From f84ae29a6169318f9c929720c49d96323d2bbab9 Mon Sep 17 00:00:00 2001 From: Hewenliang Date: Mon, 9 Dec 2019 01:35:49 -0500 Subject: [PATCH 01/24] tools lib traceevent: Fix memory leakage in filter_event It is necessary to call free_arg(arg) when add_filter_type() returns NULL in filter_event(). Signed-off-by: Hewenliang Reviewed-by: Steven Rostedt (VMware) Cc: Feilong Lin Cc: Tzvetomir Stoyanov Link: http://lore.kernel.org/lkml/20191209063549.59941-1-hewenliang4@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/traceevent/parse-filter.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/lib/traceevent/parse-filter.c b/tools/lib/traceevent/parse-filter.c index f3cbf86e51acf..20eed719542e5 100644 --- a/tools/lib/traceevent/parse-filter.c +++ b/tools/lib/traceevent/parse-filter.c @@ -1228,8 +1228,10 @@ filter_event(struct tep_event_filter *filter, struct tep_event *event, } filter_type = add_filter_type(filter, event->id); - if (filter_type == NULL) + if (filter_type == NULL) { + free_arg(arg); return TEP_ERRNO__MEM_ALLOC_FAILED; + } if (filter_type->filter) free_arg(filter_type->filter); From 0feba17bd7ee3b7e03d141f119049dcc23efa94e Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Fri, 20 Dec 2019 09:37:19 +0800 Subject: [PATCH 02/24] perf report: Fix incorrectly added dimensions as switch perf data file We observed an issue that was some extra columns displayed after switching perf data file in browser. The steps to reproduce: 1. perf record -a -e cycles,instructions -- sleep 3 2. perf report --group 3. In browser, we use hotkey 's' to switch to another perf.data 4. Now in browser, the extra columns 'Self' and 'Children' are displayed. The issue is setup_sorting() executed again after repeat path, so dimensions are added again. This patch checks the last key returned from __cmd_report(). If it's K_SWITCH_INPUT_DATA, skips the setup_sorting(). Fixes: ad0de0971b7f ("perf report: Enable the runtime switching of perf data file") Signed-off-by: Jin Yao Tested-by: Arnaldo Carvalho de Melo Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Feng Tang Cc: Jin Yao Cc: Kan Liang Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20191220013722.20592-1-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-report.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 387311c672640..de988589d99bd 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -1076,6 +1076,7 @@ int cmd_report(int argc, const char **argv) struct stat st; bool has_br_stack = false; int branch_mode = -1; + int last_key = 0; bool branch_call_mode = false; #define CALLCHAIN_DEFAULT_OPT "graph,0.5,caller,function,percent" static const char report_callchain_help[] = "Display call graph (stack chain/backtrace):\n\n" @@ -1450,7 +1451,8 @@ int cmd_report(int argc, const char **argv) sort_order = sort_tmp; } - if (setup_sorting(session->evlist) < 0) { + if ((last_key != K_SWITCH_INPUT_DATA) && + (setup_sorting(session->evlist) < 0)) { if (sort_order) parse_options_usage(report_usage, options, "s", 1); if (field_order) @@ -1530,6 +1532,7 @@ int cmd_report(int argc, const char **argv) ret = __cmd_report(&report); if (ret == K_SWITCH_INPUT_DATA) { perf_session__delete(session); + last_key = K_SWITCH_INPUT_DATA; goto repeat; } else ret = 0; From a75af86b6f344f99825cc894596804a91a2ee9cf Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 18 Dec 2019 15:23:14 -0300 Subject: [PATCH 03/24] perf map: Set kmap->kmaps backpointer for main kernel map chunks When a map is create to represent the main kernel area (vmlinux) with map__new2() we allocate an extra area to store a pointer to the 'struct maps' for the kernel maps, so that we can access that struct when loading ELF files or kallsyms, as we will need to split it in multiple maps, one per kernel module or ELF section (such as ".init.text"). So when map->dso->kernel is non-zero, it is expected that map__kmap(map)->kmaps to be set to the tree of kernel maps (modules, chunks of the main kernel, bpf progs put in place via PERF_RECORD_KSYMBOL, the main kernel). This was not the case when we were splitting the main kernel into chunks for its ELF sections, which ended up making 'perf report --children' processing a perf.data file with callchains to trip on __map__is_kernel(), when we press ENTER to see the popup menu for main histogram entries that starts at a symbol in the ".init.text" ELF section, e.g.: - 8.83% 0.00% swapper [kernel.vmlinux].init.text [k] start_kernel start_kernel cpu_startup_entry do_idle cpuidle_enter cpuidle_enter_state intel_idle Fix it. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Namhyung Kim Link: https://lore.kernel.org/lkml/20191218190120.GB13282@kernel.org/ Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol-elf.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/perf/util/symbol-elf.c b/tools/perf/util/symbol-elf.c index 6658fbf196e6a..1965aefccb022 100644 --- a/tools/perf/util/symbol-elf.c +++ b/tools/perf/util/symbol-elf.c @@ -920,6 +920,9 @@ static int dso__process_kernel_symbol(struct dso *dso, struct map *map, if (curr_map == NULL) return -1; + if (curr_dso->kernel) + map__kmap(curr_map)->kmaps = kmaps; + if (adjust_kernel_syms) { curr_map->start = shdr->sh_addr + ref_reloc(kmap); curr_map->end = curr_map->start + shdr->sh_size; From 55347ec340af401437680fd0e88df6739a967f9f Mon Sep 17 00:00:00 2001 From: Yuya Fujita Date: Thu, 19 Dec 2019 08:08:32 +0000 Subject: [PATCH 04/24] perf hists: Fix variable name's inconsistency in hists__for_each() macro Variable names are inconsistent in hists__for_each macro(). Due to this inconsistency, the macro replaces its second argument with "fmt" regardless of its original name. So far it works because only "fmt" is passed to the second argument. However, this behavior is not expected and should be fixed. Fixes: f0786af536bb ("perf hists: Introduce hists__for_each_format macro") Fixes: aa6f50af822a ("perf hists: Introduce hists__for_each_sort_list macro") Signed-off-by: Yuya Fujita Acked-by: Jiri Olsa Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/OSAPR01MB1588E1C47AC22043175DE1B2E8520@OSAPR01MB1588.jpnprd01.prod.outlook.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/hist.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index 45286900aacbf..0aa63aeb58ecf 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -339,10 +339,10 @@ static inline void perf_hpp__prepend_sort_field(struct perf_hpp_fmt *format) list_for_each_entry_safe(format, tmp, &(_list)->sorts, sort_list) #define hists__for_each_format(hists, format) \ - perf_hpp_list__for_each_format((hists)->hpp_list, fmt) + perf_hpp_list__for_each_format((hists)->hpp_list, format) #define hists__for_each_sort_list(hists, format) \ - perf_hpp_list__for_each_sort_list((hists)->hpp_list, fmt) + perf_hpp_list__for_each_sort_list((hists)->hpp_list, format) extern struct perf_hpp_fmt perf_hpp__format[]; From 8812ad412f851216d6c39488a7e563ccc5c604cc Mon Sep 17 00:00:00 2001 From: Alexey Budankov Date: Tue, 3 Dec 2019 14:43:33 +0300 Subject: [PATCH 05/24] tools bitmap: Implement bitmap_equal() operation at bitmap API Extend tools bitmap API with bitmap_equal() implementation. The implementation has been derived from the kernel. Extend tools bitmap API with bitmap_free() implementation for symmetry with bitmap_alloc() function. Signed-off-by: Alexey Budankov Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/43757993-0b28-d8af-a6c7-ede12e3a6877@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/linux/bitmap.h | 30 ++++++++++++++++++++++++++++++ tools/lib/bitmap.c | 15 +++++++++++++++ 2 files changed, 45 insertions(+) diff --git a/tools/include/linux/bitmap.h b/tools/include/linux/bitmap.h index 05dca5c203f39..477a1cae513f2 100644 --- a/tools/include/linux/bitmap.h +++ b/tools/include/linux/bitmap.h @@ -15,6 +15,8 @@ void __bitmap_or(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, int bits); int __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, const unsigned long *bitmap2, unsigned int bits); +int __bitmap_equal(const unsigned long *bitmap1, + const unsigned long *bitmap2, unsigned int bits); void bitmap_clear(unsigned long *map, unsigned int start, int len); #define BITMAP_FIRST_WORD_MASK(start) (~0UL << ((start) & (BITS_PER_LONG - 1))) @@ -123,6 +125,15 @@ static inline unsigned long *bitmap_alloc(int nbits) return calloc(1, BITS_TO_LONGS(nbits) * sizeof(unsigned long)); } +/* + * bitmap_free - Free bitmap + * @bitmap: pointer to bitmap + */ +static inline void bitmap_free(unsigned long *bitmap) +{ + free(bitmap); +} + /* * bitmap_scnprintf - print bitmap list into buffer * @bitmap: bitmap @@ -148,4 +159,23 @@ static inline int bitmap_and(unsigned long *dst, const unsigned long *src1, return __bitmap_and(dst, src1, src2, nbits); } +#ifdef __LITTLE_ENDIAN +#define BITMAP_MEM_ALIGNMENT 8 +#else +#define BITMAP_MEM_ALIGNMENT (8 * sizeof(unsigned long)) +#endif +#define BITMAP_MEM_MASK (BITMAP_MEM_ALIGNMENT - 1) +#define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) - 1)) == 0) + +static inline int bitmap_equal(const unsigned long *src1, + const unsigned long *src2, unsigned int nbits) +{ + if (small_const_nbits(nbits)) + return !((*src1 ^ *src2) & BITMAP_LAST_WORD_MASK(nbits)); + if (__builtin_constant_p(nbits & BITMAP_MEM_MASK) && + IS_ALIGNED(nbits, BITMAP_MEM_ALIGNMENT)) + return !memcmp(src1, src2, nbits / 8); + return __bitmap_equal(src1, src2, nbits); +} + #endif /* _PERF_BITOPS_H */ diff --git a/tools/lib/bitmap.c b/tools/lib/bitmap.c index 38494782be06e..5043747ef6c5f 100644 --- a/tools/lib/bitmap.c +++ b/tools/lib/bitmap.c @@ -71,3 +71,18 @@ int __bitmap_and(unsigned long *dst, const unsigned long *bitmap1, BITMAP_LAST_WORD_MASK(bits)); return result != 0; } + +int __bitmap_equal(const unsigned long *bitmap1, + const unsigned long *bitmap2, unsigned int bits) +{ + unsigned int k, lim = bits/BITS_PER_LONG; + for (k = 0; k < lim; ++k) + if (bitmap1[k] != bitmap2[k]) + return 0; + + if (bits % BITS_PER_LONG) + if ((bitmap1[k] ^ bitmap2[k]) & BITMAP_LAST_WORD_MASK(bits)) + return 0; + + return 1; +} From 9c080c0279a80057cad3dfc05d09fb283ddf72f4 Mon Sep 17 00:00:00 2001 From: Alexey Budankov Date: Tue, 3 Dec 2019 14:44:18 +0300 Subject: [PATCH 06/24] perf mmap: Declare type for cpu mask of arbitrary length Declare a dedicated struct map_cpu_mask type for cpu masks of arbitrary length. The mask is available thru bits pointer and the mask length is kept in nbits field. MMAP_CPU_MASK_BYTES() macro returns mask storage size in bytes. The mmap_cpu_mask__scnprintf() function can be used to log text representation of the mask. Committer notes: To print the 'nbits' struct member we must use %zd, since it is a size_t, this fixes the build in some toolchains/arches. Signed-off-by: Alexey Budankov Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/0fd2454f-477f-d15a-f4ee-79bcbd2585ff@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/mmap.c | 12 ++++++++++++ tools/perf/util/mmap.h | 11 +++++++++++ 2 files changed, 23 insertions(+) diff --git a/tools/perf/util/mmap.c b/tools/perf/util/mmap.c index 063d1b93c53d1..2ee4faacca21f 100644 --- a/tools/perf/util/mmap.c +++ b/tools/perf/util/mmap.c @@ -23,6 +23,18 @@ #include "mmap.h" #include "../perf.h" #include /* page_size */ +#include + +#define MASK_SIZE 1023 +void mmap_cpu_mask__scnprintf(struct mmap_cpu_mask *mask, const char *tag) +{ + char buf[MASK_SIZE + 1]; + size_t len; + + len = bitmap_scnprintf(mask->bits, mask->nbits, buf, MASK_SIZE); + buf[len] = '\0'; + pr_debug("%p: %s mask[%zd]: %s\n", mask, tag, mask->nbits, buf); +} size_t mmap__mmap_len(struct mmap *map) { diff --git a/tools/perf/util/mmap.h b/tools/perf/util/mmap.h index bee4e83f7109d..ef51667fabcbe 100644 --- a/tools/perf/util/mmap.h +++ b/tools/perf/util/mmap.h @@ -15,6 +15,15 @@ #include "event.h" struct aiocb; + +struct mmap_cpu_mask { + unsigned long *bits; + size_t nbits; +}; + +#define MMAP_CPU_MASK_BYTES(m) \ + (BITS_TO_LONGS(((struct mmap_cpu_mask *)m)->nbits) * sizeof(unsigned long)) + /** * struct mmap - perf's ring buffer mmap details * @@ -52,4 +61,6 @@ int perf_mmap__push(struct mmap *md, void *to, size_t mmap__mmap_len(struct mmap *map); +void mmap_cpu_mask__scnprintf(struct mmap_cpu_mask *mask, const char *tag); + #endif /*__PERF_MMAP_H */ From 8384a2600c7ddfc875f64e160d8b423aca4e203a Mon Sep 17 00:00:00 2001 From: Alexey Budankov Date: Tue, 3 Dec 2019 14:45:27 +0300 Subject: [PATCH 07/24] perf record: Adapt affinity to machines with #CPUs > 1K Use struct mmap_cpu_mask type for the tool's thread and mmap data buffers to overcome current 1024 CPUs mask size limitation of cpu_set_t type. Currently glibc's cpu_set_t type has an internal mask size limit of 1024 CPUs. Moving to the 'struct mmap_cpu_mask' type allows overcoming that limit. The tools bitmap API is used to manipulate objects of 'struct mmap_cpu_mask' type. Committer notes: To print the 'nbits' struct member we must use %zd, since it is a size_t, this fixes the build in some toolchains/arches. Reported-by: Andi Kleen Signed-off-by: Alexey Budankov Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/96d7e2ff-ce8b-c1e0-d52c-aa59ea96f0ea@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 28 ++++++++++++++++++++++------ tools/perf/util/mmap.c | 28 ++++++++++++++++++++++------ tools/perf/util/mmap.h | 2 +- 3 files changed, 45 insertions(+), 13 deletions(-) diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index fb19ef63cc35f..4c301466101ba 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -62,6 +62,7 @@ #include #include #include +#include struct switch_output { bool enabled; @@ -93,7 +94,7 @@ struct record { bool timestamp_boundary; struct switch_output switch_output; unsigned long long samples; - cpu_set_t affinity_mask; + struct mmap_cpu_mask affinity_mask; unsigned long output_max_size; /* = 0: unlimited */ }; @@ -961,10 +962,15 @@ static struct perf_event_header finished_round_event = { static void record__adjust_affinity(struct record *rec, struct mmap *map) { if (rec->opts.affinity != PERF_AFFINITY_SYS && - !CPU_EQUAL(&rec->affinity_mask, &map->affinity_mask)) { - CPU_ZERO(&rec->affinity_mask); - CPU_OR(&rec->affinity_mask, &rec->affinity_mask, &map->affinity_mask); - sched_setaffinity(0, sizeof(rec->affinity_mask), &rec->affinity_mask); + !bitmap_equal(rec->affinity_mask.bits, map->affinity_mask.bits, + rec->affinity_mask.nbits)) { + bitmap_zero(rec->affinity_mask.bits, rec->affinity_mask.nbits); + bitmap_or(rec->affinity_mask.bits, rec->affinity_mask.bits, + map->affinity_mask.bits, rec->affinity_mask.nbits); + sched_setaffinity(0, MMAP_CPU_MASK_BYTES(&rec->affinity_mask), + (cpu_set_t *)rec->affinity_mask.bits); + if (verbose == 2) + mmap_cpu_mask__scnprintf(&rec->affinity_mask, "thread"); } } @@ -2433,7 +2439,6 @@ int cmd_record(int argc, const char **argv) # undef REASON #endif - CPU_ZERO(&rec->affinity_mask); rec->opts.affinity = PERF_AFFINITY_SYS; rec->evlist = evlist__new(); @@ -2499,6 +2504,16 @@ int cmd_record(int argc, const char **argv) symbol__init(NULL); + if (rec->opts.affinity != PERF_AFFINITY_SYS) { + rec->affinity_mask.nbits = cpu__max_cpu(); + rec->affinity_mask.bits = bitmap_alloc(rec->affinity_mask.nbits); + if (!rec->affinity_mask.bits) { + pr_err("Failed to allocate thread mask for %zd cpus\n", rec->affinity_mask.nbits); + return -ENOMEM; + } + pr_debug2("thread mask[%zd]: empty\n", rec->affinity_mask.nbits); + } + err = record__auxtrace_init(rec); if (err) goto out; @@ -2613,6 +2628,7 @@ int cmd_record(int argc, const char **argv) err = __cmd_record(&record, argc, argv); out: + bitmap_free(rec->affinity_mask.bits); evlist__delete(rec->evlist); symbol__exit(); auxtrace_record__free(rec->itr); diff --git a/tools/perf/util/mmap.c b/tools/perf/util/mmap.c index 2ee4faacca21f..3b664fa673a6c 100644 --- a/tools/perf/util/mmap.c +++ b/tools/perf/util/mmap.c @@ -219,6 +219,8 @@ static void perf_mmap__aio_munmap(struct mmap *map __maybe_unused) void mmap__munmap(struct mmap *map) { + bitmap_free(map->affinity_mask.bits); + perf_mmap__aio_munmap(map); if (map->data != NULL) { munmap(map->data, mmap__mmap_len(map)); @@ -227,7 +229,7 @@ void mmap__munmap(struct mmap *map) auxtrace_mmap__munmap(&map->auxtrace_mmap); } -static void build_node_mask(int node, cpu_set_t *mask) +static void build_node_mask(int node, struct mmap_cpu_mask *mask) { int c, cpu, nr_cpus; const struct perf_cpu_map *cpu_map = NULL; @@ -240,17 +242,23 @@ static void build_node_mask(int node, cpu_set_t *mask) for (c = 0; c < nr_cpus; c++) { cpu = cpu_map->map[c]; /* map c index to online cpu index */ if (cpu__get_node(cpu) == node) - CPU_SET(cpu, mask); + set_bit(cpu, mask->bits); } } -static void perf_mmap__setup_affinity_mask(struct mmap *map, struct mmap_params *mp) +static int perf_mmap__setup_affinity_mask(struct mmap *map, struct mmap_params *mp) { - CPU_ZERO(&map->affinity_mask); + map->affinity_mask.nbits = cpu__max_cpu(); + map->affinity_mask.bits = bitmap_alloc(map->affinity_mask.nbits); + if (!map->affinity_mask.bits) + return -1; + if (mp->affinity == PERF_AFFINITY_NODE && cpu__max_node() > 1) build_node_mask(cpu__get_node(map->core.cpu), &map->affinity_mask); else if (mp->affinity == PERF_AFFINITY_CPU) - CPU_SET(map->core.cpu, &map->affinity_mask); + set_bit(map->core.cpu, map->affinity_mask.bits); + + return 0; } int mmap__mmap(struct mmap *map, struct mmap_params *mp, int fd, int cpu) @@ -261,7 +269,15 @@ int mmap__mmap(struct mmap *map, struct mmap_params *mp, int fd, int cpu) return -1; } - perf_mmap__setup_affinity_mask(map, mp); + if (mp->affinity != PERF_AFFINITY_SYS && + perf_mmap__setup_affinity_mask(map, mp)) { + pr_debug2("failed to alloc mmap affinity mask, error %d\n", + errno); + return -1; + } + + if (verbose == 2) + mmap_cpu_mask__scnprintf(&map->affinity_mask, "mmap"); map->core.flush = mp->flush; diff --git a/tools/perf/util/mmap.h b/tools/perf/util/mmap.h index ef51667fabcbe..9d5f589f02ae7 100644 --- a/tools/perf/util/mmap.h +++ b/tools/perf/util/mmap.h @@ -40,7 +40,7 @@ struct mmap { int nr_cblocks; } aio; #endif - cpu_set_t affinity_mask; + struct mmap_cpu_mask affinity_mask; void *data; int comp_level; }; From c30d630d1bcfad8d2f70ff0cbb2a86d5a43bc152 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Wed, 4 Dec 2019 10:39:25 -0700 Subject: [PATCH 08/24] perf sched timehist: Add support for filtering on CPU Allow user to limit output to one or more CPUs. Really helpful on systems with a large number of cpus. Committer testing: # perf sched record -a sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.765 MB perf.data (1412 samples) ] [root@quaco ~]# perf sched timehist | head Samples do not have callchains. time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) --------------- ------ ------------------------------ --------- --------- --------- 66307.802686 [0000] perf[13086] 0.000 0.000 0.000 66307.802700 [0000] migration/0[12] 0.000 0.001 0.014 66307.802766 [0001] perf[13086] 0.000 0.000 0.000 66307.802774 [0001] migration/1[15] 0.000 0.001 0.007 66307.802841 [0002] perf[13086] 0.000 0.000 0.000 66307.802849 [0002] migration/2[20] 0.000 0.001 0.008 66307.802913 [0003] perf[13086] 0.000 0.000 0.000 # # perf sched timehist --cpu 2 | head Samples do not have callchains. time cpu task name wait time sch delay run time [tid/pid] (msec) (msec) (msec) --------------- ------ ------------------------------ --------- --------- --------- 66307.802841 [0002] perf[13086] 0.000 0.000 0.000 66307.802849 [0002] migration/2[20] 0.000 0.001 0.008 66307.964485 [0002] 0.000 0.000 161.635 66307.964811 [0002] CPU 0/KVM[3589/3561] 0.000 0.056 0.325 66307.965477 [0002] 0.325 0.000 0.666 66307.965553 [0002] CPU 0/KVM[3589/3561] 0.666 0.024 0.076 66307.966456 [0002] 0.076 0.000 0.903 # Signed-off-by: David Ahern Tested-by: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lore.kernel.org/lkml/20191204173925.66976-1-dsahern@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf-sched.txt | 4 ++++ tools/perf/builtin-sched.c | 13 +++++++++++++ 2 files changed, 17 insertions(+) diff --git a/tools/perf/Documentation/perf-sched.txt b/tools/perf/Documentation/perf-sched.txt index 63f938b887dd1..5fbe42bd599bf 100644 --- a/tools/perf/Documentation/perf-sched.txt +++ b/tools/perf/Documentation/perf-sched.txt @@ -110,6 +110,10 @@ OPTIONS for 'perf sched timehist' --max-stack:: Maximum number of functions to display in backtrace, default 5. +-C=:: +--cpu=:: + Only show events for the given CPU(s) (comma separated list). + -p=:: --pid=:: Only show events for given process ID (comma separated list). diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c index 8a12d71364c30..82fcc2c15fe46 100644 --- a/tools/perf/builtin-sched.c +++ b/tools/perf/builtin-sched.c @@ -51,6 +51,9 @@ #define SYM_LEN 129 #define MAX_PID 1024000 +static const char *cpu_list; +static DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS); + struct sched_atom; struct task_desc { @@ -2008,6 +2011,9 @@ static void timehist_print_sample(struct perf_sched *sched, char nstr[30]; u64 wait_time; + if (cpu_list && !test_bit(sample->cpu, cpu_bitmap)) + return; + timestamp__scnprintf_usec(t, tstr, sizeof(tstr)); printf("%15s [%04d] ", tstr, sample->cpu); @@ -2994,6 +3000,12 @@ static int perf_sched__timehist(struct perf_sched *sched) if (IS_ERR(session)) return PTR_ERR(session); + if (cpu_list) { + err = perf_session__cpu_bitmap(session, cpu_list, cpu_bitmap); + if (err < 0) + goto out; + } + evlist = session->evlist; symbol__init(&session->header.env); @@ -3429,6 +3441,7 @@ int cmd_sched(int argc, const char **argv) "analyze events only for given process id(s)"), OPT_STRING('t', "tid", &symbol_conf.tid_list_str, "tid[,tid...]", "analyze events only for given thread id(s)"), + OPT_STRING('C', "cpu", &cpu_list, "cpu", "list of cpus to profile"), OPT_PARENT(sched_options) }; From 6ae9c10b7cd50ac9080880204f8d9ff6381b2869 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 5 Dec 2019 16:09:59 -0300 Subject: [PATCH 09/24] perf tests bp_signal: Show expected versus obtained values To help understand failures. Cc: Adrian Hunter Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-c951j3gvrgnrsyg7ki7pwkiz@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/bp_signal.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/tools/perf/tests/bp_signal.c b/tools/perf/tests/bp_signal.c index 415903b48578f..da8ec1e8e0648 100644 --- a/tools/perf/tests/bp_signal.c +++ b/tools/perf/tests/bp_signal.c @@ -263,20 +263,20 @@ int test__bp_signal(struct test *test __maybe_unused, int subtest __maybe_unused if (count1 == 11) pr_debug("failed: RF EFLAG recursion issue detected\n"); else - pr_debug("failed: wrong count for bp1%lld\n", count1); + pr_debug("failed: wrong count for bp1: %lld, expected 1\n", count1); } if (overflows != 3) - pr_debug("failed: wrong overflow hit\n"); + pr_debug("failed: wrong overflow (%d) hit, expected 3\n", overflows); if (overflows_2 != 3) - pr_debug("failed: wrong overflow_2 hit\n"); + pr_debug("failed: wrong overflow_2 (%d) hit, expected 3\n", overflows_2); if (count2 != 3) - pr_debug("failed: wrong count for bp2\n"); + pr_debug("failed: wrong count for bp2 (%lld), expected 3\n", count2); if (count3 != 2) - pr_debug("failed: wrong count for bp3\n"); + pr_debug("failed: wrong count for bp3 (%lld), expected 2\n", count3); return count1 == 1 && overflows == 3 && count2 == 3 && overflows_2 == 3 && count3 == 2 ? TEST_OK : TEST_FAIL; From 3ce311afb5583cf3d3b7f54ab088949da28aea05 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Fri, 6 Dec 2019 22:06:11 +0100 Subject: [PATCH 10/24] libperf: Move to tools/lib/perf Move libperf from its current location under tools/perf to a separate directory under tools/lib/. Also change various paths (mainly includes) to reflect the libperf move to a separate directory and add a new directory under MANIFEST. Signed-off-by: Jiri Olsa Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20191206210612.8676-2-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/{perf/lib => lib/perf}/Build | 0 tools/{perf/lib => lib/perf}/Documentation/Makefile | 0 tools/{perf/lib => lib/perf}/Documentation/man/libperf.rst | 0 .../{perf/lib => lib/perf}/Documentation/tutorial/tutorial.rst | 0 tools/{perf/lib => lib/perf}/Makefile | 2 +- tools/{perf/lib => lib/perf}/core.c | 0 tools/{perf/lib => lib/perf}/cpumap.c | 0 tools/{perf/lib => lib/perf}/evlist.c | 0 tools/{perf/lib => lib/perf}/evsel.c | 0 tools/{perf/lib => lib/perf}/include/internal/cpumap.h | 0 tools/{perf/lib => lib/perf}/include/internal/evlist.h | 0 tools/{perf/lib => lib/perf}/include/internal/evsel.h | 0 tools/{perf/lib => lib/perf}/include/internal/lib.h | 0 tools/{perf/lib => lib/perf}/include/internal/mmap.h | 0 tools/{perf/lib => lib/perf}/include/internal/tests.h | 0 tools/{perf/lib => lib/perf}/include/internal/threadmap.h | 0 tools/{perf/lib => lib/perf}/include/internal/xyarray.h | 0 tools/{perf/lib => lib/perf}/include/perf/core.h | 0 tools/{perf/lib => lib/perf}/include/perf/cpumap.h | 0 tools/{perf/lib => lib/perf}/include/perf/event.h | 0 tools/{perf/lib => lib/perf}/include/perf/evlist.h | 0 tools/{perf/lib => lib/perf}/include/perf/evsel.h | 0 tools/{perf/lib => lib/perf}/include/perf/mmap.h | 0 tools/{perf/lib => lib/perf}/include/perf/threadmap.h | 0 tools/{perf/lib => lib/perf}/internal.h | 0 tools/{perf/lib => lib/perf}/lib.c | 0 tools/{perf/lib => lib/perf}/libperf.map | 0 tools/{perf/lib => lib/perf}/libperf.pc.template | 0 tools/{perf/lib => lib/perf}/mmap.c | 0 tools/{perf/lib => lib/perf}/tests/Makefile | 2 +- tools/{perf/lib => lib/perf}/tests/test-cpumap.c | 0 tools/{perf/lib => lib/perf}/tests/test-evlist.c | 0 tools/{perf/lib => lib/perf}/tests/test-evsel.c | 0 tools/{perf/lib => lib/perf}/tests/test-threadmap.c | 0 tools/{perf/lib => lib/perf}/threadmap.c | 0 tools/{perf/lib => lib/perf}/xyarray.c | 0 tools/perf/MANIFEST | 1 + tools/perf/Makefile.config | 2 +- tools/perf/Makefile.perf | 2 +- 39 files changed, 5 insertions(+), 4 deletions(-) rename tools/{perf/lib => lib/perf}/Build (100%) rename tools/{perf/lib => lib/perf}/Documentation/Makefile (100%) rename tools/{perf/lib => lib/perf}/Documentation/man/libperf.rst (100%) rename tools/{perf/lib => lib/perf}/Documentation/tutorial/tutorial.rst (100%) rename tools/{perf/lib => lib/perf}/Makefile (99%) rename tools/{perf/lib => lib/perf}/core.c (100%) rename tools/{perf/lib => lib/perf}/cpumap.c (100%) rename tools/{perf/lib => lib/perf}/evlist.c (100%) rename tools/{perf/lib => lib/perf}/evsel.c (100%) rename tools/{perf/lib => lib/perf}/include/internal/cpumap.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/evlist.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/evsel.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/lib.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/mmap.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/tests.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/threadmap.h (100%) rename tools/{perf/lib => lib/perf}/include/internal/xyarray.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/core.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/cpumap.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/event.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/evlist.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/evsel.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/mmap.h (100%) rename tools/{perf/lib => lib/perf}/include/perf/threadmap.h (100%) rename tools/{perf/lib => lib/perf}/internal.h (100%) rename tools/{perf/lib => lib/perf}/lib.c (100%) rename tools/{perf/lib => lib/perf}/libperf.map (100%) rename tools/{perf/lib => lib/perf}/libperf.pc.template (100%) rename tools/{perf/lib => lib/perf}/mmap.c (100%) rename tools/{perf/lib => lib/perf}/tests/Makefile (93%) rename tools/{perf/lib => lib/perf}/tests/test-cpumap.c (100%) rename tools/{perf/lib => lib/perf}/tests/test-evlist.c (100%) rename tools/{perf/lib => lib/perf}/tests/test-evsel.c (100%) rename tools/{perf/lib => lib/perf}/tests/test-threadmap.c (100%) rename tools/{perf/lib => lib/perf}/threadmap.c (100%) rename tools/{perf/lib => lib/perf}/xyarray.c (100%) diff --git a/tools/perf/lib/Build b/tools/lib/perf/Build similarity index 100% rename from tools/perf/lib/Build rename to tools/lib/perf/Build diff --git a/tools/perf/lib/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile similarity index 100% rename from tools/perf/lib/Documentation/Makefile rename to tools/lib/perf/Documentation/Makefile diff --git a/tools/perf/lib/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst similarity index 100% rename from tools/perf/lib/Documentation/man/libperf.rst rename to tools/lib/perf/Documentation/man/libperf.rst diff --git a/tools/perf/lib/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst similarity index 100% rename from tools/perf/lib/Documentation/tutorial/tutorial.rst rename to tools/lib/perf/Documentation/tutorial/tutorial.rst diff --git a/tools/perf/lib/Makefile b/tools/lib/perf/Makefile similarity index 99% rename from tools/perf/lib/Makefile rename to tools/lib/perf/Makefile index 0f233638ef1fb..768dd423730b1 100644 --- a/tools/perf/lib/Makefile +++ b/tools/lib/perf/Makefile @@ -60,7 +60,7 @@ else endif INCLUDES = \ --I$(srctree)/tools/perf/lib/include \ +-I$(srctree)/tools/lib/perf/include \ -I$(srctree)/tools/lib/ \ -I$(srctree)/tools/include \ -I$(srctree)/tools/arch/$(SRCARCH)/include/ \ diff --git a/tools/perf/lib/core.c b/tools/lib/perf/core.c similarity index 100% rename from tools/perf/lib/core.c rename to tools/lib/perf/core.c diff --git a/tools/perf/lib/cpumap.c b/tools/lib/perf/cpumap.c similarity index 100% rename from tools/perf/lib/cpumap.c rename to tools/lib/perf/cpumap.c diff --git a/tools/perf/lib/evlist.c b/tools/lib/perf/evlist.c similarity index 100% rename from tools/perf/lib/evlist.c rename to tools/lib/perf/evlist.c diff --git a/tools/perf/lib/evsel.c b/tools/lib/perf/evsel.c similarity index 100% rename from tools/perf/lib/evsel.c rename to tools/lib/perf/evsel.c diff --git a/tools/perf/lib/include/internal/cpumap.h b/tools/lib/perf/include/internal/cpumap.h similarity index 100% rename from tools/perf/lib/include/internal/cpumap.h rename to tools/lib/perf/include/internal/cpumap.h diff --git a/tools/perf/lib/include/internal/evlist.h b/tools/lib/perf/include/internal/evlist.h similarity index 100% rename from tools/perf/lib/include/internal/evlist.h rename to tools/lib/perf/include/internal/evlist.h diff --git a/tools/perf/lib/include/internal/evsel.h b/tools/lib/perf/include/internal/evsel.h similarity index 100% rename from tools/perf/lib/include/internal/evsel.h rename to tools/lib/perf/include/internal/evsel.h diff --git a/tools/perf/lib/include/internal/lib.h b/tools/lib/perf/include/internal/lib.h similarity index 100% rename from tools/perf/lib/include/internal/lib.h rename to tools/lib/perf/include/internal/lib.h diff --git a/tools/perf/lib/include/internal/mmap.h b/tools/lib/perf/include/internal/mmap.h similarity index 100% rename from tools/perf/lib/include/internal/mmap.h rename to tools/lib/perf/include/internal/mmap.h diff --git a/tools/perf/lib/include/internal/tests.h b/tools/lib/perf/include/internal/tests.h similarity index 100% rename from tools/perf/lib/include/internal/tests.h rename to tools/lib/perf/include/internal/tests.h diff --git a/tools/perf/lib/include/internal/threadmap.h b/tools/lib/perf/include/internal/threadmap.h similarity index 100% rename from tools/perf/lib/include/internal/threadmap.h rename to tools/lib/perf/include/internal/threadmap.h diff --git a/tools/perf/lib/include/internal/xyarray.h b/tools/lib/perf/include/internal/xyarray.h similarity index 100% rename from tools/perf/lib/include/internal/xyarray.h rename to tools/lib/perf/include/internal/xyarray.h diff --git a/tools/perf/lib/include/perf/core.h b/tools/lib/perf/include/perf/core.h similarity index 100% rename from tools/perf/lib/include/perf/core.h rename to tools/lib/perf/include/perf/core.h diff --git a/tools/perf/lib/include/perf/cpumap.h b/tools/lib/perf/include/perf/cpumap.h similarity index 100% rename from tools/perf/lib/include/perf/cpumap.h rename to tools/lib/perf/include/perf/cpumap.h diff --git a/tools/perf/lib/include/perf/event.h b/tools/lib/perf/include/perf/event.h similarity index 100% rename from tools/perf/lib/include/perf/event.h rename to tools/lib/perf/include/perf/event.h diff --git a/tools/perf/lib/include/perf/evlist.h b/tools/lib/perf/include/perf/evlist.h similarity index 100% rename from tools/perf/lib/include/perf/evlist.h rename to tools/lib/perf/include/perf/evlist.h diff --git a/tools/perf/lib/include/perf/evsel.h b/tools/lib/perf/include/perf/evsel.h similarity index 100% rename from tools/perf/lib/include/perf/evsel.h rename to tools/lib/perf/include/perf/evsel.h diff --git a/tools/perf/lib/include/perf/mmap.h b/tools/lib/perf/include/perf/mmap.h similarity index 100% rename from tools/perf/lib/include/perf/mmap.h rename to tools/lib/perf/include/perf/mmap.h diff --git a/tools/perf/lib/include/perf/threadmap.h b/tools/lib/perf/include/perf/threadmap.h similarity index 100% rename from tools/perf/lib/include/perf/threadmap.h rename to tools/lib/perf/include/perf/threadmap.h diff --git a/tools/perf/lib/internal.h b/tools/lib/perf/internal.h similarity index 100% rename from tools/perf/lib/internal.h rename to tools/lib/perf/internal.h diff --git a/tools/perf/lib/lib.c b/tools/lib/perf/lib.c similarity index 100% rename from tools/perf/lib/lib.c rename to tools/lib/perf/lib.c diff --git a/tools/perf/lib/libperf.map b/tools/lib/perf/libperf.map similarity index 100% rename from tools/perf/lib/libperf.map rename to tools/lib/perf/libperf.map diff --git a/tools/perf/lib/libperf.pc.template b/tools/lib/perf/libperf.pc.template similarity index 100% rename from tools/perf/lib/libperf.pc.template rename to tools/lib/perf/libperf.pc.template diff --git a/tools/perf/lib/mmap.c b/tools/lib/perf/mmap.c similarity index 100% rename from tools/perf/lib/mmap.c rename to tools/lib/perf/mmap.c diff --git a/tools/perf/lib/tests/Makefile b/tools/lib/perf/tests/Makefile similarity index 93% rename from tools/perf/lib/tests/Makefile rename to tools/lib/perf/tests/Makefile index a43cd08c5c037..96841775feaf2 100644 --- a/tools/perf/lib/tests/Makefile +++ b/tools/lib/perf/tests/Makefile @@ -16,7 +16,7 @@ all: include $(srctree)/tools/scripts/Makefile.include -INCLUDE = -I$(srctree)/tools/perf/lib/include -I$(srctree)/tools/include -I$(srctree)/tools/lib +INCLUDE = -I$(srctree)/tools/lib/perf/include -I$(srctree)/tools/include -I$(srctree)/tools/lib $(TESTS_A): FORCE $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -o $@ $(subst -a,.c,$@) ../libperf.a $(LIBAPI) diff --git a/tools/perf/lib/tests/test-cpumap.c b/tools/lib/perf/tests/test-cpumap.c similarity index 100% rename from tools/perf/lib/tests/test-cpumap.c rename to tools/lib/perf/tests/test-cpumap.c diff --git a/tools/perf/lib/tests/test-evlist.c b/tools/lib/perf/tests/test-evlist.c similarity index 100% rename from tools/perf/lib/tests/test-evlist.c rename to tools/lib/perf/tests/test-evlist.c diff --git a/tools/perf/lib/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c similarity index 100% rename from tools/perf/lib/tests/test-evsel.c rename to tools/lib/perf/tests/test-evsel.c diff --git a/tools/perf/lib/tests/test-threadmap.c b/tools/lib/perf/tests/test-threadmap.c similarity index 100% rename from tools/perf/lib/tests/test-threadmap.c rename to tools/lib/perf/tests/test-threadmap.c diff --git a/tools/perf/lib/threadmap.c b/tools/lib/perf/threadmap.c similarity index 100% rename from tools/perf/lib/threadmap.c rename to tools/lib/perf/threadmap.c diff --git a/tools/perf/lib/xyarray.c b/tools/lib/perf/xyarray.c similarity index 100% rename from tools/perf/lib/xyarray.c rename to tools/lib/perf/xyarray.c diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index 4934edb5adfd1..5d7b947320fbe 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -7,6 +7,7 @@ tools/lib/traceevent tools/lib/api tools/lib/bpf tools/lib/subcmd +tools/lib/perf tools/lib/argv_split.c tools/lib/ctype.c tools/lib/hweight.c diff --git a/tools/perf/Makefile.config b/tools/perf/Makefile.config index c90f4146e5a27..80e55e796be9c 100644 --- a/tools/perf/Makefile.config +++ b/tools/perf/Makefile.config @@ -286,7 +286,7 @@ ifeq ($(DEBUG),0) endif endif -INC_FLAGS += -I$(src-perf)/lib/include +INC_FLAGS += -I$(srctree)/tools/lib/perf/include INC_FLAGS += -I$(src-perf)/util/include INC_FLAGS += -I$(src-perf)/arch/$(SRCARCH)/include INC_FLAGS += -I$(srctree)/tools/include/ diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index eae5d5e959527..3eda9d4b88e7d 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -230,7 +230,7 @@ LIB_DIR = $(srctree)/tools/lib/api/ TRACE_EVENT_DIR = $(srctree)/tools/lib/traceevent/ BPF_DIR = $(srctree)/tools/lib/bpf/ SUBCMD_DIR = $(srctree)/tools/lib/subcmd/ -LIBPERF_DIR = $(srctree)/tools/perf/lib/ +LIBPERF_DIR = $(srctree)/tools/lib/perf/ # Set FEATURE_TESTS to 'all' so all possible feature checkers are executed. # Without this setting the output feature dump file misses some features, for From 81de3bf37a8bf58ecdbef608d16ddb0f4bbb71ca Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Fri, 6 Dec 2019 22:06:12 +0100 Subject: [PATCH 11/24] libperf: Add man pages Change the man page generation to asciidoc, because it's easier to use and has been more commonly used in related projects. Remove the current rst pages. Add 3 man pages to have a base for more additions: libperf.3 - overall description libperf-counting.7 - counting basics explained on simple example libperf-sampling.7 - sampling basics explained on simple example The plan is to add more man pages to cover the basic API. The build generates html and man pages: $ cd tools/lib/perf/Documentation $ make ASCIIDOC libperf.xml XMLTO libperf.3 ASCIIDOC libperf-counting.xml XMLTO libperf-counting.7 ASCIIDOC libperf-sampling.xml XMLTO libperf-sampling.7 ASCIIDOC libperf.html ASCIIDOC libperf-counting.html ASCIIDOC libperf-sampling.html Add the following install targets: install-man - man pages install-html - html version of man pages install-examples - examples mentioned in the man pages Signed-off-by: Jiri Olsa Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20191206210612.8676-3-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/perf/Documentation/Makefile | 159 ++++++++++- tools/lib/perf/Documentation/asciidoc.conf | 120 +++++++++ .../perf/Documentation/examples/sampling.c | 119 +++++++++ .../perf/Documentation/libperf-counting.txt | 211 +++++++++++++++ .../perf/Documentation/libperf-sampling.txt | 243 +++++++++++++++++ tools/lib/perf/Documentation/libperf.txt | 246 ++++++++++++++++++ tools/lib/perf/Documentation/man/libperf.rst | 100 ------- tools/lib/perf/Documentation/manpage-1.72.xsl | 14 + tools/lib/perf/Documentation/manpage-base.xsl | 35 +++ .../Documentation/manpage-bold-literal.xsl | 17 ++ .../lib/perf/Documentation/manpage-normal.xsl | 13 + .../Documentation/manpage-suppress-sp.xsl | 21 ++ .../perf/Documentation/tutorial/tutorial.rst | 123 --------- tools/lib/perf/Makefile | 5 +- 14 files changed, 1197 insertions(+), 229 deletions(-) create mode 100644 tools/lib/perf/Documentation/asciidoc.conf create mode 100644 tools/lib/perf/Documentation/examples/sampling.c create mode 100644 tools/lib/perf/Documentation/libperf-counting.txt create mode 100644 tools/lib/perf/Documentation/libperf-sampling.txt create mode 100644 tools/lib/perf/Documentation/libperf.txt delete mode 100644 tools/lib/perf/Documentation/man/libperf.rst create mode 100644 tools/lib/perf/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/perf/Documentation/manpage-base.xsl create mode 100644 tools/lib/perf/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-normal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-suppress-sp.xsl delete mode 100644 tools/lib/perf/Documentation/tutorial/tutorial.rst diff --git a/tools/lib/perf/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile index 586425a88795e..972754082a854 100644 --- a/tools/lib/perf/Documentation/Makefile +++ b/tools/lib/perf/Documentation/Makefile @@ -1,7 +1,156 @@ -all: - rst2man man/libperf.rst > man/libperf.7 - rst2pdf tutorial/tutorial.rst +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +# Most of this file is copied from tools/perf/Documentation/Makefile + +include ../../../scripts/Makefile.include +include ../../../scripts/utilities.mak + +MAN3_TXT = libperf.txt +MAN7_TXT = libperf-counting.txt libperf-sampling.txt +MAN_EX = examples/*.c + +MAN_TXT = $(MAN3_TXT) $(MAN7_TXT) + +_MAN_XML = $(patsubst %.txt,%.xml,$(MAN_TXT)) +_MAN_HTML = $(patsubst %.txt,%.html,$(MAN_TXT)) +_MAN_3 = $(patsubst %.txt,%.3,$(MAN3_TXT)) +_MAN_7 = $(patsubst %.txt,%.7,$(MAN7_TXT)) + +MAN_XML = $(addprefix $(OUTPUT),$(_MAN_XML)) +MAN_HTML = $(addprefix $(OUTPUT),$(_MAN_HTML)) +MAN_3 = $(addprefix $(OUTPUT),$(_MAN_3)) +MAN_7 = $(addprefix $(OUTPUT),$(_MAN_7)) +MAN_X = $(MAN_3) $(MAN_7) + +# Make the path relative to DESTDIR, not prefix +ifndef DESTDIR + prefix ?=$(HOME) +endif + +mandir ?= $(prefix)/share/man +man3dir = $(mandir)/man3 +man7dir = $(mandir)/man7 + +docdir ?= $(prefix)/share/doc/libperf +htmldir = $(docdir)/html +exdir = $(docdir)/examples + +ASCIIDOC = asciidoc +ASCIIDOC_EXTRA = --unsafe -f asciidoc.conf +ASCIIDOC_HTML = xhtml11 +MANPAGE_XSL = manpage-normal.xsl +XMLTO_EXTRA = +XMLTO =xmlto + +INSTALL ?= install +RM ?= rm -f + +# For asciidoc ... +# -7.1.2, no extra settings are needed. +# 8.0-, set ASCIIDOC8. +# + +# For docbook-xsl ... +# -1.68.1, set ASCIIDOC_NO_ROFF? (based on changelog from 1.73.0) +# 1.69.0, no extra settings are needed? +# 1.69.1-1.71.0, set DOCBOOK_SUPPRESS_SP? +# 1.71.1, no extra settings are needed? +# 1.72.0, set DOCBOOK_XSL_172. +# 1.73.0-, set ASCIIDOC_NO_ROFF + +# If you had been using DOCBOOK_XSL_172 in an attempt to get rid +# of 'the ".ft C" problem' in your generated manpages, and you +# instead ended up with weird characters around callouts, try +# using ASCIIDOC_NO_ROFF instead (it works fine with ASCIIDOC8). + +ifdef ASCIIDOC8 + ASCIIDOC_EXTRA += -a asciidoc7compatible +endif +ifdef DOCBOOK_XSL_172 + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + MANPAGE_XSL = manpage-1.72.xsl +else + ifdef ASCIIDOC_NO_ROFF + # docbook-xsl after 1.72 needs the regular XSL, but will not + # pass-thru raw roff codes from asciidoc.conf, so turn them off. + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + endif +endif +ifdef MAN_BOLD_LITERAL + XMLTO_EXTRA += -m manpage-bold-literal.xsl +endif +ifdef DOCBOOK_SUPPRESS_SP + XMLTO_EXTRA += -m manpage-suppress-sp.xsl +endif + +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +export DESTDIR DESTDIR_SQ + +# Please note that there is a minor bug in asciidoc. +# The version after 6.0.3 _will_ include the patch found here: +# http://marc.theaimsgroup.com/?l=libtraceevent&m=111558757202243&w=2 +# +# Until that version is released you may have to apply the patch +# yourself - yes, all 6 characters of it! + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) + PRINT_DIR = --no-print-directory +else # "make -w" + NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) + ifneq ($(V),1) + QUIET_ASCIIDOC = @echo ' ASCIIDOC '$@; + QUIET_XMLTO = @echo ' XMLTO '$@; + endif +endif + +all: $(MAN_X) $(MAN_HTML) + +$(MAN_HTML) $(MAN_X): asciidoc.conf + +install-man: all + $(call QUIET_INSTALL, man) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man3dir); \ + $(INSTALL) -m 644 $(MAN_3) $(DESTDIR)$(man3dir); \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man7dir); \ + $(INSTALL) -m 644 $(MAN_7) $(DESTDIR)$(man7dir); + +install-html: + $(call QUIET_INSTALL, html) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(htmldir); \ + $(INSTALL) -m 644 $(MAN_HTML) $(DESTDIR)$(htmldir); \ + +install-examples: + $(call QUIET_INSTALL, examples) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(exdir); \ + $(INSTALL) -m 644 $(MAN_EX) $(DESTDIR)$(exdir); \ + +CLEAN_FILES = \ + $(MAN_XML) $(addsuffix +,$(MAN_XML)) \ + $(MAN_HTML) $(addsuffix +,$(MAN_HTML)) \ + $(MAN_X) clean: - rm -f man/libperf.7 - rm -f tutorial/tutorial.pdf + $(call QUIET_CLEAN, Documentation) $(RM) $(CLEAN_FILES) + +$(MAN_3): $(OUTPUT)%.3: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_7): $(OUTPUT)%.7: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_XML): $(OUTPUT)%.xml: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b docbook -d manpage \ + $(ASCIIDOC_EXTRA) -alibperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ + +$(MAN_HTML): $(OUTPUT)%.html: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b $(ASCIIDOC_HTML) -d manpage \ + $(ASCIIDOC_EXTRA) -aperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ diff --git a/tools/lib/perf/Documentation/asciidoc.conf b/tools/lib/perf/Documentation/asciidoc.conf new file mode 100644 index 0000000000000..9d5a5a5ee091c --- /dev/null +++ b/tools/lib/perf/Documentation/asciidoc.conf @@ -0,0 +1,120 @@ +## linktep: macro +# +# Usage: linktep:command[manpage-section] +# +# Note, {0} is the manpage section, while {target} is the command. +# +# Show TEP link as: (
); if section is defined, else just show +# the command. + +[macros] +(?su)[\\]?(?Plinktep):(?P\S*?)\[(?P.*?)\]= + +[attributes] +asterisk=* +plus=+ +caret=^ +startsb=[ +endsb=] +tilde=~ + +ifdef::backend-docbook[] +[linktep-inlinemacro] +{0%{target}} +{0#} +{0#{target}{0}} +{0#} +endif::backend-docbook[] + +ifdef::backend-docbook[] +ifndef::tep-asciidoc-no-roff[] +# "unbreak" docbook-xsl v1.68 for manpages. v1.69 works with or without this. +# v1.72 breaks with this because it replaces dots not in roff requests. +[listingblock] +{title} + +ifdef::doctype-manpage[] + .ft C +endif::doctype-manpage[] +| +ifdef::doctype-manpage[] + .ft +endif::doctype-manpage[] + +{title#} +endif::tep-asciidoc-no-roff[] + +ifdef::tep-asciidoc-no-roff[] +ifdef::doctype-manpage[] +# The following two small workarounds insert a simple paragraph after screen +[listingblock] +{title} + +| + +{title#} + +[verseblock] +{title} +{title%} +{title#} +| + +{title#} +{title%} +endif::doctype-manpage[] +endif::tep-asciidoc-no-roff[] +endif::backend-docbook[] + +ifdef::doctype-manpage[] +ifdef::backend-docbook[] +[header] +template::[header-declarations] + + +{mantitle} +{manvolnum} +libperf +{libperf_version} +libperf Manual + + + {manname1} + {manname2} + {manname3} + {manname4} + {manname5} + {manname6} + {manname7} + {manname8} + {manname9} + {manname10} + {manname11} + {manname12} + {manname13} + {manname14} + {manname15} + {manname16} + {manname17} + {manname18} + {manname19} + {manname20} + {manname21} + {manname22} + {manname23} + {manname24} + {manname25} + {manname26} + {manname27} + {manname28} + {manname29} + {manname30} + {manpurpose} + +endif::backend-docbook[] +endif::doctype-manpage[] + +ifdef::backend-xhtml11[] +[linktep-inlinemacro] +{target}{0?({0})} +endif::backend-xhtml11[] diff --git a/tools/lib/perf/Documentation/examples/sampling.c b/tools/lib/perf/Documentation/examples/sampling.c new file mode 100644 index 0000000000000..8e1a926a9cfe6 --- /dev/null +++ b/tools/lib/perf/Documentation/examples/sampling.c @@ -0,0 +1,119 @@ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +union u64_swap { + __u64 val64; + __u32 val32[2]; +}; + +int main(int argc, char **argv) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_event_attr attr = { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CPU_CYCLES, + .disabled = 1, + .freq = 1, + .sample_freq = 10, + .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + }; + int err = -1; + union perf_event *event; + + libperf_init(libperf_print); + + cpus = perf_cpu_map__new(NULL); + if (!cpus) { + fprintf(stderr, "failed to create cpus\n"); + return -1; + } + + evlist = perf_evlist__new(); + if (!evlist) { + fprintf(stderr, "failed to create evlist\n"); + goto out_cpus; + } + + evsel = perf_evsel__new(&attr); + if (!evsel) { + fprintf(stderr, "failed to create cycles\n"); + goto out_cpus; + } + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + if (err) { + fprintf(stderr, "failed to open evlist\n"); + goto out_evlist; + } + + err = perf_evlist__mmap(evlist, 4); + if (err) { + fprintf(stderr, "failed to mmap evlist\n"); + goto out_evlist; + } + + perf_evlist__enable(evlist); + sleep(3); + perf_evlist__disable(evlist); + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + int cpu, pid, tid; + __u64 ip, period, *array; + union u64_swap u; + + array = event->sample.array; + + ip = *array; + array++; + + u.val64 = *array; + pid = u.val32[0]; + tid = u.val32[1]; + array++; + + u.val64 = *array; + cpu = u.val32[0]; + array++; + + period = *array; + + fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", + cpu, pid, tid, ip, period); + + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + +out_evlist: + perf_evlist__delete(evlist); +out_cpus: + perf_cpu_map__put(cpus); + return err; +} diff --git a/tools/lib/perf/Documentation/libperf-counting.txt b/tools/lib/perf/Documentation/libperf-counting.txt new file mode 100644 index 0000000000000..cae9757f49c12 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-counting.txt @@ -0,0 +1,211 @@ +libperf-counting(7) +=================== + +NAME +---- +libperf-counting - counting interface + +DESCRIPTION +----------- +The counting interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `counting.c` example. + +It is by no means complete guide to counting, but shows libperf basic API for counting. + +The `counting.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o counting counting.c -lperf +$ sudo ./counting +count 176792, enabled 176944, run 176944 +count 176242, enabled 176242, run 176242 +-- + +It requires root access, because of the `PERF_COUNT_SW_CPU_CLOCK` event, +which is available only for root. + +The `counting.c` example monitors two events on the current process and displays their count, in a nutshel it: + +* creates events +* adds them to the event list +* opens and enables events through the event list +* does some workload +* disables events +* reads and displays event counts +* destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 8 static int libperf_print(enum libperf_print_level level, + 9 const char *fmt, va_list ap) + 10 { + 11 return vfprintf(stderr, fmt, ap); + 12 } + + 14 int main(int argc, char **argv) + 15 { + ... + 35 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining specific events using the `struct perf_event_attr`. + +We create software events for cpu and task: + +[source,c] +-- + 20 struct perf_event_attr attr1 = { + 21 .type = PERF_TYPE_SOFTWARE, + 22 .config = PERF_COUNT_SW_CPU_CLOCK, + 23 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 24 .disabled = 1, + 25 }; + 26 struct perf_event_attr attr2 = { + 27 .type = PERF_TYPE_SOFTWARE, + 28 .config = PERF_COUNT_SW_TASK_CLOCK, + 29 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 30 .disabled = 1, + 31 }; +-- + +The `read_format` setup tells perf to include timing details together with each count. + +Next step is to prepare threads map. + +In this case we will monitor current process, so we create threads map with single pid (0): + +[source,c] +-- + 37 threads = perf_thread_map__new_dummy(); + 38 if (!threads) { + 39 fprintf(stderr, "failed to create threads\n"); + 40 return -1; + 41 } + 42 + 43 perf_thread_map__set_pid(threads, 0, 0); +-- + +Now we create libperf's event list, which will serve as holder for the events we want: + +[source,c] +-- + 45 evlist = perf_evlist__new(); + 46 if (!evlist) { + 47 fprintf(stderr, "failed to create evlist\n"); + 48 goto out_threads; + 49 } +-- + +We create libperf's events for the attributes we defined earlier and add them to the list: + +[source,c] +-- + 51 evsel = perf_evsel__new(&attr1); + 52 if (!evsel) { + 53 fprintf(stderr, "failed to create evsel1\n"); + 54 goto out_evlist; + 55 } + 56 + 57 perf_evlist__add(evlist, evsel); + 58 + 59 evsel = perf_evsel__new(&attr2); + 60 if (!evsel) { + 61 fprintf(stderr, "failed to create evsel2\n"); + 62 goto out_evlist; + 63 } + 64 + 65 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the thread map and open events: + +[source,c] +-- + 67 perf_evlist__set_maps(evlist, NULL, threads); + 68 + 69 err = perf_evlist__open(evlist); + 70 if (err) { + 71 fprintf(stderr, "failed to open evsel\n"); + 72 goto out_evlist; + 73 } +-- + +Both events are created as disabled (note the `disabled = 1` assignment above), +so we need to enable the whole list explicitely (both events). + +From this moment events are counting and we can do our workload. + +When we are done we disable the events list. + +[source,c] +-- + 75 perf_evlist__enable(evlist); + 76 + 77 while (count--); + 78 + 79 perf_evlist__disable(evlist); +-- + +Now we need to get the counts from events, following code iterates throught the events list and read counts: + +[source,c] +-- + 81 perf_evlist__for_each_evsel(evlist, evsel) { + 82 perf_evsel__read(evsel, 0, 0, &counts); + 83 fprintf(stdout, "count %llu, enabled %llu, run %llu\n", + 84 counts.val, counts.ena, counts.run); + 85 } +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- + 87 perf_evlist__close(evlist); + 88 + 89 out_evlist: + 90 perf_evlist__delete(evlist); + 91 out_threads: + 92 perf_thread_map__put(threads); + 93 return err; + 94 } +-- + +REPORTING BUGS +-------------- +Report bugs to . + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-sampling(7) diff --git a/tools/lib/perf/Documentation/libperf-sampling.txt b/tools/lib/perf/Documentation/libperf-sampling.txt new file mode 100644 index 0000000000000..d71a7b4fcf5f6 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-sampling.txt @@ -0,0 +1,243 @@ +libperf-sampling(7) +=================== + +NAME +---- +libperf-sampling - sampling interface + + +DESCRIPTION +----------- +The sampling interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `sampling.c` example. + +It is by no means complete guide to sampling, but shows libperf basic API for sampling. + +The `sampling.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o sampling sampling.c -lperf +$ sudo ./sampling +cpu 0, pid 0, tid 0, ip ffffffffad06c4e6, period 1 +cpu 0, pid 4465, tid 4469, ip ffffffffad118748, period 18322959 +cpu 0, pid 0, tid 0, ip ffffffffad115722, period 33544846 +cpu 0, pid 4465, tid 4470, ip 7f84fe0cdad6, period 23687474 +cpu 0, pid 0, tid 0, ip ffffffffad9e0349, period 34255790 +cpu 0, pid 4465, tid 4469, ip ffffffffad136581, period 38664069 +cpu 0, pid 0, tid 0, ip ffffffffad9e55e2, period 21922384 +cpu 0, pid 4465, tid 4470, ip 7f84fe0ebebf, period 17655175 +... +-- + +It requires root access, because it uses hardware cycles event. + +The `sampling.c` example profiles/samples all CPUs with hardware cycles, in a nutshel it: + +- creates events +- adds them to the event list +- opens and enables events through the event list +- sleeps for 3 seconds +- disables events +- reads and displays recorded samples +- destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 12 static int libperf_print(enum libperf_print_level level, + 13 const char *fmt, va_list ap) + 14 { + 15 return vfprintf(stderr, fmt, ap); + 16 } + + 23 int main(int argc, char **argv) + 24 { + ... + 40 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining cycles event using the `struct perf_event_attr`: + +[source,c] +-- + 29 struct perf_event_attr attr = { + 30 .type = PERF_TYPE_HARDWARE, + 31 .config = PERF_COUNT_HW_CPU_CYCLES, + 32 .disabled = 1, + 33 .freq = 1, + 34 .sample_freq = 10, + 35 .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + 36 }; +-- + +Next step is to prepare cpus map. + +In this case we will monitor all the available CPUs: + +[source,c] +-- + 42 cpus = perf_cpu_map__new(NULL); + 43 if (!cpus) { + 44 fprintf(stderr, "failed to create cpus\n"); + 45 return -1; + 46 } +-- + +Now we create libperf's event list, which will serve as holder for the cycles event: + +[source,c] +-- + 48 evlist = perf_evlist__new(); + 49 if (!evlist) { + 50 fprintf(stderr, "failed to create evlist\n"); + 51 goto out_cpus; + 52 } +-- + +We create libperf's event for the cycles attribute we defined earlier and add it to the list: + +[source,c] +-- + 54 evsel = perf_evsel__new(&attr); + 55 if (!evsel) { + 56 fprintf(stderr, "failed to create cycles\n"); + 57 goto out_cpus; + 58 } + 59 + 60 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the cpus map and open event: + +[source,c] +-- + 62 perf_evlist__set_maps(evlist, cpus, NULL); + 63 + 64 err = perf_evlist__open(evlist); + 65 if (err) { + 66 fprintf(stderr, "failed to open evlist\n"); + 67 goto out_evlist; + 68 } +-- + +Once the events list is open, we can create memory maps AKA perf ring buffers: + +[source,c] +-- + 70 err = perf_evlist__mmap(evlist, 4); + 71 if (err) { + 72 fprintf(stderr, "failed to mmap evlist\n"); + 73 goto out_evlist; + 74 } +-- + +The event is created as disabled (note the `disabled = 1` assignment above), +so we need to enable the events list explicitely. + +From this moment the cycles event is sampling. + +We will sleep for 3 seconds while the ring buffers get data from all CPUs, then we disable the events list. + +[source,c] +-- + 76 perf_evlist__enable(evlist); + 77 sleep(3); + 78 perf_evlist__disable(evlist); +-- + +Following code walks through the ring buffers and reads stored events/samples: + +[source,c] +-- + 80 perf_evlist__for_each_mmap(evlist, map, false) { + 81 if (perf_mmap__read_init(map) < 0) + 82 continue; + 83 + 84 while ((event = perf_mmap__read_event(map)) != NULL) { + + /* process event */ + +108 perf_mmap__consume(map); +109 } +110 perf_mmap__read_done(map); +111 } + +-- + +Each sample needs to get parsed: + +[source,c] +-- + 85 int cpu, pid, tid; + 86 __u64 ip, period, *array; + 87 union u64_swap u; + 88 + 89 array = event->sample.array; + 90 + 91 ip = *array; + 92 array++; + 93 + 94 u.val64 = *array; + 95 pid = u.val32[0]; + 96 tid = u.val32[1]; + 97 array++; + 98 + 99 u.val64 = *array; +100 cpu = u.val32[0]; +101 array++; +102 +103 period = *array; +104 +105 fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", +106 cpu, pid, tid, ip, period); +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- +113 out_evlist: +114 perf_evlist__delete(evlist); +115 out_cpus: +116 perf_cpu_map__put(cpus); +117 return err; +118 } +-- + +REPORTING BUGS +-------------- +Report bugs to . + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/libperf.txt b/tools/lib/perf/Documentation/libperf.txt new file mode 100644 index 0000000000000..5a6bb512789d3 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf.txt @@ -0,0 +1,246 @@ +libperf(3) +========== + +NAME +---- +libperf - Linux kernel perf event library + + +SYNOPSIS +-------- +*Generic API:* + +[source,c] +-- + #include + + enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, + }; + + typedef int (*libperf_print_fn_t)(enum libperf_print_level level, + const char *, va_list ap); + + void libperf_init(libperf_print_fn_t fn); +-- + +*API to handle cpu maps:* + +[source,c] +-- + #include + + struct perf_cpu_map; + + struct perf_cpu_map *perf_cpu_map__dummy_new(void); + struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); + struct perf_cpu_map *perf_cpu_map__read(FILE *file); + struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); + struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other); + void perf_cpu_map__put(struct perf_cpu_map *map); + int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); + int perf_cpu_map__nr(const struct perf_cpu_map *cpus); + bool perf_cpu_map__empty(const struct perf_cpu_map *map); + int perf_cpu_map__max(struct perf_cpu_map *map); + + #define perf_cpu_map__for_each_cpu(cpu, idx, cpus) +-- + +*API to handle thread maps:* + +[source,c] +-- + #include + + struct perf_thread_map; + + struct perf_thread_map *perf_thread_map__new_dummy(void); + + void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); + char *perf_thread_map__comm(struct perf_thread_map *map, int thread); + int perf_thread_map__nr(struct perf_thread_map *threads); + pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); + + struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); + void perf_thread_map__put(struct perf_thread_map *map); +-- + +*API to handle event lists:* + +[source,c] +-- + #include + + struct perf_evlist; + + void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel); + void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel); + struct perf_evlist *perf_evlist__new(void); + void perf_evlist__delete(struct perf_evlist *evlist); + struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, + struct perf_evsel *evsel); + int perf_evlist__open(struct perf_evlist *evlist); + void perf_evlist__close(struct perf_evlist *evlist); + void perf_evlist__enable(struct perf_evlist *evlist); + void perf_evlist__disable(struct perf_evlist *evlist); + + #define perf_evlist__for_each_evsel(evlist, pos) + + void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + int perf_evlist__poll(struct perf_evlist *evlist, int timeout); + int perf_evlist__filter_pollfd(struct perf_evlist *evlist, + short revents_and_mask); + + int perf_evlist__mmap(struct perf_evlist *evlist, int pages); + void perf_evlist__munmap(struct perf_evlist *evlist); + + struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, + struct perf_mmap *map, + bool overwrite); + + #define perf_evlist__for_each_mmap(evlist, pos, overwrite) +-- + +*API to handle events:* + +[source,c] +-- + #include * + + struct perf_evsel; + + struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; + }; + + struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); + void perf_evsel__delete(struct perf_evsel *evsel); + int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + void perf_evsel__close(struct perf_evsel *evsel); + void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count); + int perf_evsel__enable(struct perf_evsel *evsel); + int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__disable(struct perf_evsel *evsel); + int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); + struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); + struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); + struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); +-- + +*API to handle maps (perf ring buffers):* + +[source,c] +-- + #include + + struct perf_mmap; + + void perf_mmap__consume(struct perf_mmap *map); + int perf_mmap__read_init(struct perf_mmap *map); + void perf_mmap__read_done(struct perf_mmap *map); + union perf_event *perf_mmap__read_event(struct perf_mmap *map); +-- + +*Structures to access perf API events:* + +[source,c] +-- + #include + + struct perf_record_mmap; + struct perf_record_mmap2; + struct perf_record_comm; + struct perf_record_namespaces; + struct perf_record_fork; + struct perf_record_lost; + struct perf_record_lost_samples; + struct perf_record_read; + struct perf_record_throttle; + struct perf_record_ksymbol; + struct perf_record_bpf_event; + struct perf_record_sample; + struct perf_record_switch; + struct perf_record_header_attr; + struct perf_record_record_cpu_map; + struct perf_record_cpu_map_data; + struct perf_record_cpu_map; + struct perf_record_event_update_cpus; + struct perf_record_event_update_scale; + struct perf_record_event_update; + struct perf_trace_event_type; + struct perf_record_header_event_type; + struct perf_record_header_tracing_data; + struct perf_record_header_build_id; + struct perf_record_id_index; + struct perf_record_auxtrace_info; + struct perf_record_auxtrace; + struct perf_record_auxtrace_error; + struct perf_record_aux; + struct perf_record_itrace_start; + struct perf_record_thread_map_entry; + struct perf_record_thread_map; + struct perf_record_stat_config_entry; + struct perf_record_stat_config; + struct perf_record_stat; + struct perf_record_stat_round; + struct perf_record_time_conv; + struct perf_record_header_feature; + struct perf_record_compressed; +-- + +DESCRIPTION +----------- +The libperf library provides an API to access the linux kernel perf +events subsystem. + +Following objects are key to the libperf interface: + +[horizontal] + +struct perf_cpu_map:: Provides a cpu list abstraction. + +struct perf_thread_map:: Provides a thread list abstraction. + +struct perf_evsel:: Provides an abstraction for single a perf event. + +struct perf_evlist:: Gathers several struct perf_evsel object and performs functions on all of them. + +struct perf_mmap:: Provides an abstraction for accessing perf ring buffer. + +The exported API functions bind these objects together. + +REPORTING BUGS +-------------- +Report bugs to . + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf-sampling(7), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst deleted file mode 100644 index 09a270fccb9cd..0000000000000 --- a/tools/lib/perf/Documentation/man/libperf.rst +++ /dev/null @@ -1,100 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf - -The libperf library provides an API to access the linux kernel perf -events subsystem. It provides the following high level objects: - - - struct perf_cpu_map - - struct perf_thread_map - - struct perf_evlist - - struct perf_evsel - -reference -========= -Function reference by header files: - -perf/core.h ------------ -.. code-block:: c - - typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, - const char \*, va_list ap); - - void libperf_set_print(libperf_print_fn_t fn); - -perf/cpumap.h -------------- -.. code-block:: c - - struct perf_cpu_map \*perf_cpu_map__dummy_new(void); - struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); - struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); - struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); - void perf_cpu_map__put(struct perf_cpu_map \*map); - int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); - int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); - perf_cpu_map__for_each_cpu(cpu, idx, cpus) - -perf/threadmap.h ----------------- -.. code-block:: c - - struct perf_thread_map \*perf_thread_map__new_dummy(void); - void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); - char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); - struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); - void perf_thread_map__put(struct perf_thread_map \*map); - -perf/evlist.h -------------- -.. code-block:: - - void perf_evlist__init(struct perf_evlist \*evlist); - void perf_evlist__add(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - void perf_evlist__remove(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - struct perf_evlist \*perf_evlist__new(void); - void perf_evlist__delete(struct perf_evlist \*evlist); - struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - int perf_evlist__open(struct perf_evlist \*evlist); - void perf_evlist__close(struct perf_evlist \*evlist); - void perf_evlist__enable(struct perf_evlist \*evlist); - void perf_evlist__disable(struct perf_evlist \*evlist); - perf_evlist__for_each_evsel(evlist, pos) - void perf_evlist__set_maps(struct perf_evlist \*evlist, - struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - -perf/evsel.h ------------- -.. code-block:: c - - struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; - }; - - void perf_evsel__init(struct perf_evsel \*evsel, - struct perf_event_attr \*attr); - struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); - void perf_evsel__delete(struct perf_evsel \*evsel); - int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - void perf_evsel__close(struct perf_evsel \*evsel); - int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, - struct perf_counts_values \*count); - int perf_evsel__enable(struct perf_evsel \*evsel); - int perf_evsel__disable(struct perf_evsel \*evsel); - int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); - struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); - struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); - struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/lib/perf/Documentation/manpage-1.72.xsl b/tools/lib/perf/Documentation/manpage-1.72.xsl new file mode 100644 index 0000000000000..b4d315cb8c479 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-1.72.xsl @@ -0,0 +1,14 @@ + + + + + + +▓ +⌂ + + diff --git a/tools/lib/perf/Documentation/manpage-base.xsl b/tools/lib/perf/Documentation/manpage-base.xsl new file mode 100644 index 0000000000000..a264fa616093c --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-base.xsl @@ -0,0 +1,35 @@ + + + + + + + + + + + + + + sp + + + + + + + + br + + + diff --git a/tools/lib/perf/Documentation/manpage-bold-literal.xsl b/tools/lib/perf/Documentation/manpage-bold-literal.xsl new file mode 100644 index 0000000000000..608eb5df62814 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-bold-literal.xsl @@ -0,0 +1,17 @@ + + + + + + + fB + + + fR + + + diff --git a/tools/lib/perf/Documentation/manpage-normal.xsl b/tools/lib/perf/Documentation/manpage-normal.xsl new file mode 100644 index 0000000000000..a48f5b11f3dcc --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-normal.xsl @@ -0,0 +1,13 @@ + + + + + + +\ +. + + diff --git a/tools/lib/perf/Documentation/manpage-suppress-sp.xsl b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl new file mode 100644 index 0000000000000..a63c7632a87d4 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl @@ -0,0 +1,21 @@ + + + + + + + + + + + + + + + diff --git a/tools/lib/perf/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst deleted file mode 100644 index 7be7bc27b385f..0000000000000 --- a/tools/lib/perf/Documentation/tutorial/tutorial.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf tutorial -================ - -Compile and install libperf from kernel sources -=============================================== -.. code-block:: bash - - git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git - cd linux/tools/perf/lib - make - sudo make install prefix=/usr - -Libperf object -============== -The libperf library provides several high level objects: - -struct perf_cpu_map - Provides a cpu list abstraction. - -struct perf_thread_map - Provides a thread list abstraction. - -struct perf_evsel - Provides an abstraction for single a perf event. - -struct perf_evlist - Gathers several struct perf_evsel object and performs functions on all of them. - -The exported API binds these objects together, -for full reference see the libperf.7 man page. - -Examples -======== -Examples aim to explain libperf functionality on simple use cases. -They are based in on a checked out linux kernel git tree: - -.. code-block:: bash - - $ cd tools/perf/lib/Documentation/tutorial/ - $ ls -d ex-* - ex-1-compile ex-2-evsel-stat ex-3-evlist-stat - -ex-1-compile example -==================== -This example shows the basic usage of *struct perf_cpu_map*, -how to create it and display its cpus: - -.. code-block:: bash - - $ cd ex-1-compile/ - $ make - gcc -o test test.c -lperf - $ ./test - 0 1 2 3 4 5 6 7 - - -The full code listing is here: - -.. code-block:: c - - 1 #include - 2 - 3 int main(int argc, char **Argv) - 4 { - 5 struct perf_cpu_map *cpus; - 6 int cpu, tmp; - 7 - 8 cpus = perf_cpu_map__new(NULL); - 9 - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - 12 - 13 fprintf(stdout, "\n"); - 14 - 15 perf_cpu_map__put(cpus); - 16 return 0; - 17 } - - -First you need to include the proper header to have *struct perf_cpumap* -declaration and functions: - -.. code-block:: c - - 1 #include - - -The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. -The *NULL* argument asks it to populate the object with the current online CPUs list: - -.. code-block:: c - - 8 cpus = perf_cpu_map__new(NULL); - -This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. - -.. code-block:: c - - 15 perf_cpu_map__put(cpus); - -The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* -macro which requires 3 arguments: - -- cpu - the cpu numer -- tmp - iteration helper variable -- cpus - the *struct perf_cpumap* object - -.. code-block:: c - - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - -ex-2-evsel-stat example -======================= - -TBD - -ex-3-evlist-stat example -======================== - -TBD diff --git a/tools/lib/perf/Makefile b/tools/lib/perf/Makefile index 768dd423730b1..3718d65cffac7 100644 --- a/tools/lib/perf/Makefile +++ b/tools/lib/perf/Makefile @@ -181,7 +181,10 @@ install_pkgconfig: $(LIBPERF_PC) $(call QUIET_INSTALL, $(LIBPERF_PC)) \ $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) -install: install_lib install_headers install_pkgconfig +install_doc: + $(Q)$(MAKE) -C Documentation install-man install-html install-examples + +install: install_lib install_headers install_pkgconfig install_doc FORCE: From 6794200fa3c9c3e6759dae099145f23e4310f4f7 Mon Sep 17 00:00:00 2001 From: Andrey Zhizhikin Date: Wed, 11 Dec 2019 08:01:09 +0000 Subject: [PATCH 12/24] tools lib api fs: Fix gcc9 stringop-truncation compilation error GCC9 introduced string hardening mechanisms, which exhibits the error during fs api compilation: error: '__builtin_strncpy' specified bound 4096 equals destination size [-Werror=stringop-truncation] This comes when the length of copy passed to strncpy is is equal to destination size, which could potentially lead to buffer overflow. There is a need to mitigate this potential issue by limiting the size of destination by 1 and explicitly terminate the destination with NULL. Signed-off-by: Andrey Zhizhikin Reviewed-by: Petr Mladek Acked-by: Jiri Olsa Cc: Alexei Starovoitov Cc: Andrii Nakryiko Cc: Daniel Borkmann Cc: Kefeng Wang Cc: Martin KaFai Lau Cc: Petr Mladek Cc: Sergey Senozhatsky Cc: Song Liu Cc: Yonghong Song Cc: bpf@vger.kernel.org Cc: netdev@vger.kernel.org Link: http://lore.kernel.org/lkml/20191211080109.18765-1-andrey.zhizhikin@leica-geosystems.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/api/fs/fs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/lib/api/fs/fs.c b/tools/lib/api/fs/fs.c index 11b3885e833ed..027b18f7ed8cf 100644 --- a/tools/lib/api/fs/fs.c +++ b/tools/lib/api/fs/fs.c @@ -210,6 +210,7 @@ static bool fs__env_override(struct fs *fs) size_t name_len = strlen(fs->name); /* name + "_PATH" + '\0' */ char upper_name[name_len + 5 + 1]; + memcpy(upper_name, fs->name, name_len); mem_toupper(upper_name, name_len); strcpy(&upper_name[name_len], "_PATH"); @@ -219,7 +220,8 @@ static bool fs__env_override(struct fs *fs) return false; fs->found = true; - strncpy(fs->path, override_path, sizeof(fs->path)); + strncpy(fs->path, override_path, sizeof(fs->path) - 1); + fs->path[sizeof(fs->path) - 1] = '\0'; return true; } From 3f7774033e6820d25beee5cf7aefa11d4968b951 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 16 Dec 2019 13:22:33 -0300 Subject: [PATCH 13/24] perf hists browser: Restore ESC as "Zoom out" of DSO/thread/etc We need to set actions->ms.map since 599a2f38a989 ("perf hists browser: Check sort keys before hot key actions"), as in that patch we bail out if map is NULL. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Namhyung Kim Fixes: 599a2f38a989 ("perf hists browser: Check sort keys before hot key actions") Link: https://lkml.kernel.org/n/tip-wp1ssoewy6zihwwexqpohv0j@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index d4d3558fdef42..cfc6172ecab76 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -3062,6 +3062,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, continue; } + actions->ms.map = map; top = pstack__peek(browser->pstack); if (top == &browser->hists->dso_filter) { /* From 9218a9132f83d2c08cd23c1fd8e8e9b63b47cb5f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 10:02:33 -0300 Subject: [PATCH 14/24] perf report/top: Make ENTER consistently bring up menu When callchains are present the ENTER key switches from bringing up the menu that offers Annotation, Zoom by DSO, etc to expanding/collapsing one callchain level, causing confusion, fix it by making it consistently bring up the menu and use '+' to expand/collapse one callchain level. Next patch will also add an entry to the menu to allow expanding/collapsing, so that people used to ENTER expanding one callchain level can quickly find it and use it instead. Reported-by: Linus Torvalds Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-bjz35omktig8cwn6lbj1ifns@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index cfc6172ecab76..fefa505d4fa8e 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -706,7 +706,7 @@ int hist_browser__run(struct hist_browser *browser, const char *help, browser->show_headers = !browser->show_headers; hist_browser__update_rows(browser); break; - case K_ENTER: + case '+': if (hist_browser__toggle_fold(browser)) break; /* fall thru */ @@ -2858,6 +2858,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, "For symbolic views (--sort has sym):\n\n" \ "ENTER Zoom into DSO/Threads & Annotate current symbol\n" \ "ESC Zoom out\n" \ + "+ Expand/Collapse one callchain level\n" \ "a Annotate current symbol\n" \ "C Collapse all callchains\n" \ "d Zoom into current DSO\n" \ From d5a599d9890f51cb2cabfa21f8c38bb6f51f4bb2 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 10:58:21 -0300 Subject: [PATCH 15/24] perf report/top: Add menu entry for toggling callchain expansion Since previously pressing ENTER toggled expansion/collapse of callchain entries and now brings up the same menu used when callchains are not present, add an entry so that users can quickly figure out the change in behaviour. Its worth mentioning that we also always had 'e'/'c' to expand/collapse all entries in a hist entry and 'E'/'C' for all hist entries. Suggested-by: Linus Torvalds Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-f9o03jo29fypvd8ly3j49d36@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index fefa505d4fa8e..1b5a5990dddb4 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -2527,6 +2527,26 @@ add_dso_opt(struct hist_browser *browser, struct popup_action *act, return 1; } +static int do_toggle_callchain(struct hist_browser *browser, struct popup_action *act __maybe_unused) +{ + hist_browser__toggle_fold(browser); + return 0; +} + +static int add_callchain_toggle_opt(struct hist_browser *browser, struct popup_action *act, char **optstr) +{ + struct hist_entry *he = browser->he_selection; + + if (!he->has_children) + return 0; + + if (asprintf(optstr, "Expand/Collapse callchain") < 0) + return 0; + + act->fn = do_toggle_callchain; + return 1; +} + static int do_browse_map(struct hist_browser *browser __maybe_unused, struct popup_action *act) @@ -3137,6 +3157,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, &options[nr_options], thread); nr_options += add_dso_opt(browser, &actions[nr_options], &options[nr_options], map); + nr_options += add_callchain_toggle_opt(browser, &actions[nr_options], &options[nr_options]); nr_options += add_map_opt(browser, &actions[nr_options], &options[nr_options], browser->selection ? From bdc633fec50be7e6856b9dee89af9bb7c5e9a04f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 11:48:23 -0300 Subject: [PATCH 16/24] perf report/top: Improve toggle callchain menu option Taking into account the current status of the callchain, i.e. if folded, show "Expand", otherwise "Collapse", also show the name of the entry that will be affected and mention the hotkeys for expanding/collapsing all callchains below the main entry, the one that appears with/without callchains. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-03arm6poo8463k5tfcfp7gkk@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 54 ++++++++++++++++++++++++++++++++-- tools/perf/util/sort.c | 3 +- tools/perf/util/sort.h | 2 ++ 3 files changed, 54 insertions(+), 5 deletions(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 1b5a5990dddb4..a4413d983216c 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -391,6 +391,52 @@ static void hist_entry__init_have_children(struct hist_entry *he) he->init_have_children = true; } +static bool hist_browser__selection_has_children(struct hist_browser *browser) +{ + struct hist_entry *he = browser->he_selection; + struct map_symbol *ms = browser->selection; + + if (!he || !ms) + return false; + + if (ms == &he->ms) + return he->has_children; + + return container_of(ms, struct callchain_list, ms)->has_children; +} + +static bool hist_browser__selection_unfolded(struct hist_browser *browser) +{ + struct hist_entry *he = browser->he_selection; + struct map_symbol *ms = browser->selection; + + if (!he || !ms) + return false; + + if (ms == &he->ms) + return he->unfolded; + + return container_of(ms, struct callchain_list, ms)->unfolded; +} + +static char *hist_browser__selection_sym_name(struct hist_browser *browser, char *bf, size_t size) +{ + struct hist_entry *he = browser->he_selection; + struct map_symbol *ms = browser->selection; + struct callchain_list *callchain_entry; + + if (!he || !ms) + return NULL; + + if (ms == &he->ms) { + hist_entry__sym_snprintf(he, bf, size, 0); + return bf + 4; // skip the level, e.g. '[k] ' + } + + callchain_entry = container_of(ms, struct callchain_list, ms); + return callchain_list__sym_name(callchain_entry, bf, size, browser->show_dso); +} + static bool hist_browser__toggle_fold(struct hist_browser *browser) { struct hist_entry *he = browser->he_selection; @@ -2535,12 +2581,14 @@ static int do_toggle_callchain(struct hist_browser *browser, struct popup_action static int add_callchain_toggle_opt(struct hist_browser *browser, struct popup_action *act, char **optstr) { - struct hist_entry *he = browser->he_selection; + char sym_name[512]; - if (!he->has_children) + if (!hist_browser__selection_has_children(browser)) return 0; - if (asprintf(optstr, "Expand/Collapse callchain") < 0) + if (asprintf(optstr, "%s [%s] callchain (one level, same as '+' hotkey, use 'e'/'c' for the whole main level entry)", + hist_browser__selection_unfolded(browser) ? "Collapse" : "Expand", + hist_browser__selection_sym_name(browser, sym_name, sizeof(sym_name))) < 0) return 0; act->fn = do_toggle_callchain; diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index 9fcba2872130b..ab0cfd790ad00 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -324,8 +324,7 @@ static int _hist_entry__sym_snprintf(struct map_symbol *ms, return ret; } -static int hist_entry__sym_snprintf(struct hist_entry *he, char *bf, - size_t size, unsigned int width) +int hist_entry__sym_snprintf(struct hist_entry *he, char *bf, size_t size, unsigned int width) { return _hist_entry__sym_snprintf(&he->ms, he->ip, he->level, bf, size, width); diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 5aff9542d9b79..6c862d62d0525 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -164,6 +164,8 @@ static __pure inline bool hist_entry__has_callchains(struct hist_entry *he) return he->callchain_size != 0; } +int hist_entry__sym_snprintf(struct hist_entry *he, char *bf, size_t size, unsigned int width); + static inline bool hist_entry__has_pairs(struct hist_entry *he) { return !list_empty(&he->pairs.node); From 632003f400d341592b6af8d96bd83b74a0329fe3 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 11:55:54 -0300 Subject: [PATCH 17/24] perf hists browser: Generalize the do_zoom_dso() function We'll use it to provide a top level hotkey to zoom into the kernel dso directly. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-ae9cjel6v05wjnz9r6z77b6x@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index a4413d983216c..8aba1aeea0ebd 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -2530,11 +2530,8 @@ add_thread_opt(struct hist_browser *browser, struct popup_action *act, return 1; } -static int -do_zoom_dso(struct hist_browser *browser, struct popup_action *act) +static int hists_browser__zoom_map(struct hist_browser *browser, struct map *map) { - struct map *map = act->ms.map; - if (!hists__has(browser->hists, dso) || map == NULL) return 0; @@ -2556,6 +2553,12 @@ do_zoom_dso(struct hist_browser *browser, struct popup_action *act) return 0; } +static int +do_zoom_dso(struct hist_browser *browser, struct popup_action *act) +{ + return hists_browser__zoom_map(browser, act->ms.map); +} + static int add_dso_opt(struct hist_browser *browser, struct popup_action *act, char **optstr, struct map *map) From 209f4e70a2f10bc6819eb20c5bc3988be31972c6 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 13:06:36 -0300 Subject: [PATCH 18/24] perf report/top: Add 'k' hotkey to zoom directly into the kernel map As a convenience, equivalent to pressing Enter in a line with a kernel symbol and then selecting "Zoom" into the kernel DSO. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-vbnlnrpyfvz9deqoobtc3dz7@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 8aba1aeea0ebd..6dfdd8d5a7438 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -18,7 +18,9 @@ #include "../../util/evlist.h" #include "../../util/header.h" #include "../../util/hist.h" +#include "../../util/machine.h" #include "../../util/map.h" +#include "../../util/maps.h" #include "../../util/symbol.h" #include "../../util/map_symbol.h" #include "../../util/branch.h" @@ -2566,7 +2568,7 @@ add_dso_opt(struct hist_browser *browser, struct popup_action *act, if (!hists__has(browser->hists, dso) || map == NULL) return 0; - if (asprintf(optstr, "Zoom %s %s DSO", + if (asprintf(optstr, "Zoom %s %s DSO (use the 'k' hotkey to zoom directly into the kernel)", browser->hists->dso_filter ? "out of" : "into", __map__is_kernel(map) ? "the Kernel" : map->dso->short_name) < 0) return 0; @@ -2936,6 +2938,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, "E Expand all callchains\n" \ "F Toggle percentage of filtered entries\n" \ "H Display column headers\n" \ + "k Zoom into the kernel map\n" \ "L Change percent limit\n" \ "m Display context menu\n" \ "S Zoom into current Processor Socket\n" \ @@ -3033,6 +3036,10 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, actions->ms.map = map; do_zoom_dso(browser, actions); continue; + case 'k': + if (browser->selection != NULL) + hists_browser__zoom_map(browser, browser->selection->maps->machine->vmlinux_map); + continue; case 'V': verbose = (verbose + 1) % 4; browser->show_dso = verbose > 0; From d10ec006dcd7b20b2eb7e9ef32fc6f83b0112893 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Dec 2019 15:31:40 -0300 Subject: [PATCH 19/24] perf hists browser: Allow passing an initial hotkey Sometimes we're in an outer code, like the main hists browser popup menu and the user follows a suggestion about using some hotkey, and that hotkey is really handled by hists_browser__run(), so allow for calling it with that hotkey, making it handle it instead of waiting for the user to press one. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-xv2l7i6o4urn37nv1h40ryfs@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-c2c.c | 4 +- tools/perf/ui/browsers/hists.c | 153 +++++++++++++++++---------------- tools/perf/ui/browsers/hists.h | 2 +- 3 files changed, 82 insertions(+), 77 deletions(-) diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c index e69f44941aad9..346351260c0b0 100644 --- a/tools/perf/builtin-c2c.c +++ b/tools/perf/builtin-c2c.c @@ -2384,7 +2384,7 @@ static int perf_c2c__browse_cacheline(struct hist_entry *he) c2c_browser__update_nr_entries(browser); while (1) { - key = hist_browser__run(browser, "? - help", true); + key = hist_browser__run(browser, "? - help", true, 0); switch (key) { case 's': @@ -2453,7 +2453,7 @@ static int perf_c2c__hists_browse(struct hists *hists) c2c_browser__update_nr_entries(browser); while (1) { - key = hist_browser__run(browser, "? - help", true); + key = hist_browser__run(browser, "? - help", true, 0); switch (key) { case 'q': diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 6dfdd8d5a7438..ac118aef5ed11 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -672,10 +672,81 @@ static int hist_browser__title(struct hist_browser *browser, char *bf, size_t si return browser->title ? browser->title(browser, bf, size) : 0; } +static int hist_browser__handle_hotkey(struct hist_browser *browser, bool warn_lost_event, char *title, int key) +{ + switch (key) { + case K_TIMER: { + struct hist_browser_timer *hbt = browser->hbt; + u64 nr_entries; + + WARN_ON_ONCE(!hbt); + + if (hbt) + hbt->timer(hbt->arg); + + if (hist_browser__has_filter(browser) || symbol_conf.report_hierarchy) + hist_browser__update_nr_entries(browser); + + nr_entries = hist_browser__nr_entries(browser); + ui_browser__update_nr_entries(&browser->b, nr_entries); + + if (warn_lost_event && + (browser->hists->stats.nr_lost_warned != + browser->hists->stats.nr_events[PERF_RECORD_LOST])) { + browser->hists->stats.nr_lost_warned = + browser->hists->stats.nr_events[PERF_RECORD_LOST]; + ui_browser__warn_lost_events(&browser->b); + } + + hist_browser__title(browser, title, sizeof(title)); + ui_browser__show_title(&browser->b, title); + break; + } + case 'D': { /* Debug */ + struct hist_entry *h = rb_entry(browser->b.top, struct hist_entry, rb_node); + static int seq; + + ui_helpline__pop(); + ui_helpline__fpush("%d: nr_ent=(%d,%d), etl: %d, rows=%d, idx=%d, fve: idx=%d, row_off=%d, nrows=%d", + seq++, browser->b.nr_entries, browser->hists->nr_entries, + browser->b.extra_title_lines, browser->b.rows, + browser->b.index, browser->b.top_idx, h->row_offset, h->nr_rows); + } + break; + case 'C': + /* Collapse the whole world. */ + hist_browser__set_folding(browser, false); + break; + case 'c': + /* Collapse the selected entry. */ + hist_browser__set_folding_selected(browser, false); + break; + case 'E': + /* Expand the whole world. */ + hist_browser__set_folding(browser, true); + break; + case 'e': + /* Expand the selected entry. */ + hist_browser__set_folding_selected(browser, true); + break; + case 'H': + browser->show_headers = !browser->show_headers; + hist_browser__update_rows(browser); + break; + case '+': + if (hist_browser__toggle_fold(browser)) + break; + /* fall thru */ + default: + return -1; + } + + return 0; +} + int hist_browser__run(struct hist_browser *browser, const char *help, - bool warn_lost_event) + bool warn_lost_event, int key) { - int key; char title[160]; struct hist_browser_timer *hbt = browser->hbt; int delay_secs = hbt ? hbt->refresh : 0; @@ -688,79 +759,14 @@ int hist_browser__run(struct hist_browser *browser, const char *help, if (ui_browser__show(&browser->b, title, "%s", help) < 0) return -1; + if (key && hist_browser__handle_hotkey(browser, warn_lost_event, title, key)) + goto out; + while (1) { key = ui_browser__run(&browser->b, delay_secs); - switch (key) { - case K_TIMER: { - u64 nr_entries; - - WARN_ON_ONCE(!hbt); - - if (hbt) - hbt->timer(hbt->arg); - - if (hist_browser__has_filter(browser) || - symbol_conf.report_hierarchy) - hist_browser__update_nr_entries(browser); - - nr_entries = hist_browser__nr_entries(browser); - ui_browser__update_nr_entries(&browser->b, nr_entries); - - if (warn_lost_event && - (browser->hists->stats.nr_lost_warned != - browser->hists->stats.nr_events[PERF_RECORD_LOST])) { - browser->hists->stats.nr_lost_warned = - browser->hists->stats.nr_events[PERF_RECORD_LOST]; - ui_browser__warn_lost_events(&browser->b); - } - - hist_browser__title(browser, title, sizeof(title)); - ui_browser__show_title(&browser->b, title); - continue; - } - case 'D': { /* Debug */ - static int seq; - struct hist_entry *h = rb_entry(browser->b.top, - struct hist_entry, rb_node); - ui_helpline__pop(); - ui_helpline__fpush("%d: nr_ent=(%d,%d), etl: %d, rows=%d, idx=%d, fve: idx=%d, row_off=%d, nrows=%d", - seq++, browser->b.nr_entries, - browser->hists->nr_entries, - browser->b.extra_title_lines, - browser->b.rows, - browser->b.index, - browser->b.top_idx, - h->row_offset, h->nr_rows); - } - break; - case 'C': - /* Collapse the whole world. */ - hist_browser__set_folding(browser, false); - break; - case 'c': - /* Collapse the selected entry. */ - hist_browser__set_folding_selected(browser, false); - break; - case 'E': - /* Expand the whole world. */ - hist_browser__set_folding(browser, true); + if (hist_browser__handle_hotkey(browser, warn_lost_event, title, key)) break; - case 'e': - /* Expand the selected entry. */ - hist_browser__set_folding_selected(browser, true); - break; - case 'H': - browser->show_headers = !browser->show_headers; - hist_browser__update_rows(browser); - break; - case '+': - if (hist_browser__toggle_fold(browser)) - break; - /* fall thru */ - default: - goto out; - } } out: ui_browser__hide(&browser->b); @@ -2994,8 +3000,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, nr_options = 0; - key = hist_browser__run(browser, helpline, - warn_lost_event); + key = hist_browser__run(browser, helpline, warn_lost_event, 0); if (browser->he_selection != NULL) { thread = hist_browser__selected_thread(browser); @@ -3573,7 +3578,7 @@ int block_hists_tui_browse(struct block_hist *bh, struct evsel *evsel, memset(&action, 0, sizeof(action)); while (1) { - key = hist_browser__run(browser, "? - help", true); + key = hist_browser__run(browser, "? - help", true, 0); switch (key) { case 'q': diff --git a/tools/perf/ui/browsers/hists.h b/tools/perf/ui/browsers/hists.h index 078f2f2c7abd2..1e938d9ffa5ee 100644 --- a/tools/perf/ui/browsers/hists.h +++ b/tools/perf/ui/browsers/hists.h @@ -34,7 +34,7 @@ struct hist_browser { struct hist_browser *hist_browser__new(struct hists *hists); void hist_browser__delete(struct hist_browser *browser); int hist_browser__run(struct hist_browser *browser, const char *help, - bool warn_lost_event); + bool warn_lost_event, int key); void hist_browser__init(struct hist_browser *browser, struct hists *hists); #endif /* _PERF_UI_BROWSER_HISTS_H_ */ From d07126560cab572539621702137eeeeb2a4edf30 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 16 Dec 2019 12:23:34 -0300 Subject: [PATCH 20/24] tools ui popup: Allow returning hotkeys With this patch if an optional pointer is passed to ui__popup_menu() then when any key that is not being handled (ENTER, ESC, etc) is typed, it'll record that key in the pointer and return, allowing for hotkey processing on the caller. If NULL is passed, no change in logic, unhandled keys continue to be ignored. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-6ojn19mqzgmrdm8kdoigic0m@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 4 ++-- tools/perf/ui/browsers/res_sample.c | 2 +- tools/perf/ui/browsers/scripts.c | 2 +- tools/perf/ui/tui/util.c | 12 ++++++++---- tools/perf/ui/util.h | 2 +- 5 files changed, 13 insertions(+), 9 deletions(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index ac118aef5ed11..c44b508f9e064 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -2393,7 +2393,7 @@ static int switch_data_file(void) closedir(pwd_dir); if (nr_options) { - choice = ui__popup_menu(nr_options, options); + choice = ui__popup_menu(nr_options, options, NULL); if (choice < nr_options && choice >= 0) { tmp = strdup(abs_path[choice]); if (tmp) { @@ -3279,7 +3279,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, do { struct popup_action *act; - choice = ui__popup_menu(nr_options, options); + choice = ui__popup_menu(nr_options, options, NULL); if (choice == -1 || choice >= nr_options) break; diff --git a/tools/perf/ui/browsers/res_sample.c b/tools/perf/ui/browsers/res_sample.c index 76d356a187906..7cb2d66780397 100644 --- a/tools/perf/ui/browsers/res_sample.c +++ b/tools/perf/ui/browsers/res_sample.c @@ -56,7 +56,7 @@ int res_sample_browse(struct res_sample *res_samples, int num_res, return -1; } } - choice = ui__popup_menu(num_res, names); + choice = ui__popup_menu(num_res, names, NULL); for (i = 0; i < num_res; i++) zfree(&names[i]); free(names); diff --git a/tools/perf/ui/browsers/scripts.c b/tools/perf/ui/browsers/scripts.c index fc733a6354d4d..47d2c7a8cbe13 100644 --- a/tools/perf/ui/browsers/scripts.c +++ b/tools/perf/ui/browsers/scripts.c @@ -126,7 +126,7 @@ static int list_scripts(char *script_name, bool *custom, SCRIPT_FULLPATH_LEN); if (num < 0) num = 0; - choice = ui__popup_menu(num + max_std, (char * const *)names); + choice = ui__popup_menu(num + max_std, (char * const *)names, NULL); if (choice < 0) { ret = -1; goto out; diff --git a/tools/perf/ui/tui/util.c b/tools/perf/ui/tui/util.c index b98dd0e31dc1a..0f562e2cb1e88 100644 --- a/tools/perf/ui/tui/util.c +++ b/tools/perf/ui/tui/util.c @@ -23,7 +23,7 @@ static void ui_browser__argv_write(struct ui_browser *browser, ui_browser__write_nstring(browser, *arg, browser->width); } -static int popup_menu__run(struct ui_browser *menu) +static int popup_menu__run(struct ui_browser *menu, int *keyp) { int key; @@ -45,6 +45,11 @@ static int popup_menu__run(struct ui_browser *menu) key = -1; break; default: + if (keyp) { + *keyp = key; + key = menu->nr_entries; + break; + } continue; } @@ -55,7 +60,7 @@ static int popup_menu__run(struct ui_browser *menu) return key; } -int ui__popup_menu(int argc, char * const argv[]) +int ui__popup_menu(int argc, char * const argv[], int *keyp) { struct ui_browser menu = { .entries = (void *)argv, @@ -64,8 +69,7 @@ int ui__popup_menu(int argc, char * const argv[]) .write = ui_browser__argv_write, .nr_entries = argc, }; - - return popup_menu__run(&menu); + return popup_menu__run(&menu, keyp); } int ui_browser__input_window(const char *title, const char *text, char *input, diff --git a/tools/perf/ui/util.h b/tools/perf/ui/util.h index 40891942f465d..e30cea807564f 100644 --- a/tools/perf/ui/util.h +++ b/tools/perf/ui/util.h @@ -5,7 +5,7 @@ #include int ui__getch(int delay_secs); -int ui__popup_menu(int argc, char * const argv[]); +int ui__popup_menu(int argc, char * const argv[], int *keyp); int ui__help_window(const char *text); int ui__dialog_yesno(const char *msg); void __ui__info_window(const char *title, const char *text, const char *exit_msg); From 4c8b9c0f4281c8517542c26425aade3a31988575 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 16 Dec 2019 13:27:47 -0300 Subject: [PATCH 21/24] perf report/top: Allow pressing hotkeys in the options popup menu When the users presses ENTER in the main 'perf report/top' screen a popup menu is presented, in it some hotkeys are suggested as alternatives to using the menu, or for additional features. At that point the user may try those hotkeys, so allow for that by recording the key used and exiting, the caller then can check for that possibility and process the hotkey. I.e. try pressing ENTER, and then 'k' to exit and zoom into the kernel map, using ESC then zooms out, etc. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-ujfq3fw44kf6qrtfajl5dcsp@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index c44b508f9e064..8776b1cb29b73 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -2995,12 +2995,13 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, while (1) { struct thread *thread = NULL; struct map *map = NULL; - int choice = 0; + int choice; int socked_id = -1; - nr_options = 0; - - key = hist_browser__run(browser, helpline, warn_lost_event, 0); + key = 0; // reset key +do_hotkey: // key came straight from options ui__popup_menu() + choice = nr_options = 0; + key = hist_browser__run(browser, helpline, warn_lost_event, key); if (browser->he_selection != NULL) { thread = hist_browser__selected_thread(browser); @@ -3279,10 +3280,13 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, do { struct popup_action *act; - choice = ui__popup_menu(nr_options, options, NULL); - if (choice == -1 || choice >= nr_options) + choice = ui__popup_menu(nr_options, options, &key); + if (choice == -1) break; + if (choice == nr_options) + goto do_hotkey; + act = &actions[choice]; key = act->fn(browser, act); } while (key == 1); From ea537f22f6e5b3e2026fc58419cc937d833b1a61 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 16 Dec 2019 18:21:16 -0300 Subject: [PATCH 22/24] perf report/top: Do not offer annotation for symbols without samples This can happen in the --children mode, i.e. the default mode when callchains are present, where one of the main entries may be a callchain entry with no samples. So far we were not providing any information about why an annotation couldn't be provided even offering the Annotation option in the popup menu. Work is needed to allow for no-samples "annotation', i.e. to show the disassembly anyway and allow for navigation, etc. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-0hhzj2de15o88cguy7h66zre@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 8776b1cb29b73..3bec8de898808 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -2465,7 +2465,8 @@ add_annotate_opt(struct hist_browser *browser __maybe_unused, struct popup_action *act, char **optstr, struct map_symbol *ms) { - if (ms->sym == NULL || ms->map->dso->annotate_warned) + if (ms->sym == NULL || ms->map->dso->annotate_warned || + symbol__annotation(ms->sym)->src == NULL) return 0; if (asprintf(optstr, "Annotate %s", ms->sym->name) < 0) @@ -3031,6 +3032,14 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, browser->selection->map->dso->annotate_warned) continue; + if (symbol__annotation(browser->selection->sym)->src == NULL) { + ui_browser__warning(&browser->b, delay_secs * 2, + "No samples for the \"%s\" symbol.\n\n" + "Probably appeared just in a callchain", + browser->selection->sym->name); + continue; + } + actions->ms.map = browser->selection->map; actions->ms.sym = browser->selection->sym; do_annotate(browser, actions); From e6d6abfc447a65e949c1e883e66e1450903d2fbd Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 17 Dec 2019 10:39:04 -0300 Subject: [PATCH 23/24] perf report/top: Make 'e' visible in the help and make it toggle showing callchains The 'e' and 'c' hotkeys were present for a long time, but not documented in the help window, change 'e' to be a toggle so that it gets consistent with other toggles like '+' and document it in the help window. Keep 'c' as is for people used to it but don't document, as it is easier to just use 'e' to show/hide all the callchains for a top level histogram entry. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-pmyi5x34stlqmyu81rci94x9@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 3bec8de898808..f36dee4993201 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -407,6 +407,11 @@ static bool hist_browser__selection_has_children(struct hist_browser *browser) return container_of(ms, struct callchain_list, ms)->has_children; } +static bool hist_browser__he_selection_unfolded(struct hist_browser *browser) +{ + return browser->he_selection ? browser->he_selection->unfolded : false; +} + static bool hist_browser__selection_unfolded(struct hist_browser *browser) { struct hist_entry *he = browser->he_selection; @@ -727,7 +732,7 @@ static int hist_browser__handle_hotkey(struct hist_browser *browser, bool warn_l break; case 'e': /* Expand the selected entry. */ - hist_browser__set_folding_selected(browser, true); + hist_browser__set_folding_selected(browser, !hist_browser__he_selection_unfolded(browser)); break; case 'H': browser->show_headers = !browser->show_headers; @@ -2942,6 +2947,7 @@ static int perf_evsel__hists_browse(struct evsel *evsel, int nr_events, "a Annotate current symbol\n" \ "C Collapse all callchains\n" \ "d Zoom into current DSO\n" \ + "e Expand/Collapse main entry callchains\n" \ "E Expand all callchains\n" \ "F Toggle percentage of filtered entries\n" \ "H Display column headers\n" \ From 6c4798d3f08b81c2c52936b10e0fa872590c96ae Mon Sep 17 00:00:00 2001 From: Vitaly Chikunov Date: Tue, 24 Dec 2019 20:20:29 +0300 Subject: [PATCH 24/24] tools lib: Fix builds when glibc contains strlcpy() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Disable a couple of compilation warnings (which are treated as errors) on strlcpy() definition and declaration, allowing users to compile perf and kernel (objtool) when: 1. glibc have strlcpy() (such as in ALT Linux since 2004) objtool and perf build fails with this (in gcc): In file included from exec-cmd.c:3: tools/include/linux/string.h:20:15: error: redundant redeclaration of ‘strlcpy’ [-Werror=redundant-decls] 20 | extern size_t strlcpy(char *dest, const char *src, size_t size); 2. clang ignores `-Wredundant-decls', but produces another warning when building perf: CC util/string.o ../lib/string.c:99:8: error: attribute declaration must precede definition [-Werror,-Wignored-attributes] size_t __weak strlcpy(char *dest, const char *src, size_t size) ../../tools/include/linux/compiler.h:66:34: note: expanded from macro '__weak' # define __weak __attribute__((weak)) /usr/include/bits/string_fortified.h:151:8: note: previous definition is here __NTH (strlcpy (char *__restrict __dest, const char *__restrict __src, Committer notes: The #pragma GCC diagnostic directive was introduced in gcc 4.6, so check for that as well. Fixes: ce99091 ("perf tools: Move strlcpy() from perf to tools/lib/string.c") Fixes: 0215d59 ("tools lib: Reinstate strlcpy() header guard with __UCLIBC__") Resolves: https://bugzilla.kernel.org/show_bug.cgi?id=118481 Signed-off-by: Vitaly Chikunov Reviewed-by: Dmitry Levin Cc: Dmitry Levin Cc: Josh Poimboeuf Cc: kbuild test robot Cc: Peter Zijlstra Cc: stable@vger.kernel.org Cc: Vineet Gupta Link: http://lore.kernel.org/lkml/20191224172029.19690-1-vt@altlinux.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/linux/string.h | 8 ++++++++ tools/lib/string.c | 7 +++++++ 2 files changed, 15 insertions(+) diff --git a/tools/include/linux/string.h b/tools/include/linux/string.h index 980cb9266718a..5e9e781905edc 100644 --- a/tools/include/linux/string.h +++ b/tools/include/linux/string.h @@ -17,7 +17,15 @@ int strtobool(const char *s, bool *res); * However uClibc headers also define __GLIBC__ hence the hack below */ #if defined(__GLIBC__) && !defined(__UCLIBC__) +// pragma diagnostic was introduced in gcc 4.6 +#if __GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wredundant-decls" +#endif extern size_t strlcpy(char *dest, const char *src, size_t size); +#if __GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6) +#pragma GCC diagnostic pop +#endif #endif char *str_error_r(int errnum, char *buf, size_t buflen); diff --git a/tools/lib/string.c b/tools/lib/string.c index f2ae1b87c7195..f645343815de6 100644 --- a/tools/lib/string.c +++ b/tools/lib/string.c @@ -96,6 +96,10 @@ int strtobool(const char *s, bool *res) * If libc has strlcpy() then that version will override this * implementation: */ +#ifdef __clang__ +#pragma clang diagnostic push +#pragma clang diagnostic ignored "-Wignored-attributes" +#endif size_t __weak strlcpy(char *dest, const char *src, size_t size) { size_t ret = strlen(src); @@ -107,6 +111,9 @@ size_t __weak strlcpy(char *dest, const char *src, size_t size) } return ret; } +#ifdef __clang__ +#pragma clang diagnostic pop +#endif /** * skip_spaces - Removes leading whitespace from @str.