From af9eb56bfed273a85b8c3f99d3ed7ff979c36ae0 Mon Sep 17 00:00:00 2001 From: Changbin Du Date: Tue, 18 Apr 2023 11:18:25 +0800 Subject: perf script: Add new output field 'dsoff' to print dso offset This adds a new 'dsoff' field to print dso offset for resolved symbols, and the offset is appended to dso name. Default output: $ perf script ls 2695501 3011030.487017: 500000 cycles: 152cc73ef4b5 get_common_indices.constprop.0+0x155 (/usr/lib/x86_64-linux-gnu/ld-2.31.so) ls 2695501 3011030.487018: 500000 cycles: ffffffff99045b3e [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffff9968e107 [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffffc1f54afb [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffff9968382f [unknown] ([unknown]) ls 2695501 3011030.487019: 500000 cycles: ffffffff99e00094 [unknown] ([unknown]) ls 2695501 3011030.487019: 500000 cycles: 152cc718a8d0 __errno_location@plt+0x0 (/usr/lib/x86_64-linux-gnu/libselinux.so.1) Display 'dsoff' field: $ perf script -F +dsoff ls 2695501 3011030.487017: 500000 cycles: 152cc73ef4b5 get_common_indices.constprop.0+0x155 (/usr/lib/x86_64-linux-gnu/ld-2.31.so+0x1c4b5) ls 2695501 3011030.487018: 500000 cycles: ffffffff99045b3e [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffff9968e107 [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffffc1f54afb [unknown] ([unknown]) ls 2695501 3011030.487018: 500000 cycles: ffffffff9968382f [unknown] ([unknown]) ls 2695501 3011030.487019: 500000 cycles: ffffffff99e00094 [unknown] ([unknown]) ls 2695501 3011030.487019: 500000 cycles: 152cc718a8d0 __errno_location@plt+0x0 (/usr/lib/x86_64-linux-gnu/libselinux.so.1+0x68d0) ls 2695501 3011030.487019: 500000 cycles: ffffffff992a6db0 [unknown] ([unknown]) Signed-off-by: Changbin Du Acked-by: Adrian Hunter Cc: Alexander Shishkin Cc: Hui Wang Cc: Ian Rogers Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20230418031825.1262579-4-changbin.du@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 60 ++++++++++++++++++--------------------------- 1 file changed, 24 insertions(+), 36 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index c57be48d65bb..029d5a597233 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -133,6 +133,7 @@ enum perf_output_field { PERF_OUTPUT_VCPU = 1ULL << 38, PERF_OUTPUT_CGROUP = 1ULL << 39, PERF_OUTPUT_RETIRE_LAT = 1ULL << 40, + PERF_OUTPUT_DSOFF = 1ULL << 41, }; struct perf_script { @@ -174,6 +175,7 @@ struct output_option { {.str = "ip", .field = PERF_OUTPUT_IP}, {.str = "sym", .field = PERF_OUTPUT_SYM}, {.str = "dso", .field = PERF_OUTPUT_DSO}, + {.str = "dsoff", .field = PERF_OUTPUT_DSOFF}, {.str = "addr", .field = PERF_OUTPUT_ADDR}, {.str = "symoff", .field = PERF_OUTPUT_SYMOFFSET}, {.str = "srcline", .field = PERF_OUTPUT_SRCLINE}, @@ -574,6 +576,9 @@ static void set_print_ip_opts(struct perf_event_attr *attr) if (PRINT_FIELD(DSO)) output[type].print_ip_opts |= EVSEL__PRINT_DSO; + if (PRINT_FIELD(DSOFF)) + output[type].print_ip_opts |= EVSEL__PRINT_DSOFF; + if (PRINT_FIELD(SYMOFFSET)) output[type].print_ip_opts |= EVSEL__PRINT_SYMOFFSET; @@ -627,6 +632,10 @@ static int perf_session__check_output_opt(struct perf_session *session) if (evsel == NULL) continue; + /* 'dsoff' implys 'dso' field */ + if (output[j].fields & PERF_OUTPUT_DSOFF) + output[j].fields |= PERF_OUTPUT_DSO; + set_print_ip_opts(&evsel->core.attr); tod |= output[j].fields & PERF_OUTPUT_TOD; } @@ -929,18 +938,12 @@ static int perf_sample__fprintf_brstack(struct perf_sample *sample, } printed += fprintf(fp, " 0x%"PRIx64, from); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alf.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alf.map, PRINT_FIELD(DSOFF), alf.addr, fp); printed += fprintf(fp, "/0x%"PRIx64, to); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alt.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); printed += print_bstack_flags(fp, entries + i); } @@ -972,18 +975,12 @@ static int perf_sample__fprintf_brstacksym(struct perf_sample *sample, thread__find_symbol_fb(thread, sample->cpumode, to, &alt); printed += symbol__fprintf_symname_offs(alf.sym, &alf, fp); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alf.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alf.map, PRINT_FIELD(DSOFF), alf.addr, fp); printed += fprintf(fp, "%c", '/'); printed += symbol__fprintf_symname_offs(alt.sym, &alt, fp); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alt.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); printed += print_bstack_flags(fp, entries + i); } @@ -1019,17 +1016,11 @@ static int perf_sample__fprintf_brstackoff(struct perf_sample *sample, to = map__dso_map_ip(alt.map, to); printed += fprintf(fp, " 0x%"PRIx64, from); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alf.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alf.map, PRINT_FIELD(DSOFF), alf.addr, fp); printed += fprintf(fp, "/0x%"PRIx64, to); - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, "("); - printed += map__fprintf_dsoname(alt.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); printed += print_bstack_flags(fp, entries + i); } @@ -1393,11 +1384,8 @@ static int perf_sample__fprintf_addr(struct perf_sample *sample, printed += symbol__fprintf_symname(al.sym, fp); } - if (PRINT_FIELD(DSO)) { - printed += fprintf(fp, " ("); - printed += map__fprintf_dsoname(al.map, fp); - printed += fprintf(fp, ")"); - } + if (PRINT_FIELD(DSO)) + printed += map__fprintf_dsoname_dsoff(al.map, PRINT_FIELD(DSOFF), al.addr, fp); out: return printed; } @@ -3883,7 +3871,7 @@ int cmd_script(int argc, const char **argv) "comma separated output fields prepend with 'type:'. " "+field to add and -field to remove." "Valid types: hw,sw,trace,raw,synth. " - "Fields: comm,tid,pid,time,cpu,event,trace,ip,sym,dso," + "Fields: comm,tid,pid,time,cpu,event,trace,ip,sym,dso,dsoff" "addr,symoff,srcline,period,iregs,uregs,brstack," "brstacksym,flags,data_src,weight,bpf-output,brstackinsn," "brstackinsnlen,brstackoff,callindent,insn,insnlen,synth," -- cgit From fe8e04348727f992f6fce3709639fb6d92a81137 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Wed, 31 May 2023 13:32:36 -0700 Subject: perf script: Increase PID/TID width for output On large systems, it's common that PID/TID is bigger than 5-digit and it makes the output unaligned. Let's increase the width to 7. Before: $ perf script ... swapper 0 [006] 1540823.803935: 1369324 cycles:P: ffffffff9c755588 ktime_get+0x18 ([kernel.kallsyms]) gvfsd-dnssd 95114 [004] 1540823.804164: 1643871 cycles:P: ffffffff9cfdca5c __get_user_8+0x1c ([kernel.kallsyms]) perf-exec 1558582 [000] 1540823.804209: 1018714 cycles:P: ffffffff9c924ab9 __slab_free+0x9 ([kernel.kallsyms]) nmcli 1558589 [007] 1540823.804384: 1859212 cycles:P: 7f70537a8ad8 __strchrnul_evex+0x18 (/usr/lib/x86_64-linux-gnu/libc.so.6> sleep 1558582 [000] 1540823.804456: 987425 cycles:P: 7fd35bb27b30 _dl_init+0x0 (/usr/lib/x86_64-linux-gnu/ld-linux-x86-64.so.2> dbus-daemon 3043 [003] 1540823.804575: 1564465 cycles:P: ffffffff9cb2bb70 llist_add_batch+0x0 ([kernel.kallsyms]) gdbus 1558592 [001] 1540823.804766: 1315219 cycles:P: ffffffff9c797b2e audit_filter_syscall+0x9e ([kernel.kallsyms]) NetworkManager 3452 [005] 1540823.805301: 1558782 cycles:P: 7fa957737748 g_bit_lock+0x58 (/usr/lib/x86_64-linux-gnu/libglib-2.0.so.0.7400.5> After: $ perf script ... swapper 0 [006] 1540823.803935: 1369324 cycles:P: ffffffff9c755588 ktime_get+0x18 ([kernel.kallsyms]) gvfsd-dnssd 95114 [004] 1540823.804164: 1643871 cycles:P: ffffffff9cfdca5c __get_user_8+0x1c ([kernel.kallsyms]) perf-exec 1558582 [000] 1540823.804209: 1018714 cycles:P: ffffffff9c924ab9 __slab_free+0x9 ([kernel.kallsyms]) nmcli 1558589 [007] 1540823.804384: 1859212 cycles:P: 7f70537a8ad8 __strchrnul_evex+0x18 (/usr/lib/x86_64-linux-gnu/libc.so.6> sleep 1558582 [000] 1540823.804456: 987425 cycles:P: 7fd35bb27b30 _dl_init+0x0 (/usr/lib/x86_64-linux-gnu/ld-linux-x86-64.so.2> dbus-daemon 3043 [003] 1540823.804575: 1564465 cycles:P: ffffffff9cb2bb70 llist_add_batch+0x0 ([kernel.kallsyms]) gdbus 1558592 [001] 1540823.804766: 1315219 cycles:P: ffffffff9c797b2e audit_filter_syscall+0x9e ([kernel.kallsyms]) NetworkManager 3452 [005] 1540823.805301: 1558782 cycles:P: 7fa957737748 g_bit_lock+0x58 (/usr/lib/x86_64-linux-gnu/libglib-2.0.so.0.7400.5> Reviewer notes: Adrian added: "Might be worth noting that currently the biggest PID_MAX_LIMIT is 2^22 so pids don't get bigger than 7 digits presently" $ echo $((2 ** 22)) 4194304 $ echo -n $((2 ** 22)) | wc -c 7 $ Signed-off-by: Namhyung Kim Acked-by: Adrian Hunter Tested-by: Arnaldo Carvalho de Melo Cc: Ian Rogers Cc: Jiri Olsa Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20230531203236.1602054-1-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 029d5a597233..70549fc93b12 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -801,11 +801,11 @@ static int perf_sample__fprintf_start(struct perf_script *script, } if (PRINT_FIELD(PID) && PRINT_FIELD(TID)) - printed += fprintf(fp, "%5d/%-5d ", sample->pid, sample->tid); + printed += fprintf(fp, "%7d/%-7d ", sample->pid, sample->tid); else if (PRINT_FIELD(PID)) - printed += fprintf(fp, "%5d ", sample->pid); + printed += fprintf(fp, "%7d ", sample->pid); else if (PRINT_FIELD(TID)) - printed += fprintf(fp, "%5d ", sample->tid); + printed += fprintf(fp, "%7d ", sample->tid); if (PRINT_FIELD(CPU)) { if (latency_format) -- cgit From 36d3e4138e1b6cc9ab179f3f397b5548f8b1eaae Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 6 Jun 2023 16:11:10 -0300 Subject: perf script: Fix allocation of evsel->priv related to per-event dump files When printing output we may want to generate per event files, where the --per-event-dump option should be used, creating perf.data.EVENT.dump files instead of printing to stdout. The callback thar processes event thus expects that evsel->priv->fp should point to either the per-event FILE descriptor or to stdout. The a3af66f51bd0bca7 ("perf script: Fix crash because of missing evsel->priv") changeset fixed a case where evsel->priv wasn't setup, thus set to NULL, causing a segfault when trying to access evsel->priv->fp. But it did it for the non --per-event-dump case by allocating a 'struct perf_evsel_script' just to set its ->fp to stdout. Since evsel->priv is only freed when --per-event-dump is used, we ended up with a memory leak, detected using ASAN. Fix it by using the same method as perf_script__setup_per_event_dump(), and reuse that static 'struct perf_evsel_script'. Also check if evsel_script__new() failed. Fixes: a3af66f51bd0bca7 ("perf script: Fix crash because of missing evsel->priv") Reported-by: Ian Rogers Tested-by: Ian Rogers Cc: Adrian Hunter Cc: Jiri Olsa Cc: Namhyung Kim Cc: Ravi Bangoria Link: https://lore.kernel.org/lkml/ZH+F0wGAWV14zvMP@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 70549fc93b12..b02ad386a55b 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -2410,6 +2410,9 @@ out_put: return ret; } +// Used when scr->per_event_dump is not set +static struct evsel_script es_stdout; + static int process_attr(struct perf_tool *tool, union perf_event *event, struct evlist **pevlist) { @@ -2418,7 +2421,6 @@ static int process_attr(struct perf_tool *tool, union perf_event *event, struct evsel *evsel, *pos; u64 sample_type; int err; - static struct evsel_script *es; err = perf_event__process_attr(tool, event, pevlist); if (err) @@ -2428,14 +2430,13 @@ static int process_attr(struct perf_tool *tool, union perf_event *event, evsel = evlist__last(*pevlist); if (!evsel->priv) { - if (scr->per_event_dump) { + if (scr->per_event_dump) { evsel->priv = evsel_script__new(evsel, scr->session->data); - } else { - es = zalloc(sizeof(*es)); - if (!es) + if (!evsel->priv) return -ENOMEM; - es->fp = stdout; - evsel->priv = es; + } else { // Replicate what is done in perf_script__setup_per_event_dump() + es_stdout.fp = stdout; + evsel->priv = &es_stdout; } } @@ -2741,7 +2742,6 @@ out_err_fclose: static int perf_script__setup_per_event_dump(struct perf_script *script) { struct evsel *evsel; - static struct evsel_script es_stdout; if (script->per_event_dump) return perf_script__fopen_per_event_dump(script); -- cgit From ee84a3032b74055feed192a727e872b0a18d1140 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Thu, 8 Jun 2023 16:28:00 -0700 Subject: perf thread: Add accessor functions for thread Using accessors will make it easier to add reference count checking in later patches. Committer notes: thread->nsinfo wasn't wrapped as it is used together with nsinfo__zput(), where does a trick to set the field with a refcount being dropped to NULL, and that doesn't work well with using thread__nsinfo(thread), that loses the &thread->nsinfo pointer. When refcount checking is added to 'struct thread', later in this series, nsinfo__zput(RC_CHK_ACCESS(thread)->nsinfo) will be used to check the thread pointer. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ali Saidi Cc: Andi Kleen Cc: Athira Rajeev Cc: Brian Robbins Cc: Changbin Du Cc: Dmitrii Dolgov <9erthalion6@gmail.com> Cc: Fangrui Song Cc: German Gomez Cc: Ingo Molnar Cc: Ivan Babrou Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: K Prateek Nayak Cc: Kan Liang Cc: Leo Yan Cc: Liam Howlett Cc: Mark Rutland Cc: Miguel Ojeda Cc: Mike Leach Cc: Namhyung Kim Cc: Naveen N. Rao Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Sean Christopherson Cc: Steinar H. Gunderson Cc: Suzuki Poulouse Cc: Wenyu Liu Cc: Will Deacon Cc: Yang Jihong Cc: Ye Xingchen Cc: Yuan Can Cc: coresight@lists.linaro.org Cc: linux-arm-kernel@lists.infradead.org Link: https://lore.kernel.org/r/20230608232823.4027869-4-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index b02ad386a55b..e756290de2ac 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -1142,7 +1142,7 @@ static int print_srccode(struct thread *thread, u8 cpumode, uint64_t addr) if (!al.map) return 0; ret = map__fprintf_srccode(al.map, al.addr, stdout, - &thread->srccode_state); + thread__srccode_state(thread)); if (ret) ret += printf("\n"); return ret; @@ -1439,7 +1439,7 @@ static int perf_sample__fprintf_callindent(struct perf_sample *sample, * The 'return' has already been popped off the stack so the depth has * to be adjusted to match the 'call'. */ - if (thread->ts && sample->flags & PERF_IP_FLAG_RETURN) + if (thread__ts(thread) && sample->flags & PERF_IP_FLAG_RETURN) depth += 1; name = resolve_branch_sym(sample, evsel, thread, al, addr_al, &ip); @@ -1577,7 +1577,7 @@ static int perf_sample__fprintf_bts(struct perf_sample *sample, printed += fprintf(fp, "\n"); if (PRINT_FIELD(SRCCODE)) { int ret = map__fprintf_srccode(al->map, al->addr, stdout, - &thread->srccode_state); + thread__srccode_state(thread)); if (ret) { printed += ret; printed += printf("\n"); @@ -2086,9 +2086,9 @@ static bool show_event(struct perf_sample *sample, if (!symbol_conf.graph_function) return true; - if (thread->filter) { - if (depth <= thread->filter_entry_depth) { - thread->filter = false; + if (thread__filter(thread)) { + if (depth <= thread__filter_entry_depth(thread)) { + thread__set_filter(thread, false); return false; } return true; @@ -2105,8 +2105,8 @@ static bool show_event(struct perf_sample *sample, while (*s) { unsigned len = strcspn(s, ","); if (nlen == len && !strncmp(name, s, len)) { - thread->filter = true; - thread->filter_entry_depth = depth; + thread__set_filter(thread, true); + thread__set_filter_entry_depth(thread, depth); return true; } s += len; @@ -2186,7 +2186,7 @@ static void process_event(struct perf_script *script, struct callchain_cursor *cursor = NULL; if (script->stitch_lbr) - al->thread->lbr_stitch_enable = true; + thread__set_lbr_stitch_enable(al->thread, true); if (symbol_conf.use_callchain && sample->callchain && thread__resolve_callchain(al->thread, &callchain_cursor, evsel, @@ -2241,7 +2241,7 @@ static void process_event(struct perf_script *script, if (PRINT_FIELD(SRCCODE)) { if (map__fprintf_srccode(al->map, al->addr, stdout, - &thread->srccode_state)) + thread__srccode_state(thread))) printf("\n"); } -- cgit From 0dd5041c9a0eaf8c5c3fd46df4ee60f877799f44 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Thu, 8 Jun 2023 16:28:03 -0700 Subject: perf addr_location: Add init/exit/copy functions struct addr_location holds references to multiple reference counted objects. Add init/exit functions to make maintenance of those more consistent with the rest of the code and to try to avoid leaks. Modification of thread reference counts isn't included in this change. Committer notes: I needed to initialize result to sample->ip to make sure is set to something, fixing a compile time error, mostly keeping the previous logic as build_alloc_func_list() already does debugging/error prints about what went wrong if it takes the 'goto out'. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ali Saidi Cc: Andi Kleen Cc: Athira Rajeev Cc: Brian Robbins Cc: Changbin Du Cc: Dmitrii Dolgov <9erthalion6@gmail.com> Cc: Fangrui Song Cc: German Gomez Cc: Ingo Molnar Cc: Ivan Babrou Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: K Prateek Nayak Cc: Kan Liang Cc: Leo Yan Cc: Liam Howlett Cc: Mark Rutland Cc: Miguel Ojeda Cc: Mike Leach Cc: Namhyung Kim Cc: Naveen N. Rao Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Sean Christopherson Cc: Steinar H. Gunderson Cc: Suzuki Poulouse Cc: Wenyu Liu Cc: Will Deacon Cc: Yang Jihong Cc: Ye Xingchen Cc: Yuan Can Cc: coresight@lists.linaro.org Cc: linux-arm-kernel@lists.infradead.org Link: https://lore.kernel.org/r/20230608232823.4027869-7-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 77 +++++++++++++++++++++++++++------------------ 1 file changed, 46 insertions(+), 31 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index e756290de2ac..784d478c2e05 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -919,7 +919,6 @@ static int perf_sample__fprintf_brstack(struct perf_sample *sample, { struct branch_stack *br = sample->branch_stack; struct branch_entry *entries = perf_sample__branch_entries(sample); - struct addr_location alf, alt; u64 i, from, to; int printed = 0; @@ -930,20 +929,22 @@ static int perf_sample__fprintf_brstack(struct perf_sample *sample, from = entries[i].from; to = entries[i].to; + printed += fprintf(fp, " 0x%"PRIx64, from); if (PRINT_FIELD(DSO)) { - memset(&alf, 0, sizeof(alf)); - memset(&alt, 0, sizeof(alt)); + struct addr_location alf, alt; + + addr_location__init(&alf); + addr_location__init(&alt); thread__find_map_fb(thread, sample->cpumode, from, &alf); thread__find_map_fb(thread, sample->cpumode, to, &alt); - } - printed += fprintf(fp, " 0x%"PRIx64, from); - if (PRINT_FIELD(DSO)) printed += map__fprintf_dsoname_dsoff(alf.map, PRINT_FIELD(DSOFF), alf.addr, fp); - - printed += fprintf(fp, "/0x%"PRIx64, to); - if (PRINT_FIELD(DSO)) + printed += fprintf(fp, "/0x%"PRIx64, to); printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); + addr_location__exit(&alt); + addr_location__exit(&alf); + } else + printed += fprintf(fp, "/0x%"PRIx64, to); printed += print_bstack_flags(fp, entries + i); } @@ -957,7 +958,6 @@ static int perf_sample__fprintf_brstacksym(struct perf_sample *sample, { struct branch_stack *br = sample->branch_stack; struct branch_entry *entries = perf_sample__branch_entries(sample); - struct addr_location alf, alt; u64 i, from, to; int printed = 0; @@ -965,9 +965,10 @@ static int perf_sample__fprintf_brstacksym(struct perf_sample *sample, return 0; for (i = 0; i < br->nr; i++) { + struct addr_location alf, alt; - memset(&alf, 0, sizeof(alf)); - memset(&alt, 0, sizeof(alt)); + addr_location__init(&alf); + addr_location__init(&alt); from = entries[i].from; to = entries[i].to; @@ -982,6 +983,8 @@ static int perf_sample__fprintf_brstacksym(struct perf_sample *sample, if (PRINT_FIELD(DSO)) printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); printed += print_bstack_flags(fp, entries + i); + addr_location__exit(&alt); + addr_location__exit(&alf); } return printed; @@ -993,7 +996,6 @@ static int perf_sample__fprintf_brstackoff(struct perf_sample *sample, { struct branch_stack *br = sample->branch_stack; struct branch_entry *entries = perf_sample__branch_entries(sample); - struct addr_location alf, alt; u64 i, from, to; int printed = 0; @@ -1001,9 +1003,10 @@ static int perf_sample__fprintf_brstackoff(struct perf_sample *sample, return 0; for (i = 0; i < br->nr; i++) { + struct addr_location alf, alt; - memset(&alf, 0, sizeof(alf)); - memset(&alt, 0, sizeof(alt)); + addr_location__init(&alf); + addr_location__init(&alt); from = entries[i].from; to = entries[i].to; @@ -1022,6 +1025,8 @@ static int perf_sample__fprintf_brstackoff(struct perf_sample *sample, if (PRINT_FIELD(DSO)) printed += map__fprintf_dsoname_dsoff(alt.map, PRINT_FIELD(DSOFF), alt.addr, fp); printed += print_bstack_flags(fp, entries + i); + addr_location__exit(&alt); + addr_location__exit(&alf); } return printed; @@ -1036,6 +1041,7 @@ static int grab_bb(u8 *buffer, u64 start, u64 end, struct addr_location al; bool kernel; struct dso *dso; + int ret = 0; if (!start || !end) return 0; @@ -1057,7 +1063,6 @@ static int grab_bb(u8 *buffer, u64 start, u64 end, return -ENXIO; } - memset(&al, 0, sizeof(al)); if (end - start > MAXBB - MAXINSN) { if (last) pr_debug("\tbrstack does not reach to final jump (%" PRIx64 "-%" PRIx64 ")\n", start, end); @@ -1066,13 +1071,14 @@ static int grab_bb(u8 *buffer, u64 start, u64 end, return 0; } + addr_location__init(&al); if (!thread__find_map(thread, *cpumode, start, &al) || (dso = map__dso(al.map)) == NULL) { pr_debug("\tcannot resolve %" PRIx64 "-%" PRIx64 "\n", start, end); - return 0; + goto out; } if (dso->data.status == DSO_DATA_STATUS_ERROR) { pr_debug("\tcannot resolve %" PRIx64 "-%" PRIx64 "\n", start, end); - return 0; + goto out; } /* Load maps to ensure dso->is_64_bit has been updated */ @@ -1086,7 +1092,10 @@ static int grab_bb(u8 *buffer, u64 start, u64 end, if (len <= 0) pr_debug("\tcannot fetch code for block at %" PRIx64 "-%" PRIx64 "\n", start, end); - return len; + ret = len; +out: + addr_location__exit(&al); + return ret; } static int map__fprintf_srccode(struct map *map, u64 addr, FILE *fp, struct srccode_state *state) @@ -1137,14 +1146,16 @@ static int print_srccode(struct thread *thread, u8 cpumode, uint64_t addr) struct addr_location al; int ret = 0; - memset(&al, 0, sizeof(al)); + addr_location__init(&al); thread__find_map(thread, cpumode, addr, &al); if (!al.map) - return 0; + goto out; ret = map__fprintf_srccode(al.map, al.addr, stdout, thread__srccode_state(thread)); if (ret) ret += printf("\n"); +out: + addr_location__exit(&al); return ret; } @@ -1179,14 +1190,13 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread, struct perf_event_attr *attr, FILE *fp) { struct addr_location al; - int off, printed = 0; - - memset(&al, 0, sizeof(al)); + int off, printed = 0, ret = 0; + addr_location__init(&al); thread__find_map(thread, cpumode, addr, &al); if ((*lastsym) && al.addr >= (*lastsym)->start && al.addr < (*lastsym)->end) - return 0; + goto out; al.cpu = cpu; al.sym = NULL; @@ -1194,7 +1204,7 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread, al.sym = map__find_symbol(al.map, al.addr); if (!al.sym) - return 0; + goto out; if (al.addr < al.sym->end) off = al.addr - al.sym->start; @@ -1209,7 +1219,10 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread, printed += fprintf(fp, "\n"); *lastsym = al.sym; - return printed; + ret = printed; +out: + addr_location__exit(&al); + return ret; } static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample, @@ -1371,6 +1384,7 @@ static int perf_sample__fprintf_addr(struct perf_sample *sample, struct addr_location al; int printed = fprintf(fp, "%16" PRIx64, sample->addr); + addr_location__init(&al); if (!sample_addr_correlates_sym(attr)) goto out; @@ -1387,6 +1401,7 @@ static int perf_sample__fprintf_addr(struct perf_sample *sample, if (PRINT_FIELD(DSO)) printed += map__fprintf_dsoname_dsoff(al.map, PRINT_FIELD(DSOFF), al.addr, fp); out: + addr_location__exit(&al); return printed; } @@ -2338,8 +2353,8 @@ static int process_sample_event(struct perf_tool *tool, int ret = 0; /* Set thread to NULL to indicate addr_al and al are not initialized */ - addr_al.thread = NULL; - al.thread = NULL; + addr_location__init(&al); + addr_location__init(&addr_al); ret = dlfilter__filter_event_early(dlfilter, event, sample, evsel, machine, &al, &addr_al); if (ret) { @@ -2405,8 +2420,8 @@ static int process_sample_event(struct perf_tool *tool, } out_put: - if (al.thread) - addr_location__put(&al); + addr_location__exit(&addr_al); + addr_location__exit(&al); return ret; } -- cgit From 8ab12a2038e36beda4062a8e7562a8cfe9655553 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Thu, 8 Jun 2023 16:28:21 -0700 Subject: perf callchain: Use pthread keys for tls callchain_cursor Pthread keys are more portable than __thread and allow the association of a destructor with the key. Use the destructor to clean up TLS callchain cursors to aid understanding memory leaks. Committer notes: Had to fixup a series of unconverted places and also check for the return of get_tls_callchain_cursor() as it may fail and return NULL. In that unlikely case we now either print something to a file, if the caller was expecting to print a callchain, or return an error code to state that resolving the callchain isn't possible. In some cases this was made easier because thread__resolve_callchain() already can fail for other reasons, so this new one (cursor == NULL) can be added and the callers don't have to explicitely check for this new condition. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ali Saidi Cc: Andi Kleen Cc: Athira Rajeev Cc: Brian Robbins Cc: Changbin Du Cc: Dmitrii Dolgov <9erthalion6@gmail.com> Cc: Fangrui Song Cc: German Gomez Cc: Ingo Molnar Cc: Ivan Babrou Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: K Prateek Nayak Cc: Kan Liang Cc: Leo Yan Cc: Liam Howlett Cc: Mark Rutland Cc: Miguel Ojeda Cc: Mike Leach Cc: Namhyung Kim Cc: Naveen N. Rao Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Sean Christopherson Cc: Steinar H. Gunderson Cc: Suzuki Poulouse Cc: Wenyu Liu Cc: Will Deacon Cc: Yang Jihong Cc: Ye Xingchen Cc: Yuan Can Cc: coresight@lists.linaro.org Cc: linux-arm-kernel@lists.infradead.org Link: https://lore.kernel.org/r/20230608232823.4027869-25-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 24 ++++++++++++++---------- 1 file changed, 14 insertions(+), 10 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 784d478c2e05..e3f435e6a7d0 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -1557,11 +1557,13 @@ static int perf_sample__fprintf_bts(struct perf_sample *sample, unsigned int print_opts = output[type].print_ip_opts; struct callchain_cursor *cursor = NULL; - if (symbol_conf.use_callchain && sample->callchain && - thread__resolve_callchain(al->thread, &callchain_cursor, evsel, - sample, NULL, NULL, scripting_max_stack) == 0) - cursor = &callchain_cursor; - + if (symbol_conf.use_callchain && sample->callchain) { + cursor = get_tls_callchain_cursor(); + if (thread__resolve_callchain(al->thread, cursor, evsel, + sample, NULL, NULL, + scripting_max_stack)) + cursor = NULL; + } if (cursor == NULL) { printed += fprintf(fp, " "); if (print_opts & EVSEL__PRINT_SRCLINE) { @@ -2203,11 +2205,13 @@ static void process_event(struct perf_script *script, if (script->stitch_lbr) thread__set_lbr_stitch_enable(al->thread, true); - if (symbol_conf.use_callchain && sample->callchain && - thread__resolve_callchain(al->thread, &callchain_cursor, evsel, - sample, NULL, NULL, scripting_max_stack) == 0) - cursor = &callchain_cursor; - + if (symbol_conf.use_callchain && sample->callchain) { + cursor = get_tls_callchain_cursor(); + if (thread__resolve_callchain(al->thread, cursor, evsel, + sample, NULL, NULL, + scripting_max_stack)) + cursor = NULL; + } fputc(cursor ? '\n' : ' ', fp); sample__fprintf_sym(sample, al, 0, output[type].print_ip_opts, cursor, symbol_conf.bt_stop_list, fp); -- cgit From e590e46b548e0de3df52a8a093639ce67002fae0 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Fri, 26 May 2023 20:43:20 -0700 Subject: perf script: Remove some large stack allocations Some char buffers are stack allocated but in total they come to 24kb. Avoid Wstack-usage warnings by moving the arrays to being dynamically allocated. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20230527034324.2597593-3-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) (limited to 'tools/perf/builtin-script.c') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index e3f435e6a7d0..200b3e7ea8da 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -3318,14 +3318,21 @@ static int list_available_scripts(const struct option *opt __maybe_unused, int unset __maybe_unused) { struct dirent *script_dirent, *lang_dirent; - char scripts_path[MAXPATHLEN]; + char *buf, *scripts_path, *script_path, *lang_path, *first_half; DIR *scripts_dir, *lang_dir; - char script_path[MAXPATHLEN]; - char lang_path[MAXPATHLEN]; struct script_desc *desc; - char first_half[BUFSIZ]; char *script_root; + buf = malloc(3 * MAXPATHLEN + BUFSIZ); + if (!buf) { + pr_err("malloc failed\n"); + exit(-1); + } + scripts_path = buf; + script_path = buf + MAXPATHLEN; + lang_path = buf + 2 * MAXPATHLEN; + first_half = buf + 3 * MAXPATHLEN; + snprintf(scripts_path, MAXPATHLEN, "%s/scripts", get_argv_exec_path()); scripts_dir = opendir(scripts_path); @@ -3334,6 +3341,7 @@ static int list_available_scripts(const struct option *opt __maybe_unused, "open(%s) failed.\n" "Check \"PERF_EXEC_PATH\" env to set scripts dir.\n", scripts_path); + free(buf); exit(-1); } @@ -3364,6 +3372,7 @@ static int list_available_scripts(const struct option *opt __maybe_unused, desc->half_liner ? desc->half_liner : ""); } + free(buf); exit(0); } -- cgit