fromAl Grant1-1/+8
Commit 42bbabed09ce6208 ("perf tools: Add hw_idx in struct branch_stack") changed the format of branch stacks in perf samples. When samples use this new format, a flag must be set in the corresponding event. Synthesized branch stacks generated from Intel PT were using the new format, but not setting the event attribute, leading to consumers seeing corrupt data. This patch fixes the issue by setting the event attribute to indicate use of the new format. Fixes: 42bbabed09ce6208 ("perf tools: Add hw_idx in struct branch_stack") Signed-off-by: Al Grant <[email protected]> Acked-by: Adrian Hunter <[email protected]> Reviewed-by: Mathieu Poirier <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Kan Liang <[email protected]> Cc: Leo Yan <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Mike Leach <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Suzuki Poulouse <[email protected]> Cc: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Leo Yan <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf cs-etm: Fix corrupt data after perf inject fromAl Grant1-1/+8
Commit 42bbabed09ce6208 ("perf tools: Add hw_idx in struct branch_stack") changed the format of branch stacks in perf samples. When samples use this new format, a flag must be set in the corresponding event. Synthesized branch stacks generated from CoreSight ETM trace were using the new format, but not setting the event attribute, leading to consumers seeing corrupt data. This patch fixes the issue by setting the event attribute to indicate use of the new format. Fixes: 42bbabed09ce6208 ("perf tools: Add hw_idx in struct branch_stack") Signed-off-by: Al Grant <[email protected]> Reviewed-by: Andrea Brunato <[email protected]> Cc: Adrian Hunter <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Kan Liang <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Mathieu Poirier <[email protected]> Cc: Mike Leach <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Suzuki Poulouse <[email protected]> Cc: [email protected] Signed-off-by: Leo Yan <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf top/report: Fix infinite loop in the TUI for grouped eventsArnaldo Carvalho de Melo1-1/+2
For a while we need to have a dummy event for doing things like receiving PERF_RECORD_COMM, PERF_RECORD_EXEC, etc for threads being created and dying while we synthesize the pre-existing ones at tool start. This 'dummy' event is needed for keeping track of thread lifetime events early in the session but are uninteresting otherwise, i.e. no need to have it in a initial events menu for the non-grouped case, i.e. for: # perf top -e cycles,instructions or even for plain: # perf top When 'cycles' and that 'dummy' event are in place. The code to remove that 'dummy' event ended up creating an endless loop for the grouped case, i.e.: # perf top -e '{cycles,instructions}' Fix it. Fixes: bee9ca1c8a237ca1 ("perf report TUI: Remove needless 'dummy' event from menu") Cc: Adrian Hunter <[email protected]> Cc: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf parse-events: Avoid an uninitialized read when using fake PMUsIan Rogers1-13/+17
With a fake_pmu the pmu_info isn't populated by perf_pmu__check_alias. In this case, don't try to copy the uninitialized values to the evsel. Signed-off-by: Ian Rogers <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Adrian Hunter <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jin Yao <[email protected]> Cc: Kajol Jain <[email protected]> Cc: Kan Liang <[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]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf stat: Fix out of bounds array access in the print_counters() evlist methodThomas Richter1-1/+1
Fix a compile error on F32 and gcc version 10.1 on s390 in file utils/stat-display.c. The error does not show up with make DEBUG=y. In fact the issue shows up when using both compiler options -O6 and -D_FORTIFY_SOURCE=2 (which are omitted with DEBUG=Y). This is the offending call chain: print_counter_aggr() printout(config, -1, 0, ...) with 2nd parm id set to -1 aggr_printout(config, x, id --> -1, ...) which leads to this code: case AGGR_NONE: if (evsel->percore && !config->percore_show_thread) { .... } else { fprintf(config->output, "CPU%*d%s", config->csv_output ? 0 : -7, evsel__cpus(evsel)->map[id], ^^ id is -1 !!!! config->csv_sep); } This is a compiler inlining issue which is detected on s390 but not on other plattforms. Output before: # make util/stat-display.o ..... util/stat-display.c: In function ‘perf_evlist__print_counters’: util/stat-display.c:121:4: error: array subscript -1 is below array bounds of ‘int[]’ [-Werror=array-bounds] 121 | fprintf(config->output, "CPU%*d%s", | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 122 | config->csv_output ? 0 : -7, | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 123 | evsel__cpus(evsel)->map[id], | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 124 | config->csv_sep); | ~~~~~~~~~~~~~~~~ In file included from util/evsel.h:13, from util/evlist.h:13, from util/stat-display.c:9: /root/linux/tools/lib/perf/include/internal/cpumap.h:10:7: note: while referencing ‘map’ 10 | int map[]; | ^~~ cc1: all warnings being treated as errors mv: cannot stat 'util/.stat-display.o.tmp': No such file or directory make[3]: *** [/root/linux/tools/build/Makefile.build:97: util/stat-display.o] Error 1 make[2]: *** [Makefile.perf:716: util/stat-display.o] Error 2 make[1]: *** [Makefile.perf:231: sub-make] Error 2 make: *** [Makefile:110: util/stat-display.o] Error 2 [root@t35lp46 perf]# Output after: # make util/stat-display.o ..... CC util/stat-display.o [root@t35lp46 perf]# Committer notes: Removed the removal of {} enclosing the multiline else block, as pointed out by Jiri Olsa. Suggested-by: Jiri Olsa <[email protected]> Signed-off-by: Thomas Richter <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Heiko Carstens <[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]>
2020-09-01perf test: Set NULL sentinel in pmu_events table in "Parse and process ↵Thomas Richter1-0/+3
metrics" test Linux 5.9 introduced perf test case "Parse and process metrics" and on s390 this test case always dumps core: [root@t35lp67 perf]# ./perf test -vvvv -F 67 67: Parse and process metrics : --- start --- metric expr inst_retired.any / cpu_clk_unhalted.thread for IPC parsing metric: inst_retired.any / cpu_clk_unhalted.thread Segmentation fault (core dumped) [root@t35lp67 perf]# I debugged this core dump and gdb shows this call chain: (gdb) where #0 0x000003ffabc3192a in __strnlen_c_1 () from /lib64/libc.so.6 #1 0x000003ffabc293de in strcasestr () from /lib64/libc.so.6 #2 0x0000000001102ba2 in match_metric(list=0x1e6ea20 "inst_retired.any", n=<optimized out>) at util/metricgroup.c:368 #3 find_metric (map=<optimized out>, map=<optimized out>, metric=0x1e6ea20 "inst_retired.any") at util/metricgroup.c:765 #4 __resolve_metric (ids=0x0, map=<optimized out>, metric_list=0x0, metric_no_group=<optimized out>, m=<optimized out>) at util/metricgroup.c:844 #5 resolve_metric (ids=0x0, map=0x0, metric_list=0x0, metric_no_group=<optimized out>) at util/metricgroup.c:881 #6 metricgroup__add_metric (metric=<optimized out>, metric_no_group=metric_no_group@entry=false, events=<optimized out>, events@entry=0x3ffd84fb878, metric_list=0x0, metric_list@entry=0x3ffd84fb868, map=0x0) at util/metricgroup.c:943 #7 0x00000000011034ae in metricgroup__add_metric_list (map=0x13f9828 <map>, metric_list=0x3ffd84fb868, events=0x3ffd84fb878, metric_no_group=<optimized out>, list=<optimized out>) at util/metricgroup.c:988 #8 parse_groups (perf_evlist=perf_evlist@entry=0x1e70260, str=str@entry=0x12f34b2 "IPC", metric_no_group=<optimized out>, metric_no_merge=<optimized out>, fake_pmu=fake_pmu@entry=0x1462f18 <perf_pmu.fake>, metric_events=0x3ffd84fba58, map=0x1) at util/metricgroup.c:1040 #9 0x0000000001103eb2 in metricgroup__parse_groups_test( evlist=evlist@entry=0x1e70260, map=map@entry=0x13f9828 <map>, str=str@entry=0x12f34b2 "IPC", metric_no_group=metric_no_group@entry=false, metric_no_merge=metric_no_merge@entry=false, metric_events=0x3ffd84fba58) at util/metricgroup.c:1082 #10 0x00000000010c84d8 in __compute_metric (ratio2=0x0, name2=0x0, ratio1=<synthetic pointer>, name1=0x12f34b2 "IPC", vals=0x3ffd84fbad8, name=0x12f34b2 "IPC") at tests/parse-metric.c:159 #11 compute_metric (ratio=<synthetic pointer>, vals=0x3ffd84fbad8, name=0x12f34b2 "IPC") at tests/parse-metric.c:189 #12 test_ipc () at tests/parse-metric.c:208 ..... ..... omitted many more lines This test case was added with commit 218ca91df477 ("perf tests: Add parse metric test for frontend metric"). When I compile with make DEBUG=y it works fine and I do not get a core dump. It turned out that the above listed function call chain worked on a struct pmu_event array which requires a trailing element with zeroes which was missing. The marco map_for_each_event() loops over that array tests for members metric_expr/metric_name/metric_group being non-NULL. Adding this element fixes the issue. Output after: [root@t35lp46 perf]# ./perf test 67 67: Parse and process metrics : Ok [root@t35lp46 perf]# Committer notes: As Ian remarks, this is not s390 specific: <quote Ian> This also shows up with address sanitizer on all architectures (perhaps change the patch title) and perhaps add a "Fixes: <commit>" tag. ================================================================= ==4718==ERROR: AddressSanitizer: global-buffer-overflow on address 0x55c93b4d59e8 at pc 0x55c93a1541e2 bp 0x7ffd24327c60 sp 0x7ffd24327c58 READ of size 8 at 0x55c93b4d59e8 thread T0 #0 0x55c93a1541e1 in find_metric tools/perf/util/metricgroup.c:764:2 #1 0x55c93a153e6c in __resolve_metric tools/perf/util/metricgroup.c:844:9 #2 0x55c93a152f18 in resolve_metric tools/perf/util/metricgroup.c:881:9 #3 0x55c93a1528db in metricgroup__add_metric tools/perf/util/metricgroup.c:943:9 #4 0x55c93a151996 in metricgroup__add_metric_list tools/perf/util/metricgroup.c:988:9 #5 0x55c93a1511b9 in parse_groups tools/perf/util/metricgroup.c:1040:8 #6 0x55c93a1513e1 in metricgroup__parse_groups_test tools/perf/util/metricgroup.c:1082:9 #7 0x55c93a0108ae in __compute_metric tools/perf/tests/parse-metric.c:159:8 #8 0x55c93a010744 in compute_metric tools/perf/tests/parse-metric.c:189:9 #9 0x55c93a00f5ee in test_ipc tools/perf/tests/parse-metric.c:208:2 #10 0x55c93a00f1e8 in test__parse_metric tools/perf/tests/parse-metric.c:345:2 #11 0x55c939fd7202 in run_test tools/perf/tests/builtin-test.c:410:9 #12 0x55c939fd6736 in test_and_print tools/perf/tests/builtin-test.c:440:9 #13 0x55c939fd58c3 in __cmd_test tools/perf/tests/builtin-test.c:661:4 #14 0x55c939fd4e02 in cmd_test tools/perf/tests/builtin-test.c:807:9 #15 0x55c939e4763d in run_builtin tools/perf/perf.c:313:11 #16 0x55c939e46475 in handle_internal_command tools/perf/perf.c:365:8 #17 0x55c939e4737e in run_argv tools/perf/perf.c:409:2 #18 0x55c939e45f7e in main tools/perf/perf.c:539:3 0x55c93b4d59e8 is located 0 bytes to the right of global variable 'pme_test' defined in 'tools/perf/tests/parse-metric.c:17:25' (0x55c93b4d54a0) of size 1352 SUMMARY: AddressSanitizer: global-buffer-overflow tools/perf/util/metricgroup.c:764:2 in find_metric Shadow bytes around the buggy address: 0x0ab9a7692ae0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0ab9a7692af0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0ab9a7692b00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0ab9a7692b10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0ab9a7692b20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 =>0x0ab9a7692b30: 00 00 00 00 00 00 00 00 00 00 00 00 00[f9]f9 f9 0x0ab9a7692b40: f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 0x0ab9a7692b50: f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 0x0ab9a7692b60: f9 f9 f9 f9 f9 f9 f9 f9 00 00 00 00 00 00 00 00 0x0ab9a7692b70: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0ab9a7692b80: f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 f9 Shadow byte legend (one shadow byte represents 8 application bytes): Addressable: 00 Partially addressable: 01 02 03 04 05 06 07 Heap left redzone: fa Freed heap region: fd Stack left redzone: f1 Stack mid redzone: f2 Stack right redzone: f3 Stack after return: f5 Stack use after scope: f8 Global redzone: f9 Global init order: f6 Poisoned by user: f7 Container overflow: fc Array cookie: ac Intra object redzone: bb ASan internal: fe Left alloca redzone: ca Right alloca redzone: cb Shadow gap: cc </quote> I'm also adding the missing "Fixes" tag and setting just .name to NULL, as doing it that way is more compact (the compiler will zero out everything else) and the table iterators look for .name being NULL as the sentinel marking the end of the table. Fixes: 0a507af9c681ac2a ("perf tests: Add parse metric test for ipc metric") Signed-off-by: Thomas Richter <[email protected]> Reviewed-by: Sumanth Korikkar <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Heiko Carstens <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[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]>
2020-09-01perf parse-events: Set exclude_guest=1 for user-space countingJin Yao2-2/+5
Currently if we run 'perf record -e cycles:u', exclude_guest=0. But it doesn't make sense in most cases that we request for user-space counting but we also get the guest report. Of course, we also need to consider 'perf kvm' usage case that authorized perf users on the host may only want to count guest user space events. For example, # perf kvm --guest record -e cycles:u When we have 'exclude_guest=1' for 'perf kvm' usage, we may get nothing from guest events. To keep perf semantics consistent and clear, this patch sets exclude_guest=1 for user-space counting but except for 'perf kvm' usage. Before: perf record -e cycles:u ./div perf evlist -v cycles:u: ..., exclude_kernel: 1, exclude_hv: 1, ... After: perf record -e cycles:u ./div perf evlist -v cycles:u: ..., exclude_kernel: 1, exclude_hv: 1, exclude_guest: 1, ... Before: perf kvm --guest record -e cycles:u -vvv perf_event_attr: size 120 { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|ID|CPU|PERIOD read_format ID disabled 1 inherit 1 exclude_kernel 1 exclude_hv 1 freq 1 sample_id_all 1 After: perf kvm --guest record -e cycles:u -vvv perf_event_attr: size 120 { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|ID|CPU|PERIOD read_format ID disabled 1 inherit 1 exclude_kernel 1 exclude_hv 1 freq 1 sample_id_all 1 For Before/After, exclude_guest are both 0 for perf kvm usage. perf test 6 6: Parse event definition strings : Ok Signed-off-by: Jin Yao <[email protected]> Tested-by: Like Xu <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Jin Yao <[email protected]> Cc: Jiri Olsa <[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]>
2020-09-01perf record: Correct the help info of option "--no-bpf-event"Wei Li1-1/+1
The help info of option "--no-bpf-event" is wrongly described as "record bpf events", correct it. Committer testing: $ perf record -h bpf Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>] --clang-opt <clang options> options passed to clang when compiling BPF scriptlets --clang-path <clang path> clang binary to use for compiling BPF scriptlets --no-bpf-event do not record bpf events $ Fixes: 71184c6ab7e6 ("perf record: Replace option --bpf-event with --no-bpf-event") Signed-off-by: Wei Li <[email protected]> Acked-by: Song Liu <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Hanjun Guo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Li Bin <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-09-01perf tools: Use %zd for size_t printf formats on 32-bitChris Wilson2-2/+2
A couple of trivial fixes for using %zd for size_t in the code supporting the ZSTD compression library. Signed-off-by: Chris Wilson <[email protected]> Acked-by: Jiri Olsa <[email protected]> Cc: Adrian Hunter <[email protected]> Cc: Alexey Budankov <[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]>
2020-08-21perf: arm-spe: Fix check error when synthesizing eventsWei Li1-3/+3
In arm_spe_read_record(), when we are processing an events packet, 'decoder->packet.index' is the length of payload, which has been transformed in payloadlen(). So correct the check of 'idx'. Signed-off-by: Wei Li <[email protected]> Reviewed-by: Leo Yan <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Hanjun Guo <[email protected]> Cc: James Clark <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Mathieu Poirier <[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]>
2020-08-21perf symbols: Add mwait_idle_with_hints.constprop.0 to the list of idle symbolsArnaldo Carvalho de Melo1-0/+1
The "mwait_idle_with_hints" one was already there, some compiler artifact now adds this ".constprop.0" suffix, cover that one too. At some point we need to put these in a special bucket and show it somewhere on the screen. Noticed building the kernel on a fedora:32 system using: gcc version 10.2.1 20200723 (Red Hat 10.2.1-1) (GCC) Cc: Adrian Hunter <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-21perf top: Skip side-band event setup if HAVE_LIBBPF_SUPPORT is not setTiezhu Yang1-0/+2
When I execute 'perf top' without HAVE_LIBBPF_SUPPORT, there exists the following segmentation fault, skip the side-band event setup to fix it, this is similar with commit 1101c872c8c7 ("perf record: Skip side-band event setup if HAVE_LIBBPF_SUPPORT is not set"). [yangtiezhu@linux perf]$ ./perf top <SNIP> perf: Segmentation fault Obtained 6 stack frames. ./perf(sighandler_dump_stack+0x5c) [0x12011b604] [0xffffffc010] ./perf(perf_mmap__read_init+0x3e) [0x1201feeae] ./perf() [0x1200d715c] /lib64/libpthread.so.0(+0xab9c) [0xffee10ab9c] /lib64/libc.so.6(+0x128f4c) [0xffedc08f4c] Segmentation fault [yangtiezhu@linux perf]$ I use git bisect to find commit b38d85ef49cf ("perf bpf: Decouple creating the evlist from adding the SB event") is the first bad commit, so also add the Fixes tag. Committer testing: First build perf explicitely disabling libbpf: $ make NO_LIBBPF=1 O=/tmp/build/perf -C tools/perf install-bin && perf test python Now make sure it isn't linked: $ perf -vv | grep -w bpf bpf: [ OFF ] # HAVE_LIBBPF_SUPPORT $ $ nm ~/bin/perf | grep libbpf $ And now try to run 'perf top': # perf top perf: Segmentation fault -------- backtrace -------- perf[0x5bcd6d] /lib64/libc.so.6(+0x3ca6f)[0x7fd0f5a66a6f] perf(perf_mmap__read_init+0x1e)[0x5e1afe] perf[0x4cc468] /lib64/libpthread.so.0(+0x9431)[0x7fd0f645a431] /lib64/libc.so.6(clone+0x42)[0x7fd0f5b2b912] # Applying this patch fixes the issue. Fixes: b38d85ef49cf ("perf bpf: Decouple creating the evlist from adding the SB event") Signed-off-by: Tiezhu Yang <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[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]>
2020-08-21perf sched timehist: Fix use of CPU list with summary optionDavid Ahern1-1/+5
Do not update thread stats or show idle summary unless CPU is in the list of interest. Fixes: c30d630d1bcfad8d ("perf sched timehist: Add support for filtering on CPU") Signed-off-by: David Ahern <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-21perf test: Fix basic bpf filtering testSumanth Korikkar1-1/+1
BPF basic filtering test fails on s390x (when vmlinux debuginfo is utilized instead of /proc/kallsyms) Info: - bpf_probe_load installs the bpf code at do_epoll_wait. - For s390x, do_epoll_wait resolves to 3 functions including inlines. found inline addr: 0x43769e Probe point found: __s390_sys_epoll_wait+6 found inline addr: 0x437290 Probe point found: do_epoll_wait+0 found inline addr: 0x4375d6 Probe point found: __se_sys_epoll_wait+6 - add_bpf_event creates evsel for every probe in a BPF object. This results in 3 evsels. Solution: - Expected result = 50% of the samples to be collected from epoll_wait * number of entries present in the evlist. Committer testing: # perf test 42 42: BPF filter : 42.1: Basic BPF filtering : Ok 42.2: BPF pinning : Ok 42.3: BPF prologue generation : Ok 42.4: BPF relocation checker : Ok # Signed-off-by: Sumanth Korikkar <[email protected]> Reviewed-by: Thomas Richter <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: [email protected] Cc: Heiko Carstens <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Sven Schnelle <[email protected]> LPU-Reference: [email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-18tools: Remove feature-libelf-mmap feature detectionAndrii Nakryiko2-5/+1
It's trivial to handle missing ELF_C_MMAP_READ support in libelf the way that objtool has solved it in ("774bec3fddcc objtool: Add fallback from ELF_C_READ_MMAP to ELF_C_READ"). So instead of having an entire feature detector for that, just do what objtool does for perf and libbpf. And keep their Makefiles a bit simpler. Signed-off-by: Andrii Nakryiko <[email protected]> Signed-off-by: Alexei Starovoitov <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
2020-08-15Merge tag 'perf-tools-2020-08-14' of ↵Linus Torvalds27-229/+1209
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux Pull more perf tools updates from Arnaldo Carvalho de Melo: "Fixes: - Fixes for 'perf bench numa'. - Always memset source before memcpy in 'perf bench mem'. - Quote CC and CXX for their arguments to fix build in environments using those variables to pass more than just the compiler names. - Fix module symbol processing, addressing regression detected via "perf test". - Allow multiple probes in record+script_probe_vfs_getname.sh 'perf test' entry. Improvements: - Add script to autogenerate socket family name id->string table from copy of kernel header, used so far in 'perf trace'. - 'perf ftrace' improvements to provide similar options for this utility so that one can go from 'perf record', 'perf trace', etc to 'perf ftrace' just by changing the name of the subcommand. - Prefer new "sched:sched_waking" trace event when it exists in 'perf sched' post processing. - Update POWER9 metrics to utilize other metrics. - Fall back to querying debuginfod if debuginfo not found locally. Miscellaneous: - Sync various kvm headers with kernel sources" * tag 'perf-tools-2020-08-14' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (40 commits) perf ftrace: Make option description initials all capital letters perf build-ids: Fall back to debuginfod query if debuginfo not found perf bench numa: Remove dead code in parse_nodes_opt() perf stat: Update POWER9 metrics to utilize other metrics perf ftrace: Add change log perf: ftrace: Add set_tracing_options() to set all trace options perf ftrace: Add option --tid to filter by thread id perf ftrace: Add option -D/--delay to delay tracing perf: ftrace: Allow set graph depth by '--graph-opts' perf ftrace: Add support for trace option tracing_thresh perf ftrace: Add option 'verbose' to show more info for graph tracer perf ftrace: Add support for tracing option 'irq-info' perf ftrace: Add support for trace option funcgraph-irqs perf ftrace: Add support for trace option sleep-time perf ftrace: Add support for tracing option 'func_stack_trace' perf tools: Add general function to parse sublevel options perf ftrace: Add option '--inherit' to trace children processes perf ftrace: Show trace column header perf ftrace: Add option '-m/--buffer-size' to set per-cpu buffer size perf ftrace: Factor out function write_tracing_file_int() ...
2020-08-14all arch: remove system call sys_sysctlXiaoming Ni3-3/+3
Since commit 61a47c1ad3a4dc ("sysctl: Remove the sysctl system call"), sys_sysctl is actually unavailable: any input can only return an error. We have been warning about people using the sysctl system call for years and believe there are no more users. Even if there are users of this interface if they have not complained or fixed their code by now they probably are not going to, so there is no point in warning them any longer. So completely remove sys_sysctl on all architectures. [[email protected]: s390: fix build error for sys_call_table_emu] Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Xiaoming Ni <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Acked-by: Will Deacon <[email protected]> [arm/arm64] Acked-by: "Eric W. Biederman" <[email protected]> Cc: Aleksa Sarai <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Al Viro <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Andrew Morton <[email protected]> Cc: Andy Lutomirski <[email protected]> Cc: Arnaldo Carvalho de Melo <[email protected]> Cc: Arnd Bergmann <[email protected]> Cc: Benjamin Herrenschmidt <[email protected]> Cc: Bin Meng <[email protected]> Cc: Borislav Petkov <[email protected]> Cc: Brian Gerst <[email protected]> Cc: Catalin Marinas <[email protected]> Cc: chenzefeng <[email protected]> Cc: Christian Borntraeger <[email protected]> Cc: Christian Brauner <[email protected]> Cc: Chris Zankel <[email protected]> Cc: David Howells <[email protected]> Cc: David S. Miller <[email protected]> Cc: Diego Elio Pettenò <[email protected]> Cc: Dmitry Vyukov <[email protected]> Cc: Dominik Brodowski <[email protected]> Cc: Fenghua Yu <[email protected]> Cc: Geert Uytterhoeven <[email protected]> Cc: Heiko Carstens <[email protected]> Cc: Helge Deller <[email protected]> Cc: "H. Peter Anvin" <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Iurii Zaikin <[email protected]> Cc: Ivan Kokshaysky <[email protected]> Cc: James Bottomley <[email protected]> Cc: Jens Axboe <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Kars de Jong <[email protected]> Cc: Kees Cook <[email protected]> Cc: Krzysztof Kozlowski <[email protected]> Cc: Luis Chamberlain <[email protected]> Cc: Marco Elver <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Martin K. Petersen <[email protected]> Cc: Masahiro Yamada <[email protected]> Cc: Matt Turner <[email protected]> Cc: Max Filippov <[email protected]> Cc: Michael Ellerman <[email protected]> Cc: Michal Simek <[email protected]> Cc: Miklos Szeredi <[email protected]> Cc: Minchan Kim <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Naveen N. Rao <[email protected]> Cc: Nick Piggin <[email protected]> Cc: Oleg Nesterov <[email protected]> Cc: Olof Johansson <[email protected]> Cc: Paul Burton <[email protected]> Cc: "Paul E. McKenney" <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra (Intel) <[email protected]> Cc: Randy Dunlap <[email protected]> Cc: Ravi Bangoria <[email protected]> Cc: Richard Henderson <[email protected]> Cc: Rich Felker <[email protected]> Cc: Russell King <[email protected]> Cc: Sami Tolvanen <[email protected]> Cc: Sargun Dhillon <[email protected]> Cc: Stephen Rothwell <[email protected]> Cc: Sudeep Holla <[email protected]> Cc: Sven Schnelle <[email protected]> Cc: Thiago Jung Bauermann <[email protected]> Cc: Thomas Bogendoerfer <[email protected]> Cc: Thomas Gleixner <[email protected]> Cc: Tony Luck <[email protected]> Cc: Vasily Gorbik <[email protected]> Cc: Vlastimil Babka <[email protected]> Cc: Yoshinori Sato <[email protected]> Cc: Zhou Yanjie <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Linus Torvalds <[email protected]>
2020-08-14perf ftrace: Make option description initials all capital lettersArnaldo Carvalho de Melo1-14/+14
And improve a bit the -m description to state that a B/K/M/G suffix is needed. Cc: Changbin Du <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf build-ids: Fall back to debuginfod query if debuginfo not foundFrank Ch. Eigler3-0/+29
During a perf-record, use the -ldebuginfod API to query a debuginfod server, should the debug data not be found in the usual system locations. If successful, the usual $HOME/.debug dir is populated. Tested with: $ find . . ./ctags-debuginfo-5.8-26.fc31.x86_64.rpm ./usr ./usr/lib ./usr/lib/debug ./usr/lib/debug/.build-id ./usr/lib/debug/.build-id/ca ./usr/lib/debug/.build-id/ca/46f6ae6a0cee57d85abc1d461c49074248908d ./usr/lib/debug/.build-id/ca/46f6ae6a0cee57d85abc1d461c49074248908d.debug ./usr/lib/debug/usr ./usr/lib/debug/usr/bin ./usr/lib/debug/usr/bin/ctags-5.8-26.fc31.x86_64.debug $ debuginfod -F . ... $ rm -rf ~/.debug/ ; mkdir ~/.debug $ perf record make tags BUILD: Doing 'make -j8' parallel build GEN tags [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.107 MB perf.data (1483 samples) ] $ find ~/.debug | grep ctags /home/jolsa/.debug/usr/bin/ctags /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d/elf /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d/probes $ rm -rf ~/.debug/ ; mkdir ~/.debug $ DEBUGINFOD_URLS=http://localhost:8002 perf record make tags BUILD: Doing 'make -j8' parallel build GEN tags [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.108 MB perf.data (1531 samples) ] $ find ~/.debug | grep ctag /home/jolsa/.debug/usr/bin/ctags /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d/debug /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d/elf /home/jolsa/.debug/usr/bin/ctags/ca46f6ae6a0cee57d85abc1d461c49074248908d/probes Note the 'debug' file is created in the last run. Note that currently the debuginfo data are downloaded only on record path, we still need add this support to perf build-id/report.. and test ;-) Tested-by: Jiri Olsa <[email protected]> Signed-off-by: Jiri Olsa <[email protected]> Signed-off-by: Frank Ch. Eigler <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf bench numa: Remove dead code in parse_nodes_opt()Peng Fan1-2/+0
In the function parse_nodes_opt(), the statement "return 0;" is dead code, remove it. Signed-off-by: Peng Fan <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Jiri Olsa <[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]>
2020-08-14perf stat: Update POWER9 metrics to utilize other metricsPaul A. Clarke1-24/+24
These changes take advantage of the new capability added in merge commit 00e4db51259a5f936fec1424b884f029479d3981 "Allow using computed metrics in calculating other metrics". The net is a simplification of the expressions for a handful of metrics, but no functional change. Signed-off-by: Paul Clarke <[email protected]> Reviewed-by: Kajol Jain <[email protected]> Acked-by: Ian Rogers <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Madhavan Srinivasan <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add change logChangbin Du1-0/+1
Add a change log after previous enhancements. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf: ftrace: Add set_tracing_options() to set all trace optionsChangbin Du1-55/+63
Now the __cmd_ftrace() becomes a bit long. This moves the trace option setting code to a separate function set_tracing_options(). Suggested-by: Namhyung Kim <[email protected]> Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option --tid to filter by thread idChangbin Du2-0/+6
This allows us to trace single thread instead of the whole process. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option -D/--delay to delay tracingChangbin Du2-3/+20
This adds an option '-D/--delay' to allow us to start tracing some times later after workload is launched. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf: ftrace: Allow set graph depth by '--graph-opts'Changbin Du2-7/+3
This is to have a consistent view of all graph tracer options. The original option '--graph-depth' is marked as deprecated. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add support for trace option tracing_threshChangbin Du2-1/+26
This adds an option '--graph-opts thresh' to setup trace duration threshold for funcgraph tracer. $ sudo ./perf ftrace -G '*' --graph-opts thresh=100 3) ! 184.060 us | } /* schedule */ 3) ! 185.600 us | } /* exit_to_usermode_loop */ 2) ! 225.989 us | } /* schedule_idle */ 2) # 4140.051 us | } /* do_idle */ Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option 'verbose' to show more info for graph tracerChangbin Du2-1/+29
Sometimes we want ftrace display more and longer information about the trace. $ sudo perf ftrace -G '*' 2) 0.979 us | mutex_unlock(); 2) 1.540 us | __fsnotify_parent(); 2) 0.433 us | fsnotify(); $ sudo perf ftrace -G '*' --graph-opts verbose 14160.770883 | 0) <...>-47814 | .... | 1.289 us | mutex_unlock(); 14160.770886 | 0) <...>-47814 | .... | 1.624 us | __fsnotify_parent(); 14160.770887 | 0) <...>-47814 | .... | 0.636 us | fsnotify(); 14160.770888 | 0) <...>-47814 | .... | 0.328 us | __sb_end_write(); 14160.770888 | 0) <...>-47814 | d... | 0.430 us | fpregs_assert_state_consistent(); 14160.770889 | 0) <...>-47814 | d... | | do_syscall_64() { 14160.770889 | 0) <...>-47814 | .... | | __x64_sys_close() { Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add support for tracing option 'irq-info'Changbin Du2-1/+21
This adds support to display irq context info for function tracer. To do this, just specify a '--func-opts irq-info' option. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add support for trace option funcgraph-irqsChangbin Du2-1/+21
This adds an option '--graph-opts noirqs' to filter out functions executed in irq context. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add support for trace option sleep-timeChangbin Du2-0/+45
This adds an option '--graph-opts nosleep-time' which allow us only to measure on-CPU time. This option is function_graph tracer only. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add support for tracing option 'func_stack_trace'Changbin Du2-0/+46
This adds support to display call trace for function tracer. To do this, just specify a '--func-opts call-graph' option. Example: $ sudo perf ftrace -T vfs_read --func-opts call-graph iio-sensor-prox-855 [003] 6168.369657: vfs_read <-ksys_read iio-sensor-prox-855 [003] 6168.369677: <stack trace> => vfs_read => ksys_read => __x64_sys_read => do_syscall_64 => entry_SYSCALL_64_after_hwframe ... Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf tools: Add general function to parse sublevel optionsChangbin Du4-44/+99
This factors out a general function perf_parse_sublevel_options() to parse sublevel options. The 'sublevel' options is something like the '--debug' options which allow more sublevel options. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option '--inherit' to trace children processesChangbin Du2-0/+41
This adds an option '--inherit' to allow us trace children processes spawned by our target. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Show trace column headerChangbin Du1-0/+3
This makes 'perf ftrace' display column header before printing trace. $ sudo perf ftrace # tracer: function # # entries-in-buffer/entries-written: 0/0 #P:8 # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | <...>-9246 [006] 10726.262760: mutex_unlock <-rb_simple_write <...>-9246 [006] 10726.262764: __fsnotify_parent <-vfs_write <...>-9246 [006] 10726.262765: fsnotify <-vfs_write <...>-9246 [006] 10726.262766: __sb_end_write <-vfs_write <...>-9246 [006] 10726.262767: fpregs_assert_state_consistent <-do_syscall_64 Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option '-m/--buffer-size' to set per-cpu buffer sizeChangbin Du2-0/+60
This adds an option '-m/--buffer-size' to allow us set the size of per-cpu tracing buffer. Committer testing: Before running with this option: # find /sys/kernel/tracing/ -name buffer_size_kb | xargs cat 1408 1408 1408 1408 1408 1408 1408 1408 1408 # Then, run: # perf ftrace -m 2048K | head -10 2) | mutex_unlock() { 2) ==========> | 2) | smp_irq_work_interrupt() { 2) | irq_enter() { 2) 0.121 us | rcu_irq_enter(); 2) 0.128 us | irqtime_account_irq(); 2) 0.719 us | } 2) | __wake_up() { 2) | __wake_up_common_lock() { 2) 0.105 us | _raw_spin_lock_irqsave(); # Now look at those tracefs knobs: # find /sys/kernel/tracing/ -name buffer_size_kb | xargs cat 2048 2048 2048 2048 2048 2048 2048 2048 2048 # This should be similar to the -m option in the other perf tools, such as 'perf record', 'perf trace', etc. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Factor out function write_tracing_file_int()Changbin Du1-5/+12
We will reuse this function later. Signed-off-by: Changbin Du <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Add option '-F/--funcs' to list available functionsChangbin Du2-0/+50
This adds an option '-F/--funcs' to list all available functions to trace, which is read from tracing file 'available_filter_functions'. $ sudo ./perf ftrace -F | head trace_initcall_finish_cb initcall_blacklisted do_one_initcall do_one_initcall trace_initcall_start_cb run_init_process try_to_run_init_process match_dev_by_label match_dev_by_uuid rootfs_init_fs_context $ Committer notes: This is the same command line option and for the same purpose as in 'perf probe'. Signed-off-by: Changbin Du <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-14perf ftrace: Select function/function_graph tracer automaticallyChangbin Du3-23/+48
The '-g/-G' options have already implied function_graph tracer should be used instead of function tracer. So we don't need extra option '--tracer' in this case. This patch changes the behavior as below: - If '-g' or '-G' option is on, then function_graph tracer is used. - If '-T' or '-N' option is on, then function tracer is used. - The function_graph has priority over function tracer. - The option '--tracer' only take effect if neither -g/-G nor -T/-N is specified. Here are some examples. This will start tracing all functions using default tracer: $ sudo perf ftrace This will trace all functions using function graph tracer: $ sudo perf ftrace -G '*' This will trace function vfs_read using function graph tracer: $ sudo perf ftrace -G vfs_read This will trace function vfs_read using function tracer: $ sudo perf ftrace -T vfs_read Committer notes: Using '-h -G' will tell what that option is about, so to further clarify the above examples: # perf ftrace -h -G -G, --graph-funcs <func> Set graph filter on given functions # perf ftrace -h -g -g, --nograph-funcs <func> Set nograph filter on given functions # perf ftrace -h -T -T, --trace-funcs <func> trace given functions only # perf ftrace -h -N -N, --notrace-funcs <func> do not trace given functions # Signed-off-by: Changbin Du <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Steven Rostedt (VMware) <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf bench numa: Use numa_node_to_cpus() to bind tasks to nodesAlexander Gordeev1-17/+16
It is currently assumed that each node contains at most nr_cpus/nr_nodes CPUs and nodes' CPU ranges do not overlap. That assumption is generally incorrect as there are archs where a CPU number does not depend on to its node number. This update removes the described assumption by simply calling numa_node_to_cpus() interface and using the returned mask for binding CPUs to nodes. Also, variable types and names made consistent in functions using cpumask. Signed-off-by: Alexander Gordeev <[email protected]> Reviewed-by: Srikar Dronamraju <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Balamuruhan S <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Naveen N. Rao <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Satheesh Rajendran <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf bench numa: Fix cpumask memory leak in node_has_cpus()Alexander Gordeev1-4/+9
Couple numa_allocate_cpumask() and numa_free_cpumask() functions Signed-off-by: Alexander Gordeev <[email protected]> Reviewed-by: Srikar Dronamraju <[email protected]> Cc: Alexander Shishkin <[email protected]> Cc: Balamuruhan S <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Mark Rutland <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Naveen N. Rao <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Satheesh Rajendran <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf tools: Fix module symbol processingJiri Olsa2-5/+4
The 'dso->kernel' condition is true also for kernel modules now, and there are several places that were omited by the initial change: - we need to identify modules separately in dso__process_kernel_symbol - we need to set 'dso->kernel' for module from buildid table - there's no need to use 'dso->kernel || kmodule' in one condition Committer testing: Before: # perf test -v object <SNIP> Objdump command is: objdump -z -d --start-address=0xffffffff813e682f --stop-address=0xffffffff813e68af /usr/lib/debug/lib/modules/5.7.14-200.fc32.x86_64/vmlinux Bytes read match those read by objdump Reading object code for memory address: 0xffffffffc02dc257 File is: /lib/modules/5.7.14-200.fc32.x86_64/kernel/arch/x86/crypto/crc32c-intel.ko.xz On file address is: 0xffffffffc02dc2e7 dso__data_read_offset failed test child finished with -1 ---- end ---- Object code reading: FAILED! # After: # perf test object 26: Object code reading : Ok # perf test object 26: Object code reading : Ok # perf test object 26: Object code reading : Ok # perf test object 26: Object code reading : Ok # perf test object 26: Object code reading : Ok # Fixes: 02213cec64bb ("perf maps: Mark module DSOs with kernel type") Reported-by: Arnaldo Carvalho de Melo <[email protected]> Tested-by: Arnaldo Carvalho de Melo <[email protected]> Cc: Adrian Hunter <[email protected]> Cc: Namhyung Kim <[email protected]> Signed-off-by: Jiri Olsa <[email protected]>
2020-08-13perf tools: Rename 'enum dso_kernel_type' to 'enum dso_space_type'Jiri Olsa7-35/+35
Rename enum dso_kernel_type to enum dso_space_type, which seems like better fit. Committer notes: This is used with 'struct dso'->kernel, which once was a boolean, so DSO_SPACE__USER is zero, !zero means some sort of kernel space, be it the host kernel space or a guest kernel space. Signed-off-by: Jiri Olsa <[email protected]> Cc: Adrian Hunter <[email protected]> Cc: Jiri Olsa <[email protected]> Cc: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf test: Allow multiple probes in record+script_probe_vfs_getname.shMichael Petlan1-2/+2
Sometimes when adding a kprobe by perf, it results in multiple probe points, such as the following: # ./perf probe -l probe:vfs_getname (on getname_flags:73@fs/namei.c with pathname) probe:vfs_getname_1 (on getname_flags:73@fs/namei.c with pathname) probe:vfs_getname_2 (on getname_flags:73@fs/namei.c with pathname) # cat /sys/kernel/debug/tracing/kprobe_events p:probe/vfs_getname _text+5501804 pathname=+0(+0(%gpr31)):string p:probe/vfs_getname_1 _text+5505388 pathname=+0(+0(%gpr31)):string p:probe/vfs_getname_2 _text+5508396 pathname=+0(+0(%gpr31)):string In this test, we need to record all of them and expect any of them in the perf-script output, since it's not clear which one will be used for the desired syscall: # perf stat -e probe:vfs_getname\* -- touch /tmp/nic Performance counter stats for 'touch /tmp/nic': 31 probe:vfs_getname_2 0 probe:vfs_getname_1 1 probe:vfs_getname 0.001421826 seconds time elapsed 0.001506000 seconds user 0.000000000 seconds sys If the test relies only on probe:vfs_getname, it might easily miss the relevant data. Signed-off-by: Michael Petlan <[email protected]> Cc: Jiri Olsa <[email protected]> LPU-Reference: [email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf bench mem: Always memset source before memcpyVincent Whitchurch1-10/+11
For memcpy, the source pages are memset to zero only when --cycles is used. This leads to wildly different results with or without --cycles, since all sources pages are likely to be mapped to the same zero page without explicit writes. Before this fix: $ export cmd="./perf stat -e LLC-loads -- ./perf bench \ mem memcpy -s 1024MB -l 100 -f default" $ $cmd 2,935,826 LLC-loads 3.821677452 seconds time elapsed $ $cmd --cycles 217,533,436 LLC-loads 8.616725985 seconds time elapsed After this fix: $ $cmd 214,459,686 LLC-loads 8.674301124 seconds time elapsed $ $cmd --cycles 214,758,651 LLC-loads 8.644480006 seconds time elapsed Fixes: 47b5757bac03c338 ("perf bench mem: Move boilerplate memory allocation to the infrastructure") Signed-off-by: Vincent Whitchurch <[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: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-13perf sched: Prefer sched_waking event when it existsDavid Ahern1-3/+29
Commit fbd705a0c618 ("sched: Introduce the 'trace_sched_waking' tracepoint") added sched_waking tracepoint which should be preferred over sched_wakeup when analyzing scheduling delays. Update 'perf sched record' to collect sched_waking events if it exists and fallback to sched_wakeup if it does not. Similarly, update timehist command to skip sched_wakeup events if the session includes sched_waking (ie., sched_waking is preferred over sched_wakeup). Signed-off-by: David Ahern <[email protected]> Acked-by: Namhyung Kim <[email protected]> Cc: Jiri Olsa <[email protected]> Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2020-08-12perf bench: Fix a couple of spelling mistakes in options textColin Ian King1-2/+2
There are a couple of spelling mistakes in the text. Fix these. Signed-off-by: Colin King <[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: [email protected] Link: http://lore.kernel.org/lkml/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.