Refactor perf cpumap
Subsystems (4)
Ian Rogers
reviewerv2 cover Refactor perf cpumap
Perf cpu map has various functions where a cpumap and index are passed in order to load the cpu. A problem with this is that the wrong index may be passed for the cpumap, causing problems like aggregation on the wrong CPU:
https://lore.kernel.org/lkml/[email protected]/
This patch set refactors the cpu map API, reducing it and explicitly passing the cpu (rather than the pair) to functions that need it. Comments are added at the same time. Changes modify the same file/function more than once as refactoring and fixes are broken apart for the sake of bisection.
v2. Incorproates fixes suggested Jiri Olsa, rewrites the evlist CPU
iterator in part in a way suggested by Riccardo Mancini. The new fixes start at patch 23. The final change was suggested by John Garry to make the CPUs have their own struct wrapper.
Ian Rogers (48):
libperf: Add comments to perf_cpu_map.
perf stat: Add aggr creators that are passed a cpu. perf stat: Correct aggregation CPU map
perf stat: Switch aggregation to use for_each loop perf stat: Switch to cpu version of cpu_map__get
perf cpumap: Switch cpu_map__build_map to cpu function perf cpumap: Remove map+index get_socket
perf cpumap: Remove map+index get_die
perf cpumap: Remove map+index get_core
perf cpumap: Remove map+index get_node
perf cpumap: Add comments to aggr_cpu_id
perf cpumap: Remove unused cpu_map__socket
perf cpumap: Simplify equal function name.
perf cpumap: Rename empty functions.
perf cpumap: Document cpu__get_node and remove redundant function perf cpumap: Remove map from function names that don't use a map. perf cpumap: Remove cpu_map__cpu, use libperf function. perf cpumap: Refactor cpu_map__build_map
perf cpumap: Rename cpu_map__get_X_aggr_by_cpu functions perf cpumap: Move 'has' function to libperf
perf cpumap: Add some comments to cpu_aggr_map
perf cpumap: Trim the cpu_aggr_map
perf stat: Fix memory leak in check_per_pkg
perf cpumap: Add CPU to aggr_cpu_id
perf stat-display: Avoid use of core for CPU.
perf evsel: Derive CPUs and threads in alloc_counts
libperf: Switch cpu to more accurate cpu_map_idx
libperf: Use cpu not index for evsel mmap
perf counts: Switch name cpu to cpu_map_idx
perf stat: Rename aggr_data cpu to imply it's an index perf stat: Use perf_cpu_map__for_each_cpu
perf script: Use for each cpu to aid readability
libperf: Allow NULL in perf_cpu_map__idx
perf evlist: Refactor evlist__for_each_cpu.
perf evsel: Pass cpu not cpu map index to synthesize perf stat: Correct variable name for read counter
perf evsel: Rename CPU around get_group_fd
perf evsel: Reduce scope of evsel__ignore_missing_thread perf evsel: Rename variable cpu to index
perf test: Use perf_cpu_map__for_each_cpu
perf stat: Correct check_per_pkg cpu
perf stat: Swap variable name cpu to index
libperf: Sync evsel documentation
perf bpf: Rename cpu to cpu_map_idx
perf c2c: Use more intention revealing iterator
perf script: Fix flipped index and cpu
perf stat: Correct first_shadow_cpu to return index perf cpumap: Give CPUs their own type.
tools/lib/perf/Documentation/libperf.txt | 10 +- tools/lib/perf/cpumap.c | 131 +++-- tools/lib/perf/evlist.c | 4 +- tools/lib/perf/evsel.c | 92 ++-- tools/lib/perf/include/internal/cpumap.h | 18 +- tools/lib/perf/include/internal/evlist.h | 3 +- tools/lib/perf/include/internal/evsel.h | 4 +- tools/lib/perf/include/internal/mmap.h | 5 +- tools/lib/perf/include/perf/cpumap.h | 8 +- tools/lib/perf/include/perf/evsel.h | 10 +- tools/lib/perf/mmap.c | 2 +- tools/perf/arch/arm/util/cs-etm.c | 16 +- tools/perf/bench/epoll-ctl.c | 2 +- tools/perf/bench/epoll-wait.c | 2 +- tools/perf/bench/futex-hash.c | 2 +- tools/perf/bench/futex-lock-pi.c | 2 +- tools/perf/bench/futex-requeue.c | 2 +- tools/perf/bench/futex-wake-parallel.c | 2 +- tools/perf/bench/futex-wake.c | 2 +- tools/perf/builtin-c2c.c | 15 +- tools/perf/builtin-ftrace.c | 2 +- tools/perf/builtin-kmem.c | 2 +- tools/perf/builtin-record.c | 2 +- tools/perf/builtin-sched.c | 71 +-- tools/perf/builtin-script.c | 10 +- tools/perf/builtin-stat.c | 516 +++++++++--------- tools/perf/tests/attr.c | 6 +- tools/perf/tests/bitmap.c | 2 +- tools/perf/tests/cpumap.c | 6 +- tools/perf/tests/event_update.c | 6 +- tools/perf/tests/mem2node.c | 2 +- tools/perf/tests/mmap-basic.c | 4 +- tools/perf/tests/openat-syscall-all-cpus.c | 39 +- tools/perf/tests/stat.c | 3 +- tools/perf/tests/topology.c | 43 +- tools/perf/util/affinity.c | 2 +- tools/perf/util/auxtrace.c | 12 +- tools/perf/util/auxtrace.h | 5 +- tools/perf/util/bpf_counter.c | 16 +- tools/perf/util/bpf_counter.h | 4 +- tools/perf/util/counts.c | 8 +- tools/perf/util/counts.h | 14 +- tools/perf/util/cpumap.c | 250 ++++----- tools/perf/util/cpumap.h | 116 ++-- tools/perf/util/cputopo.c | 6 +- tools/perf/util/env.c | 29 +- tools/perf/util/env.h | 3 +- tools/perf/util/evlist.c | 148 ++--- tools/perf/util/evlist.h | 50 +- tools/perf/util/evsel.c | 143 ++--- tools/perf/util/evsel.h | 27 +- tools/perf/util/expr.c | 2 +- tools/perf/util/header.c | 6 +- tools/perf/util/mmap.c | 19 +- tools/perf/util/mmap.h | 3 +- tools/perf/util/perf_api_probe.c | 15 +- tools/perf/util/python.c | 2 +- tools/perf/util/record.c | 11 +- .../scripting-engines/trace-event-python.c | 6 +- tools/perf/util/session.c | 10 +- tools/perf/util/stat-display.c | 138 ++--- tools/perf/util/stat-shadow.c | 308 +++++------ tools/perf/util/stat.c | 47 +- tools/perf/util/stat.h | 9 +- tools/perf/util/svghelper.c | 6 +- tools/perf/util/synthetic-events.c | 12 +- tools/perf/util/synthetic-events.h | 3 +- tools/perf/util/util.h | 5 +- 68 files changed, 1327 insertions(+), 1154 deletions(-)
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 1/48 libperf: Add comments to perf_cpu_map.
A particular observed problem is confusing the index with the CPU value, documentation should hopefully reduce this type of problem.
--- tools/lib/perf/include/internal/cpumap.h | 9 +++++++++ 1 file changed, 9 insertions(+)
tools/lib/perf/include/internal/cpumap.h +9 -0
@@ -4,9 +4,18 @@#include <linux/refcount.h> +/** +* A sized, reference counted, sorted array of integers representing CPU +* numbers. This is commonly used to capture which CPUs a PMU is associated +* with. The indices into the cpumap are frequently used as they avoid having +* gaps if CPU numbers were used. For events associated with a pid, rather than +* a CPU, a single dummy map with an entry of -1 is used. +*/ struct perf_cpu_map {refcount_t refcnt; +/** Length of the map array. */ int nr; +/** The CPU values. */ int map[]; };
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 2/48 perf stat: Add aggr creators that are passed a cpu.
In reply to Ian RogersThe cpu_map and index can get confused. Add variants of the cpu_map__get routines that are passed a cpu. Make the existing cpu_map__get routines use the new functions with a view to remove them when no longer used.
--- tools/perf/util/cpumap.c | 79 +++++++++++++++++++++++----------------- tools/perf/util/cpumap.h | 6 ++- 2 files changed, 51 insertions(+), 34 deletions(-)
tools/perf/util/cpumap.c +46 -33
@@ -128,21 +128,23 @@ int cpu_map__get_socket_id(int cpu)return ret ?: value; } -struct aggr_cpu_id cpu_map__get_socket (struct perf_cpu_map *map,int idx, -void *data__maybe_unused ) +struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data__maybe_unused ) { -int cpu;struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); -if (idx > map->nr) -return id; - - cpu = map->map[idx]; - id.socket =cpu_map__get_socket_id (cpu);return id; } +struct aggr_cpu_id cpu_map__get_socket (struct perf_cpu_map *map,int idx, +void *data) +{ +if (idx <0 || idx > map->nr) +return cpu_map__empty_aggr_cpu_id (); + +return cpu_map__get_socket_aggr_by_cpu (map->map[idx], data); +} +static int cmp_aggr_cpu_id (const void *a_pointer,const void *b_pointer) {struct aggr_cpu_id *a = (struct aggr_cpu_id *)a_pointer;
@@ -200,15 +202,10 @@ int cpu_map__get_die_id(int cpu)return ret ?: value; } -struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx,void *data) +struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data) { -int cpu, die; -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); - -if (idx > map->nr) -return id; - - cpu = map->map[idx]; +struct aggr_cpu_id id; +int die; die =cpu_map__get_die_id (cpu);/* There is no die_id on legacy system. */
@@ -220,7 +217,7 @@ struct aggr_cpu_id cpu_map__get_die(struct perf_cpu_map *map, int idx, void *dat* with the socket ID and then add die to * make a unique ID. */ - id =cpu_map__get_socket (map, idx, data); + id =cpu_map__get_socket_aggr_by_cpu (cpu, data);if (cpu_map__aggr_cpu_id_is_empty (id))return id;
@@ -228,6 +225,15 @@ struct aggr_cpu_id cpu_map__get_die(struct perf_cpu_map *map, int idx, void *datreturn id; } +struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx, +void *data) +{ +if (idx <0 || idx > map->nr) +return cpu_map__empty_aggr_cpu_id (); + +return cpu_map__get_die_aggr_by_cpu (map->map[idx], data); +} +int cpu_map__get_core_id (int cpu) {int value, ret =cpu__get_topology_int (cpu,"core_id" , &value);
@@ -239,20 +245,13 @@ int cpu_map__get_node_id(int cpu)return cpu__get_node (cpu); } -struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data) +struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data) { -int cpu; -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); - -if (idx > map->nr) -return id; - - cpu = map->map[idx]; - - cpu =cpu_map__get_core_id (cpu); +struct aggr_cpu_id id; +int core =cpu_map__get_core_id (cpu);/* cpu_map__get_die returns a struct with socket and die set*/ - id =cpu_map__get_die (map, idx, data); + id =cpu_map__get_die_aggr_by_cpu (cpu, data);if (cpu_map__aggr_cpu_id_is_empty (id))return id;
@@ -260,19 +259,33 @@ struct aggr_cpu_id cpu_map__get_core(struct perf_cpu_map *map, int idx, void *da* core_id is relative to socket and die, we need a global id. * So we combine the result from cpu_map__get_die with the core id */ - id.core = cpu; + id.core = core;return id; + } -struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data__maybe_unused ) +struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data) +{ +if (idx <0 || idx > map->nr) +return cpu_map__empty_aggr_cpu_id (); + +return cpu_map__get_core_aggr_by_cpu (map->map[idx], data); +} + +struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); + id.node =cpu_map__get_node_id (cpu); +return id; +} + +struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data) +{if (idx <0 || idx >= map->nr) -return id; +return cpu_map__empty_aggr_cpu_id (); - id.node =cpu_map__get_node_id (map->map[idx]); -return id; +return cpu_map__get_node_aggr_by_cpu (map->map[idx], data); }int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp)
tools/perf/util/cpumap.h +5 -1
@@ -31,13 +31,17 @@ size_t cpu_map__snprint(struct perf_cpu_map *map, char *buf, size_t size);size_t cpu_map__snprint_mask (struct perf_cpu_map *map,char *buf,size_t size);size_t cpu_map__fprintf (struct perf_cpu_map *map,FILE *fp);int cpu_map__get_socket_id (int cpu); +struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_socket (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_die_id (int cpu); +struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_core_id (int cpu); +struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_node_id (int cpu); -struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data); +struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data); +struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data);int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp);int cpu_map__build_die_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **diep);int cpu_map__build_core_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **corep);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 3/48 perf stat: Correct aggregation CPU map
In reply to Ian RogersSwitch the perf_cpu_map in aggr_update_shadow from
the evlist to the counter's cpu map, so the index is appropriate. This addresses a problem where uncore counts, with a cpumap like:
$ cat /sys/devices/uncore_imc_0/cpumask 0,18 Don't aggregate counts in CPUs based on the index of those values in the cpumap (0 and 1) but on the actual CPU (0 and 18). Thereby correcting metric calculations in per-socket mode for counters without a full cpumask.
On a SkylakeX with a tweaked DRAM_BW_Use metric, to remove unnecessary scaling, this gives:
Before:
$ /perf stat --per-socket -M DRAM_BW_Use -I 1000
1.001102293 S0 1 27.01 MiB uncore_imc/cas_count_write/ # 103.00 DRAM_BW_Use
1.001102293 S0 1 30.22 MiB uncore_imc/cas_count_read/
1.001102293 S0 1 1,001,102,293 ns duration_time
1.001102293 S1 1 20.10 MiB uncore_imc/cas_count_write/ # 0.00 DRAM_BW_Use
1.001102293 S1 1 32.74 MiB uncore_imc/cas_count_read/
1.001102293 S1 0 <not counted> ns duration_time
2.003517973 S0 1 83.04 MiB uncore_imc/cas_count_write/ # 920.00 DRAM_BW_Use
2.003517973 S0 1 145.95 MiB uncore_imc/cas_count_read/
2.003517973 S0 1 1,002,415,680 ns duration_time
2.003517973 S1 1 302.45 MiB uncore_imc/cas_count_write/ # 0.00 DRAM_BW_Use
2.003517973 S1 1 290.99 MiB uncore_imc/cas_count_read/
2.003517973 S1 0 <not counted> ns duration_timeAfter:
$ perf stat --per-socket -M DRAM_BW_Use -I 1000
1.001080840 S0 1 24.96 MiB uncore_imc/cas_count_write/ # 54.00 DRAM_BW_Use
1.001080840 S0 1 33.64 MiB uncore_imc/cas_count_read/
1.001080840 S0 1 1,001,080,840 ns duration_time
1.001080840 S1 1 42.43 MiB uncore_imc/cas_count_write/ # 84.00 DRAM_BW_Use
1.001080840 S1 1 47.05 MiB uncore_imc/cas_count_read/
1.001080840 S1 0 <not counted> ns duration_time--- tools/perf/util/stat-display.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/stat-display.c +1 -1
@@ -526,7 +526,7 @@ static void aggr_update_shadow(struct perf_stat_config *config,evlist__for_each_entry (evlist, counter) { val =0 ;for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { - s2 = config->aggr_get_id (config, evlist->core.cpus, cpu); + s2 = config->aggr_get_id (config,evsel__cpus (counter), cpu);if (!cpu_map__compare_aggr_cpu_id (s2, id))continue ; val +=perf_counts (counter->counts, cpu,0 )->val;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 4/48 perf stat: Switch aggregation to use for_each loop
In reply to Ian RogersTidy up the use of cpu and index to hopefully make the code less error prone. Avoid unused warnings with (void) which will be removed in a later patch.
--- tools/perf/util/stat-display.c | 48 +++++++++++++++++++--------------- 1 file changed, 27 insertions(+), 21 deletions(-)
tools/perf/util/stat-display.c +27 -21
@@ -330,8 +330,8 @@ static void print_metric_header(struct perf_stat_config *config,static int first_shadow_cpu (struct perf_stat_config *config,struct evsel *evsel,struct aggr_cpu_id id) { -struct evlist *evlist = evsel->evlist; -int i; +struct perf_cpu_map *cpus; +int cpu, idx;if (config->aggr_mode ==AGGR_NONE )return id.core;
@@ -339,14 +339,11 @@ static int first_shadow_cpu(struct perf_stat_config *config,if (!config->aggr_get_id)return 0 ; -for (i =0 ; i <evsel__nr_cpus (evsel); i++) { -int cpu2 =evsel__cpus (evsel)->map[i]; - -if (cpu_map__compare_aggr_cpu_id ( - config->aggr_get_id (config, evlist->core.cpus, cpu2), - id)) { -return cpu2; - } + cpus =evsel__cpus (evsel); +perf_cpu_map__for_each_cpu (cpu, idx, cpus) { +if (cpu_map__compare_aggr_cpu_id (config->aggr_get_id (config, cpus, idx), + id)) +return cpu; }return 0 ; }
@@ -516,20 +513,23 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, intstatic void aggr_update_shadow (struct perf_stat_config *config,struct evlist *evlist) { -int cpu, s; +int cpu, idx, s;struct aggr_cpu_id s2, id;u64 val;struct evsel *counter; +struct perf_cpu_map *cpus;for (s =0 ; s < config->aggr_map->nr; s++) { id = config->aggr_map->map[s];evlist__for_each_entry (evlist, counter) { + cpus =evsel__cpus (counter); val =0 ; -for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { - s2 = config->aggr_get_id (config,evsel__cpus (counter), cpu); +perf_cpu_map__for_each_cpu (cpu, idx, cpus) { + (void )cpu; + s2 = config->aggr_get_id (config, cpus, idx);if (!cpu_map__compare_aggr_cpu_id (s2, id))continue ; - val +=perf_counts (counter->counts, cpu,0 )->val; + val +=perf_counts (counter->counts, idx,0 )->val; } perf_stat__update_shadow_stats(counter, val, first_shadow_cpu(config, counter, id),
@@ -634,18 +634,21 @@ static void aggr_cb(struct perf_stat_config *config, struct evsel *counter, void *data, bool first) { struct aggr_data *ad = data; -int cpu; +int idx, cpu; +struct perf_cpu_map *cpus;struct aggr_cpu_id s2; -for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { + cpus =evsel__cpus (counter); +perf_cpu_map__for_each_cpu (cpu, idx, cpus) {struct perf_counts_values *counts; - s2 = config->aggr_get_id (config,evsel__cpus (counter), cpu); + (void )cpu; + s2 = config->aggr_get_id (config, cpus, idx);if (!cpu_map__compare_aggr_cpu_id (s2, ad->id))continue ;if (first) ad->nr++; - counts =perf_counts (counter->counts, cpu,0 ); + counts =perf_counts (counter->counts, idx,0 );/* * When any result is bad, make them all to give * consistent output in interval mode.
@@ -1208,10 +1211,13 @@ static void print_percore_thread(struct perf_stat_config *config, {int s;struct aggr_cpu_id s2, id; +struct perf_cpu_map *cpus;bool first = true; +int idx, cpu; -for (int i =0 ; i <evsel__nr_cpus (counter); i++) { - s2 = config->aggr_get_id (config,evsel__cpus (counter), i); + cpus =evsel__cpus (counter); +perf_cpu_map__for_each_cpu (cpu, idx, cpus) { + s2 = config->aggr_get_id (config, cpus, idx);for (s =0 ; s < config->aggr_map->nr; s++) { id = config->aggr_map->map[s];if (cpu_map__compare_aggr_cpu_id (s2, id))
@@ -1220,7 +1226,7 @@ static void print_percore_thread(struct perf_stat_config *config,print_counter_aggrdata (config, counter, s, prefix, false, - &first, i); + &first, cpu); } }
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 5/48 perf stat: Switch to cpu version of cpu_map__get
In reply to Ian RogersAvoid bugs where the wrong index is passed with the cpu_map.
--- tools/perf/builtin-stat.c | 93 +++++++++++++++++++--------------- tools/perf/util/stat-display.c | 11 ++-- tools/perf/util/stat.h | 3 +- 3 files changed, 57 insertions(+), 50 deletions(-)
tools/perf/builtin-stat.c +52 -41
@@ -1323,69 +1323,63 @@ static struct option stat_options[] = { };static struct aggr_cpu_id perf_stat__get_socket (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int cpu) +int cpu) { -return cpu_map__get_socket (map, cpu,NULL ); +return cpu_map__get_socket_aggr_by_cpu (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_die (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int cpu) +int cpu) { -return cpu_map__get_die (map, cpu,NULL ); +return cpu_map__get_die_aggr_by_cpu (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_core (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int cpu) +int cpu) { -return cpu_map__get_core (map, cpu,NULL ); +return cpu_map__get_core_aggr_by_cpu (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_node (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int cpu) +int cpu) { -return cpu_map__get_node (map, cpu,NULL ); +return cpu_map__get_node_aggr_by_cpu (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_aggr (struct perf_stat_config *config, -aggr_get_id_t get_id,struct perf_cpu_map *map,int idx) +aggr_get_id_t get_id,int cpu) { -int cpu;struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); -if (idx >= map->nr) -return id; - - cpu = map->map[idx]; -if (cpu_map__aggr_cpu_id_is_empty (config->cpus_aggr_map->map[cpu])) - config->cpus_aggr_map->map[cpu] =get_id (config, map, idx); + config->cpus_aggr_map->map[cpu] =get_id (config, cpu); id = config->cpus_aggr_map->map[cpu];return id; }static struct aggr_cpu_id perf_stat__get_socket_cached (struct perf_stat_config *config, -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_stat__get_aggr (config, perf_stat__get_socket, map, idx); +return perf_stat__get_aggr (config, perf_stat__get_socket, cpu); }static struct aggr_cpu_id perf_stat__get_die_cached (struct perf_stat_config *config, -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_stat__get_aggr (config, perf_stat__get_die, map, idx); +return perf_stat__get_aggr (config, perf_stat__get_die, cpu); }static struct aggr_cpu_id perf_stat__get_core_cached (struct perf_stat_config *config, -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_stat__get_aggr (config, perf_stat__get_core, map, idx); +return perf_stat__get_aggr (config, perf_stat__get_core, cpu); }static struct aggr_cpu_id perf_stat__get_node_cached (struct perf_stat_config *config, -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_stat__get_aggr (config, perf_stat__get_node, map, idx); +return perf_stat__get_aggr (config, perf_stat__get_node, cpu); }static bool term_percore_set (void )
@@ -1483,8 +1477,9 @@ static void perf_stat__exit_aggr_mode(void) stat_config.cpus_aggr_map =NULL ; } -static inline int perf_env__get_cpu (struct perf_env *env,struct perf_cpu_map *map,int idx) +static inline int perf_env__get_cpu (void *data,struct perf_cpu_map *map,int idx) { +struct perf_env *env = data;int cpu;if (idx > map->nr)
@@ -1498,10 +1493,9 @@ static inline int perf_env__get_cpu(struct perf_env *env, struct perf_cpu_map *mreturn cpu; } -static struct aggr_cpu_id perf_env__get_socket (struct perf_cpu_map *map,int idx,void *data) +static struct aggr_cpu_id perf_env__get_socket_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data; -int cpu =perf_env__get_cpu (env, map, idx);struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id ();if (cpu !=-1 )
@@ -1510,11 +1504,15 @@ static struct aggr_cpu_id perf_env__get_socket(struct perf_cpu_map *map, int idxreturn id; } -static struct aggr_cpu_id perf_env__get_die (struct perf_cpu_map *map,int idx,void *data) +static struct aggr_cpu_id perf_env__get_socket (struct perf_cpu_map *map,int idx,void *data) +{ +return perf_env__get_socket_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); +} + +static struct aggr_cpu_id perf_env__get_die_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data;struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); -int cpu =perf_env__get_cpu (env, map, idx);if (cpu !=-1 ) {/*
@@ -1529,11 +1527,15 @@ static struct aggr_cpu_id perf_env__get_die(struct perf_cpu_map *map, int idx, vreturn id; } -static struct aggr_cpu_id perf_env__get_core (struct perf_cpu_map *map,int idx,void *data) +static struct aggr_cpu_id perf_env__get_die (struct perf_cpu_map *map,int idx,void *data) +{ +return perf_env__get_die_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); +} + +static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data;struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); -int cpu =perf_env__get_cpu (env, map, idx);if (cpu !=-1 ) {/*
@@ -1549,15 +1551,24 @@ static struct aggr_cpu_id perf_env__get_core(struct perf_cpu_map *map, int idx,return id; } -static struct aggr_cpu_id perf_env__get_node (struct perf_cpu_map *map,int idx,void *data) +static struct aggr_cpu_id perf_env__get_core (struct perf_cpu_map *map,int idx,void *data) +{ +return perf_env__get_core_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); +} + +static struct aggr_cpu_id perf_env__get_node_aggr_by_cpu (int cpu,void *data) { -int cpu =perf_env__get_cpu (data, map, idx);struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); id.node =perf_env__numa_node (data, cpu);return id; } +static struct aggr_cpu_id perf_env__get_node (struct perf_cpu_map *map,int idx,void *data) +{ +return perf_env__get_node_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); +} +static int perf_env__build_socket_map (struct perf_env *env,struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp) {
@@ -1583,26 +1594,26 @@ static int perf_env__build_node_map(struct perf_env *env, struct perf_cpu_map *c }static struct aggr_cpu_id perf_stat__get_socket_file (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_env__get_socket (map, idx, &perf_stat.session->header.env); +return perf_env__get_socket_aggr_by_cpu (cpu, &perf_stat.session->header.env); }static struct aggr_cpu_id perf_stat__get_die_file (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_env__get_die (map, idx, &perf_stat.session->header.env); +return perf_env__get_die_aggr_by_cpu (cpu, &perf_stat.session->header.env); }static struct aggr_cpu_id perf_stat__get_core_file (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_env__get_core (map, idx, &perf_stat.session->header.env); +return perf_env__get_core_aggr_by_cpu (cpu, &perf_stat.session->header.env); }static struct aggr_cpu_id perf_stat__get_node_file (struct perf_stat_config *config__maybe_unused , -struct perf_cpu_map *map,int idx) +int cpu) { -return perf_env__get_node (map, idx, &perf_stat.session->header.env); +return perf_env__get_node_aggr_by_cpu (cpu, &perf_stat.session->header.env); }static int perf_stat_init_aggr_mode_file (struct perf_stat *st)
tools/perf/util/stat-display.c +4 -7
@@ -341,8 +341,7 @@ static int first_shadow_cpu(struct perf_stat_config *config, cpus =evsel__cpus (evsel);perf_cpu_map__for_each_cpu (cpu, idx, cpus) { -if (cpu_map__compare_aggr_cpu_id (config->aggr_get_id (config, cpus, idx), - id)) +if (cpu_map__compare_aggr_cpu_id (config->aggr_get_id (config, cpu), id))return cpu; }return 0 ;
@@ -525,8 +524,7 @@ static void aggr_update_shadow(struct perf_stat_config *config, cpus =evsel__cpus (counter); val =0 ;perf_cpu_map__for_each_cpu (cpu, idx, cpus) { - (void )cpu; - s2 = config->aggr_get_id (config, cpus, idx); + s2 = config->aggr_get_id (config, cpu);if (!cpu_map__compare_aggr_cpu_id (s2, id))continue ; val +=perf_counts (counter->counts, idx,0 )->val;
@@ -642,8 +640,7 @@ static void aggr_cb(struct perf_stat_config *config,perf_cpu_map__for_each_cpu (cpu, idx, cpus) {struct perf_counts_values *counts; - (void )cpu; - s2 = config->aggr_get_id (config, cpus, idx); + s2 = config->aggr_get_id (config, cpu);if (!cpu_map__compare_aggr_cpu_id (s2, ad->id))continue ;if (first)
@@ -1217,7 +1214,7 @@ static void print_percore_thread(struct perf_stat_config *config, cpus =evsel__cpus (counter);perf_cpu_map__for_each_cpu (cpu, idx, cpus) { - s2 = config->aggr_get_id (config, cpus, idx); + s2 = config->aggr_get_id (config, cpu);for (s =0 ; s < config->aggr_map->nr; s++) { id = config->aggr_map->map[s];if (cpu_map__compare_aggr_cpu_id (s2, id))
tools/perf/util/stat.h +1 -2
@@ -108,8 +108,7 @@ struct runtime_stat {struct rblist value_list; }; -typedef struct aggr_cpu_id (*aggr_get_id_t )(struct perf_stat_config *config, -struct perf_cpu_map *m,int cpu); +typedef struct aggr_cpu_id (*aggr_get_id_t )(struct perf_stat_config *config,int cpu);struct perf_stat_config {enum aggr_mode aggr_mode;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 6/48 perf cpumap: Switch cpu_map__build_map to cpu function
In reply to Ian RogersAvoid error prone cpu_map + idx variant. Remove now unused functions.
--- tools/perf/builtin-stat.c | 28 ++++------------------------ tools/perf/util/cpumap.c | 12 ++++++------ tools/perf/util/cpumap.h | 2 +- 3 files changed, 11 insertions(+), 31 deletions(-)
tools/perf/builtin-stat.c +4 -24
@@ -1504,11 +1504,6 @@ static struct aggr_cpu_id perf_env__get_socket_aggr_by_cpu(int cpu, void *data)return id; } -static struct aggr_cpu_id perf_env__get_socket (struct perf_cpu_map *map,int idx,void *data) -{ -return perf_env__get_socket_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); -} -static struct aggr_cpu_id perf_env__get_die_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data;
@@ -1527,11 +1522,6 @@ static struct aggr_cpu_id perf_env__get_die_aggr_by_cpu(int cpu, void *data)return id; } -static struct aggr_cpu_id perf_env__get_die (struct perf_cpu_map *map,int idx,void *data) -{ -return perf_env__get_die_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); -} -static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data;
@@ -1551,11 +1541,6 @@ static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu(int cpu, void *data)return id; } -static struct aggr_cpu_id perf_env__get_core (struct perf_cpu_map *map,int idx,void *data) -{ -return perf_env__get_core_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); -} -static struct aggr_cpu_id perf_env__get_node_aggr_by_cpu (int cpu,void *data) {struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id ();
@@ -1564,33 +1549,28 @@ static struct aggr_cpu_id perf_env__get_node_aggr_by_cpu(int cpu, void *data)return id; } -static struct aggr_cpu_id perf_env__get_node (struct perf_cpu_map *map,int idx,void *data) -{ -return perf_env__get_node_aggr_by_cpu (perf_env__get_cpu (data, map, idx), data); -} -static int perf_env__build_socket_map (struct perf_env *env,struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp) { -return cpu_map__build_map (cpus, sockp, perf_env__get_socket, env); +return cpu_map__build_map (cpus, sockp, perf_env__get_socket_aggr_by_cpu, env); }static int perf_env__build_die_map (struct perf_env *env,struct perf_cpu_map *cpus,struct cpu_aggr_map **diep) { -return cpu_map__build_map (cpus, diep, perf_env__get_die, env); +return cpu_map__build_map (cpus, diep, perf_env__get_die_aggr_by_cpu, env); }static int perf_env__build_core_map (struct perf_env *env,struct perf_cpu_map *cpus,struct cpu_aggr_map **corep) { -return cpu_map__build_map (cpus, corep, perf_env__get_core, env); +return cpu_map__build_map (cpus, corep, perf_env__get_core_aggr_by_cpu, env); }static int perf_env__build_node_map (struct perf_env *env,struct perf_cpu_map *cpus,struct cpu_aggr_map **nodep) { -return cpu_map__build_map (cpus, nodep, perf_env__get_node, env); +return cpu_map__build_map (cpus, nodep, perf_env__get_node_aggr_by_cpu, env); } static struct aggr_cpu_id perf_stat__get_socket_file(struct perf_stat_config *config __maybe_unused,
tools/perf/util/cpumap.c +6 -6
@@ -163,7 +163,7 @@ static int cmp_aggr_cpu_id(const void *a_pointer, const void *b_pointer) }int cpu_map__build_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **res, -struct aggr_cpu_id (*f)(struct perf_cpu_map *map,int cpu,void *data), +struct aggr_cpu_id (*f)(int cpu,void *data),void *data) {int nr = cpus->nr;
@@ -178,7 +178,7 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res, c->nr =0 ;for (cpu =0 ; cpu < nr; cpu++) { - s1 =f (cpus, cpu, data); + s1 =f (cpu, data);for (s2 =0 ; s2 < c->nr; s2++) {if (cpu_map__compare_aggr_cpu_id (s1, c->map[s2]))break ;
@@ -290,22 +290,22 @@ struct aggr_cpu_id cpu_map__get_node(struct perf_cpu_map *map, int idx, void *daint cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp) { -return cpu_map__build_map (cpus, sockp, cpu_map__get_socket,NULL ); +return cpu_map__build_map (cpus, sockp, cpu_map__get_socket_aggr_by_cpu,NULL ); }int cpu_map__build_die_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **diep) { -return cpu_map__build_map (cpus, diep, cpu_map__get_die,NULL ); +return cpu_map__build_map (cpus, diep, cpu_map__get_die_aggr_by_cpu,NULL ); }int cpu_map__build_core_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **corep) { -return cpu_map__build_map (cpus, corep, cpu_map__get_core,NULL ); +return cpu_map__build_map (cpus, corep, cpu_map__get_core_aggr_by_cpu,NULL ); }int cpu_map__build_node_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **numap) { -return cpu_map__build_map (cpus, numap, cpu_map__get_node,NULL ); +return cpu_map__build_map (cpus, numap, cpu_map__get_node_aggr_by_cpu,NULL ); }/* setup simple routines to easily access node numbers given a cpu number */
tools/perf/util/cpumap.h +1 -1
@@ -63,7 +63,7 @@ int cpu__max_present_cpu(void);int cpu__get_node (int cpu);int cpu_map__build_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **res, -struct aggr_cpu_id (*f)(struct perf_cpu_map *map,int cpu,void *data), +struct aggr_cpu_id (*f)(int cpu,void *data),void *data);int cpu_map__cpu (struct perf_cpu_map *cpus,int idx);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 7/48 perf cpumap: Remove map+index get_socket
In reply to Ian RogersMigrate final users to appropriate cpu variant.
--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 9 --------- tools/perf/util/cpumap.h | 1 - tools/perf/util/stat.c | 2 +- 4 files changed, 2 insertions(+), 12 deletions(-)
tools/perf/tests/topology.c +1 -1
@@ -150,7 +150,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that socket ID contains only socket for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_socket (map, i,NULL ); + id =cpu_map__get_socket_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Socket map - Socket ID doesn't match" , session->header.env.cpu[map->map[i]].socket_id == id.socket);
tools/perf/util/cpumap.c +0 -9
@@ -136,15 +136,6 @@ struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu(int cpu, void *data __maybe_ureturn id; } -struct aggr_cpu_id cpu_map__get_socket (struct perf_cpu_map *map,int idx, -void *data) -{ -if (idx <0 || idx > map->nr) -return cpu_map__empty_aggr_cpu_id (); - -return cpu_map__get_socket_aggr_by_cpu (map->map[idx], data); -} -static int cmp_aggr_cpu_id (const void *a_pointer,const void *b_pointer) {struct aggr_cpu_id *a = (struct aggr_cpu_id *)a_pointer;
tools/perf/util/cpumap.h +0 -1
@@ -32,7 +32,6 @@ size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size);size_t cpu_map__fprintf (struct perf_cpu_map *map,FILE *fp);int cpu_map__get_socket_id (int cpu);struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_socket (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_die_id (int cpu);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx,void *data);
tools/perf/util/stat.c +1 -1
@@ -328,7 +328,7 @@ static int check_per_pkg(struct evsel *counter,if (!(vals->run && vals->ena))return 0 ; - s =cpu_map__get_socket (cpus, cpu,NULL ).socket; + s =cpu_map__get_socket_id (cpu);if (s <0 )return -1 ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 8/48 perf cpumap: Remove map+index get_die
In reply to Ian RogersMigrate final users to appropriate cpu variant.
--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 9 --------- tools/perf/util/cpumap.h | 1 - tools/perf/util/stat.c | 2 +- 4 files changed, 2 insertions(+), 12 deletions(-)
tools/perf/tests/topology.c +1 -1
@@ -136,7 +136,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that die ID contains socket and die for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_die (map, i,NULL ); + id =cpu_map__get_die_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Die map - Socket ID doesn't match" , session->header.env.cpu[map->map[i]].socket_id == id.socket);
tools/perf/util/cpumap.c +0 -9
@@ -216,15 +216,6 @@ struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu(int cpu, void *data)return id; } -struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx, -void *data) -{ -if (idx <0 || idx > map->nr) -return cpu_map__empty_aggr_cpu_id (); - -return cpu_map__get_die_aggr_by_cpu (map->map[idx], data); -} -int cpu_map__get_core_id (int cpu) {int value, ret =cpu__get_topology_int (cpu,"core_id" , &value);
tools/perf/util/cpumap.h +0 -1
@@ -34,7 +34,6 @@ int cpu_map__get_socket_id(int cpu);struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data);int cpu_map__get_die_id (int cpu);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_die (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_core_id (int cpu);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data);
tools/perf/util/stat.c +1 -1
@@ -336,7 +336,7 @@ static int check_per_pkg(struct evsel *counter,* On multi-die system, die_id > 0. On no-die system, die_id = 0. * We use hashmap(socket, die) to check the used socket+die pair. */ - d =cpu_map__get_die (cpus, cpu,NULL ).die; + d =cpu_map__get_die_id (cpu);if (d <0 )return -1 ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 9/48 perf cpumap: Remove map+index get_core
In reply to Ian RogersMigrate final users to appropriate cpu variant.
--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 8 -------- tools/perf/util/cpumap.h | 1 - 3 files changed, 1 insertion(+), 10 deletions(-)
tools/perf/tests/topology.c +1 -1
@@ -121,7 +121,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that core ID contains socket, die and core for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_core (map, i,NULL ); + id =cpu_map__get_core_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Core map - Core ID doesn't match" , session->header.env.cpu[map->map[i]].core_id == id.core);
tools/perf/util/cpumap.c +0 -8
@@ -246,14 +246,6 @@ struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu(int cpu, void *data) } -struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data) -{ -if (idx <0 || idx > map->nr) -return cpu_map__empty_aggr_cpu_id (); - -return cpu_map__get_core_aggr_by_cpu (map->map[idx], data); -} -struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id ();
tools/perf/util/cpumap.h +0 -1
@@ -36,7 +36,6 @@ int cpu_map__get_die_id(int cpu);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data);int cpu_map__get_core_id (int cpu);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_core (struct perf_cpu_map *map,int idx,void *data);int cpu_map__get_node_id (int cpu);struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 10/48 perf cpumap: Remove map+index get_node
In reply to Ian RogersMigrate final users to appropriate cpu variant.
--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 8 -------- tools/perf/util/cpumap.h | 1 - 3 files changed, 1 insertion(+), 10 deletions(-)
tools/perf/tests/topology.c +1 -1
@@ -162,7 +162,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that node ID contains only node for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_node (map, i,NULL ); + id =cpu_map__get_node_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Node map - Node ID doesn't match" ,cpu__get_node (map->map[i]) == id.node);TEST_ASSERT_VAL ("Node map - Socket is set" , id.socket ==-1 );
tools/perf/util/cpumap.c +0 -8
@@ -254,14 +254,6 @@ struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu(int cpu, void *data __maybe_unureturn id; } -struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data) -{ -if (idx <0 || idx >= map->nr) -return cpu_map__empty_aggr_cpu_id (); - -return cpu_map__get_node_aggr_by_cpu (map->map[idx], data); -} -int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp) {return cpu_map__build_map (cpus, sockp, cpu_map__get_socket_aggr_by_cpu,NULL );
tools/perf/util/cpumap.h +0 -1
@@ -38,7 +38,6 @@ int cpu_map__get_core_id(int cpu);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data);int cpu_map__get_node_id (int cpu);struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_node (struct perf_cpu_map *map,int idx,void *data);int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp);int cpu_map__build_die_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **diep);int cpu_map__build_core_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **corep);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 11/48 perf cpumap: Add comments to aggr_cpu_id
In reply to Ian RogersThis code is already tested in topology.c.
--- tools/perf/util/cpumap.h | 9 +++++++++ 1 file changed, 9 insertions(+)
tools/perf/util/cpumap.h +9 -0
@@ -7,11 +7,20 @@#include <internal/cpumap.h> #include <perf/cpumap.h> +/** Identify where counts are aggregated, -1 implies not to aggregate. */ struct aggr_cpu_id { +/** A value in the range 0 to number of threads. */ int thread; +/** The numa node X as read from /sys/devices/system/node/nodeX. */ int node; +/** +* The socket number as read from +* /sys/devices/system/cpu/cpuX/topology/physical_package_id. +*/ int socket; +/** The die id as read from /sys/devices/system/cpu/cpuX/topology/die_id. */ int die; +/** The core id as read from /sys/devices/system/cpu/cpuX/topology/core_id. */ int core; };
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 12/48 perf cpumap: Remove unused cpu_map__socket
In reply to Ian RogersUnused function so remove.
--- tools/perf/util/cpumap.h | 7 ------- 1 file changed, 7 deletions(-)
tools/perf/util/cpumap.h +0 -7
@@ -53,13 +53,6 @@ int cpu_map__build_core_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **corint cpu_map__build_node_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **nodep);const struct perf_cpu_map *cpu_map__online (void );/* thread unsafe */ -static inline int cpu_map__socket (struct perf_cpu_map *sock,int s) -{ -if (!sock || s > sock->nr || s <0 ) -return 0 ; -return sock->map[s]; -} -int cpu__setup_cpunode_map (void );int cpu__max_node (void );
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 13/48 perf cpumap: Simplify equal function name.
In reply to Ian RogersRename cpu_map__compare_aggr_cpu_id to aggr_cpu_id__equal, the cpu_map part of the name is misleading. Equal better describes the function than compare.
Switch to const pointer rather than value as struct given the number of variables in aggr_cpu_id.
--- tools/perf/util/cpumap.c | 14 +++++++------- tools/perf/util/cpumap.h | 2 +- tools/perf/util/stat-display.c | 18 ++++++++++-------- 3 files changed, 18 insertions(+), 16 deletions(-)
tools/perf/util/cpumap.c +7 -7
@@ -171,7 +171,7 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,for (cpu =0 ; cpu < nr; cpu++) { s1 =f (cpu, data);for (s2 =0 ; s2 < c->nr; s2++) { -if (cpu_map__compare_aggr_cpu_id (s1, c->map[s2])) +if (aggr_cpu_id__equal (&s1, &c->map[s2]))break ; }if (s2 == c->nr) {
@@ -593,13 +593,13 @@ const struct perf_cpu_map *cpu_map__online(void) /* thread unsafe */return online; } -bool cpu_map__compare_aggr_cpu_id (struct aggr_cpu_id a,struct aggr_cpu_id b) +bool aggr_cpu_id__equal (const struct aggr_cpu_id *a,const struct aggr_cpu_id *b) { -return a.thread == b.thread && - a.node == b.node && - a.socket == b.socket && - a.die == b.die && - a.core == b.core; +return a->thread == b->thread && + a->node == b->node && + a->socket == b->socket && + a->die == b->die && + a->core == b->core; }bool cpu_map__aggr_cpu_id_is_empty (struct aggr_cpu_id a)
tools/perf/util/cpumap.h +1 -1
@@ -67,7 +67,7 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,int cpu_map__cpu (struct perf_cpu_map *cpus,int idx);bool cpu_map__has (struct perf_cpu_map *cpus,int cpu); -bool cpu_map__compare_aggr_cpu_id (struct aggr_cpu_id a,struct aggr_cpu_id b); +bool aggr_cpu_id__equal (const struct aggr_cpu_id *a,const struct aggr_cpu_id *b);bool cpu_map__aggr_cpu_id_is_empty (struct aggr_cpu_id a);struct aggr_cpu_id cpu_map__empty_aggr_cpu_id (void );
tools/perf/util/stat-display.c +10 -8
@@ -328,20 +328,22 @@ static void print_metric_header(struct perf_stat_config *config, }static int first_shadow_cpu (struct perf_stat_config *config, -struct evsel *evsel,struct aggr_cpu_id id) +struct evsel *evsel,const struct aggr_cpu_id *id) {struct perf_cpu_map *cpus;int cpu, idx;if (config->aggr_mode ==AGGR_NONE ) -return id.core; +return id->core;if (!config->aggr_get_id)return 0 ; cpus =evsel__cpus (evsel);perf_cpu_map__for_each_cpu (cpu, idx, cpus) { -if (cpu_map__compare_aggr_cpu_id (config->aggr_get_id (config, cpu), id)) +struct aggr_cpu_id cpu_id = config->aggr_get_id (config, cpu); + +if (aggr_cpu_id__equal (&cpu_id, id))return cpu; }return 0 ;
@@ -501,7 +503,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int }perf_stat__print_shadow_stats (config, counter, uval, -first_shadow_cpu (config, counter, id), +first_shadow_cpu (config, counter, &id), &out, &config->metric_events, st);if (!config->csv_output && !config->metric_only) {print_noise (config, counter, noise);
@@ -525,12 +527,12 @@ static void aggr_update_shadow(struct perf_stat_config *config, val =0 ;perf_cpu_map__for_each_cpu (cpu, idx, cpus) { s2 = config->aggr_get_id (config, cpu); -if (!cpu_map__compare_aggr_cpu_id (s2, id)) +if (!aggr_cpu_id__equal (&s2, &id))continue ; val +=perf_counts (counter->counts, idx,0 )->val; }perf_stat__update_shadow_stats (counter, val, -first_shadow_cpu (config, counter, id), +first_shadow_cpu (config, counter, &id), &rt_stat); } }
@@ -641,7 +643,7 @@ static void aggr_cb(struct perf_stat_config *config,struct perf_counts_values *counts; s2 = config->aggr_get_id (config, cpu); -if (!cpu_map__compare_aggr_cpu_id (s2, ad->id)) +if (!aggr_cpu_id__equal (&s2, &ad->id))continue ;if (first) ad->nr++;
@@ -1217,7 +1219,7 @@ static void print_percore_thread(struct perf_stat_config *config, s2 = config->aggr_get_id (config, cpu);for (s =0 ; s < config->aggr_map->nr; s++) { id = config->aggr_map->map[s]; -if (cpu_map__compare_aggr_cpu_id (s2, id)) +if (aggr_cpu_id__equal (&s2, &id))break ; }
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 14/48 perf cpumap: Rename empty functions.
In reply to Ian RogersRemove cpu_map from name as a cpu_map isn't used. Pass a const pointer rather than by value to avoid unnecessary copying.
--- tools/perf/builtin-stat.c | 12 ++++++------ tools/perf/util/cpumap.c | 24 ++++++++++++------------ tools/perf/util/cpumap.h | 4 ++-- tools/perf/util/stat-display.c | 10 +++++----- 4 files changed, 25 insertions(+), 25 deletions(-)
tools/perf/builtin-stat.c +6 -6
@@ -1349,9 +1349,9 @@ static struct aggr_cpu_id perf_stat__get_node(struct perf_stat_config *config __static struct aggr_cpu_id perf_stat__get_aggr (struct perf_stat_config *config,aggr_get_id_t get_id,int cpu) { -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty (); -if (cpu_map__aggr_cpu_id_is_empty (config->cpus_aggr_map->map[cpu])) +if (aggr_cpu_id__is_empty (&config->cpus_aggr_map->map[cpu])) config->cpus_aggr_map->map[cpu] =get_id (config, cpu); id = config->cpus_aggr_map->map[cpu];
@@ -1496,7 +1496,7 @@ static inline int perf_env__get_cpu(void *data, struct perf_cpu_map *map, int idstatic struct aggr_cpu_id perf_env__get_socket_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data; -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty ();if (cpu !=-1 ) id.socket = env->cpu[cpu].socket_id;
@@ -1507,7 +1507,7 @@ static struct aggr_cpu_id perf_env__get_socket_aggr_by_cpu(int cpu, void *data)static struct aggr_cpu_id perf_env__get_die_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data; -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty ();if (cpu !=-1 ) {/*
@@ -1525,7 +1525,7 @@ static struct aggr_cpu_id perf_env__get_die_aggr_by_cpu(int cpu, void *data)static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu (int cpu,void *data) {struct perf_env *env = data; -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty ();if (cpu !=-1 ) {/*
@@ -1543,7 +1543,7 @@ static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu(int cpu, void *data)static struct aggr_cpu_id perf_env__get_node_aggr_by_cpu (int cpu,void *data) { -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty (); id.node =perf_env__numa_node (data, cpu);return id;
tools/perf/util/cpumap.c +12 -12
@@ -104,7 +104,7 @@ struct cpu_aggr_map *cpu_aggr_map__empty_new(int nr) cpus->nr = nr;for (i =0 ; i < nr; i++) - cpus->map[i] =cpu_map__empty_aggr_cpu_id (); + cpus->map[i] =aggr_cpu_id__empty ();refcount_set (&cpus->refcnt,1 ); }
@@ -130,7 +130,7 @@ int cpu_map__get_socket_id(int cpu)struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data__maybe_unused ) { -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty (); id.socket =cpu_map__get_socket_id (cpu);return id;
@@ -209,7 +209,7 @@ struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu(int cpu, void *data)* make a unique ID. */ id =cpu_map__get_socket_aggr_by_cpu (cpu, data); -if (cpu_map__aggr_cpu_id_is_empty (id)) +if (aggr_cpu_id__is_empty (&id))return id; id.die = die;
@@ -234,7 +234,7 @@ struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu(int cpu, void *data)/* cpu_map__get_die returns a struct with socket and die set*/ id =cpu_map__get_die_aggr_by_cpu (cpu, data); -if (cpu_map__aggr_cpu_id_is_empty (id)) +if (aggr_cpu_id__is_empty (&id))return id;/*
@@ -248,7 +248,7 @@ struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu(int cpu, void *data)struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data__maybe_unused ) { -struct aggr_cpu_id id =cpu_map__empty_aggr_cpu_id (); +struct aggr_cpu_id id =aggr_cpu_id__empty (); id.node =cpu_map__get_node_id (cpu);return id;
@@ -602,16 +602,16 @@ bool aggr_cpu_id__equal(const struct aggr_cpu_id *a, const struct aggr_cpu_id *b a->core == b->core; } -bool cpu_map__aggr_cpu_id_is_empty (struct aggr_cpu_id a) +bool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a) { -return a.thread ==-1 && - a.node ==-1 && - a.socket ==-1 && - a.die ==-1 && - a.core ==-1 ; +return a->thread ==-1 && + a->node ==-1 && + a->socket ==-1 && + a->die ==-1 && + a->core ==-1 ; } -struct aggr_cpu_id cpu_map__empty_aggr_cpu_id (void ) +struct aggr_cpu_id aggr_cpu_id__empty (void ) {struct aggr_cpu_id ret = { .thread =-1 ,
tools/perf/util/cpumap.h +2 -2
@@ -68,7 +68,7 @@ int cpu_map__cpu(struct perf_cpu_map *cpus, int idx);bool cpu_map__has (struct perf_cpu_map *cpus,int cpu);bool aggr_cpu_id__equal (const struct aggr_cpu_id *a,const struct aggr_cpu_id *b); -bool cpu_map__aggr_cpu_id_is_empty (struct aggr_cpu_id a); -struct aggr_cpu_id cpu_map__empty_aggr_cpu_id (void ); +bool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a); +struct aggr_cpu_id aggr_cpu_id__empty (void );#endif /* __PERF_CPUMAP_H */
tools/perf/util/stat-display.c +5 -5
@@ -698,7 +698,7 @@ static void print_counter_aggrdata(struct perf_stat_config *config, uval = val * counter->scale;if (cpu !=-1 ) { - id =cpu_map__empty_aggr_cpu_id (); + id =aggr_cpu_id__empty (); id.core = cpu; } printout(config, id, nr, counter, uval,
@@ -780,7 +780,7 @@ static struct perf_aggr_thread_value *sort_aggr_thread(continue ; buf[i].counter = counter; - buf[i].id =cpu_map__empty_aggr_cpu_id (); + buf[i].id =aggr_cpu_id__empty (); buf[i].id.thread = thread; buf[i].uval = uval; buf[i].val = val;
@@ -868,7 +868,7 @@ static void print_counter_aggr(struct perf_stat_config *config,fprintf (output,"%s" , prefix); uval = cd.avg * counter->scale; -printout (config,cpu_map__empty_aggr_cpu_id (),0 , counter, uval, prefix, cd.avg_running, +printout (config,aggr_cpu_id__empty (),0 , counter, uval, prefix, cd.avg_running, cd.avg_enabled, cd.avg, &rt_stat);if (!metric_only)fprintf (output,"\n" );
@@ -911,7 +911,7 @@ static void print_counter(struct perf_stat_config *config,fprintf (output,"%s" , prefix); uval = val * counter->scale; - id =cpu_map__empty_aggr_cpu_id (); + id =aggr_cpu_id__empty (); id.core = cpu;printout (config, id,0 , counter, uval, prefix, run, ena,1.0 , &rt_stat);
@@ -938,7 +938,7 @@ static void print_no_aggr_metric(struct perf_stat_config *config,if (prefix)fputs (prefix, config->output);evlist__for_each_entry (evlist, counter) { - id =cpu_map__empty_aggr_cpu_id (); + id =aggr_cpu_id__empty (); id.core = cpu;if (first) {aggr_printout (config, counter, id,0 );
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 15/48 perf cpumap: Document cpu__get_node and remove redundant function
In reply to Ian Rogerscpu_map__get_node_id isn't used externally and merely delegates to cpu__get_node.
--- tools/perf/util/cpumap.c | 11 +++++------ tools/perf/util/cpumap.h | 5 ++++- 2 files changed, 9 insertions(+), 7 deletions(-)
tools/perf/util/cpumap.c +5 -6
@@ -16,6 +16,10 @@static int max_cpu_num;static int max_present_cpu_num;static int max_node_num; +/** +* The numa node X as read from /sys/devices/system/node/nodeX indexed by the +* CPU number. +*/ static int *cpunode_map;static struct perf_cpu_map *cpu_map__from_entries (struct cpu_map_entries *cpus)
@@ -222,11 +226,6 @@ int cpu_map__get_core_id(int cpu)return ret ?: value; } -int cpu_map__get_node_id (int cpu) -{ -return cpu__get_node (cpu); -} -struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data) {struct aggr_cpu_id id;
@@ -250,7 +249,7 @@ struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu(int cpu, void *data __maybe_unu {struct aggr_cpu_id id =aggr_cpu_id__empty (); - id.node =cpu_map__get_node_id (cpu); + id.node =cpu__get_node (cpu);return id; }
tools/perf/util/cpumap.h +4 -1
@@ -45,7 +45,6 @@ int cpu_map__get_die_id(int cpu);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data);int cpu_map__get_core_id (int cpu);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data); -int cpu_map__get_node_id (int cpu);struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data);int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp);int cpu_map__build_die_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **diep);
@@ -58,6 +57,10 @@ int cpu__setup_cpunode_map(void);int cpu__max_node (void );int cpu__max_cpu (void );int cpu__max_present_cpu (void ); +/** +* cpu__get_node - Returns the numa node X as read from +* /sys/devices/system/node/nodeX for the given CPU. +*/ int cpu__get_node (int cpu); int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 16/48 perf cpumap: Remove map from function names that don't use a map.
In reply to Ian RogersMove to the cpu name and document for consistency.
--- tools/perf/util/cpumap.c | 12 ++++++------ tools/perf/util/cpumap.h | 19 ++++++++++++++++--- tools/perf/util/env.c | 6 +++--- tools/perf/util/stat.c | 4 ++-- 4 files changed, 27 insertions(+), 14 deletions(-)
tools/perf/util/cpumap.c +6 -6
@@ -126,7 +126,7 @@ static int cpu__get_topology_int(int cpu, const char *name, int *value)return sysfs__read_int (path, value); } -int cpu_map__get_socket_id (int cpu) +int cpu__get_socket_id (int cpu) {int value, ret =cpu__get_topology_int (cpu,"physical_package_id" , &value);return ret ?: value;
@@ -136,7 +136,7 @@ struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu(int cpu, void *data __maybe_u {struct aggr_cpu_id id =aggr_cpu_id__empty (); - id.socket =cpu_map__get_socket_id (cpu); + id.socket =cpu__get_socket_id (cpu);return id; }
@@ -190,7 +190,7 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,return 0 ; } -int cpu_map__get_die_id (int cpu) +int cpu__get_die_id (int cpu) {int value, ret =cpu__get_topology_int (cpu,"die_id" , &value);
@@ -202,7 +202,7 @@ struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu(int cpu, void *data)struct aggr_cpu_id id;int die; - die =cpu_map__get_die_id (cpu); + die =cpu__get_die_id (cpu);/* There is no die_id on legacy system. */ if (die ==-1 ) die =0 ;
@@ -220,7 +220,7 @@ struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu(int cpu, void *data)return id; } -int cpu_map__get_core_id (int cpu) +int cpu__get_core_id (int cpu) {int value, ret =cpu__get_topology_int (cpu,"core_id" , &value);return ret ?: value;
@@ -229,7 +229,7 @@ int cpu_map__get_core_id(int cpu)struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data) {struct aggr_cpu_id id; -int core =cpu_map__get_core_id (cpu); +int core =cpu__get_core_id (cpu);/* cpu_map__get_die returns a struct with socket and die set*/ id =cpu_map__get_die_aggr_by_cpu (cpu, data);
tools/perf/util/cpumap.h +16 -3
@@ -39,11 +39,8 @@ struct perf_cpu_map *cpu_map__new_data(struct perf_record_cpu_map_data *data);size_t cpu_map__snprint (struct perf_cpu_map *map,char *buf,size_t size);size_t cpu_map__snprint_mask (struct perf_cpu_map *map,char *buf,size_t size);size_t cpu_map__fprintf (struct perf_cpu_map *map,FILE *fp); -int cpu_map__get_socket_id (int cpu);struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data); -int cpu_map__get_die_id (int cpu);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data); -int cpu_map__get_core_id (int cpu);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data);int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp);
@@ -62,6 +59,22 @@ int cpu__max_present_cpu(void);* /sys/devices/system/node/nodeX for the given CPU. */ int cpu__get_node (int cpu); +/** +* cpu__get_socket_id - Returns the socket number as read from +* /sys/devices/system/cpu/cpuX/topology/physical_package_id for the given CPU. +*/ +int cpu__get_socket_id (int cpu); +/** +* cpu__get_die_id - Returns the die id as read from +* /sys/devices/system/cpu/cpuX/topology/die_id for the given CPU. +*/ +int cpu__get_die_id (int cpu); +/** +* cpu__get_core_id - Returns the core id as read from +* /sys/devices/system/cpu/cpuX/topology/core_id for the given CPU. +*/ +int cpu__get_core_id (int cpu); + int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res, struct aggr_cpu_id (*f)(int cpu, void *data),
tools/perf/util/env.c +3 -3
@@ -302,9 +302,9 @@ int perf_env__read_cpu_topology_map(struct perf_env *env)return -ENOMEM ;for (cpu =0 ; cpu < nr_cpus; ++cpu) { - env->cpu[cpu].core_id =cpu_map__get_core_id (cpu); - env->cpu[cpu].socket_id =cpu_map__get_socket_id (cpu); - env->cpu[cpu].die_id =cpu_map__get_die_id (cpu); + env->cpu[cpu].core_id =cpu__get_core_id (cpu); + env->cpu[cpu].socket_id =cpu__get_socket_id (cpu); + env->cpu[cpu].die_id =cpu__get_die_id (cpu); } env->nr_cpus_avail = nr_cpus;
tools/perf/util/stat.c +2 -2
@@ -328,7 +328,7 @@ static int check_per_pkg(struct evsel *counter,if (!(vals->run && vals->ena))return 0 ; - s =cpu_map__get_socket_id (cpu); + s =cpu__get_socket_id (cpu);if (s <0 )return -1 ;
@@ -336,7 +336,7 @@ static int check_per_pkg(struct evsel *counter,* On multi-die system, die_id > 0. On no-die system, die_id = 0. * We use hashmap(socket, die) to check the used socket+die pair. */ - d =cpu_map__get_die_id (cpu); + d =cpu__get_die_id (cpu);if (d <0 )return -1 ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 17/48 perf cpumap: Remove cpu_map__cpu, use libperf function.
In reply to Ian RogersSwitch the remaining few users of cpu_map__cpu to perf_cpu_map__cpu and remove the function.
--- tools/perf/builtin-ftrace.c | 2 +- tools/perf/util/cpumap.c | 9 ++------- tools/perf/util/cpumap.h | 1 - 3 files changed, 3 insertions(+), 9 deletions(-)
tools/perf/builtin-ftrace.c +1 -1
@@ -303,7 +303,7 @@ static int set_tracing_cpumask(struct perf_cpu_map *cpumap)int ret;int last_cpu; - last_cpu =cpu_map__cpu (cpumap, cpumap->nr -1 ); + last_cpu =perf_cpu_map__cpu (cpumap, cpumap->nr -1 ); mask_size = last_cpu /4 +2 ;/* one more byte for EOS */ mask_size += last_cpu /32 ;/* ',' is needed for every 32th cpus */
tools/perf/util/cpumap.c +2 -7
@@ -485,11 +485,6 @@ bool cpu_map__has(struct perf_cpu_map *cpus, int cpu)return perf_cpu_map__idx (cpus, cpu) !=-1 ; } -int cpu_map__cpu (struct perf_cpu_map *cpus,int idx) -{ -return cpus->map[idx]; -} -size_t cpu_map__snprint (struct perf_cpu_map *map,char *buf,size_t size) {int i, cpu, start =-1 ;
@@ -547,7 +542,7 @@ size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size)int i, cpu;char *ptr = buf;unsigned char *bitmap; -int last_cpu =cpu_map__cpu (map, map->nr -1 ); +int last_cpu =perf_cpu_map__cpu (map, map->nr -1 );if (buf ==NULL )return 0 ;
@@ -559,7 +554,7 @@ size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size) }for (i =0 ; i < map->nr; i++) { - cpu =cpu_map__cpu (map, i); + cpu =perf_cpu_map__cpu (map, i); bitmap[cpu /8 ] |=1 << (cpu %8 ); }
tools/perf/util/cpumap.h +0 -1
@@ -80,7 +80,6 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,struct aggr_cpu_id (*f)(int cpu,void *data), void *data); -int cpu_map__cpu (struct perf_cpu_map *cpus,int idx);bool cpu_map__has (struct perf_cpu_map *cpus,int cpu);bool aggr_cpu_id__equal (const struct aggr_cpu_id *a,const struct aggr_cpu_id *b);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 18/48 perf cpumap: Refactor cpu_map__build_map
In reply to Ian RogersTurn it into a cpu_aggr_map__new. Pass helper functions. Refactor builtin-stat calls to manually pass function pointers. Try to reduce some copy-paste code.
--- tools/perf/builtin-stat.c | 187 ++++++++++++++++++++------------------ tools/perf/util/cpumap.c | 59 +++++------- tools/perf/util/cpumap.h | 16 ++-- 3 files changed, 130 insertions(+), 132 deletions(-)
tools/perf/builtin-stat.c +100 -87
@@ -1322,6 +1322,17 @@ static struct option stat_options[] = { OPT_END() }; +static const char *const aggr_mode__string[] = { + [AGGR_CORE ] ="core" , + [AGGR_DIE ] ="die" , + [AGGR_GLOBAL ] ="global" , + [AGGR_NODE ] ="node" , + [AGGR_NONE ] ="none" , + [AGGR_SOCKET ] ="socket" , + [AGGR_THREAD ] ="thread" , + [AGGR_UNSET ] ="unset" , +}; +static struct aggr_cpu_id perf_stat__get_socket (struct perf_stat_config *config__maybe_unused ,int cpu) {
@@ -1394,54 +1405,67 @@ static bool term_percore_set(void)return false; } -static int perf_stat_init_aggr_mode (void ) +static aggr_cpu_id_get_t aggr_mode__get_aggr (enum aggr_mode aggr_mode) { -int nr; +switch (aggr_mode) { +case AGGR_SOCKET : +return cpu_map__get_socket_aggr_by_cpu; +case AGGR_DIE : +return cpu_map__get_die_aggr_by_cpu; +case AGGR_CORE : +return cpu_map__get_core_aggr_by_cpu; +case AGGR_NODE : +return cpu_map__get_node_aggr_by_cpu; +case AGGR_NONE : +if (term_percore_set ()) +return cpu_map__get_core_aggr_by_cpu; + +return NULL ; +case AGGR_GLOBAL : +case AGGR_THREAD : +case AGGR_UNSET : +default : +return NULL ; + } +} -switch (stat_config.aggr_mode) { +static aggr_get_id_t aggr_mode__get_id (enum aggr_mode aggr_mode) +{ +switch (aggr_mode) {case AGGR_SOCKET : -if (cpu_map__build_socket_map (evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build socket map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_socket_cached; -break ; +return perf_stat__get_socket_cached;case AGGR_DIE : -if (cpu_map__build_die_map (evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build die map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_die_cached; -break ; +return perf_stat__get_die_cached;case AGGR_CORE : -if (cpu_map__build_core_map (evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build core map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_core_cached; -break ; +return perf_stat__get_core_cached;case AGGR_NODE : -if (cpu_map__build_node_map (evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build core map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_node_cached; -break ; +return perf_stat__get_node_cached;case AGGR_NONE :if (term_percore_set ()) { -if (cpu_map__build_core_map (evsel_list->core.cpus, - &stat_config.aggr_map)) { -perror ("cannot build core map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_core_cached; +return perf_stat__get_core_cached; } -break ; +return NULL ;case AGGR_GLOBAL :case AGGR_THREAD :case AGGR_UNSET :default : -break ; +return NULL ; + } +} + +static int perf_stat_init_aggr_mode (void ) +{ +int nr; +aggr_cpu_id_get_t get_id =aggr_mode__get_aggr (stat_config.aggr_mode); + +if (get_id) { + stat_config.aggr_map =cpu_aggr_map__new (evsel_list->core.cpus, + get_id,/*data=*/ NULL ); +if (!stat_config.aggr_map) { +pr_err ("cannot build %s map" , aggr_mode__string[stat_config.aggr_mode]); +return -1 ; + } + stat_config.aggr_get_id =aggr_mode__get_id (stat_config.aggr_mode); }/*
@@ -1549,30 +1573,6 @@ static struct aggr_cpu_id perf_env__get_node_aggr_by_cpu(int cpu, void *data)return id; } -static int perf_env__build_socket_map (struct perf_env *env,struct perf_cpu_map *cpus, -struct cpu_aggr_map **sockp) -{ -return cpu_map__build_map (cpus, sockp, perf_env__get_socket_aggr_by_cpu, env); -} - -static int perf_env__build_die_map (struct perf_env *env,struct perf_cpu_map *cpus, -struct cpu_aggr_map **diep) -{ -return cpu_map__build_map (cpus, diep, perf_env__get_die_aggr_by_cpu, env); -} - -static int perf_env__build_core_map (struct perf_env *env,struct perf_cpu_map *cpus, -struct cpu_aggr_map **corep) -{ -return cpu_map__build_map (cpus, corep, perf_env__get_core_aggr_by_cpu, env); -} - -static int perf_env__build_node_map (struct perf_env *env,struct perf_cpu_map *cpus, -struct cpu_aggr_map **nodep) -{ -return cpu_map__build_map (cpus, nodep, perf_env__get_node_aggr_by_cpu, env); -} -static struct aggr_cpu_id perf_stat__get_socket_file (struct perf_stat_config *config__maybe_unused ,int cpu) {
@@ -1596,47 +1596,60 @@ static struct aggr_cpu_id perf_stat__get_node_file(struct perf_stat_config *confreturn perf_env__get_node_aggr_by_cpu (cpu, &perf_stat.session->header.env); } -static int perf_stat_init_aggr_mode_file (struct perf_stat *st) +static aggr_cpu_id_get_t aggr_mode__get_aggr_file (enum aggr_mode aggr_mode) { -struct perf_env *env = &st->session->header.env; +switch (aggr_mode) { +case AGGR_SOCKET : +return perf_env__get_socket_aggr_by_cpu; +case AGGR_DIE : +return perf_env__get_die_aggr_by_cpu; +case AGGR_CORE : +return perf_env__get_core_aggr_by_cpu; +case AGGR_NODE : +return perf_env__get_node_aggr_by_cpu; +case AGGR_NONE : +case AGGR_GLOBAL : +case AGGR_THREAD : +case AGGR_UNSET : +default : +return NULL ; + } +} -switch (stat_config.aggr_mode) { +static aggr_get_id_t aggr_mode__get_id_file (enum aggr_mode aggr_mode) +{ +switch (aggr_mode) {case AGGR_SOCKET : -if (perf_env__build_socket_map (env, evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build socket map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_socket_file; -break ; +return perf_stat__get_socket_file;case AGGR_DIE : -if (perf_env__build_die_map (env, evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build die map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_die_file; -break ; +return perf_stat__get_die_file;case AGGR_CORE : -if (perf_env__build_core_map (env, evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build core map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_core_file; -break ; +return perf_stat__get_core_file;case AGGR_NODE : -if (perf_env__build_node_map (env, evsel_list->core.cpus, &stat_config.aggr_map)) { -perror ("cannot build core map" ); -return -1 ; - } - stat_config.aggr_get_id = perf_stat__get_node_file; -break ; +return perf_stat__get_node_file;case AGGR_NONE :case AGGR_GLOBAL :case AGGR_THREAD :case AGGR_UNSET :default : -break ; +return NULL ; } +} + +static int perf_stat_init_aggr_mode_file (struct perf_stat *st) +{ +struct perf_env *env = &st->session->header.env; +aggr_cpu_id_get_t get_id =aggr_mode__get_aggr_file (stat_config.aggr_mode); +if (!get_id) +return 0 ; + + stat_config.aggr_map =cpu_aggr_map__new (evsel_list->core.cpus, get_id, env); +if (!stat_config.aggr_map) { +pr_err ("cannot build %s map" , aggr_mode__string[stat_config.aggr_mode]); +return -1 ; + } + stat_config.aggr_get_id =aggr_mode__get_id_file (stat_config.aggr_mode);return 0 ; }
tools/perf/util/cpumap.c +21 -38
91 lines not on this page · raw message
tools/perf/util/cpumap.h +9 -7
@@ -43,10 +43,6 @@ struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu(int cpu, void *data);struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data);struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data); -int cpu_map__build_socket_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **sockp); -int cpu_map__build_die_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **diep); -int cpu_map__build_core_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **corep); -int cpu_map__build_node_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **nodep);const struct perf_cpu_map *cpu_map__online (void );/* thread unsafe */ int cpu__setup_cpunode_map (void );
@@ -75,10 +71,16 @@ int cpu__get_die_id(int cpu);*/ int cpu__get_core_id (int cpu); +typedef struct aggr_cpu_id (*aggr_cpu_id_get_t )(int cpu,void *data); -int cpu_map__build_map (struct perf_cpu_map *cpus,struct cpu_aggr_map **res, -struct aggr_cpu_id (*f)(int cpu,void *data), -void *data); +/** +* cpu_aggr_map__new - Create a cpu_aggr_map with an aggr_cpu_id for each cpu in +* cpus. The aggr_cpu_id is created with 'f' that may have a data value passed +* to it. The cpu_aggr_map is sorted with duplicate values removed. +*/ +struct cpu_aggr_map *cpu_aggr_map__new (const struct perf_cpu_map *cpus, +aggr_cpu_id_get_t f, +void *data);bool cpu_map__has (struct perf_cpu_map *cpus,int cpu);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 19/48 perf cpumap: Rename cpu_map__get_X_aggr_by_cpu functions
In reply to Ian RogersThe functions don't use a cpu_map so reduce them to being like constructors of aggr_cpu_id.
--- tools/perf/builtin-stat.c | 18 +++++++++--------- tools/perf/tests/topology.c | 8 ++++---- tools/perf/util/cpumap.c | 14 +++++++------- tools/perf/util/cpumap.h | 29 +++++++++++++++++++++++++---- 4 files changed, 45 insertions(+), 24 deletions(-)
tools/perf/builtin-stat.c +9 -9
@@ -1336,25 +1336,25 @@ static const char *const aggr_mode__string[] = {static struct aggr_cpu_id perf_stat__get_socket (struct perf_stat_config *config__maybe_unused ,int cpu) { -return cpu_map__get_socket_aggr_by_cpu (cpu,/*data=*/ NULL ); +return aggr_cpu_id__socket (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_die (struct perf_stat_config *config__maybe_unused ,int cpu) { -return cpu_map__get_die_aggr_by_cpu (cpu,/*data=*/ NULL ); +return aggr_cpu_id__die (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_core (struct perf_stat_config *config__maybe_unused ,int cpu) { -return cpu_map__get_core_aggr_by_cpu (cpu,/*data=*/ NULL ); +return aggr_cpu_id__core (cpu,/*data=*/ NULL ); }static struct aggr_cpu_id perf_stat__get_node (struct perf_stat_config *config__maybe_unused ,int cpu) { -return cpu_map__get_node_aggr_by_cpu (cpu,/*data=*/ NULL ); +return aggr_cpu_id__node (cpu,/*data=*/ NULL ); } static struct aggr_cpu_id perf_stat__get_aggr(struct perf_stat_config *config,
@@ -1409,16 +1409,16 @@ static aggr_cpu_id_get_t aggr_mode__get_aggr(enum aggr_mode aggr_mode) {switch (aggr_mode) {case AGGR_SOCKET : -return cpu_map__get_socket_aggr_by_cpu; +return aggr_cpu_id__socket;case AGGR_DIE : -return cpu_map__get_die_aggr_by_cpu; +return aggr_cpu_id__die;case AGGR_CORE : -return cpu_map__get_core_aggr_by_cpu; +return aggr_cpu_id__core;case AGGR_NODE : -return cpu_map__get_node_aggr_by_cpu; +return aggr_cpu_id__node;case AGGR_NONE :if (term_percore_set ()) -return cpu_map__get_core_aggr_by_cpu; +return aggr_cpu_id__core;return NULL ;case AGGR_GLOBAL :
tools/perf/tests/topology.c +4 -4
@@ -121,7 +121,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that core ID contains socket, die and core for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_core_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL ); + id =aggr_cpu_id__core (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Core map - Core ID doesn't match" , session->header.env.cpu[map->map[i]].core_id == id.core);
@@ -136,7 +136,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that die ID contains socket and die for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_die_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL ); + id =aggr_cpu_id__die (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Die map - Socket ID doesn't match" , session->header.env.cpu[map->map[i]].socket_id == id.socket);
@@ -150,7 +150,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that socket ID contains only socket for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_socket_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL ); + id =aggr_cpu_id__socket (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Socket map - Socket ID doesn't match" , session->header.env.cpu[map->map[i]].socket_id == id.socket);
@@ -162,7 +162,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that node ID contains only node for (i =0 ; i < map->nr; i++) { - id =cpu_map__get_node_aggr_by_cpu (perf_cpu_map__cpu (map, i),NULL ); + id =aggr_cpu_id__node (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Node map - Node ID doesn't match" ,cpu__get_node (map->map[i]) == id.node);TEST_ASSERT_VAL ("Node map - Socket is set" , id.socket ==-1 );
tools/perf/util/cpumap.c +7 -7
@@ -132,7 +132,7 @@ int cpu__get_socket_id(int cpu)return ret ?: value; } -struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data__maybe_unused ) +struct aggr_cpu_id aggr_cpu_id__socket (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =aggr_cpu_id__empty ();
@@ -200,7 +200,7 @@ int cpu__get_die_id(int cpu)return ret ?: value; } -struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data) +struct aggr_cpu_id aggr_cpu_id__die (int cpu,void *data) {struct aggr_cpu_id id;int die;
@@ -215,7 +215,7 @@ struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu(int cpu, void *data)* with the socket ID and then add die to * make a unique ID. */ - id =cpu_map__get_socket_aggr_by_cpu (cpu, data); + id =aggr_cpu_id__socket (cpu, data);if (aggr_cpu_id__is_empty (&id))return id;
@@ -229,13 +229,13 @@ int cpu__get_core_id(int cpu)return ret ?: value; } -struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data) +struct aggr_cpu_id aggr_cpu_id__core (int cpu,void *data) {struct aggr_cpu_id id;int core =cpu__get_core_id (cpu); -/* cpu_map__get_die returns a struct with socket and die set*/ - id =cpu_map__get_die_aggr_by_cpu (cpu, data); +/* aggr_cpu_id__die returns a struct with socket and die set*/ + id =aggr_cpu_id__die (cpu, data);if (aggr_cpu_id__is_empty (&id))return id;
@@ -248,7 +248,7 @@ struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu(int cpu, void *data) } -struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data__maybe_unused ) +struct aggr_cpu_id aggr_cpu_id__node (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =aggr_cpu_id__empty ();
tools/perf/util/cpumap.h +25 -4
@@ -39,10 +39,6 @@ struct perf_cpu_map *cpu_map__new_data(struct perf_record_cpu_map_data *data);size_t cpu_map__snprint (struct perf_cpu_map *map,char *buf,size_t size);size_t cpu_map__snprint_mask (struct perf_cpu_map *map,char *buf,size_t size);size_t cpu_map__fprintf (struct perf_cpu_map *map,FILE *fp); -struct aggr_cpu_id cpu_map__get_socket_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_die_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_core_aggr_by_cpu (int cpu,void *data); -struct aggr_cpu_id cpu_map__get_node_aggr_by_cpu (int cpu,void *data);const struct perf_cpu_map *cpu_map__online (void );/* thread unsafe */ int cpu__setup_cpunode_map (void );
@@ -88,4 +84,29 @@ bool aggr_cpu_id__equal(const struct aggr_cpu_id *a, const struct aggr_cpu_id *bbool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a);struct aggr_cpu_id aggr_cpu_id__empty (void ); + +/** +* aggr_cpu_id__socket - Create an aggr_cpu_id with the socket populated with +* the socket for cpu. The function signature is compatible with +* aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__socket (int cpu,void *data); +/** +* aggr_cpu_id__die - Create an aggr_cpu_id with the die and socket populated +* with the die and socket for cpu. The function signature is compatible with +* aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__die (int cpu,void *data); +/** +* aggr_cpu_id__core - Create an aggr_cpu_id with the core, die and socket +* populated with the core, die and socket for cpu. The function signature is +* compatible with aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__core (int cpu,void *data); +/** +* aggr_cpu_id__node - Create an aggr_cpu_id with the numa node populated for +* cpu. The function signature is compatible with aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__node (int cpu,void *data); +#endif /* __PERF_CPUMAP_H */
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 20/48 perf cpumap: Move 'has' function to libperf
In reply to Ian RogersMake the cpu map argument const for consistency with the rest of the API. Modify cpu_map__idx accordingly.
--- tools/lib/perf/cpumap.c | 7 ++++++- tools/lib/perf/include/internal/cpumap.h | 2 +- tools/lib/perf/include/perf/cpumap.h | 1 + tools/perf/arch/arm/util/cs-etm.c | 16 ++++++++-------- tools/perf/builtin-sched.c | 6 +++--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 5 ----- tools/perf/util/cpumap.h | 2 -- tools/perf/util/cputopo.c | 2 +- 9 files changed, 21 insertions(+), 22 deletions(-)
tools/lib/perf/cpumap.c +6 -1
@@ -268,7 +268,7 @@ bool perf_cpu_map__empty(const struct perf_cpu_map *map)return map ? map->map[0 ] ==-1 : true; } -int perf_cpu_map__idx (struct perf_cpu_map *cpus,int cpu) +int perf_cpu_map__idx (const struct perf_cpu_map *cpus,int cpu) {int low =0 , high = cpus->nr;
@@ -288,6 +288,11 @@ int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu)return -1 ; } +bool perf_cpu_map__has (const struct perf_cpu_map *cpus,int cpu) +{ +return perf_cpu_map__idx (cpus, cpu) !=-1 ; +} +int perf_cpu_map__max (struct perf_cpu_map *map) {// cpu_map__trim_new() qsort()s it, cpu_map__default_new() sorts it as well.
tools/lib/perf/include/internal/cpumap.h +1 -1
@@ -23,6 +23,6 @@ struct perf_cpu_map {#define MAX_NR_CPUS 2048#endif -int perf_cpu_map__idx (struct perf_cpu_map *cpus,int cpu); +int perf_cpu_map__idx (const struct perf_cpu_map *cpus,int cpu);#endif /* __LIBPERF_INTERNAL_CPUMAP_H */
tools/lib/perf/include/perf/cpumap.h +1 -0
@@ -20,6 +20,7 @@ LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); +LIBPERF_API bool perf_cpu_map__has(const struct perf_cpu_map *map, int cpu);#define perf_cpu_map__for_each_cpu (cpu, idx, cpus) \ for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \
tools/perf/arch/arm/util/cs-etm.c +8 -8
@@ -204,8 +204,8 @@ static int cs_etm_set_option(struct auxtrace_record *itr,/* Set option of each CPU we have */ for (i =0 ; i <cpu__max_cpu (); i++) { -if (!cpu_map__has (event_cpus, i) || - !cpu_map__has (online_cpus, i)) +if (!perf_cpu_map__has (event_cpus, i) || + !perf_cpu_map__has (online_cpus, i))continue ;if (option &BIT (ETM_OPT_CTXTID )) {
@@ -542,8 +542,8 @@ cs_etm_info_priv_size(struct auxtrace_record *itr __maybe_unused,/* cpu map is not empty, we have specific CPUs to work with */ if (!perf_cpu_map__empty (event_cpus)) {for (i =0 ; i <cpu__max_cpu (); i++) { -if (!cpu_map__has (event_cpus, i) || - !cpu_map__has (online_cpus, i)) +if (!perf_cpu_map__has (event_cpus, i) || + !perf_cpu_map__has (online_cpus, i))continue ;if (cs_etm_is_ete (itr, i))
@@ -556,7 +556,7 @@ cs_etm_info_priv_size(struct auxtrace_record *itr __maybe_unused, }else {/* get configuration for all CPUs in the system */ for (i =0 ; i <cpu__max_cpu (); i++) { -if (!cpu_map__has (online_cpus, i)) +if (!perf_cpu_map__has (online_cpus, i))continue ;if (cs_etm_is_ete (itr, i))
@@ -741,8 +741,8 @@ static int cs_etm_info_fill(struct auxtrace_record *itr, }else {/* Make sure all specified CPUs are online */ for (i =0 ; i <perf_cpu_map__nr (event_cpus); i++) { -if (cpu_map__has (event_cpus, i) && - !cpu_map__has (online_cpus, i)) +if (perf_cpu_map__has (event_cpus, i) && + !perf_cpu_map__has (online_cpus, i))return -EINVAL ; }
@@ -763,7 +763,7 @@ static int cs_etm_info_fill(struct auxtrace_record *itr, offset =CS_ETM_SNAPSHOT +1 ;for (i =0 ; i <cpu__max_cpu () && offset < priv_size; i++) -if (cpu_map__has (cpu_map, i)) +if (perf_cpu_map__has (cpu_map, i))cs_etm_get_metadata (i, &offset, itr, info);perf_cpu_map__put (online_cpus);
tools/perf/builtin-sched.c +3 -3
@@ -1617,10 +1617,10 @@ static int map_switch_event(struct perf_sched *sched, struct evsel *evsel,if (curr_thread &&thread__has_color (curr_thread)) pid_color =COLOR_PIDS ; -if (sched->map.cpus && !cpu_map__has (sched->map.cpus, cpu)) +if (sched->map.cpus && !perf_cpu_map__has (sched->map.cpus, cpu))continue ; -if (sched->map.color_cpus &&cpu_map__has (sched->map.color_cpus, cpu)) +if (sched->map.color_cpus &&perf_cpu_map__has (sched->map.color_cpus, cpu)) cpu_color =COLOR_CPUS ;if (cpu != this_cpu)
@@ -1639,7 +1639,7 @@ static int map_switch_event(struct perf_sched *sched, struct evsel *evsel,color_fprintf (stdout, color," " ); } -if (sched->map.cpus && !cpu_map__has (sched->map.cpus, this_cpu)) +if (sched->map.cpus && !perf_cpu_map__has (sched->map.cpus, this_cpu)) goto out;timestamp__scnprintf_usec (timestamp, stimestamp,sizeof (stimestamp));
tools/perf/tests/topology.c +1 -1
@@ -112,7 +112,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Session header CPU map not set" , session->header.env.cpu);for (i =0 ; i < session->header.env.nr_cpus_avail; i++) { -if (!cpu_map__has (map, i)) +if (!perf_cpu_map__has (map, i))continue ; pr_debug("CPU %d, core %d, socket %d\n", i, session->header.env.cpu[i].core_id,
tools/perf/util/cpumap.c +0 -5
@@ -463,11 +463,6 @@ int cpu__setup_cpunode_map(void)return 0 ; } -bool cpu_map__has (struct perf_cpu_map *cpus,int cpu) -{ -return perf_cpu_map__idx (cpus, cpu) !=-1 ; -} -size_t cpu_map__snprint (struct perf_cpu_map *map,char *buf,size_t size) {int i, cpu, start =-1 ;
tools/perf/util/cpumap.h +0 -2
@@ -78,8 +78,6 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus,aggr_cpu_id_get_t f, void *data); -bool cpu_map__has (struct perf_cpu_map *cpus,int cpu); -bool aggr_cpu_id__equal (const struct aggr_cpu_id *a,const struct aggr_cpu_id *b);bool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a);struct aggr_cpu_id aggr_cpu_id__empty (void );
tools/perf/util/cputopo.c +1 -1
@@ -218,7 +218,7 @@ struct cpu_topology *cpu_topology__new(void) tp->core_cpus_list = addr;for (i =0 ; i < nr; i++) { -if (!cpu_map__has (map, i)) +if (!perf_cpu_map__has (map, i))continue ; ret =build_cpu_topology (tp, i);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 20/48] perf cpumap: Move 'has' function to libperf
Hi Ian,
Quoted 10 lines · Ian Rogers
Make the cpu map argument const for consistency with the rest of the API. Modify
cpu_map__idxaccordingly.Reviewed-by James Clark <[email protected]>Signed-off-by Ian Rogers <[email protected]>--- tools/lib/perf/cpumap.c | 7 ++++++- tools/lib/perf/include/internal/cpumap.h | 2 +- tools/lib/perf/include/perf/cpumap.h | 1 +
You need to update tools/lib/perf/libperf.map to have the new API.
Thanks,
Namhyung
Quoted 8 lines
tools/perf/arch/arm/util/cs-etm.c | 16 ++++++++-------- tools/perf/builtin-sched.c | 6 +++--- tools/perf/tests/topology.c | 2 +- tools/perf/util/cpumap.c | 5 ----- tools/perf/util/cpumap.h | 2 -- tools/perf/util/cputopo.c | 2 +- 9 files changed, 21 insertions(+), 22 deletions(-)
Ian Rogers
reviewerPATCH v2 21/48 perf cpumap: Add some comments to cpu_aggr_map
In reply to Ian RogersMove cpu_aggr_map__empty_new to be with other cpu_aggr_map function.
--- tools/perf/util/cpumap.h | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
tools/perf/util/cpumap.h +9 -1
@@ -24,16 +24,18 @@ struct aggr_cpu_id {int core; }; +/** A collection of aggr_cpu_id values, the "built" version is sorted and uniqued. */ struct cpu_aggr_map {refcount_t refcnt; +/** Number of valid entries. */ int nr; +/** The entries. */ struct aggr_cpu_id map[]; };struct perf_record_cpu_map_data ;struct perf_cpu_map *perf_cpu_map__empty_new (int nr); -struct cpu_aggr_map *cpu_aggr_map__empty_new (int nr);struct perf_cpu_map *cpu_map__new_data (struct perf_record_cpu_map_data *data);size_t cpu_map__snprint (struct perf_cpu_map *map,char *buf,size_t size);
@@ -67,6 +69,12 @@ int cpu__get_die_id(int cpu);*/ int cpu__get_core_id (int cpu); +/** +* cpu_aggr_map__empty_new - Create a cpu_aggr_map of size nr with every entry +* being empty. +*/ +struct cpu_aggr_map *cpu_aggr_map__empty_new (int nr); +typedef struct aggr_cpu_id (*aggr_cpu_id_get_t )(int cpu,void *data);/**
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 22/48 perf cpumap: Trim the cpu_aggr_map
In reply to Ian Rogerscpu_aggr_map__new removes duplicates, when this happens shrink the array.
--- tools/perf/util/cpumap.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
tools/perf/util/cpumap.c +6 -1
@@ -185,7 +185,12 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus, c->nr++; } } - +/* Trim. */ +if (c->nr != cpus->nr) { + c =realloc (c,sizeof (struct cpu_aggr_map ) +sizeof (struct aggr_cpu_id ) * c->nr); +if (!c) +return NULL ; + }/* ensure we process id in increasing order */ qsort (c->map, c->nr,sizeof (struct aggr_cpu_id ), aggr_cpu_id__cmp);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 22/48] perf cpumap: Trim the cpu_aggr_map
Quoted 15 lines · Ian Rogers
cpu_aggr_map__newremoves duplicates, when this happens shrink the array.Reviewed-by James Clark <[email protected]>Signed-off-by Ian Rogers <[email protected]>--- tools/perf/util/cpumap.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)tools/perf/util/cpumap.c +0
-0
tools/perf/util/cpumap.c struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus, · line 192
Quoted 9 lines
tools/perf/util/cpumap.c +5
-1@@ -185,7 +185,12 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus, c->nr++; } } - +/* Trim. */ +if (c->nr != cpus->nr) { + c =realloc (c,sizeof (struct cpu_aggr_map ) +sizeof (struct aggr_cpu_id ) * c->nr); +if (!c) +return NULL ;
This leaks the original 'c'. Maybe we can just use it in case of a failure.
Thanks,
Namhyung
Quoted 7 lines
tools/perf/util/cpumap.c +1
-1+ }/* ensure we process id in increasing order */ qsort (c->map, c->nr,sizeof (struct aggr_cpu_id ), aggr_cpu_id__cmp); --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 23/48 perf stat: Fix memory leak in check_per_pkg
In reply to Ian RogersIf the key is already present then free the key used for lookup.
Found with:
$ perf stat -M IO_Read_BW /bin/true
==1749112==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 32 byte(s) in 4 object(s) allocated from:
#0 0x7f6f6fa7d7cf in __interceptor_malloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:145
#1 0x55acecd9d7a6 in check_per_pkg util/stat.c:343
#2 0x55acecd9d9c5 in process_counter_values util/stat.c:365
#3 0x55acecd9e0ab in process_counter_maps util/stat.c:421
#4 0x55acecd9e292 in perf_stat_process_counter util/stat.c:443
#5 0x55aceca8553e in read_counters ./tools/perf/builtin-stat.c:470
#6 0x55aceca88fe3 in __run_perf_stat ./tools/perf/builtin-stat.c:1023
#7 0x55aceca89146 in run_perf_stat ./tools/perf/builtin-stat.c:1048
#8 0x55aceca90858 in cmd_stat ./tools/perf/builtin-stat.c:2555
#9 0x55acecc05fa5 in run_builtin ./tools/perf/perf.c:313
#10 0x55acecc064fe in handle_internal_command ./tools/perf/perf.c:365
#11 0x55acecc068bb in run_argv ./tools/perf/perf.c:409
#12 0x55acecc070aa in main ./tools/perf/perf.c:539--- tools/perf/util/stat.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
tools/perf/util/stat.c +3 -2
@@ -345,9 +345,10 @@ static int check_per_pkg(struct evsel *counter,return -ENOMEM ; *key = (uint64_t )d <<32 | s; -if (hashmap__find (mask, (void *)key,NULL )) +if (hashmap__find (mask, (void *)key,NULL )) { *skip = true; -else +free (key); + }else ret =hashmap__add (mask, (void *)key, (void *)1 );return ret;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 24/48 perf cpumap: Add CPU to aggr_cpu_id
In reply to Ian RogersWith no aggregration, such as 'perf stat -A', the aggr_cpu_id lacks a way to describe per CPU aggregation and the core is set to the CPU in places like print_counter_aggrdata in stat-display.c. Setting the core to the CPU is undesirable as the CPU will exceed valid core values and lead to confusion. Add a CPU variable to address this.
--- tools/perf/tests/topology.c | 19 +++++++++++++++++++ tools/perf/util/cpumap.c | 23 ++++++++++++++++++++--- tools/perf/util/cpumap.h | 8 ++++++++ 3 files changed, 47 insertions(+), 3 deletions(-)
tools/perf/tests/topology.c +19 -0
@@ -119,6 +119,22 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map) session->header.env.cpu[i].socket_id); } +// Test that CPU ID contains socket, die, core and CPU +for (i =0 ; i < map->nr; i++) { + id =aggr_cpu_id__cpu (perf_cpu_map__cpu (map, i),NULL ); +TEST_ASSERT_VAL ("Cpu map - CPU ID doesn't match" , map->map[i] == id.cpu); + +TEST_ASSERT_VAL ("Cpu map - Core ID doesn't match" , + session->header.env.cpu[map->map[i]].core_id == id.core); +TEST_ASSERT_VAL ("Cpu map - Socket ID doesn't match" , + session->header.env.cpu[map->map[i]].socket_id == id.socket); + +TEST_ASSERT_VAL ("Cpu map - Die ID doesn't match" , + session->header.env.cpu[map->map[i]].die_id == id.die); +TEST_ASSERT_VAL ("Cpu map - Node ID is set" , id.node ==-1 ); +TEST_ASSERT_VAL ("Cpu map - Thread is set" , id.thread ==-1 ); + } +// Test that core ID contains socket, die and core for (i =0 ; i < map->nr; i++) { id =aggr_cpu_id__core (perf_cpu_map__cpu (map, i),NULL );
@@ -145,6 +161,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Die map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Die map - Core is set" , id.core ==-1 ); +TEST_ASSERT_VAL ("Die map - CPU is set" , id.cpu ==-1 );TEST_ASSERT_VAL ("Die map - Thread is set" , id.thread ==-1 ); }
@@ -157,6 +174,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Socket map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Socket map - Die ID is set" , id.die ==-1 );TEST_ASSERT_VAL ("Socket map - Core is set" , id.core ==-1 ); +TEST_ASSERT_VAL ("Socket map - CPU is set" , id.cpu ==-1 );TEST_ASSERT_VAL ("Socket map - Thread is set" , id.thread ==-1 ); }
@@ -168,6 +186,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Node map - Socket is set" , id.socket ==-1 );TEST_ASSERT_VAL ("Node map - Die ID is set" , id.die ==-1 );TEST_ASSERT_VAL ("Node map - Core is set" , id.core ==-1 ); +TEST_ASSERT_VAL ("Node map - CPU is set" , id.cpu ==-1 );TEST_ASSERT_VAL ("Node map - Thread is set" , id.thread ==-1 ); }perf_session__delete (session);
tools/perf/util/cpumap.c +20 -3
@@ -253,6 +253,20 @@ struct aggr_cpu_id aggr_cpu_id__core(int cpu, void *data) } +struct aggr_cpu_id aggr_cpu_id__cpu (int cpu,void *data) +{ +struct aggr_cpu_id id; + +/* aggr_cpu_id__die returns a struct with socket and die set*/ + id =aggr_cpu_id__core (cpu, data); +if (aggr_cpu_id__is_empty (&id)) +return id; + + id.cpu = cpu; +return id; + +} +struct aggr_cpu_id aggr_cpu_id__node (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =aggr_cpu_id__empty ();
@@ -576,7 +590,8 @@ bool aggr_cpu_id__equal(const struct aggr_cpu_id *a, const struct aggr_cpu_id *b a->node == b->node && a->socket == b->socket && a->die == b->die && - a->core == b->core; + a->core == b->core && + a->cpu == b->cpu; }bool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a)
@@ -585,7 +600,8 @@ bool aggr_cpu_id__is_empty(const struct aggr_cpu_id *a) a->node ==-1 && a->socket ==-1 && a->die ==-1 && - a->core ==-1 ; + a->core ==-1 && + a->cpu ==-1 ; }struct aggr_cpu_id aggr_cpu_id__empty (void )
@@ -595,7 +611,8 @@ struct aggr_cpu_id aggr_cpu_id__empty(void) .node = -1, .socket = -1, .die = -1, - .core = -1 + .core = -1, + .cpu = -1 };return ret; }
tools/perf/util/cpumap.h +8 -0
@@ -22,6 +22,8 @@ struct aggr_cpu_id {int die;/** The core id as read from /sys/devices/system/cpu/cpuX/topology/core_id. */ int core; +/** CPU aggregation, note there is one CPU for each SMT thread. */ +int cpu; };/** A collection of aggr_cpu_id values, the "built" version is sorted and uniqued. */
@@ -109,6 +111,12 @@ struct aggr_cpu_id aggr_cpu_id__die(int cpu, void *data);* compatible with aggr_cpu_id_get_t. */ struct aggr_cpu_id aggr_cpu_id__core (int cpu,void *data); +/** +* aggr_cpu_id__core - Create an aggr_cpu_id with the cpu, core, die and socket +* populated with the cpu, core, die and socket for cpu. The function signature +* is compatible with aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__cpu (int cpu,void *data);/** * aggr_cpu_id__node - Create an aggr_cpu_id with the numa node populated for * cpu. The function signature is compatible with aggr_cpu_id_get_t.
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 24/48] perf cpumap: Add CPU to aggr_cpu_id
Quoted 9 lines · Ian Rogers
With no aggregration, such as 'perf stat -A', the
aggr_cpu_idlacks a way to describe per CPU aggregation and the core is set to the CPU in places likeprint_counter_aggrdatainstat-display.c. Setting the core to the CPU is undesirable as the CPU will exceed valid core values and lead to confusion. Add a CPU variable to address this.Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
tools/perf/util/cpumap.c struct aggr_cpu_id aggr_cpu_id__core(int cpu, void *data) · line 260
Quoted 13 lines
tools/perf/util/cpumap.c +5
-0@@ -253,6 +253,20 @@ struct aggr_cpu_id aggr_cpu_id__core(int cpu, void *data) } +struct aggr_cpu_id aggr_cpu_id__cpu (int cpu,void *data) +{ +struct aggr_cpu_id id; + +/* aggr_cpu_id__die returns a struct with socket and die set*/
A copy and paste error?
Thanks,
Namhyung
Quoted 71 lines
tools/perf/util/cpumap.c +15
-3+ id =aggr_cpu_id__core (cpu, data); +if (aggr_cpu_id__is_empty (&id)) +return id; + + id.cpu = cpu; +return id; + +} +struct aggr_cpu_id aggr_cpu_id__node (int cpu,void *data__maybe_unused ) {struct aggr_cpu_id id =aggr_cpu_id__empty ();@@ -576,7 +590,8 @@ bool aggr_cpu_id__equal(const struct aggr_cpu_id *a, const struct aggr_cpu_id *b a->node == b->node && a->socket == b->socket && a->die == b->die && - a->core == b->core; + a->core == b->core && + a->cpu == b->cpu; }bool aggr_cpu_id__is_empty (const struct aggr_cpu_id *a)@@ -585,7 +600,8 @@ bool aggr_cpu_id__is_empty(const struct aggr_cpu_id *a) a->node ==-1 && a->socket ==-1 && a->die ==-1 && - a->core ==-1 ; + a->core ==-1 && + a->cpu ==-1 ; }struct aggr_cpu_id aggr_cpu_id__empty (void )@@ -595,7 +611,8 @@ struct aggr_cpu_id aggr_cpu_id__empty(void) .node = -1, .socket = -1, .die = -1, - .core = -1 + .core = -1, + .cpu = -1 };return ret; }tools/perf/util/cpumap.h +8
-1@@ -22,6 +22,8 @@ struct aggr_cpu_id {int die;/** The core id as read from /sys/devices/system/cpu/cpuX/topology/core_id. */ int core; +/** CPU aggregation, note there is one CPU for each SMT thread. */ +int cpu; };/** A collection of aggr_cpu_id values, the "built" version is sorted and uniqued. */ @@ -109,6 +111,12 @@ struct aggr_cpu_id aggr_cpu_id__die(int cpu, void *data);* compatible with aggr_cpu_id_get_t. */ struct aggr_cpu_id aggr_cpu_id__core (int cpu,void *data); +/** +* aggr_cpu_id__core - Create an aggr_cpu_id with the cpu, core, die and socket +* populated with the cpu, core, die and socket for cpu. The function signature +* is compatible with aggr_cpu_id_get_t. +*/ +struct aggr_cpu_id aggr_cpu_id__cpu (int cpu,void *data);/** * aggr_cpu_id__node - Create an aggr_cpu_id with the numa node populated for * cpu. The function signature is compatible with aggr_cpu_id_get_t. -- 2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 25/48 perf stat-display: Avoid use of core for CPU.
In reply to Ian RogersCorrect use of cpumap index in print_no_aggr_metric.
--- tools/perf/util/stat-display.c | 42 +++++++++++++++++----------------- 1 file changed, 21 insertions(+), 21 deletions(-)
tools/perf/util/stat-display.c +21 -21
@@ -120,11 +120,10 @@ static void aggr_printout(struct perf_stat_config *config, id.die, config->csv_output ?0 :-3 , id.core, config->csv_sep); - }else if (id.core >-1 ) { + }else if (id.cpu >-1 ) {fprintf (config->output,"CPU%*d%s" , config->csv_output ?0 :-7 , -evsel__cpus (evsel)->map[id.core], - config->csv_sep); + id.cpu, config->csv_sep); }break ;case AGGR_THREAD :
@@ -334,7 +333,7 @@ static int first_shadow_cpu(struct perf_stat_config *config,int cpu, idx;if (config->aggr_mode ==AGGR_NONE ) -return id->core; +return id->cpu;if (!config->aggr_get_id)return 0 ;
@@ -697,10 +696,9 @@ static void print_counter_aggrdata(struct perf_stat_config *config,fprintf (output,"%s" , prefix); uval = val * counter->scale; -if (cpu !=-1 ) { - id =aggr_cpu_id__empty (); - id.core = cpu; - } +if (cpu !=-1 ) + id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL ); +printout (config, id, nr, counter, uval, prefix, run, ena,1.0 , &rt_stat);if (!metric_only)
@@ -911,8 +909,7 @@ static void print_counter(struct perf_stat_config *config,fprintf (output,"%s" , prefix); uval = val * counter->scale; - id =aggr_cpu_id__empty (); - id.core = cpu; + id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL );printout (config, id,0 , counter, uval, prefix, run, ena,1.0 , &rt_stat);
@@ -924,29 +921,32 @@ static void print_no_aggr_metric(struct perf_stat_config *config,struct evlist *evlist, char *prefix) { - int cpu; -int nrcpus =0 ; -struct evsel *counter; -u64 ena, run, val; -double uval; -struct aggr_cpu_id id; + int cpu, nrcpus; nrcpus = evlist->core.cpus->nr;for (cpu =0 ; cpu < nrcpus; cpu++) { +struct evsel *counter;bool first = true;if (prefix)fputs (prefix, config->output);evlist__for_each_entry (evlist, counter) { - id =aggr_cpu_id__empty (); - id.core = cpu; +u64 ena, run, val; +double uval; +struct aggr_cpu_id id; +int idx =perf_cpu_map__idx (evsel__cpus (counter), cpu); + +if (idx <0 ) +continue ; + + id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL );if (first) {aggr_printout (config, counter, id,0 ); first = false; } - val =perf_counts (counter->counts, cpu,0 )->val; - ena =perf_counts (counter->counts, cpu,0 )->ena; - run =perf_counts (counter->counts, cpu,0 )->run; + val =perf_counts (counter->counts, idx,0 )->val; + ena =perf_counts (counter->counts, idx,0 )->ena; + run =perf_counts (counter->counts, idx,0 )->run; uval = val * counter->scale; printout(config, id, 0, counter, uval, prefix,
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 25/48] perf stat-display: Avoid use of core for CPU.
Quoted 5 lines · Ian Rogers
Correct use of cpumap index in
print_no_aggr_metric.Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
Quoted 16 lines
tools/perf/util/stat-display.c +2
-6@@ -924,29 +921,32 @@ static void print_no_aggr_metric(struct perf_stat_config *config,struct evlist *evlist, char *prefix) { - int cpu; -int nrcpus =0 ; -struct evsel *counter; -u64 ena, run, val; -double uval; -struct aggr_cpu_id id; + int cpu, nrcpus; nrcpus = evlist->core.cpus->nr;for (cpu =0 ; cpu < nrcpus; cpu++) { +struct evsel *counter;bool first = true;
tools/perf/util/stat-display.c static void print_no_aggr_metric(struct perf_stat_config *config, · line 937
Quoted 10 lines
tools/perf/util/stat-display.c +4
-2if (prefix)fputs (prefix, config->output);evlist__for_each_entry (evlist, counter) { - id =aggr_cpu_id__empty (); - id.core = cpu; +u64 ena, run, val; +double uval; +struct aggr_cpu_id id; +int idx =perf_cpu_map__idx (evsel__cpus (counter), cpu);
Not sure about this. Here the 'cpu' is an index for the evlist->core.cpus, not a CPU number. But the perf_cpu_map__idx() requires a CPU number, right?
Thanks,
Namhyung
Quoted 21 lines
tools/perf/util/stat-display.c +8
-4+ +if (idx <0 ) +continue ; + + id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL );if (first) {aggr_printout (config, counter, id,0 ); first = false; } - val =perf_counts (counter->counts, cpu,0 )->val; - ena =perf_counts (counter->counts, cpu,0 )->ena; - run =perf_counts (counter->counts, cpu,0 )->run; + val =perf_counts (counter->counts, idx,0 )->val; + ena =perf_counts (counter->counts, idx,0 )->ena; + run =perf_counts (counter->counts, idx,0 )->run; uval = val * counter->scale; printout(config, id, 0, counter, uval, prefix, --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerRe: [PATCH v2 25/48] perf stat-display: Avoid use of core for CPU.
Quoted 25 lines · Namhyung Kim
Quoted 5 lines · Ian Rogers
Correct use of cpumap index in
print_no_aggr_metric.Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
Quoted 17 lines
tools/perf/util/stat-display.c +2
-6@@ -924,29 +921,32 @@ static void print_no_aggr_metric(struct perf_stat_config *config,struct evlist *evlist, char *prefix) { - int cpu; -int nrcpus =0 ; -struct evsel *counter; -u64 ena, run, val; -double uval; -struct aggr_cpu_id id; + int cpu, nrcpus; nrcpus = evlist->core.cpus->nr;for (cpu =0 ; cpu < nrcpus; cpu++) { +struct evsel *counter;bool first = true;
Quoted 13 lines
tools/perf/util/stat-display.c static void print_no_aggr_metric(struct perf_stat_config *config, · line 937
Quoted 9 lines
tools/perf/util/stat-display.c +4
-2if (prefix)fputs (prefix, config->output);evlist__for_each_entry (evlist, counter) { - id =aggr_cpu_id__empty (); - id.core = cpu; +u64 ena, run, val; +double uval; +struct aggr_cpu_id id; +int idx =perf_cpu_map__idx (evsel__cpus (counter), cpu);Not sure about this. Here the 'cpu' is an index for the
evlist->core.cpus, not a CPU number. But theperf_cpu_map__idx()requires a CPU number, right?
Thanks for the reviews! You are right, I think it makes sense two have two indices here, evlist and counter. I will change the outer loop to use perf_cpu_map__for_each_cpu. v3 will have the other fixes you've pointed out too.
Thanks,
Ian
Quoted 25 lines
Thanks,
NamhyungQuoted 21 lines
tools/perf/util/stat-display.c +8
-4+ +if (idx <0 ) +continue ; + + id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL );if (first) {aggr_printout (config, counter, id,0 ); first = false; } - val =perf_counts (counter->counts, cpu,0 )->val; - ena =perf_counts (counter->counts, cpu,0 )->ena; - run =perf_counts (counter->counts, cpu,0 )->run; + val =perf_counts (counter->counts, idx,0 )->val; + ena =perf_counts (counter->counts, idx,0 )->ena; + run =perf_counts (counter->counts, idx,0 )->run; uval = val * counter->scale; printout(config, id, 0, counter, uval, prefix, --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 26/48 perf evsel: Derive CPUs and threads in alloc_counts
In reply to Ian RogersPassing the number of CPUs and threads allows for an evsel's counts to be mismatched to its cpu map. To avoid this always derive the counts size from the cpu map. Change openaat-syscall-all-cpus to set the cpus to allow for this to work.
--- tools/perf/tests/openat-syscall-all-cpus.c | 10 +--------- tools/perf/util/counts.c | 8 ++++++-- tools/perf/util/counts.h | 2 +- tools/perf/util/evsel.c | 2 +- tools/perf/util/stat.c | 13 ++++++------- 5 files changed, 15 insertions(+), 20 deletions(-)
tools/perf/tests/openat-syscall-all-cpus.c +1 -9
@@ -85,15 +85,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __maybCPU_CLR (cpus->map[cpu], &cpu_set); } -/* -* Here we need to explicitly preallocate the counts, as if -* we use the auto allocation it will allocate just for 1 cpu, -* as we start by cpu 0. -*/ -if (evsel__alloc_counts (evsel, cpus->nr,1 ) <0 ) { -pr_debug ("evsel__alloc_counts(ncpus=%d)\n" , cpus->nr); - goto out_close_fd; - } + evsel->core.cpus =perf_cpu_map__get (cpus); err =0 ;
tools/perf/util/counts.c +6 -2
@@ -4,6 +4,7 @@#include <string.h> #include "evsel.h" #include "counts.h" +#include <perf/threadmap.h> #include <linux/zalloc.h> struct perf_counts *perf_counts__new (int ncpus,int nthreads)
@@ -55,9 +56,12 @@ void evsel__reset_counts(struct evsel *evsel)perf_counts__reset (evsel->counts); } -int evsel__alloc_counts (struct evsel *evsel,int ncpus,int nthreads) +int evsel__alloc_counts (struct evsel *evsel) { - evsel->counts =perf_counts__new (ncpus, nthreads); +struct perf_cpu_map *cpus =evsel__cpus (evsel); +int nthreads =perf_thread_map__nr (evsel->core.threads); + + evsel->counts =perf_counts__new (cpus ? cpus->nr :1 , nthreads);return evsel->counts !=NULL ?0 : -ENOMEM ; }
tools/perf/util/counts.h +1 -1
@@ -40,7 +40,7 @@ void perf_counts__delete(struct perf_counts *counts);void perf_counts__reset (struct perf_counts *counts);void evsel__reset_counts (struct evsel *evsel); -int evsel__alloc_counts (struct evsel *evsel,int ncpus,int nthreads); +int evsel__alloc_counts (struct evsel *evsel);void evsel__free_counts (struct evsel *evsel);#endif /* __PERF_COUNTS_H */
tools/perf/util/evsel.c +1 -1
@@ -1578,7 +1578,7 @@ int __evsel__read_on_cpu(struct evsel *evsel, int cpu, int thread, bool scale)if (FD (evsel, cpu, thread) <0 )return -EINVAL ; -if (evsel->counts ==NULL &&evsel__alloc_counts (evsel, cpu +1 , thread +1 ) <0 ) +if (evsel->counts ==NULL &&evsel__alloc_counts (evsel) <0 )return -ENOMEM ;if (readn (FD (evsel, cpu, thread), &count, nv *sizeof (u64)) <=0 )
tools/perf/util/stat.c +6 -7
@@ -152,11 +152,13 @@ static void evsel__free_stat_priv(struct evsel *evsel)zfree (&evsel->stats); } -static int evsel__alloc_prev_raw_counts (struct evsel *evsel,int ncpus,int nthreads) +static int evsel__alloc_prev_raw_counts (struct evsel *evsel) { +int cpu_map_nr =evsel__nr_cpus (evsel); +int nthreads =perf_thread_map__nr (evsel->core.threads);struct perf_counts *counts; - counts =perf_counts__new (ncpus, nthreads); + counts =perf_counts__new (cpu_map_nr, nthreads);if (counts) evsel->prev_raw_counts = counts;
@@ -177,12 +179,9 @@ static void evsel__reset_prev_raw_counts(struct evsel *evsel)static int evsel__alloc_stats (struct evsel *evsel,bool alloc_raw) { -int ncpus =evsel__nr_cpus (evsel); -int nthreads =perf_thread_map__nr (evsel->core.threads); -if (evsel__alloc_stat_priv (evsel) <0 || -evsel__alloc_counts (evsel, ncpus, nthreads) <0 || - (alloc_raw &&evsel__alloc_prev_raw_counts (evsel, ncpus, nthreads) <0 )) +evsel__alloc_counts (evsel) <0 || + (alloc_raw &&evsel__alloc_prev_raw_counts (evsel) <0 ))return -ENOMEM ;return 0 ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 26/48] perf evsel: Derive CPUs and threads in alloc_counts
Quoted 4 lines · Ian Rogers
Passing the number of CPUs and threads allows for an evsel's counts to be mismatched to its cpu map. To avoid this always derive the counts size from the cpu map. Change openaat-syscall-all-cpus to set the cpus
s/openaat/openat/
Quoted 19 lines
to allow for this to work.
Signed-off-by Ian Rogers <[email protected]>--- tools/perf/tests/openat-syscall-all-cpus.c | 10 +--------- tools/perf/util/counts.c | 8 ++++++-- tools/perf/util/counts.h | 2 +- tools/perf/util/evsel.c | 2 +- tools/perf/util/stat.c | 13 ++++++------- 5 files changed, 15 insertions(+), 20 deletions(-)tools/perf/tests/openat-syscall-all-cpus.c +0
-0@@ -85,15 +85,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __maybCPU_CLR (cpus->map[cpu], &cpu_set); }
tools/perf/tests/openat-syscall-all-cpus.c static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb · line 88
Quoted 10 lines
tools/perf/tests/openat-syscall-all-cpus.c +1
-9-/* -* Here we need to explicitly preallocate the counts, as if -* we use the auto allocation it will allocate just for 1 cpu, -* as we start by cpu 0. -*/ -if (evsel__alloc_counts (evsel, cpus->nr,1 ) <0 ) { -pr_debug ("evsel__alloc_counts(ncpus=%d)\n" , cpus->nr); - goto out_close_fd; - } + evsel->core.cpus =perf_cpu_map__get (cpus);
I was about to write that you need to call evsel__alloc_counts() after this. But it seems evsel__read_on_cpu() can do the job later. So I'm fine with this. :)
Thanks,
Namhyung
Quoted 94 lines
tools/perf/tests/openat-syscall-all-cpus.c +0
-0err =0 ;tools/perf/util/counts.c +6
-2@@ -4,6 +4,7 @@#include <string.h> #include "evsel.h" #include "counts.h" +#include <perf/threadmap.h> #include <linux/zalloc.h> struct perf_counts *perf_counts__new (int ncpus,int nthreads)@@ -55,9 +56,12 @@ void evsel__reset_counts(struct evsel *evsel)perf_counts__reset (evsel->counts); } -int evsel__alloc_counts (struct evsel *evsel,int ncpus,int nthreads) +int evsel__alloc_counts (struct evsel *evsel) { - evsel->counts =perf_counts__new (ncpus, nthreads); +struct perf_cpu_map *cpus =evsel__cpus (evsel); +int nthreads =perf_thread_map__nr (evsel->core.threads); + + evsel->counts =perf_counts__new (cpus ? cpus->nr :1 , nthreads);return evsel->counts !=NULL ?0 : -ENOMEM ; }tools/perf/util/counts.h +1
-1@@ -40,7 +40,7 @@ void perf_counts__delete(struct perf_counts *counts);void perf_counts__reset (struct perf_counts *counts);void evsel__reset_counts (struct evsel *evsel); -int evsel__alloc_counts (struct evsel *evsel,int ncpus,int nthreads); +int evsel__alloc_counts (struct evsel *evsel);void evsel__free_counts (struct evsel *evsel);#endif /* __PERF_COUNTS_H */ tools/perf/util/evsel.c +1
-1@@ -1578,7 +1578,7 @@ int __evsel__read_on_cpu(struct evsel *evsel, int cpu, int thread, bool scale)if (FD (evsel, cpu, thread) <0 )return -EINVAL ; -if (evsel->counts ==NULL &&evsel__alloc_counts (evsel, cpu +1 , thread +1 ) <0 ) +if (evsel->counts ==NULL &&evsel__alloc_counts (evsel) <0 )return -ENOMEM ;if (readn (FD (evsel, cpu, thread), &count, nv *sizeof (u64)) <=0 )tools/perf/util/stat.c +6
-8@@ -152,11 +152,13 @@ static void evsel__free_stat_priv(struct evsel *evsel)zfree (&evsel->stats); } -static int evsel__alloc_prev_raw_counts (struct evsel *evsel,int ncpus,int nthreads) +static int evsel__alloc_prev_raw_counts (struct evsel *evsel) { +int cpu_map_nr =evsel__nr_cpus (evsel); +int nthreads =perf_thread_map__nr (evsel->core.threads);struct perf_counts *counts; - counts =perf_counts__new (ncpus, nthreads); + counts =perf_counts__new (cpu_map_nr, nthreads);if (counts) evsel->prev_raw_counts = counts;@@ -177,12 +179,9 @@ static void evsel__reset_prev_raw_counts(struct evsel *evsel)static int evsel__alloc_stats (struct evsel *evsel,bool alloc_raw) { -int ncpus =evsel__nr_cpus (evsel); -int nthreads =perf_thread_map__nr (evsel->core.threads); -if (evsel__alloc_stat_priv (evsel) <0 || -evsel__alloc_counts (evsel, ncpus, nthreads) <0 || - (alloc_raw &&evsel__alloc_prev_raw_counts (evsel, ncpus, nthreads) <0 )) +evsel__alloc_counts (evsel) <0 || + (alloc_raw &&evsel__alloc_prev_raw_counts (evsel) <0 ))return -ENOMEM ;return 0 ; --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 27/48 libperf: Switch cpu to more accurate cpu_map_idx
In reply to Ian RogersModify variable names and adopt perf_cpu_map__for_each_cpu in perf_evsel__open. Renaming is done by looking for consistency in API usage.
--- tools/lib/perf/evsel.c | 88 +++++++++++++++-------------- tools/lib/perf/include/perf/evsel.h | 10 ++-- 2 files changed, 50 insertions(+), 48 deletions(-)
tools/lib/perf/evsel.c +45 -43
@@ -43,18 +43,22 @@ void perf_evsel__delete(struct perf_evsel *evsel)free (evsel); } -#define FD (e, x, y) ((int *) xyarray__entry(e->fd, x, y)) -#define MMAP (e, x, y) (e->mmap ? ((struct perf_mmap *) xyarray__entry(e->mmap, x, y)) : NULL) +#define FD (_evsel, _cpu_map_idx, _thread) \ + ((int *)xyarray__entry(_evsel->fd, _cpu_map_idx, _thread)) +#define MMAP (_evsel, _cpu_map_idx, _thread) \ + (_evsel->mmap ? ((struct perf_mmap *) xyarray__entry(_evsel->mmap, _cpu_map_idx, _thread)) \ + : NULL)int perf_evsel__alloc_fd (struct perf_evsel *evsel,int ncpus,int nthreads) { evsel->fd =xyarray__new (ncpus, nthreads,sizeof (int ));if (evsel->fd) { -int cpu, thread; -for (cpu =0 ; cpu < ncpus; cpu++) { +int idx, thread; + +for (idx =0 ; idx < ncpus; idx++) {for (thread =0 ; thread < nthreads; thread++) { -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, idx, thread);if (fd) *fd =-1 ;
@@ -80,7 +84,7 @@ sys_perf_event_open(struct perf_event_attr *attr,return syscall (__NR_perf_event_open, attr, pid, cpu, group_fd, flags); } -static int get_group_fd (struct perf_evsel *evsel,int cpu,int thread,int *group_fd) +static int get_group_fd (struct perf_evsel *evsel,int cpu_map_idx,int thread,int *group_fd) {struct perf_evsel *leader = evsel->leader;int *fd;
@@ -97,7 +101,7 @@ static int get_group_fd(struct perf_evsel *evsel, int cpu, int thread, int *grouif (!leader->fd)return -ENOTCONN ; - fd =FD (leader, cpu, thread); + fd =FD (leader, cpu_map_idx, thread);if (fd ==NULL || *fd ==-1 )return -EBADF ;
@@ -109,7 +113,7 @@ static int get_group_fd(struct perf_evsel *evsel, int cpu, int thread, int *grouint perf_evsel__open (struct perf_evsel *evsel,struct perf_cpu_map *cpus,struct perf_thread_map *threads) { -int cpu, thread, err =0 ; +int cpu, idx, thread, err =0 ;if (cpus ==NULL ) {static struct perf_cpu_map *empty_cpu_map;
@@ -139,21 +143,21 @@ int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0)return -ENOMEM ; -for (cpu =0 ; cpu < cpus->nr; cpu++) { +perf_cpu_map__for_each_cpu (cpu, idx, cpus) {for (thread =0 ; thread < threads->nr; thread++) {int fd, group_fd, *evsel_fd; - evsel_fd =FD (evsel, cpu, thread); + evsel_fd =FD (evsel, idx, thread);if (evsel_fd ==NULL )return -EINVAL ; - err =get_group_fd (evsel, cpu, thread, &group_fd); + err =get_group_fd (evsel, idx, thread, &group_fd);if (err <0 )return err; fd =sys_perf_event_open (&evsel->attr, threads->map[thread].pid, - cpus->map[cpu], group_fd,0 ); + cpu, group_fd,0 );if (fd <0 )return -errno;
@@ -165,12 +169,12 @@ int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus,return err; } -static void perf_evsel__close_fd_cpu (struct perf_evsel *evsel,int cpu) +static void perf_evsel__close_fd_cpu (struct perf_evsel *evsel,int cpu_map_idx) {int thread;for (thread =0 ; thread <xyarray__max_y (evsel->fd); ++thread) { -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, cpu_map_idx, thread);if (fd && *fd >=0 ) {close (*fd);
@@ -181,10 +185,8 @@ static void perf_evsel__close_fd_cpu(struct perf_evsel *evsel, int cpu)void perf_evsel__close_fd (struct perf_evsel *evsel) { -int cpu; - -for (cpu =0 ; cpu <xyarray__max_x (evsel->fd); cpu++) -perf_evsel__close_fd_cpu (evsel, cpu); +for (int idx =0 ; idx <xyarray__max_x (evsel->fd); idx++) +perf_evsel__close_fd_cpu (evsel, idx); }void perf_evsel__free_fd (struct perf_evsel *evsel)
@@ -202,29 +204,29 @@ void perf_evsel__close(struct perf_evsel *evsel)perf_evsel__free_fd (evsel); } -void perf_evsel__close_cpu (struct perf_evsel *evsel,int cpu) +void perf_evsel__close_cpu (struct perf_evsel *evsel,int cpu_map_idx) {if (evsel->fd ==NULL )return ; -perf_evsel__close_fd_cpu (evsel, cpu); +perf_evsel__close_fd_cpu (evsel, cpu_map_idx); }void perf_evsel__munmap (struct perf_evsel *evsel) { -int cpu, thread; +int idx, thread;if (evsel->fd ==NULL || evsel->mmap ==NULL )return ; -for (cpu =0 ; cpu <xyarray__max_x (evsel->fd); cpu++) { +for (idx =0 ; idx <xyarray__max_x (evsel->fd); idx++) {for (thread =0 ; thread <xyarray__max_y (evsel->fd); thread++) { -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, idx, thread);if (fd ==NULL || *fd <0 )continue ; -perf_mmap__munmap (MMAP (evsel, cpu, thread)); +perf_mmap__munmap (MMAP (evsel, idx, thread)); } }
@@ -234,7 +236,7 @@ void perf_evsel__munmap(struct perf_evsel *evsel)int perf_evsel__mmap (struct perf_evsel *evsel,int pages) { -int ret, cpu, thread; +int ret, idx, thread;struct perf_mmap_param mp = { .prot =PROT_READ |PROT_WRITE , .mask = (pages * page_size) -1 ,
@@ -246,18 +248,18 @@ int perf_evsel__mmap(struct perf_evsel *evsel, int pages)if (perf_evsel__alloc_mmap (evsel,xyarray__max_x (evsel->fd),xyarray__max_y (evsel->fd)) <0 )return -ENOMEM ; -for (cpu =0 ; cpu <xyarray__max_x (evsel->fd); cpu++) { +for (idx =0 ; idx <xyarray__max_x (evsel->fd); idx++) {for (thread =0 ; thread <xyarray__max_y (evsel->fd); thread++) { -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, idx, thread);struct perf_mmap *map;if (fd ==NULL || *fd <0 )continue ; - map =MMAP (evsel, cpu, thread); + map =MMAP (evsel, idx, thread);perf_mmap__init (map,NULL , false,NULL ); - ret =perf_mmap__mmap (map, &mp, *fd, cpu); + ret =perf_mmap__mmap (map, &mp, *fd, idx);if (ret) {perf_evsel__munmap (evsel);return ret;
@@ -268,14 +270,14 @@ int perf_evsel__mmap(struct perf_evsel *evsel, int pages)return 0 ; } -void *perf_evsel__mmap_base (struct perf_evsel *evsel,int cpu,int thread) +void *perf_evsel__mmap_base (struct perf_evsel *evsel,int cpu_map_idx,int thread) { -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, cpu_map_idx, thread); -if (fd ==NULL || *fd <0 ||MMAP (evsel, cpu, thread) ==NULL ) +if (fd ==NULL || *fd <0 ||MMAP (evsel, cpu_map_idx, thread) ==NULL )return NULL ; -return MMAP (evsel, cpu, thread)->base; +return MMAP (evsel, cpu_map_idx, thread)->base; }int perf_evsel__read_size (struct perf_evsel *evsel)
@@ -303,19 +305,19 @@ int perf_evsel__read_size(struct perf_evsel *evsel)return size; } -int perf_evsel__read (struct perf_evsel *evsel,int cpu,int thread, +int perf_evsel__read (struct perf_evsel *evsel,int cpu_map_idx,int thread,struct perf_counts_values *count) {size_t size =perf_evsel__read_size (evsel); -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, cpu_map_idx, thread);memset (count,0 ,sizeof (*count));if (fd ==NULL || *fd <0 )return -EINVAL ; -if (MMAP (evsel, cpu, thread) && - !perf_mmap__read_self (MMAP (evsel, cpu, thread), count)) +if (MMAP (evsel, cpu_map_idx, thread) && + !perf_mmap__read_self (MMAP (evsel, cpu_map_idx, thread), count))return 0 ;if (readn (*fd, count->values, size) <=0 )
@@ -326,13 +328,13 @@ int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread,static int perf_evsel__run_ioctl (struct perf_evsel *evsel,int ioc,void *arg, -int cpu) +int cpu_map_idx) {int thread;for (thread =0 ; thread <xyarray__max_y (evsel->fd); thread++) {int err; -int *fd =FD (evsel, cpu, thread); +int *fd =FD (evsel, cpu_map_idx, thread);if (fd ==NULL || *fd <0 )return -1 ;
@@ -346,9 +348,9 @@ static int perf_evsel__run_ioctl(struct perf_evsel *evsel,return 0 ; } -int perf_evsel__enable_cpu (struct perf_evsel *evsel,int cpu) +int perf_evsel__enable_cpu (struct perf_evsel *evsel,int cpu_map_idx) { -return perf_evsel__run_ioctl (evsel,PERF_EVENT_IOC_ENABLE ,NULL , cpu); +return perf_evsel__run_ioctl (evsel,PERF_EVENT_IOC_ENABLE ,NULL , cpu_map_idx); }int perf_evsel__enable (struct perf_evsel *evsel)
@@ -361,9 +363,9 @@ int perf_evsel__enable(struct perf_evsel *evsel)return err; } -int perf_evsel__disable_cpu (struct perf_evsel *evsel,int cpu) +int perf_evsel__disable_cpu (struct perf_evsel *evsel,int cpu_map_idx) { -return perf_evsel__run_ioctl (evsel,PERF_EVENT_IOC_DISABLE ,NULL , cpu); +return perf_evsel__run_ioctl (evsel,PERF_EVENT_IOC_DISABLE ,NULL , cpu_map_idx); }int perf_evsel__disable (struct perf_evsel *evsel)
tools/lib/perf/include/perf/evsel.h +5 -5
@@ -28,16 +28,16 @@ LIBPERF_API void perf_evsel__delete(struct perf_evsel *evsel); LIBPERF_API int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus,struct perf_thread_map *threads); LIBPERF_API void perf_evsel__close(struct perf_evsel *evsel); -LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu_map_idx); LIBPERF_API int perf_evsel__mmap(struct perf_evsel *evsel, int pages); LIBPERF_API void perf_evsel__munmap(struct perf_evsel *evsel); -LIBPERF_API void *perf_evsel__mmap_base(struct perf_evsel *evsel, int cpu, int thread); -LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, +LIBPERF_API void *perf_evsel__mmap_base(struct perf_evsel *evsel, int cpu_map_idx, int thread); +LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu_map_idx, int thread,struct perf_counts_values *count); LIBPERF_API int perf_evsel__enable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu_map_idx); LIBPERF_API int perf_evsel__disable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu_map_idx); LIBPERF_API struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); LIBPERF_API struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); LIBPERF_API struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 28/48 libperf: Use cpu not index for evsel mmap
In reply to Ian RogersFix issue where evsel's CPU map index was being used as the mmap cpu.
--- tools/lib/perf/evsel.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
tools/lib/perf/evsel.c +2 -1
@@ -252,6 +252,7 @@ int perf_evsel__mmap(struct perf_evsel *evsel, int pages)for (thread =0 ; thread <xyarray__max_y (evsel->fd); thread++) {int *fd =FD (evsel, idx, thread);struct perf_mmap *map; +int cpu =perf_cpu_map__cpu (evsel->cpus, idx);if (fd ==NULL || *fd <0 )continue ;
@@ -259,7 +260,7 @@ int perf_evsel__mmap(struct perf_evsel *evsel, int pages) map =MMAP (evsel, idx, thread);perf_mmap__init (map,NULL , false,NULL ); - ret =perf_mmap__mmap (map, &mp, *fd, idx); + ret =perf_mmap__mmap (map, &mp, *fd, cpu);if (ret) {perf_evsel__munmap (evsel);return ret;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 29/48 perf counts: Switch name cpu to cpu_map_idx
In reply to Ian RogersTry to reduce confusion in particular when the cpu map doesn't contain an entry for every CPU.
--- tools/perf/util/counts.h | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
tools/perf/util/counts.h +6 -6
@@ -18,21 +18,21 @@ struct perf_counts {static inline struct perf_counts_values * -perf_counts (struct perf_counts *counts,int cpu,int thread) +perf_counts (struct perf_counts *counts,int cpu_map_idx,int thread) { -return xyarray__entry (counts->values, cpu, thread); +return xyarray__entry (counts->values, cpu_map_idx, thread); }static inline bool -perf_counts__is_loaded (struct perf_counts *counts,int cpu,int thread) +perf_counts__is_loaded (struct perf_counts *counts,int cpu_map_idx,int thread) { -return *((bool *)xyarray__entry (counts->loaded, cpu, thread)); +return *((bool *)xyarray__entry (counts->loaded, cpu_map_idx, thread)); }static inline void -perf_counts__set_loaded (struct perf_counts *counts,int cpu,int thread,bool loaded) +perf_counts__set_loaded (struct perf_counts *counts,int cpu_map_idx,int thread,bool loaded) { - *((bool *)xyarray__entry (counts->loaded, cpu, thread)) = loaded; + *((bool *)xyarray__entry (counts->loaded, cpu_map_idx, thread)) = loaded; }struct perf_counts *perf_counts__new (int ncpus,int nthreads);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 30/48 perf stat: Rename aggr_data cpu to imply it's an index
In reply to Ian RogersTrying to make cpu maps less error prone.
--- tools/perf/util/python.c | 2 +- tools/perf/util/stat-display.c | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-)
tools/perf/util/python.c +1 -1
@@ -1057,7 +1057,7 @@ static struct mmap *get_md(struct evlist *evlist, int cpu)for (i =0 ; i < evlist->core.nr_mmaps; i++) {struct mmap *md = &evlist->mmap[i]; -if (md->core.cpu == cpu) +if (md->core.cpu_map_idx == cpu)return md; }
tools/perf/util/stat-display.c +5 -5
@@ -626,7 +626,7 @@ struct aggr_data {u64 ena, run, val;struct aggr_cpu_id id;int nr; -int cpu; +int cpu_map_idx; }; static void aggr_cb(struct perf_stat_config *config,
@@ -878,9 +878,9 @@ static void counter_cb(struct perf_stat_config *config __maybe_unused, {struct aggr_data *ad = data; - ad->val +=perf_counts (counter->counts, ad->cpu,0 )->val; - ad->ena +=perf_counts (counter->counts, ad->cpu,0 )->ena; - ad->run +=perf_counts (counter->counts, ad->cpu,0 )->run; + ad->val +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->val; + ad->ena +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->ena; + ad->run +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->run; }/*
@@ -897,7 +897,7 @@ static void print_counter(struct perf_stat_config *config,struct aggr_cpu_id id;for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { -struct aggr_data ad = { .cpu = cpu }; +struct aggr_data ad = { .cpu_map_idx = cpu };if (!collect_data (config, counter, counter_cb, &ad))return ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 30/48] perf stat: Rename aggr_data cpu to imply it's an index
Quoted 14 lines · Ian Rogers
Trying to make cpu maps less error prone.
Signed-off-by Ian Rogers <[email protected]>--- tools/perf/util/python.c | 2 +- tools/perf/util/stat-display.c | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-)tools/perf/util/python.c +0
-0
tools/perf/util/python.c static struct mmap *get_md(struct evlist *evlist, int cpu) · line 1060
Quoted 5 lines
tools/perf/util/python.c +1
-1@@ -1057,7 +1057,7 @@ static struct mmap *get_md(struct evlist *evlist, int cpu)for (i =0 ; i < evlist->core.nr_mmaps; i++) {struct mmap *md = &evlist->mmap[i]; -if (md->core.cpu == cpu) +if (md->core.cpu_map_idx == cpu)
It doesn't seem to belong to this change..
Thanks,
Namhyung
Quoted 41 lines
tools/perf/util/python.c +0
-0return md; }tools/perf/util/stat-display.c +5
-6@@ -626,7 +626,7 @@ struct aggr_data {u64 ena, run, val;struct aggr_cpu_id id;int nr; -int cpu; +int cpu_map_idx; }; static void aggr_cb(struct perf_stat_config *config,@@ -878,9 +878,9 @@ static void counter_cb(struct perf_stat_config *config __maybe_unused, {struct aggr_data *ad = data; - ad->val +=perf_counts (counter->counts, ad->cpu,0 )->val; - ad->ena +=perf_counts (counter->counts, ad->cpu,0 )->ena; - ad->run +=perf_counts (counter->counts, ad->cpu,0 )->run; + ad->val +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->val; + ad->ena +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->ena; + ad->run +=perf_counts (counter->counts, ad->cpu_map_idx,0 )->run; }/* @@ -897,7 +897,7 @@ static void print_counter(struct perf_stat_config *config,struct aggr_cpu_id id;for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { -struct aggr_data ad = { .cpu = cpu }; +struct aggr_data ad = { .cpu_map_idx = cpu };if (!collect_data (config, counter, counter_cb, &ad))return ; --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 31/48 perf stat: Use perf_cpu_map__for_each_cpu
In reply to Ian RogersCorrect in print_counter where an index was being used as a cpu.
--- tools/perf/util/stat-display.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)
tools/perf/util/stat-display.c +8 -8
@@ -4,6 +4,7 @@#include <linux/string.h> #include <linux/time64.h> #include <math.h> +#include <perf/cpumap.h> #include "color.h" #include "counts.h" #include "evlist.h"
@@ -732,7 +733,7 @@ static void print_aggr(struct perf_stat_config *config,evlist__for_each_entry (evlist, counter) {print_counter_aggrdata (config, counter, s, prefix, metric_only, - &first,-1 ); + &first,/*cpu=*/ -1 ); }if (metric_only)fputc ('\n' , output);
@@ -893,11 +894,11 @@ static void print_counter(struct perf_stat_config *config,FILE *output = config->output;u64 ena, run, val;double uval; -int cpu; +int idx, cpu;struct aggr_cpu_id id; -for (cpu =0 ; cpu <evsel__nr_cpus (counter); cpu++) { -struct aggr_data ad = { .cpu_map_idx = cpu }; +perf_cpu_map__for_each_cpu (cpu, idx,evsel__cpus (counter)) { +struct aggr_data ad = { .cpu_map_idx = idx };if (!collect_data (config, counter, counter_cb, &ad))return ;
@@ -921,10 +922,9 @@ static void print_no_aggr_metric(struct perf_stat_config *config,struct evlist *evlist, char *prefix) { - int cpu, nrcpus; + int all_idx, cpu; - nrcpus = evlist->core.cpus->nr; -for (cpu =0 ; cpu < nrcpus; cpu++) { +perf_cpu_map__for_each_cpu (cpu, all_idx, evlist->core.cpus) {struct evsel *counter;bool first = true;
@@ -1249,7 +1249,7 @@ static void print_percore(struct perf_stat_config *config,print_counter_aggrdata (config, counter, s, prefix, metric_only, - &first,-1 ); + &first,/*cpu=*/ -1 ); }if (metric_only)
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 32/48 perf script: Use for each cpu to aid readability
In reply to Ian RogersUse perf_cpu_map__for_each_cpu to help with readability.
--- tools/perf/builtin-script.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
tools/perf/builtin-script.c +3 -4
@@ -2113,8 +2113,7 @@ static struct scripting_ops *scripting_ops;static void __process_stat (struct evsel *counter,u64 tstamp) {int nthreads =perf_thread_map__nr (counter->core.threads); -int ncpus =evsel__nr_cpus (counter); -int cpu, thread; +int idx, cpu, thread;static int header_printed;if (counter->core.system_wide)
@@ -2127,13 +2126,13 @@ static void __process_stat(struct evsel *counter, u64 tstamp) }for (thread =0 ; thread < nthreads; thread++) { -for (cpu =0 ; cpu < ncpus; cpu++) { +perf_cpu_map__for_each_cpu (cpu, idx,evsel__cpus (counter)) {struct perf_counts_values *counts; counts =perf_counts (counter->counts, cpu, thread); printf("%3d %8d %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %s\n", - counter->core.cpus->map[cpu], + cpu, perf_thread_map__pid(counter->core.threads, thread), counts->val, counts->ena,
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 33/48 libperf: Allow NULL in perf_cpu_map__idx
In reply to Ian RogersReturn -1, not found, if NULL is passed.
--- tools/lib/perf/cpumap.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
tools/lib/perf/cpumap.c +6 -1
@@ -270,8 +270,13 @@ bool perf_cpu_map__empty(const struct perf_cpu_map *map)int perf_cpu_map__idx (const struct perf_cpu_map *cpus,int cpu) { -int low =0 , high = cpus->nr; +int low, high; +if (!cpus) +return -1 ; + + low =0 ; + high = cpus->nr;while (low < high) {int idx = (low + high) /2 , cpu_at_idx = cpus->map[idx];
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 34/48 perf evlist: Refactor evlist__for_each_cpu.
In reply to Ian RogersPreviously evlist__for_each_cpu needed to iterate over the evlist in an inner loop and call "skip" routines. Refactor this so that the iteratr is smarter and the next function can update both the current CPU and evsel.
By using a cpu map index, fix apparent off-by-1 in __run_perf_stat's call to perf_evsel__close_cpu.
--- tools/perf/builtin-stat.c | 179 ++++++++++++++++++-------------------- tools/perf/util/evlist.c | 146 +++++++++++++++++-------------- tools/perf/util/evlist.h | 50 +++++++++-- tools/perf/util/evsel.h | 1 - 4 files changed, 210 insertions(+), 166 deletions(-)
tools/perf/builtin-stat.c +86 -93
@@ -405,36 +405,33 @@ static int read_counter_cpu(struct evsel *counter, struct timespec *rs, int cpu)static int read_affinity_counters (struct timespec *rs) { -struct evsel *counter; -struct affinity affinity; -int i, ncpus, cpu; +struct evlist_cpu_iterator evlist_cpu_itr; +struct affinity saved_affinity, *affinity;if (all_counters_use_bpf)return 0 ; -if (affinity__setup (&affinity) <0 ) +if (!target__has_cpu (&target) ||target__has_per_thread (&target)) + affinity =NULL ; +else if (affinity__setup (&saved_affinity) <0 )return -1 ; +else + affinity = &saved_affinity; - ncpus =perf_cpu_map__nr (evsel_list->core.all_cpus); -if (!target__has_cpu (&target) ||target__has_per_thread (&target)) - ncpus =1 ; -evlist__for_each_cpu (evsel_list, i, cpu) { -if (i >= ncpus) -break ; -affinity__set (&affinity, cpu); +evlist__for_each_cpu (evlist_cpu_itr, evsel_list, affinity) { +struct evsel *counter = evlist_cpu_itr.evsel; -evlist__for_each_entry (evsel_list, counter) { -if (evsel__cpu_iter_skip (counter, cpu)) -continue ; -if (evsel__is_bpf (counter)) -continue ; -if (!counter->err) { - counter->err =read_counter_cpu (counter, rs, - counter->cpu_iter -1 ); - } +if (evsel__is_bpf (counter)) +continue ; + +if (!counter->err) { + counter->err =read_counter_cpu (counter, rs, + evlist_cpu_itr.cpu_map_idx); } } -affinity__cleanup (&affinity); +if (affinity) +affinity__cleanup (&saved_affinity); +return 0 ; }
@@ -788,8 +785,9 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx)int status =0 ;const bool forks = (argc >0 );bool is_pipe =STAT_RECORD ? perf_stat.data.is_pipe : false; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int i, cpu, err; +int err;bool second_pass = false;if (forks) {
@@ -813,102 +811,97 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx) all_counters_use_bpf = false; } - evlist__for_each_cpu (evsel_list, i, cpu) { +evlist__for_each_cpu (evlist_cpu_itr, evsel_list, &affinity) { + counter = evlist_cpu_itr.evsel; +/* * bperf calls evsel__open_per_cpu() in bperf__load(), so * no need to call it again here. */ if (target.use_bpf)break ; -affinity__set (&affinity, cpu); -evlist__for_each_entry (evsel_list, counter) { -if (evsel__cpu_iter_skip (counter, cpu)) +if (counter->reset_group || counter->errored) +continue ; +if (evsel__is_bpf (counter)) +continue ; +try_again: +if (create_perf_stat_counter (counter, &stat_config, &target, + evlist_cpu_itr.cpu_map_idx) <0 ) { + +/* +* Weak group failed. We cannot just undo this here +* because earlier CPUs might be in group mode, and the kernel +* doesn't support mixing group and non group reads. Defer +* it to later. +* Don't close here because we're in the wrong affinity. +*/ +if ((errno ==EINVAL || errno ==EBADF ) && +evsel__leader (counter) != counter && + counter->weak_group) { +evlist__reset_weak_group (evsel_list, counter, false); +assert (counter->reset_group); + second_pass = true;continue ; -if (counter->reset_group || counter->errored) + } + +switch (stat_handle_error (counter)) { +case COUNTER_FATAL : +return -1 ; +case COUNTER_RETRY : + goto try_again; +case COUNTER_SKIP :continue ; -if (evsel__is_bpf (counter)) +default : +break ; + } + + } + counter->supported = true; + } + +if (second_pass) { +/* +* Now redo all the weak group after closing them, +* and also close errored counters. +*/ + +/* First close errored or weak retry */ +evlist__for_each_cpu (evlist_cpu_itr, evsel_list, &affinity) { + counter = evlist_cpu_itr.evsel; + +if (!counter->reset_group && !counter->errored)continue ; -try_again: + +perf_evsel__close_cpu (&counter->core, evlist_cpu_itr.cpu_map_idx); + } +/* Now reopen weak */ +evlist__for_each_cpu (evlist_cpu_itr, evsel_list, &affinity) { + counter = evlist_cpu_itr.evsel; + +if (!counter->reset_group && !counter->errored) +continue ; +if (!counter->reset_group) +continue ; +try_again_reset: +pr_debug2 ("reopening weak %s\n" ,evsel__name (counter));if (create_perf_stat_counter (counter, &stat_config, &target, - counter->cpu_iter -1 ) <0 ) { - -/* -* Weak group failed. We cannot just undo this here -* because earlier CPUs might be in group mode, and the kernel -* doesn't support mixing group and non group reads. Defer -* it to later. -* Don't close here because we're in the wrong affinity. -*/ -if ((errno ==EINVAL || errno ==EBADF ) && -evsel__leader (counter) != counter && - counter->weak_group) { -evlist__reset_weak_group (evsel_list, counter, false); -assert (counter->reset_group); - second_pass = true; -continue ; - } + evlist_cpu_itr.cpu_map_idx) <0 ) {switch (stat_handle_error (counter)) {case COUNTER_FATAL :return -1 ;case COUNTER_RETRY : - goto try_again; + goto try_again_reset;case COUNTER_SKIP :continue ;default :break ; } - } counter->supported = true; } } - -if (second_pass) { -/* -* Now redo all the weak group after closing them, -* and also close errored counters. -*/ - -evlist__for_each_cpu (evsel_list, i, cpu) { -affinity__set (&affinity, cpu); -/* First close errored or weak retry */ -evlist__for_each_entry (evsel_list, counter) { -if (!counter->reset_group && !counter->errored) -continue ; -if (evsel__cpu_iter_skip_no_inc (counter, cpu)) -continue ; -perf_evsel__close_cpu (&counter->core, counter->cpu_iter); - } -/* Now reopen weak */ -evlist__for_each_entry (evsel_list, counter) { -if (!counter->reset_group && !counter->errored) -continue ; -if (evsel__cpu_iter_skip (counter, cpu)) -continue ; -if (!counter->reset_group) -continue ; -try_again_reset: -pr_debug2 ("reopening weak %s\n" ,evsel__name (counter)); -if (create_perf_stat_counter (counter, &stat_config, &target, - counter->cpu_iter -1 ) <0 ) { - -switch (stat_handle_error (counter)) { -case COUNTER_FATAL : -return -1 ; -case COUNTER_RETRY : - goto try_again_reset; -case COUNTER_SKIP : -continue ; -default : -break ; - } - } - counter->supported = true; - } - } - }affinity__cleanup (&affinity);evlist__for_each_entry (evsel_list, counter) {
tools/perf/util/evlist.c +81 -65
195 lines not on this page · raw message
tools/perf/util/evlist.h +43 -7
60 lines not on this page · raw message
tools/perf/util/evsel.h +0 -1
@@ -121,7 +121,6 @@ struct evsel {bool errored;struct hashmap *per_pkg_mask;int err; -int cpu_iter;struct {evsel__sb_cb_t *cb;void *data;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Namhyung Kim
maintainerRe: [PATCH v2 34/48] perf evlist: Refactor evlist__for_each_cpu.
Quoted 11 lines · Ian Rogers
Previously
evlist__for_each_cpuneeded to iterate over the evlist in an inner loop and call "skip" routines. Refactor this so that the iteratr is smarter and the next function can update both the current CPU and evsel.By using a cpu map index, fix apparent off-by-1 in __run_perf_stat's call to
perf_evsel__close_cpu.Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
Quoted 12 lines
tools/perf/util/evlist.c +0
-4@@ -342,36 +342,65 @@ static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel)return perf_thread_map__nr (evlist->core.threads); } -void evlist__cpu_iter_start (struct evlist *evlist) -{ -struct evsel *pos; -
tools/perf/util/evlist.c static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel) · line 345
Quoted 10 lines
tools/perf/util/evlist.c +1
-9-/* -* Reset the per evsel cpu_iter. This is needed because -* each evsel's cpumap may have a different index space, -* and some operations need the index to modify -* the FD xyarray (e.g. open, close) -*/ -evlist__for_each_entry (evlist, pos) - pos->cpu_iter =0 ; -} +struct evlist_cpu_iterator evlist__cpu_begin (struct evlist *evlist,struct affinity *affinity)
I think we can rename it to evlist_cpu_iterator__begin().
Quoted 12 lines
tools/perf/util/evlist.c +10
-1+{ +struct evlist_cpu_iterator itr = { + .container = evlist, + .evsel =evlist__first (evlist), + .cpu_map_idx =0 , + .evlist_cpu_map_idx =0 , + .evlist_cpu_map_nr =perf_cpu_map__nr (evlist->core.all_cpus), + .cpu =-1 , + .affinity = affinity, + }; -bool evsel__cpu_iter_skip_no_inc (struct evsel *ev,int cpu)
tools/perf/util/evlist.c static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel) · line 360
Quoted 10 lines
tools/perf/util/evlist.c +4
-6-{ -if (ev->cpu_iter >= ev->core.cpus->nr) -return true; -if (cpu >=0 && ev->core.cpus->map[ev->cpu_iter] != cpu) -return true; -return false; +if (itr.affinity) { + itr.cpu =perf_cpu_map__cpu (evlist->core.all_cpus,0 ); +affinity__set (itr.affinity, itr.cpu); + itr.cpu_map_idx =perf_cpu_map__idx (itr.evsel->core.cpus, itr.cpu);
This doesn't match to the evlist_cpu_iterator__next() when affinity is NULL. It should set the cpu and cpu_map_idx.
Other than that, I like this! :)
Thanks,
Namhyung
Quoted 240 lines
tools/perf/util/evlist.c +66
-45+/* +* If this CPU isn't in the evsel's cpu map then advance through +* the list. +*/ +if (itr.cpu_map_idx ==-1 ) +evlist_cpu_iterator__next (&itr); + } +return itr; +} + +void evlist_cpu_iterator__next (struct evlist_cpu_iterator *evlist_cpu_itr) +{ +while (evlist_cpu_itr->evsel !=evlist__last (evlist_cpu_itr->container)) { + evlist_cpu_itr->evsel =evsel__next (evlist_cpu_itr->evsel); + evlist_cpu_itr->cpu_map_idx = +perf_cpu_map__idx (evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); +if (evlist_cpu_itr->cpu_map_idx !=-1 ) +return ; + } + evlist_cpu_itr->evlist_cpu_map_idx++; +if (evlist_cpu_itr->evlist_cpu_map_idx < evlist_cpu_itr->evlist_cpu_map_nr) { + evlist_cpu_itr->evsel =evlist__first (evlist_cpu_itr->container); + evlist_cpu_itr->cpu = +perf_cpu_map__cpu (evlist_cpu_itr->container->core.all_cpus, + evlist_cpu_itr->evlist_cpu_map_idx); +if (evlist_cpu_itr->affinity) +affinity__set (evlist_cpu_itr->affinity, evlist_cpu_itr->cpu); + evlist_cpu_itr->cpu_map_idx = +perf_cpu_map__idx (evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); +/* +* If this CPU isn't in the evsel's cpu map then advance through +* the list. +*/ +if (evlist_cpu_itr->cpu_map_idx ==-1 ) +evlist_cpu_iterator__next (evlist_cpu_itr); + } } -bool evsel__cpu_iter_skip (struct evsel *ev,int cpu) +bool evlist_cpu_iterator__end (const struct evlist_cpu_iterator *evlist_cpu_itr) { -if (!evsel__cpu_iter_skip_no_inc (ev, cpu)) { - ev->cpu_iter++; -return false; - } -return true; +return evlist_cpu_itr->evlist_cpu_map_idx >= evlist_cpu_itr->evlist_cpu_map_nr; }static int evsel__strcmp (struct evsel *pos,char *evsel_name)@@ -400,31 +429,26 @@ static int evlist__is_enabled(struct evlist *evlist)static void __evlist__disable (struct evlist *evlist,char *evsel_name) {struct evsel *pos; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i, imm =0 ;bool has_imm = false;if (affinity__setup (&affinity) <0 )return ;/* Disable 'immediate' events last */ -for (imm =0 ; imm <=1 ; imm++) { -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); - -evlist__for_each_entry (evlist, pos) { -if (evsel__strcmp (pos, evsel_name)) -continue ; -if (evsel__cpu_iter_skip (pos, cpu)) -continue ; -if (pos->disabled || !evsel__is_group_leader (pos) || !pos->core.fd) -continue ; -if (pos->immediate) - has_imm = true; -if (pos->immediate != imm) -continue ; -evsel__disable_cpu (pos, pos->cpu_iter -1 ); - } +for (int imm =0 ; imm <=1 ; imm++) { +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { + pos = evlist_cpu_itr.evsel; +if (evsel__strcmp (pos, evsel_name)) +continue ; +if (pos->disabled || !evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +if (pos->immediate) + has_imm = true; +if (pos->immediate != imm) +continue ; +evsel__disable_cpu (pos, evlist_cpu_itr.cpu_map_idx); }if (!has_imm)break ;@@ -462,24 +486,19 @@ void evlist__disable_evsel(struct evlist *evlist, char *evsel_name)static void __evlist__enable (struct evlist *evlist,char *evsel_name) {struct evsel *pos; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i;if (affinity__setup (&affinity) <0 )return ; -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); - -evlist__for_each_entry (evlist, pos) { -if (evsel__strcmp (pos, evsel_name)) -continue ; -if (evsel__cpu_iter_skip (pos, cpu)) -continue ; -if (!evsel__is_group_leader (pos) || !pos->core.fd) -continue ; -evsel__enable_cpu (pos, pos->cpu_iter -1 ); - } +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { + pos = evlist_cpu_itr.evsel; +if (evsel__strcmp (pos, evsel_name)) +continue ; +if (!evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +evsel__enable_cpu (pos, evlist_cpu_itr.cpu_map_idx); }affinity__cleanup (&affinity);evlist__for_each_entry (evlist, pos) {@@ -1264,8 +1283,8 @@ void evlist__set_selected(struct evlist *evlist, struct evsel *evsel)void evlist__close (struct evlist *evlist) {struct evsel *evsel; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i;/* * With perf record core.cpus is usually NULL. @@ -1279,15 +1298,12 @@ void evlist__close(struct evlist *evlist)if (affinity__setup (&affinity) <0 )return ; -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); -evlist__for_each_entry_reverse (evlist, evsel) { -if (evsel__cpu_iter_skip (evsel, cpu)) -continue ; -perf_evsel__close_cpu (&evsel->core, evsel->cpu_iter -1 ); - } +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { +perf_evsel__close_cpu (&evlist_cpu_itr.evsel->core, + evlist_cpu_itr.cpu_map_idx); } +affinity__cleanup (&affinity);evlist__for_each_entry_reverse (evlist, evsel) {perf_evsel__free_fd (&evsel->core);tools/perf/util/evlist.h +43
-7@@ -327,17 +327,53 @@ void evlist__to_front(struct evlist *evlist, struct evsel *move_evsel);#define evlist__for_each_entry_safe (evlist, tmp, evsel) \ __evlist__for_each_entry_safe(&(evlist)->core.entries, tmp, evsel) -#define evlist__for_each_cpu (evlist, index, cpu) \ - evlist__cpu_iter_start(evlist); \ - perf_cpu_map__for_each_cpu (cpu, index, (evlist)->core.all_cpus) +/** Iterator state for evlist__for_each_cpu */ +struct evlist_cpu_iterator { +/** The list being iterated through. */ +struct evlist *container; +/** The current evsel of the iterator. */ +struct evsel *evsel; +/** The CPU map index corresponding to the evsel->core.cpus for the current CPU. */ +int cpu_map_idx; +/** +* The CPU map index corresponding to evlist->core.all_cpus for the +* current CPU. Distinct from cpu_map_idx as the evsel's cpu map may +* contain fewer entries. +*/ +int evlist_cpu_map_idx; +/** The number of CPU map entries in evlist->core.all_cpus. */ +int evlist_cpu_map_nr; +/** The current CPU of the iterator. */ +int cpu; +/** If present, used to set the affinity when switching between CPUs. */ +struct affinity *affinity; +}; + +/** +* evlist__for_each_cpu - without affinity, iterate over the evlist. With +* affinity, iterate over all CPUs and then the evlist +* for each evsel on that CPU. When switching between +* CPUs the affinity is set to the CPU to avoid IPIs +* during syscalls. +* @evlist_cpu_itr: the iterator instance. +* @evlist: evlist instance to iterate. +* @affinity: NULL or used to set the affinity to the current CPU. +*/ +#define evlist__for_each_cpu (evlist_cpu_itr, evlist, affinity) \ + for ((evlist_cpu_itr) = evlist__cpu_begin(evlist, affinity); \ + !evlist_cpu_iterator__end(&evlist_cpu_itr); \ + evlist_cpu_iterator__next(&evlist_cpu_itr)) + +/** Returns an iterator set to the first CPU/evsel of evlist. */ +struct evlist_cpu_iterator evlist__cpu_begin (struct evlist *evlist,struct affinity *affinity); +/** Move to next element in iterator, updating CPU, evsel and the affinity. */ +void evlist_cpu_iterator__next (struct evlist_cpu_iterator *evlist_cpu_itr); +/** Returns true when iterator is at the end of the CPUs and evlist. */ +bool evlist_cpu_iterator__end (const struct evlist_cpu_iterator *evlist_cpu_itr);struct evsel *evlist__get_tracking_event (struct evlist *evlist);void evlist__set_tracking_event (struct evlist *evlist,struct evsel *tracking_evsel); -void evlist__cpu_iter_start (struct evlist *evlist); -bool evsel__cpu_iter_skip (struct evsel *ev,int cpu); -bool evsel__cpu_iter_skip_no_inc (struct evsel *ev,int cpu); -struct evsel *evlist__find_evsel_by_str (struct evlist *evlist,const char *str);struct evsel *evlist__event2evsel (struct evlist *evlist,union perf_event *event);tools/perf/util/evsel.h +0
-2@@ -121,7 +121,6 @@ struct evsel {bool errored;struct hashmap *per_pkg_mask;int err; -int cpu_iter;struct {evsel__sb_cb_t *cb;void *data; --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerRe: [PATCH v2 34/48] perf evlist: Refactor evlist__for_each_cpu.
Quoted 22 lines · Namhyung Kim
Quoted 11 lines · Ian Rogers
Previously
evlist__for_each_cpuneeded to iterate over the evlist in an inner loop and call "skip" routines. Refactor this so that the iteratr is smarter and the next function can update both the current CPU and evsel.By using a cpu map index, fix apparent off-by-1 in __run_perf_stat's call to
perf_evsel__close_cpu.Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
Quoted 8 lines
tools/perf/util/evlist.c +0
-0@@ -342,36 +342,65 @@ static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel)return perf_thread_map__nr (evlist->core.threads); }
Quoted 16 lines
Quoted 4 lines
tools/perf/util/evlist.c +0
-4-void evlist__cpu_iter_start (struct evlist *evlist) -{ -struct evsel *pos; -tools/perf/util/evlist.c static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel) · line 345
Quoted 10 lines
tools/perf/util/evlist.c +1
-9-/* -* Reset the per evsel cpu_iter. This is needed because -* each evsel's cpumap may have a different index space, -* and some operations need the index to modify -* the FD xyarray (e.g. open, close) -*/ -evlist__for_each_entry (evlist, pos) - pos->cpu_iter =0 ; -} +struct evlist_cpu_iterator evlist__cpu_begin (struct evlist *evlist,struct affinity *affinity)I think we can rename it to
evlist_cpu_iterator__begin().
Apologies, I missed this with v3. The begin/end is intended to match how STL will have a begin/end on the thing you foreach loop on.
Quoted 11 lines
Quoted 11 lines
tools/perf/util/evlist.c +10
-0+{ +struct evlist_cpu_iterator itr = { + .container = evlist, + .evsel =evlist__first (evlist), + .cpu_map_idx =0 , + .evlist_cpu_map_idx =0 , + .evlist_cpu_map_nr =perf_cpu_map__nr (evlist->core.all_cpus), + .cpu =-1 , + .affinity = affinity, + };
Quoted 14 lines
tools/perf/util/evlist.c static int evlist__nr_threads(struct evlist *evlist, struct evsel *evsel) · line 360
Quoted 11 lines
tools/perf/util/evlist.c +4
-7-bool evsel__cpu_iter_skip_no_inc (struct evsel *ev,int cpu) -{ -if (ev->cpu_iter >= ev->core.cpus->nr) -return true; -if (cpu >=0 && ev->core.cpus->map[ev->cpu_iter] != cpu) -return true; -return false; +if (itr.affinity) { + itr.cpu =perf_cpu_map__cpu (evlist->core.all_cpus,0 ); +affinity__set (itr.affinity, itr.cpu); + itr.cpu_map_idx =perf_cpu_map__idx (itr.evsel->core.cpus, itr.cpu);This doesn't match to the
evlist_cpu_iterator__next()when affinity is NULL. It should set the cpu andcpu_map_idx.
Agreed this is complex, the pattern is trying to match current usage. When the affinity is NULL in !per-CPU and per-thread mode, the CPU is intentionally always -1 and the cpu maps of the events is expected to just match the dummy (ie an array containing just -1). Next will advance through the evsels keeping the CPU as -1 and the cpu_map_idx as 0. When the end of the evlist is reached in next, the cpu_map_idx will become 1, which is larger than the dummy map nr, and end will become true.
So I think the code is correct here. It could more explicitly state the dummy map assumptions when !affinity. Setting CPU doesn't really make sense for the !per-CPU and per-thread mode case and the map lookup of index 0 is correct here.
Thanks,
Ian
Quoted 246 lines
Other than that, I like this! :)
Thanks,
NamhyungQuoted 240 lines
tools/perf/util/evlist.c +66
-45+/* +* If this CPU isn't in the evsel's cpu map then advance through +* the list. +*/ +if (itr.cpu_map_idx ==-1 ) +evlist_cpu_iterator__next (&itr); + } +return itr; +} + +void evlist_cpu_iterator__next (struct evlist_cpu_iterator *evlist_cpu_itr) +{ +while (evlist_cpu_itr->evsel !=evlist__last (evlist_cpu_itr->container)) { + evlist_cpu_itr->evsel =evsel__next (evlist_cpu_itr->evsel); + evlist_cpu_itr->cpu_map_idx = +perf_cpu_map__idx (evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); +if (evlist_cpu_itr->cpu_map_idx !=-1 ) +return ; + } + evlist_cpu_itr->evlist_cpu_map_idx++; +if (evlist_cpu_itr->evlist_cpu_map_idx < evlist_cpu_itr->evlist_cpu_map_nr) { + evlist_cpu_itr->evsel =evlist__first (evlist_cpu_itr->container); + evlist_cpu_itr->cpu = +perf_cpu_map__cpu (evlist_cpu_itr->container->core.all_cpus, + evlist_cpu_itr->evlist_cpu_map_idx); +if (evlist_cpu_itr->affinity) +affinity__set (evlist_cpu_itr->affinity, evlist_cpu_itr->cpu); + evlist_cpu_itr->cpu_map_idx = +perf_cpu_map__idx (evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); +/* +* If this CPU isn't in the evsel's cpu map then advance through +* the list. +*/ +if (evlist_cpu_itr->cpu_map_idx ==-1 ) +evlist_cpu_iterator__next (evlist_cpu_itr); + } } -bool evsel__cpu_iter_skip (struct evsel *ev,int cpu) +bool evlist_cpu_iterator__end (const struct evlist_cpu_iterator *evlist_cpu_itr) { -if (!evsel__cpu_iter_skip_no_inc (ev, cpu)) { - ev->cpu_iter++; -return false; - } -return true; +return evlist_cpu_itr->evlist_cpu_map_idx >= evlist_cpu_itr->evlist_cpu_map_nr; }static int evsel__strcmp (struct evsel *pos,char *evsel_name)@@ -400,31 +429,26 @@ static int evlist__is_enabled(struct evlist *evlist)static void __evlist__disable (struct evlist *evlist,char *evsel_name) {struct evsel *pos; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i, imm =0 ;bool has_imm = false;if (affinity__setup (&affinity) <0 )return ;/* Disable 'immediate' events last */ -for (imm =0 ; imm <=1 ; imm++) { -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); - -evlist__for_each_entry (evlist, pos) { -if (evsel__strcmp (pos, evsel_name)) -continue ; -if (evsel__cpu_iter_skip (pos, cpu)) -continue ; -if (pos->disabled || !evsel__is_group_leader (pos) || !pos->core.fd) -continue ; -if (pos->immediate) - has_imm = true; -if (pos->immediate != imm) -continue ; -evsel__disable_cpu (pos, pos->cpu_iter -1 ); - } +for (int imm =0 ; imm <=1 ; imm++) { +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { + pos = evlist_cpu_itr.evsel; +if (evsel__strcmp (pos, evsel_name)) +continue ; +if (pos->disabled || !evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +if (pos->immediate) + has_imm = true; +if (pos->immediate != imm) +continue ; +evsel__disable_cpu (pos, evlist_cpu_itr.cpu_map_idx); }if (!has_imm)break ;@@ -462,24 +486,19 @@ void evlist__disable_evsel(struct evlist *evlist, char *evsel_name)static void __evlist__enable (struct evlist *evlist,char *evsel_name) {struct evsel *pos; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i;if (affinity__setup (&affinity) <0 )return ; -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); - -evlist__for_each_entry (evlist, pos) { -if (evsel__strcmp (pos, evsel_name)) -continue ; -if (evsel__cpu_iter_skip (pos, cpu)) -continue ; -if (!evsel__is_group_leader (pos) || !pos->core.fd) -continue ; -evsel__enable_cpu (pos, pos->cpu_iter -1 ); - } +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { + pos = evlist_cpu_itr.evsel; +if (evsel__strcmp (pos, evsel_name)) +continue ; +if (!evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +evsel__enable_cpu (pos, evlist_cpu_itr.cpu_map_idx); }affinity__cleanup (&affinity);evlist__for_each_entry (evlist, pos) {@@ -1264,8 +1283,8 @@ void evlist__set_selected(struct evlist *evlist, struct evsel *evsel)void evlist__close (struct evlist *evlist) {struct evsel *evsel; +struct evlist_cpu_iterator evlist_cpu_itr;struct affinity affinity; -int cpu, i;/* * With perf record core.cpus is usually NULL. @@ -1279,15 +1298,12 @@ void evlist__close(struct evlist *evlist)if (affinity__setup (&affinity) <0 )return ; -evlist__for_each_cpu (evlist, i, cpu) { -affinity__set (&affinity, cpu); -evlist__for_each_entry_reverse (evlist, evsel) { -if (evsel__cpu_iter_skip (evsel, cpu)) -continue ; -perf_evsel__close_cpu (&evsel->core, evsel->cpu_iter -1 ); - } +evlist__for_each_cpu (evlist_cpu_itr, evlist, &affinity) { +perf_evsel__close_cpu (&evlist_cpu_itr.evsel->core, + evlist_cpu_itr.cpu_map_idx); } +affinity__cleanup (&affinity);evlist__for_each_entry_reverse (evlist, evsel) {perf_evsel__free_fd (&evsel->core);tools/perf/util/evlist.h +43
-7@@ -327,17 +327,53 @@ void evlist__to_front(struct evlist *evlist, struct evsel *move_evsel);#define evlist__for_each_entry_safe (evlist, tmp, evsel) \ __evlist__for_each_entry_safe(&(evlist)->core.entries, tmp, evsel) -#define evlist__for_each_cpu (evlist, index, cpu) \ - evlist__cpu_iter_start(evlist); \ - perf_cpu_map__for_each_cpu (cpu, index, (evlist)->core.all_cpus) +/** Iterator state for evlist__for_each_cpu */ +struct evlist_cpu_iterator { +/** The list being iterated through. */ +struct evlist *container; +/** The current evsel of the iterator. */ +struct evsel *evsel; +/** The CPU map index corresponding to the evsel->core.cpus for the current CPU. */ +int cpu_map_idx; +/** +* The CPU map index corresponding to evlist->core.all_cpus for the +* current CPU. Distinct from cpu_map_idx as the evsel's cpu map may +* contain fewer entries. +*/ +int evlist_cpu_map_idx; +/** The number of CPU map entries in evlist->core.all_cpus. */ +int evlist_cpu_map_nr; +/** The current CPU of the iterator. */ +int cpu; +/** If present, used to set the affinity when switching between CPUs. */ +struct affinity *affinity; +}; + +/** +* evlist__for_each_cpu - without affinity, iterate over the evlist. With +* affinity, iterate over all CPUs and then the evlist +* for each evsel on that CPU. When switching between +* CPUs the affinity is set to the CPU to avoid IPIs +* during syscalls. +* @evlist_cpu_itr: the iterator instance. +* @evlist: evlist instance to iterate. +* @affinity: NULL or used to set the affinity to the current CPU. +*/ +#define evlist__for_each_cpu (evlist_cpu_itr, evlist, affinity) \ + for ((evlist_cpu_itr) = evlist__cpu_begin(evlist, affinity); \ + !evlist_cpu_iterator__end(&evlist_cpu_itr); \ + evlist_cpu_iterator__next(&evlist_cpu_itr)) + +/** Returns an iterator set to the first CPU/evsel of evlist. */ +struct evlist_cpu_iterator evlist__cpu_begin (struct evlist *evlist,struct affinity *affinity); +/** Move to next element in iterator, updating CPU, evsel and the affinity. */ +void evlist_cpu_iterator__next (struct evlist_cpu_iterator *evlist_cpu_itr); +/** Returns true when iterator is at the end of the CPUs and evlist. */ +bool evlist_cpu_iterator__end (const struct evlist_cpu_iterator *evlist_cpu_itr);struct evsel *evlist__get_tracking_event (struct evlist *evlist);void evlist__set_tracking_event (struct evlist *evlist,struct evsel *tracking_evsel); -void evlist__cpu_iter_start (struct evlist *evlist); -bool evsel__cpu_iter_skip (struct evsel *ev,int cpu); -bool evsel__cpu_iter_skip_no_inc (struct evsel *ev,int cpu); -struct evsel *evlist__find_evsel_by_str (struct evlist *evlist,const char *str);struct evsel *evlist__event2evsel (struct evlist *evlist,union perf_event *event);tools/perf/util/evsel.h +0
-2@@ -121,7 +121,6 @@ struct evsel {bool errored;struct hashmap *per_pkg_mask;int err; -int cpu_iter;struct {evsel__sb_cb_t *cb;void *data; --2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 35/48 perf evsel: Pass cpu not cpu map index to synthesize
In reply to Ian Rogersevsel__write_stat_event was incorrectly passing a cpu map index rather than a CPU to perf_event__synthesize_stat.
--- tools/perf/builtin-stat.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
tools/perf/builtin-stat.c +3 -2
@@ -327,10 +327,11 @@ static int write_stat_round_event(u64 tm, u64 type)#define SID (e, x, y) xyarray__entry(e->core.sample_id, x, y) -static int evsel__write_stat_event (struct evsel *counter,u32 cpu,u32 thread, +static int evsel__write_stat_event (struct evsel *counter,int cpu_map_idx,u32 thread,struct perf_counts_values *count) { -struct perf_sample_id *sid =SID (counter, cpu, thread); +struct perf_sample_id *sid =SID (counter, cpu_map_idx, thread); +int cpu =perf_cpu_map__cpu (evsel__cpus (counter), cpu_map_idx);return perf_event__synthesize_stat (NULL , cpu, thread, sid->id, count, process_synthesized_event,NULL );
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 36/48 perf stat: Correct variable name for read counter
In reply to Ian RogersSwitch from cpu to cpu_map_idx to reduce confusion.
--- tools/perf/builtin-stat.c | 24 ++++++++++++------------ tools/perf/util/evsel.c | 30 +++++++++++++++--------------- tools/perf/util/evsel.h | 16 ++++++++-------- 3 files changed, 35 insertions(+), 35 deletions(-)
tools/perf/builtin-stat.c +12 -12
@@ -337,25 +337,25 @@ static int evsel__write_stat_event(struct evsel *counter, int cpu_map_idx, u32 t process_synthesized_event, NULL); } -static int read_single_counter (struct evsel *counter,int cpu, +static int read_single_counter (struct evsel *counter,int cpu_map_idx,int thread,struct timespec *rs) {if (counter->tool_event ==PERF_TOOL_DURATION_TIME ) {u64 val = rs->tv_nsec + rs->tv_sec*1000000000ULL ;struct perf_counts_values *count = -perf_counts (counter->counts, cpu, thread); +perf_counts (counter->counts, cpu_map_idx, thread); count->ena = count->run = val; count->val = val;return 0 ; } -return evsel__read_counter (counter, cpu, thread); +return evsel__read_counter (counter, cpu_map_idx, thread); }/* * Read out the results of a single counter: * do not aggregate counts across CPUs in system-wide mode */ -static int read_counter_cpu (struct evsel *counter,struct timespec *rs,int cpu) +static int read_counter_cpu (struct evsel *counter,struct timespec *rs,int cpu_map_idx) {int nthreads =perf_thread_map__nr (evsel_list->core.threads);int thread;
@@ -369,24 +369,24 @@ static int read_counter_cpu(struct evsel *counter, struct timespec *rs, int cpu)for (thread =0 ; thread < nthreads; thread++) {struct perf_counts_values *count; - count =perf_counts (counter->counts, cpu, thread); + count =perf_counts (counter->counts, cpu_map_idx, thread);/* * The leader's group read loads data into its group members * (via evsel__read_counter()) and sets their count->loaded. */ -if (!perf_counts__is_loaded (counter->counts, cpu, thread) && -read_single_counter (counter, cpu, thread, rs)) { +if (!perf_counts__is_loaded (counter->counts, cpu_map_idx, thread) && +read_single_counter (counter, cpu_map_idx, thread, rs)) { counter->counts->scaled =-1 ; -perf_counts (counter->counts, cpu, thread)->ena =0 ; -perf_counts (counter->counts, cpu, thread)->run =0 ; +perf_counts (counter->counts, cpu_map_idx, thread)->ena =0 ; +perf_counts (counter->counts, cpu_map_idx, thread)->run =0 ;return -1 ; } -perf_counts__set_loaded (counter->counts, cpu, thread, false); +perf_counts__set_loaded (counter->counts, cpu_map_idx, thread, false);if (STAT_RECORD ) { -if (evsel__write_stat_event (counter, cpu, thread, count)) { +if (evsel__write_stat_event (counter, cpu_map_idx, thread, count)) {pr_err ("failed to write stat event\n" );return -1 ; }
@@ -396,7 +396,7 @@ static int read_counter_cpu(struct evsel *counter, struct timespec *rs, int cpu)fprintf (stat_config.output,"%s: %d: %" PRIu64" %" PRIu64" %" PRIu64"\n" ,evsel__name (counter), - cpu, +perf_cpu_map__cpu (evsel__cpus (counter), cpu_map_idx), count->val, count->ena, count->run); } }
tools/perf/util/evsel.c +15 -15
@@ -1476,11 +1476,11 @@ void evsel__compute_deltas(struct evsel *evsel, int cpu, int thread, count->run = count->run - tmp.run; } -static int evsel__read_one (struct evsel *evsel,int cpu,int thread) +static int evsel__read_one (struct evsel *evsel,int cpu_map_idx,int thread) { -struct perf_counts_values *count =perf_counts (evsel->counts, cpu, thread); +struct perf_counts_values *count =perf_counts (evsel->counts, cpu_map_idx, thread); -return perf_evsel__read (&evsel->core, cpu, thread, count); +return perf_evsel__read (&evsel->core, cpu_map_idx, thread, count); }static void evsel__set_count (struct evsel *counter,int cpu,int thread,u64 val,u64 ena,u64 run)
@@ -1530,7 +1530,7 @@ static int evsel__process_group_data(struct evsel *leader, int cpu, int thread,return 0 ; } -static int evsel__read_group (struct evsel *leader,int cpu,int thread) +static int evsel__read_group (struct evsel *leader,int cpu_map_idx,int thread) {struct perf_stat_evsel *ps = leader->stats;u64 read_format = leader->core.attr.read_format;
@@ -1551,42 +1551,42 @@ static int evsel__read_group(struct evsel *leader, int cpu, int thread) ps->group_data = data; } -if (FD (leader, cpu, thread) <0 ) +if (FD (leader, cpu_map_idx, thread) <0 )return -EINVAL ; -if (readn (FD (leader, cpu, thread), data, size) <=0 ) +if (readn (FD (leader, cpu_map_idx, thread), data, size) <=0 )return -errno; -return evsel__process_group_data (leader, cpu, thread, data); +return evsel__process_group_data (leader, cpu_map_idx, thread, data); } -int evsel__read_counter (struct evsel *evsel,int cpu,int thread) +int evsel__read_counter (struct evsel *evsel,int cpu_map_idx,int thread) {u64 read_format = evsel->core.attr.read_format;if (read_format &PERF_FORMAT_GROUP ) -return evsel__read_group (evsel, cpu, thread); +return evsel__read_group (evsel, cpu_map_idx, thread); -return evsel__read_one (evsel, cpu, thread); +return evsel__read_one (evsel, cpu_map_idx, thread); } -int __evsel__read_on_cpu (struct evsel *evsel,int cpu,int thread,bool scale) +int __evsel__read_on_cpu (struct evsel *evsel,int cpu_map_idx,int thread,bool scale) {struct perf_counts_values count;size_t nv = scale ?3 :1 ; -if (FD (evsel, cpu, thread) <0 ) +if (FD (evsel, cpu_map_idx, thread) <0 )return -EINVAL ;if (evsel->counts ==NULL &&evsel__alloc_counts (evsel) <0 )return -ENOMEM ; -if (readn (FD (evsel, cpu, thread), &count, nv *sizeof (u64)) <=0 ) +if (readn (FD (evsel, cpu_map_idx, thread), &count, nv *sizeof (u64)) <=0 )return -errno; -evsel__compute_deltas (evsel, cpu, thread, &count); +evsel__compute_deltas (evsel, cpu_map_idx, thread, &count);perf_counts_values__scale (&count, scale,NULL ); - *perf_counts (evsel->counts, cpu, thread) = count; + *perf_counts (evsel->counts, cpu_map_idx, thread) = count;return 0 ; }
tools/perf/util/evsel.h +8 -8
@@ -333,32 +333,32 @@ static inline bool evsel__match2(struct evsel *e1, struct evsel *e2) (e1->core.attr.config == e2->core.attr.config); } -int evsel__read_counter (struct evsel *evsel,int cpu,int thread); +int evsel__read_counter (struct evsel *evsel,int cpu_map_idx,int thread); -int __evsel__read_on_cpu (struct evsel *evsel,int cpu,int thread,bool scale); +int __evsel__read_on_cpu (struct evsel *evsel,int cpu_map_idx,int thread,bool scale);/** * evsel__read_on_cpu - Read out the results on a CPU and thread * * @evsel - event selector to read value -* @cpu - CPU of interest +* @cpu_map_idx - CPU of interest * @thread - thread of interest */ -static inline int evsel__read_on_cpu (struct evsel *evsel,int cpu,int thread) +static inline int evsel__read_on_cpu (struct evsel *evsel,int cpu_map_idx,int thread) { -return __evsel__read_on_cpu (evsel, cpu, thread, false); +return __evsel__read_on_cpu (evsel, cpu_map_idx, thread, false); }/** * evsel__read_on_cpu_scaled - Read out the results on a CPU and thread, scaled * * @evsel - event selector to read value -* @cpu - CPU of interest +* @cpu_map_idx - CPU of interest * @thread - thread of interest */ -static inline int evsel__read_on_cpu_scaled (struct evsel *evsel,int cpu,int thread) +static inline int evsel__read_on_cpu_scaled (struct evsel *evsel,int cpu_map_idx,int thread) { -return __evsel__read_on_cpu (evsel, cpu, thread, true); +return __evsel__read_on_cpu (evsel, cpu_map_idx, thread, true); } int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 37/48 perf evsel: Rename CPU around get_group_fd
In reply to Ian RogersCPU is really a cpu map index, change names to make code more intention revealing.
--- tools/perf/util/evsel.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-)
tools/perf/util/evsel.c +11 -11
@@ -1591,27 +1591,27 @@ int __evsel__read_on_cpu(struct evsel *evsel, int cpu_map_idx, int thread, bool }static int evsel__match_other_cpu (struct evsel *evsel,struct evsel *other, -int cpu) +int cpu_map_idx) { -int cpuid; +int cpu; - cpuid =perf_cpu_map__cpu (evsel->core.cpus, cpu); -return perf_cpu_map__idx (other->core.cpus, cpuid); + cpu =perf_cpu_map__cpu (evsel->core.cpus, cpu_map_idx); +return perf_cpu_map__idx (other->core.cpus, cpu); } -static int evsel__hybrid_group_cpu (struct evsel *evsel,int cpu) +static int evsel__hybrid_group_cpu_map_idx (struct evsel *evsel,int cpu_map_idx) {struct evsel *leader =evsel__leader (evsel);if ((evsel__is_hybrid (evsel) && !evsel__is_hybrid (leader)) || (!evsel__is_hybrid (evsel) &&evsel__is_hybrid (leader))) { -return evsel__match_other_cpu (evsel, leader, cpu); +return evsel__match_other_cpu (evsel, leader, cpu_map_idx); } -return cpu; +return cpu_map_idx; } -static int get_group_fd (struct evsel *evsel,int cpu,int thread) +static int get_group_fd (struct evsel *evsel,int cpu_map_idx,int thread) {struct evsel *leader =evsel__leader (evsel);int fd;
@@ -1625,11 +1625,11 @@ static int get_group_fd(struct evsel *evsel, int cpu, int thread)*/ BUG_ON (!leader->core.fd); - cpu =evsel__hybrid_group_cpu (evsel, cpu); -if (cpu ==-1 ) + cpu_map_idx =evsel__hybrid_group_cpu_map_idx (evsel, cpu_map_idx); +if (cpu_map_idx ==-1 )return -1 ; - fd =FD (leader, cpu, thread); + fd =FD (leader, cpu_map_idx, thread);BUG_ON (fd ==-1 );return fd;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 38/48 perf evsel: Reduce scope of evsel__ignore_missing_thread
In reply to Ian RogersMove to being static.
--- tools/perf/util/evsel.c | 8 ++++---- tools/perf/util/evsel.h | 4 ---- 2 files changed, 4 insertions(+), 8 deletions(-)
tools/perf/util/evsel.c +4 -4
@@ -1666,10 +1666,10 @@ static int update_fds(struct evsel *evsel,return 0 ; } -bool evsel__ignore_missing_thread (struct evsel *evsel, -int nr_cpus,int cpu, -struct perf_thread_map *threads, -int thread,int err) +static bool evsel__ignore_missing_thread (struct evsel *evsel, +int nr_cpus,int cpu, +struct perf_thread_map *threads, +int thread,int err) {pid_t ignore_pid =perf_thread_map__pid (threads, thread);
tools/perf/util/evsel.h +0 -4
@@ -301,10 +301,6 @@ bool evsel__detect_missing_features(struct evsel *evsel);enum rlimit_action {NO_CHANGE ,SET_TO_MAX ,INCREASED_MAX };bool evsel__increase_rlimit (enum rlimit_action *set_rlimit); -bool evsel__ignore_missing_thread (struct evsel *evsel, -int nr_cpus,int cpu, -struct perf_thread_map *threads, -int thread,int err);bool evsel__precise_ip_fallback (struct evsel *evsel);struct perf_sample ;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 39/48 perf evsel: Rename variable cpu to index
In reply to Ian RogersMake naming less error prone.
--- tools/perf/util/evsel.c | 83 +++++++++++++++++++++-------------------- tools/perf/util/evsel.h | 6 +-- tools/perf/util/stat.c | 4 +- tools/perf/util/stat.h | 2 +- 4 files changed, 48 insertions(+), 47 deletions(-)
tools/perf/util/evsel.c +42 -41
@@ -1372,9 +1372,9 @@ int evsel__append_addr_filter(struct evsel *evsel, const char *filter) }/* Caller has to clear disabled after going through all CPUs. */ -int evsel__enable_cpu (struct evsel *evsel,int cpu) +int evsel__enable_cpu (struct evsel *evsel,int cpu_map_idx) { -return perf_evsel__enable_cpu (&evsel->core, cpu); +return perf_evsel__enable_cpu (&evsel->core, cpu_map_idx); }int evsel__enable (struct evsel *evsel)
@@ -1387,9 +1387,9 @@ int evsel__enable(struct evsel *evsel) }/* Caller has to set disabled after going through all CPUs. */ -int evsel__disable_cpu (struct evsel *evsel,int cpu) +int evsel__disable_cpu (struct evsel *evsel,int cpu_map_idx) { -return perf_evsel__disable_cpu (&evsel->core, cpu); +return perf_evsel__disable_cpu (&evsel->core, cpu_map_idx); }int evsel__disable (struct evsel *evsel)
@@ -1455,7 +1455,7 @@ void evsel__delete(struct evsel *evsel)free (evsel); } -void evsel__compute_deltas (struct evsel *evsel,int cpu,int thread, +void evsel__compute_deltas (struct evsel *evsel,int cpu_map_idx,int thread,struct perf_counts_values *count) {struct perf_counts_values tmp;
@@ -1463,12 +1463,12 @@ void evsel__compute_deltas(struct evsel *evsel, int cpu, int thread,if (!evsel->prev_raw_counts)return ; -if (cpu ==-1 ) { +if (cpu_map_idx ==-1 ) { tmp = evsel->prev_raw_counts->aggr; evsel->prev_raw_counts->aggr = *count; }else { - tmp = *perf_counts (evsel->prev_raw_counts, cpu, thread); - *perf_counts (evsel->prev_raw_counts, cpu, thread) = *count; + tmp = *perf_counts (evsel->prev_raw_counts, cpu_map_idx, thread); + *perf_counts (evsel->prev_raw_counts, cpu_map_idx, thread) = *count; } count->val = count->val - tmp.val;
@@ -1483,20 +1483,21 @@ static int evsel__read_one(struct evsel *evsel, int cpu_map_idx, int thread)return perf_evsel__read (&evsel->core, cpu_map_idx, thread, count); } -static void evsel__set_count (struct evsel *counter,int cpu,int thread,u64 val,u64 ena,u64 run) +static void evsel__set_count (struct evsel *counter,int cpu_map_idx,int thread, +u64 val,u64 ena,u64 run) {struct perf_counts_values *count; - count =perf_counts (counter->counts, cpu, thread); + count =perf_counts (counter->counts, cpu_map_idx, thread); count->val = val; count->ena = ena; count->run = run; -perf_counts__set_loaded (counter->counts, cpu, thread, true); +perf_counts__set_loaded (counter->counts, cpu_map_idx, thread, true); } -static int evsel__process_group_data (struct evsel *leader,int cpu,int thread,u64 *data) +static int evsel__process_group_data (struct evsel *leader,int cpu_map_idx,int thread,u64 *data) {u64 read_format = leader->core.attr.read_format;struct sample_read_value *v;
@@ -1515,7 +1516,7 @@ static int evsel__process_group_data(struct evsel *leader, int cpu, int thread, v = (struct sample_read_value *) data; -evsel__set_count (leader, cpu, thread, v[0 ].value, ena, run); +evsel__set_count (leader, cpu_map_idx, thread, v[0 ].value, ena, run);for (i =1 ; i < nr; i++) {struct evsel *counter;
@@ -1524,7 +1525,7 @@ static int evsel__process_group_data(struct evsel *leader, int cpu, int thread,if (!counter)return -EINVAL ; -evsel__set_count (counter, cpu, thread, v[i].value, ena, run); +evsel__set_count (counter, cpu_map_idx, thread, v[i].value, ena, run); }return 0 ;
@@ -1643,16 +1644,16 @@ static void evsel__remove_fd(struct evsel *pos, int nr_cpus, int nr_threads, int }static int update_fds (struct evsel *evsel, -int nr_cpus,int cpu_idx, +int nr_cpus,int cpu_map_idx,int nr_threads,int thread_idx) {struct evsel *pos; -if (cpu_idx >= nr_cpus || thread_idx >= nr_threads) +if (cpu_map_idx >= nr_cpus || thread_idx >= nr_threads)return -EINVAL ;evlist__for_each_entry (evsel->evlist, pos) { - nr_cpus = pos != evsel ? nr_cpus : cpu_idx; + nr_cpus = pos != evsel ? nr_cpus : cpu_map_idx;evsel__remove_fd (pos, nr_cpus, nr_threads, thread_idx);
@@ -1667,7 +1668,7 @@ static int update_fds(struct evsel *evsel, }static bool evsel__ignore_missing_thread (struct evsel *evsel, -int nr_cpus,int cpu, +int nr_cpus,int cpu_map_idx,struct perf_thread_map *threads,int thread,int err) {
@@ -1692,7 +1693,7 @@ static bool evsel__ignore_missing_thread(struct evsel *evsel,* We should remove fd for missing_thread first * because thread_map__remove() will decrease threads->nr. */ -if (update_fds (evsel, nr_cpus, cpu, threads->nr, thread)) +if (update_fds (evsel, nr_cpus, cpu_map_idx, threads->nr, thread))return false;if (thread_map__remove (threads, thread))
@@ -1974,9 +1975,9 @@ bool evsel__increase_rlimit(enum rlimit_action *set_rlimit)static int evsel__open_cpu (struct evsel *evsel,struct perf_cpu_map *cpus,struct perf_thread_map *threads, -int start_cpu,int end_cpu) +int start_cpu_map_idx,int end_cpu_map_idx) { -int cpu, thread, nthreads; +int idx, thread, nthreads;int pid =-1 , err, old_errno;enum rlimit_action set_rlimit =NO_CHANGE ;
@@ -2003,7 +2004,7 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,display_attr (&evsel->core.attr); -for (cpu = start_cpu; cpu < end_cpu; cpu++) { +for (idx = start_cpu_map_idx; idx < end_cpu_map_idx; idx++) {for (thread =0 ; thread < nthreads; thread++) {int fd, group_fd;
@@ -2014,17 +2015,17 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,if (!evsel->cgrp && !evsel->core.system_wide) pid =perf_thread_map__pid (threads, thread); - group_fd =get_group_fd (evsel, cpu, thread); + group_fd =get_group_fd (evsel, idx, thread);test_attr__ready ();pr_debug2_peo ("sys_perf_event_open: pid %d cpu %d group_fd %d flags %#lx" , - pid, cpus->map[cpu], group_fd, evsel->open_flags); + pid, cpus->map[idx], group_fd, evsel->open_flags); - fd =sys_perf_event_open (&evsel->core.attr, pid, cpus->map[cpu], + fd =sys_perf_event_open (&evsel->core.attr, pid, cpus->map[idx], group_fd, evsel->open_flags); -FD (evsel, cpu, thread) = fd; +FD (evsel, idx, thread) = fd;if (fd <0 ) { err = -errno;
@@ -2034,10 +2035,10 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus, goto try_fallback; } -bpf_counter__install_pe (evsel, cpu, fd); +bpf_counter__install_pe (evsel, idx, fd);if (unlikely (test_attr__enabled)) { -test_attr__open (&evsel->core.attr, pid, cpus->map[cpu], +test_attr__open (&evsel->core.attr, pid, cpus->map[idx], fd, group_fd, evsel->open_flags); }
@@ -2078,7 +2079,7 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,if (evsel__precise_ip_fallback (evsel)) goto retry_open; -if (evsel__ignore_missing_thread (evsel, cpus->nr, cpu, threads, thread, err)) { +if (evsel__ignore_missing_thread (evsel, cpus->nr, idx, threads, thread, err)) {/* We just removed 1 thread, so lower the upper nthreads limit. */ nthreads--;
@@ -2093,7 +2094,7 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,if (err == -EMFILE &&evsel__increase_rlimit (&set_rlimit)) goto retry_open; -if (err != -EINVAL || cpu >0 || thread >0 ) +if (err != -EINVAL || idx >0 || thread >0 ) goto out_close;if (evsel__detect_missing_features (evsel))
@@ -2105,12 +2106,12 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus, old_errno = errno;do {while (--thread >=0 ) { -if (FD (evsel, cpu, thread) >=0 ) -close (FD (evsel, cpu, thread)); -FD (evsel, cpu, thread) =-1 ; +if (FD (evsel, idx, thread) >=0 ) +close (FD (evsel, idx, thread)); +FD (evsel, idx, thread) =-1 ; } thread = nthreads; - }while (--cpu >=0 ); + }while (--idx >=0 ); errno = old_errno;return err; }
@@ -2127,13 +2128,13 @@ void evsel__close(struct evsel *evsel)perf_evsel__free_id (&evsel->core); } -int evsel__open_per_cpu (struct evsel *evsel,struct perf_cpu_map *cpus,int cpu) +int evsel__open_per_cpu (struct evsel *evsel,struct perf_cpu_map *cpus,int cpu_map_idx) { -if (cpu ==-1 ) +if (cpu_map_idx ==-1 )return evsel__open_cpu (evsel, cpus,NULL ,0 , cpus ? cpus->nr :1 ); -return evsel__open_cpu (evsel, cpus,NULL , cpu, cpu +1 ); +return evsel__open_cpu (evsel, cpus,NULL , cpu_map_idx, cpu_map_idx +1 ); }int evsel__open_per_thread (struct evsel *evsel,struct perf_thread_map *threads)
@@ -2954,15 +2955,15 @@ struct perf_env *evsel__env(struct evsel *evsel)static int store_evsel_ids (struct evsel *evsel,struct evlist *evlist) { -int cpu, thread; +int cpu_map_idx, thread; -for (cpu =0 ; cpu <xyarray__max_x (evsel->core.fd); cpu++) { +for (cpu_map_idx =0 ; cpu_map_idx <xyarray__max_x (evsel->core.fd); cpu_map_idx++) {for (thread =0 ; thread <xyarray__max_y (evsel->core.fd); thread++) { -int fd =FD (evsel, cpu, thread); +int fd =FD (evsel, cpu_map_idx, thread);if (perf_evlist__id_add_fd (&evlist->core, &evsel->core, - cpu, thread, fd) <0 ) + cpu_map_idx, thread, fd) <0 )return -1 ; } }
tools/perf/util/evsel.h +3 -3
@@ -284,12 +284,12 @@ void arch_evsel__fixup_new_cycles(struct perf_event_attr *attr);int evsel__set_filter (struct evsel *evsel,const char *filter);int evsel__append_tp_filter (struct evsel *evsel,const char *filter);int evsel__append_addr_filter (struct evsel *evsel,const char *filter); -int evsel__enable_cpu (struct evsel *evsel,int cpu); +int evsel__enable_cpu (struct evsel *evsel,int cpu_map_idx);int evsel__enable (struct evsel *evsel);int evsel__disable (struct evsel *evsel); -int evsel__disable_cpu (struct evsel *evsel,int cpu); +int evsel__disable_cpu (struct evsel *evsel,int cpu_map_idx); -int evsel__open_per_cpu (struct evsel *evsel,struct perf_cpu_map *cpus,int cpu); +int evsel__open_per_cpu (struct evsel *evsel,struct perf_cpu_map *cpus,int cpu_map_idx);int evsel__open_per_thread (struct evsel *evsel,struct perf_thread_map *threads);int evsel__open (struct evsel *evsel,struct perf_cpu_map *cpus,struct perf_thread_map *threads);
tools/perf/util/stat.c +2 -2
@@ -531,7 +531,7 @@ size_t perf_event__fprintf_stat_config(union perf_event *event, FILE *fp)int create_perf_stat_counter (struct evsel *evsel,struct perf_stat_config *config,struct target *target, -int cpu) +int cpu_map_idx) {struct perf_event_attr *attr = &evsel->core.attr;struct evsel *leader =evsel__leader (evsel);
@@ -585,7 +585,7 @@ int create_perf_stat_counter(struct evsel *evsel, }if (target__has_cpu (target) && !target__has_per_thread (target)) -return evsel__open_per_cpu (evsel,evsel__cpus (evsel), cpu); +return evsel__open_per_cpu (evsel,evsel__cpus (evsel), cpu_map_idx);return evsel__open_per_thread (evsel, evsel->core.threads); }
tools/perf/util/stat.h +1 -1
@@ -248,7 +248,7 @@ size_t perf_event__fprintf_stat_config(union perf_event *event, FILE *fp);int create_perf_stat_counter (struct evsel *evsel,struct perf_stat_config *config,struct target *target, -int cpu); +int cpu_map_idx);void evlist__print_counters (struct evlist *evlist,struct perf_stat_config *config,struct target *_target,struct timespec *ts,int argc,const char **argv);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 40/48 perf test: Use perf_cpu_map__for_each_cpu
In reply to Ian RogersClean up variable naming to make cpu and index clearer.
--- tools/perf/tests/openat-syscall-all-cpus.c | 28 +++++++++++----------- 1 file changed, 14 insertions(+), 14 deletions(-)
tools/perf/tests/openat-syscall-all-cpus.c +14 -14
@@ -22,7 +22,7 @@static int test__openat_syscall_event_on_all_cpus (struct test_suite *test__maybe_unused ,int subtest__maybe_unused ) { -int err =-1 , fd, cpu; +int err =-1 , fd, idx, cpu;struct perf_cpu_map *cpus;struct evsel *evsel;unsigned int nr_openat_calls =111 , i;
@@ -58,23 +58,23 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb goto out_evsel_delete; } -for (cpu =0 ; cpu < cpus->nr; ++cpu) { -unsigned int ncalls = nr_openat_calls + cpu; +perf_cpu_map__for_each_cpu (cpu, idx, cpus) { +unsigned int ncalls = nr_openat_calls + idx;/* * XXX eventually lift this restriction in a way that * keeps perf building on older glibc installations * without CPU_ALLOC. 1024 cpus in 2010 still seems * a reasonable upper limit tho :-) */ -if (cpus->map[cpu] >=CPU_SETSIZE ) { -pr_debug ("Ignoring CPU %d\n" , cpus->map[cpu]); +if (cpu >=CPU_SETSIZE ) { +pr_debug ("Ignoring CPU %d\n" , cpu);continue ; } -CPU_SET (cpus->map[cpu], &cpu_set); +CPU_SET (cpu, &cpu_set);if (sched_setaffinity (0 ,sizeof (cpu_set), &cpu_set) <0 ) {pr_debug ("sched_setaffinity() failed on CPU %d: %s " , - cpus->map[cpu], + cpu,str_error_r (errno, sbuf,sizeof (sbuf))); goto out_close_fd; }
@@ -82,29 +82,29 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb fd =openat (0 ,"/etc/passwd" ,O_RDONLY );close (fd); } -CPU_CLR (cpus->map[cpu], &cpu_set); +CPU_CLR (cpu, &cpu_set); } evsel->core.cpus =perf_cpu_map__get (cpus); err =0 ; -for (cpu =0 ; cpu < cpus->nr; ++cpu) { +perf_cpu_map__for_each_cpu (cpu, idx, cpus) {unsigned int expected; -if (cpus->map[cpu] >=CPU_SETSIZE ) +if (cpu >=CPU_SETSIZE )continue ; -if (evsel__read_on_cpu (evsel, cpu,0 ) <0 ) { +if (evsel__read_on_cpu (evsel, idx,0 ) <0 ) {pr_debug ("evsel__read_on_cpu\n" ); err =-1 ;break ; } - expected = nr_openat_calls + cpu; -if (perf_counts (evsel->counts, cpu,0 )->val != expected) { + expected = nr_openat_calls + idx; +if (perf_counts (evsel->counts, idx,0 )->val != expected) {pr_debug ("evsel__read_on_cpu: expected to intercept %d calls on cpu %d, got %" PRIu64"\n" , - expected, cpus->map[cpu],perf_counts (evsel->counts, cpu,0 )->val); + expected, cpu,perf_counts (evsel->counts, idx,0 )->val); err =-1 ; } }
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 41/48 perf stat: Correct check_per_pkg cpu
In reply to Ian RogersCode was incorrectly using the cpu map index as the CPU.
--- tools/perf/util/stat.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
tools/perf/util/stat.c +3 -2
@@ -292,11 +292,12 @@ static bool pkg_id_equal(const void *__key1, const void *__key2,return *key1 == *key2; } -static int check_per_pkg (struct evsel *counter, -struct perf_counts_values *vals,int cpu,bool *skip) +static int check_per_pkg (struct evsel *counter,struct perf_counts_values *vals, +int cpu_map_idx,bool *skip) {struct hashmap *mask = counter->per_pkg_mask;struct perf_cpu_map *cpus =evsel__cpus (counter); +int cpu =perf_cpu_map__cpu (cpus, cpu_map_idx);int s, d, ret =0 ;uint64_t *key;
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 42/48 perf stat: Swap variable name cpu to index
In reply to Ian RogersThe use of CPU is error prone, switch to cpu_map_idx.
--- tools/perf/util/stat-shadow.c | 308 +++++++++++++++++----------------- tools/perf/util/stat.c | 16 +- tools/perf/util/stat.h | 4 +- 3 files changed, 164 insertions(+), 164 deletions(-)
tools/perf/util/stat-shadow.c +154 -154
@@ -32,7 +32,7 @@ struct saved_value {struct evsel *evsel;enum stat_type type;int ctx; -int cpu; +int cpu_map_idx;struct cgroup *cgrp;struct runtime_stat *stat;struct stats stats;
@@ -47,8 +47,8 @@ static int saved_value_cmp(struct rb_node *rb_node, const void *entry) rb_node);const struct saved_value *b = entry; -if (a->cpu != b->cpu) -return a->cpu - b->cpu; +if (a->cpu_map_idx != b->cpu_map_idx) +return a->cpu_map_idx - b->cpu_map_idx;/* * Previously the rbtree was used to link generic metrics.
@@ -105,7 +105,7 @@ static void saved_value_delete(struct rblist *rblist __maybe_unused, } static struct saved_value *saved_value_lookup(struct evsel *evsel, - int cpu, + int cpu_map_idx, bool create, enum stat_type type, int ctx,
@@ -115,7 +115,7 @@ static struct saved_value *saved_value_lookup(struct evsel *evsel,struct rblist *rblist;struct rb_node *nd;struct saved_value dm = { - .cpu = cpu, + .cpu_map_idx = cpu_map_idx, .evsel = evsel, .type = type, .ctx = ctx,
@@ -213,10 +213,10 @@ struct runtime_stat_data {static void update_runtime_stat (struct runtime_stat *st,enum stat_type type, -int cpu,u64 count, +int cpu_map_idx,u64 count,struct runtime_stat_data *rsd) { -struct saved_value *v =saved_value_lookup (NULL , cpu, true, type, +struct saved_value *v =saved_value_lookup (NULL , cpu_map_idx, true, type, rsd->ctx, st, rsd->cgrp);if (v)
@@ -229,7 +229,7 @@ static void update_runtime_stat(struct runtime_stat *st,* instruction rates, etc: */ void perf_stat__update_shadow_stats (struct evsel *counter,u64 count, -int cpu,struct runtime_stat *st) +int cpu_map_idx,struct runtime_stat *st) {u64 count_ns = count;struct saved_value *v;
@@ -241,88 +241,88 @@ void perf_stat__update_shadow_stats(struct evsel *counter, u64 count, count *= counter->scale;if (evsel__is_clock (counter)) -update_runtime_stat (st,STAT_NSECS , cpu, count_ns, &rsd); +update_runtime_stat (st,STAT_NSECS , cpu_map_idx, count_ns, &rsd);else if (evsel__match (counter,HARDWARE ,HW_CPU_CYCLES )) -update_runtime_stat (st,STAT_CYCLES , cpu, count, &rsd); +update_runtime_stat (st,STAT_CYCLES , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,CYCLES_IN_TX )) -update_runtime_stat (st,STAT_CYCLES_IN_TX , cpu, count, &rsd); +update_runtime_stat (st,STAT_CYCLES_IN_TX , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TRANSACTION_START )) -update_runtime_stat (st,STAT_TRANSACTION , cpu, count, &rsd); +update_runtime_stat (st,STAT_TRANSACTION , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,ELISION_START )) -update_runtime_stat (st,STAT_ELISION , cpu, count, &rsd); +update_runtime_stat (st,STAT_ELISION , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_TOTAL_SLOTS ))update_runtime_stat (st,STAT_TOPDOWN_TOTAL_SLOTS , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_SLOTS_ISSUED ))update_runtime_stat (st,STAT_TOPDOWN_SLOTS_ISSUED , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_SLOTS_RETIRED ))update_runtime_stat (st,STAT_TOPDOWN_SLOTS_RETIRED , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_FETCH_BUBBLES ))update_runtime_stat (st,STAT_TOPDOWN_FETCH_BUBBLES , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_RECOVERY_BUBBLES ))update_runtime_stat (st,STAT_TOPDOWN_RECOVERY_BUBBLES , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_RETIRING ))update_runtime_stat (st,STAT_TOPDOWN_RETIRING , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_BAD_SPEC ))update_runtime_stat (st,STAT_TOPDOWN_BAD_SPEC , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_FE_BOUND ))update_runtime_stat (st,STAT_TOPDOWN_FE_BOUND , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_BE_BOUND ))update_runtime_stat (st,STAT_TOPDOWN_BE_BOUND , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_HEAVY_OPS ))update_runtime_stat (st,STAT_TOPDOWN_HEAVY_OPS , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_BR_MISPREDICT ))update_runtime_stat (st,STAT_TOPDOWN_BR_MISPREDICT , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_FETCH_LAT ))update_runtime_stat (st,STAT_TOPDOWN_FETCH_LAT , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,TOPDOWN_MEM_BOUND ))update_runtime_stat (st,STAT_TOPDOWN_MEM_BOUND , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (evsel__match (counter,HARDWARE ,HW_STALLED_CYCLES_FRONTEND ))update_runtime_stat (st,STAT_STALLED_CYCLES_FRONT , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (evsel__match (counter,HARDWARE ,HW_STALLED_CYCLES_BACKEND ))update_runtime_stat (st,STAT_STALLED_CYCLES_BACK , - cpu, count, &rsd); + cpu_map_idx, count, &rsd);else if (evsel__match (counter,HARDWARE ,HW_BRANCH_INSTRUCTIONS )) -update_runtime_stat (st,STAT_BRANCHES , cpu, count, &rsd); +update_runtime_stat (st,STAT_BRANCHES , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HARDWARE ,HW_CACHE_REFERENCES )) -update_runtime_stat (st,STAT_CACHEREFS , cpu, count, &rsd); +update_runtime_stat (st,STAT_CACHEREFS , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HW_CACHE ,HW_CACHE_L1D )) -update_runtime_stat (st,STAT_L1_DCACHE , cpu, count, &rsd); +update_runtime_stat (st,STAT_L1_DCACHE , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HW_CACHE ,HW_CACHE_L1I )) -update_runtime_stat (st,STAT_L1_ICACHE , cpu, count, &rsd); +update_runtime_stat (st,STAT_L1_ICACHE , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HW_CACHE ,HW_CACHE_LL )) -update_runtime_stat (st,STAT_LL_CACHE , cpu, count, &rsd); +update_runtime_stat (st,STAT_LL_CACHE , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HW_CACHE ,HW_CACHE_DTLB )) -update_runtime_stat (st,STAT_DTLB_CACHE , cpu, count, &rsd); +update_runtime_stat (st,STAT_DTLB_CACHE , cpu_map_idx, count, &rsd);else if (evsel__match (counter,HW_CACHE ,HW_CACHE_ITLB )) -update_runtime_stat (st,STAT_ITLB_CACHE , cpu, count, &rsd); +update_runtime_stat (st,STAT_ITLB_CACHE , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,SMI_NUM )) -update_runtime_stat (st,STAT_SMI_NUM , cpu, count, &rsd); +update_runtime_stat (st,STAT_SMI_NUM , cpu_map_idx, count, &rsd);else if (perf_stat_evsel__is (counter,APERF )) -update_runtime_stat (st,STAT_APERF , cpu, count, &rsd); +update_runtime_stat (st,STAT_APERF , cpu_map_idx, count, &rsd);if (counter->collect_stat) { - v =saved_value_lookup (counter, cpu, true,STAT_NONE ,0 , st, + v =saved_value_lookup (counter, cpu_map_idx, true,STAT_NONE ,0 , st, rsd.cgrp);update_stats (&v->stats, count);if (counter->metric_leader) v->metric_total += count; }else if (counter->metric_leader) { v =saved_value_lookup (counter->metric_leader, - cpu, true,STAT_NONE ,0 , st, rsd.cgrp); + cpu_map_idx, true,STAT_NONE ,0 , st, rsd.cgrp); v->metric_total += count; v->metric_other++; }
@@ -464,12 +464,12 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list) }static double runtime_stat_avg (struct runtime_stat *st, -enum stat_type type,int cpu, +enum stat_type type,int cpu_map_idx,struct runtime_stat_data *rsd) {struct saved_value *v; - v =saved_value_lookup (NULL , cpu, false, type, rsd->ctx, st, rsd->cgrp); + v =saved_value_lookup (NULL , cpu_map_idx, false, type, rsd->ctx, st, rsd->cgrp);if (!v)return 0.0 ;
@@ -477,12 +477,12 @@ static double runtime_stat_avg(struct runtime_stat *st, }static double runtime_stat_n (struct runtime_stat *st, -enum stat_type type,int cpu, +enum stat_type type,int cpu_map_idx,struct runtime_stat_data *rsd) {struct saved_value *v; - v =saved_value_lookup (NULL , cpu, false, type, rsd->ctx, st, rsd->cgrp); + v =saved_value_lookup (NULL , cpu_map_idx, false, type, rsd->ctx, st, rsd->cgrp);if (!v)return 0.0 ;
@@ -490,7 +490,7 @@ static double runtime_stat_n(struct runtime_stat *st, }static void print_stalled_cycles_frontend (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -498,7 +498,7 @@ static void print_stalled_cycles_frontend(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_CYCLES , cpu, rsd); + total =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -513,7 +513,7 @@ static void print_stalled_cycles_frontend(struct perf_stat_config *config, }static void print_stalled_cycles_backend (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -521,7 +521,7 @@ static void print_stalled_cycles_backend(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_CYCLES , cpu, rsd); + total =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -532,7 +532,7 @@ static void print_stalled_cycles_backend(struct perf_stat_config *config, }static void print_branch_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -540,7 +540,7 @@ static void print_branch_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_BRANCHES , cpu, rsd); + total =runtime_stat_avg (st,STAT_BRANCHES , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -551,7 +551,7 @@ static void print_branch_misses(struct perf_stat_config *config, }static void print_l1_dcache_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -559,7 +559,7 @@ static void print_l1_dcache_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_L1_DCACHE , cpu, rsd); + total =runtime_stat_avg (st,STAT_L1_DCACHE , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -570,7 +570,7 @@ static void print_l1_dcache_misses(struct perf_stat_config *config, }static void print_l1_icache_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -578,7 +578,7 @@ static void print_l1_icache_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_L1_ICACHE , cpu, rsd); + total =runtime_stat_avg (st,STAT_L1_ICACHE , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -588,7 +588,7 @@ static void print_l1_icache_misses(struct perf_stat_config *config, }static void print_dtlb_cache_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -596,7 +596,7 @@ static void print_dtlb_cache_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_DTLB_CACHE , cpu, rsd); + total =runtime_stat_avg (st,STAT_DTLB_CACHE , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -606,7 +606,7 @@ static void print_dtlb_cache_misses(struct perf_stat_config *config, }static void print_itlb_cache_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -614,7 +614,7 @@ static void print_itlb_cache_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_ITLB_CACHE , cpu, rsd); + total =runtime_stat_avg (st,STAT_ITLB_CACHE , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -624,7 +624,7 @@ static void print_itlb_cache_misses(struct perf_stat_config *config, }static void print_ll_cache_misses (struct perf_stat_config *config, -int cpu,double avg, +int cpu_map_idx,double avg,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -632,7 +632,7 @@ static void print_ll_cache_misses(struct perf_stat_config *config,double total, ratio =0.0 ;const char *color; - total =runtime_stat_avg (st,STAT_LL_CACHE , cpu, rsd); + total =runtime_stat_avg (st,STAT_LL_CACHE , cpu_map_idx, rsd);if (total) ratio = avg / total *100.0 ;
@@ -690,61 +690,61 @@ static double sanitize_val(double x)return x; } -static double td_total_slots (int cpu,struct runtime_stat *st, +static double td_total_slots (int cpu_map_idx,struct runtime_stat *st,struct runtime_stat_data *rsd) { -return runtime_stat_avg (st,STAT_TOPDOWN_TOTAL_SLOTS , cpu, rsd); +return runtime_stat_avg (st,STAT_TOPDOWN_TOTAL_SLOTS , cpu_map_idx, rsd); } -static double td_bad_spec (int cpu,struct runtime_stat *st, +static double td_bad_spec (int cpu_map_idx,struct runtime_stat *st,struct runtime_stat_data *rsd) {double bad_spec =0 ;double total_slots;double total; - total =runtime_stat_avg (st,STAT_TOPDOWN_SLOTS_ISSUED , cpu, rsd) - -runtime_stat_avg (st,STAT_TOPDOWN_SLOTS_RETIRED , cpu, rsd) + -runtime_stat_avg (st,STAT_TOPDOWN_RECOVERY_BUBBLES , cpu, rsd); + total =runtime_stat_avg (st,STAT_TOPDOWN_SLOTS_ISSUED , cpu_map_idx, rsd) - +runtime_stat_avg (st,STAT_TOPDOWN_SLOTS_RETIRED , cpu_map_idx, rsd) + +runtime_stat_avg (st,STAT_TOPDOWN_RECOVERY_BUBBLES , cpu_map_idx, rsd); - total_slots =td_total_slots (cpu, st, rsd); + total_slots =td_total_slots (cpu_map_idx, st, rsd);if (total_slots) bad_spec = total / total_slots;return sanitize_val (bad_spec); } -static double td_retiring (int cpu,struct runtime_stat *st, +static double td_retiring (int cpu_map_idx,struct runtime_stat *st,struct runtime_stat_data *rsd) {double retiring =0 ; -double total_slots =td_total_slots (cpu, st, rsd); +double total_slots =td_total_slots (cpu_map_idx, st, rsd);double ret_slots =runtime_stat_avg (st,STAT_TOPDOWN_SLOTS_RETIRED , - cpu, rsd); + cpu_map_idx, rsd);if (total_slots) retiring = ret_slots / total_slots;return retiring; } -static double td_fe_bound (int cpu,struct runtime_stat *st, +static double td_fe_bound (int cpu_map_idx,struct runtime_stat *st,struct runtime_stat_data *rsd) {double fe_bound =0 ; -double total_slots =td_total_slots (cpu, st, rsd); +double total_slots =td_total_slots (cpu_map_idx, st, rsd);double fetch_bub =runtime_stat_avg (st,STAT_TOPDOWN_FETCH_BUBBLES , - cpu, rsd); + cpu_map_idx, rsd);if (total_slots) fe_bound = fetch_bub / total_slots;return fe_bound; } -static double td_be_bound (int cpu,struct runtime_stat *st, +static double td_be_bound (int cpu_map_idx,struct runtime_stat *st,struct runtime_stat_data *rsd) { -double sum = (td_fe_bound (cpu, st, rsd) + -td_bad_spec (cpu, st, rsd) + -td_retiring (cpu, st, rsd)); +double sum = (td_fe_bound (cpu_map_idx, st, rsd) + +td_bad_spec (cpu_map_idx, st, rsd) + +td_retiring (cpu_map_idx, st, rsd));if (sum ==0 )return 0 ;return sanitize_val (1.0 - sum);
@@ -755,15 +755,15 @@ static double td_be_bound(int cpu, struct runtime_stat *st,* the ratios we need to recreate the sum. */ -static double td_metric_ratio (int cpu,enum stat_type type, +static double td_metric_ratio (int cpu_map_idx,enum stat_type type,struct runtime_stat *stat,struct runtime_stat_data *rsd) { -double sum =runtime_stat_avg (stat,STAT_TOPDOWN_RETIRING , cpu, rsd) + -runtime_stat_avg (stat,STAT_TOPDOWN_FE_BOUND , cpu, rsd) + -runtime_stat_avg (stat,STAT_TOPDOWN_BE_BOUND , cpu, rsd) + -runtime_stat_avg (stat,STAT_TOPDOWN_BAD_SPEC , cpu, rsd); -double d =runtime_stat_avg (stat, type, cpu, rsd); +double sum =runtime_stat_avg (stat,STAT_TOPDOWN_RETIRING , cpu_map_idx, rsd) + +runtime_stat_avg (stat,STAT_TOPDOWN_FE_BOUND , cpu_map_idx, rsd) + +runtime_stat_avg (stat,STAT_TOPDOWN_BE_BOUND , cpu_map_idx, rsd) + +runtime_stat_avg (stat,STAT_TOPDOWN_BAD_SPEC , cpu_map_idx, rsd); +double d =runtime_stat_avg (stat, type, cpu_map_idx, rsd);if (sum)return d / sum;
@@ -775,23 +775,23 @@ static double td_metric_ratio(int cpu, enum stat_type type,* We allow two missing. */ -static bool full_td (int cpu,struct runtime_stat *stat, +static bool full_td (int cpu_map_idx,struct runtime_stat *stat,struct runtime_stat_data *rsd) {int c =0 ; -if (runtime_stat_avg (stat,STAT_TOPDOWN_RETIRING , cpu, rsd) >0 ) +if (runtime_stat_avg (stat,STAT_TOPDOWN_RETIRING , cpu_map_idx, rsd) >0 ) c++; -if (runtime_stat_avg (stat,STAT_TOPDOWN_BE_BOUND , cpu, rsd) >0 ) +if (runtime_stat_avg (stat,STAT_TOPDOWN_BE_BOUND , cpu_map_idx, rsd) >0 ) c++; -if (runtime_stat_avg (stat,STAT_TOPDOWN_FE_BOUND , cpu, rsd) >0 ) +if (runtime_stat_avg (stat,STAT_TOPDOWN_FE_BOUND , cpu_map_idx, rsd) >0 ) c++; -if (runtime_stat_avg (stat,STAT_TOPDOWN_BAD_SPEC , cpu, rsd) >0 ) +if (runtime_stat_avg (stat,STAT_TOPDOWN_BAD_SPEC , cpu_map_idx, rsd) >0 ) c++;return c >=2 ; } -static void print_smi_cost (struct perf_stat_config *config,int cpu, +static void print_smi_cost (struct perf_stat_config *config,int cpu_map_idx,struct perf_stat_output_ctx *out,struct runtime_stat *st,struct runtime_stat_data *rsd)
@@ -799,9 +799,9 @@ static void print_smi_cost(struct perf_stat_config *config, int cpu,double smi_num, aperf, cycles, cost =0.0 ;const char *color =NULL ; - smi_num =runtime_stat_avg (st,STAT_SMI_NUM , cpu, rsd); - aperf =runtime_stat_avg (st,STAT_APERF , cpu, rsd); - cycles =runtime_stat_avg (st,STAT_CYCLES , cpu, rsd); + smi_num =runtime_stat_avg (st,STAT_SMI_NUM , cpu_map_idx, rsd); + aperf =runtime_stat_avg (st,STAT_APERF , cpu_map_idx, rsd); + cycles =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, rsd);if ((cycles ==0 ) || (aperf ==0 ))return ;
@@ -818,7 +818,7 @@ static void print_smi_cost(struct perf_stat_config *config, int cpu,static int prepare_metric (struct evsel **metric_events,struct metric_ref *metric_refs,struct expr_parse_ctx *pctx, -int cpu, +int cpu_map_idx,struct runtime_stat *st) {double scale;
@@ -836,7 +836,7 @@ static int prepare_metric(struct evsel **metric_events, scale =1e-9 ; source_count =1 ; }else { - v =saved_value_lookup (metric_events[i], cpu, false, + v =saved_value_lookup (metric_events[i], cpu_map_idx, false,STAT_NONE ,0 , st, metric_events[i]->cgrp);if (!v)
@@ -874,7 +874,7 @@ static void generic_metric(struct perf_stat_config *config,const char *metric_name, const char *metric_unit, int runtime, - int cpu, + int cpu_map_idx, struct perf_stat_output_ctx *out, struct runtime_stat *st) {
@@ -889,7 +889,7 @@ static void generic_metric(struct perf_stat_config *config,return ; pctx->runtime = runtime; - i =prepare_metric (metric_events, metric_refs, pctx, cpu, st); + i =prepare_metric (metric_events, metric_refs, pctx, cpu_map_idx, st);if (i <0 ) {expr__ctx_free (pctx);return ;
@@ -934,7 +934,7 @@ static void generic_metric(struct perf_stat_config *config,expr__ctx_free (pctx); } -double test_generic_metric (struct metric_expr *mexp,int cpu,struct runtime_stat *st) +double test_generic_metric (struct metric_expr *mexp,int cpu_map_idx,struct runtime_stat *st) {struct expr_parse_ctx *pctx;double ratio =0.0 ;
@@ -943,7 +943,7 @@ double test_generic_metric(struct metric_expr *mexp, int cpu, struct runtime_staif (!pctx)return NAN ; -if (prepare_metric (mexp->metric_events, mexp->metric_refs, pctx, cpu, st) <0 ) +if (prepare_metric (mexp->metric_events, mexp->metric_refs, pctx, cpu_map_idx, st) <0 ) goto out;if (expr__parse (&ratio, pctx, mexp->metric_expr))
@@ -956,7 +956,7 @@ double test_generic_metric(struct metric_expr *mexp, int cpu, struct runtime_stavoid perf_stat__print_shadow_stats (struct perf_stat_config *config,struct evsel *evsel, -double avg,int cpu, +double avg,int cpu_map_idx,struct perf_stat_output_ctx *out,struct rblist *metric_events,struct runtime_stat *st)
@@ -975,7 +975,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,if (config->iostat_run) {iostat_print_metric (config, evsel, out); }else if (evsel__match (evsel,HARDWARE ,HW_INSTRUCTIONS )) { - total =runtime_stat_avg (st,STAT_CYCLES , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, &rsd);if (total) { ratio = avg / total;
@@ -985,11 +985,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp,NULL ,NULL ,"insn per cycle" ,0 ); } - total =runtime_stat_avg (st,STAT_STALLED_CYCLES_FRONT , cpu, &rsd); + total =runtime_stat_avg (st,STAT_STALLED_CYCLES_FRONT , cpu_map_idx, &rsd); total =max (total,runtime_stat_avg (st,STAT_STALLED_CYCLES_BACK , - cpu, &rsd)); + cpu_map_idx, &rsd));if (total && avg) { out->new_line (config, ctxp);
@@ -999,8 +999,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ratio); } }else if (evsel__match (evsel,HARDWARE ,HW_BRANCH_MISSES )) { -if (runtime_stat_n (st,STAT_BRANCHES , cpu, &rsd) !=0 ) -print_branch_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_BRANCHES , cpu_map_idx, &rsd) !=0 ) +print_branch_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all branches" ,0 ); } else if (
@@ -1009,8 +1009,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { -if (runtime_stat_n (st,STAT_L1_DCACHE , cpu, &rsd) !=0 ) -print_l1_dcache_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_L1_DCACHE , cpu_map_idx, &rsd) !=0 ) +print_l1_dcache_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all L1-dcache accesses" ,0 ); } else if (
@@ -1019,8 +1019,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { -if (runtime_stat_n (st,STAT_L1_ICACHE , cpu, &rsd) !=0 ) -print_l1_icache_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_L1_ICACHE , cpu_map_idx, &rsd) !=0 ) +print_l1_icache_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all L1-icache accesses" ,0 ); } else if (
@@ -1029,8 +1029,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { -if (runtime_stat_n (st,STAT_DTLB_CACHE , cpu, &rsd) !=0 ) -print_dtlb_cache_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_DTLB_CACHE , cpu_map_idx, &rsd) !=0 ) +print_dtlb_cache_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all dTLB cache accesses" ,0 ); } else if (
@@ -1039,8 +1039,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { -if (runtime_stat_n (st,STAT_ITLB_CACHE , cpu, &rsd) !=0 ) -print_itlb_cache_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_ITLB_CACHE , cpu_map_idx, &rsd) !=0 ) +print_itlb_cache_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all iTLB cache accesses" ,0 ); } else if (
@@ -1049,27 +1049,27 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { -if (runtime_stat_n (st,STAT_LL_CACHE , cpu, &rsd) !=0 ) -print_ll_cache_misses (config, cpu, avg, out, st, &rsd); +if (runtime_stat_n (st,STAT_LL_CACHE , cpu_map_idx, &rsd) !=0 ) +print_ll_cache_misses (config, cpu_map_idx, avg, out, st, &rsd);else print_metric (config, ctxp,NULL ,NULL ,"of all LL-cache accesses" ,0 ); } else if (evsel__match(evsel, HARDWARE, HW_CACHE_MISSES)) { - total =runtime_stat_avg (st,STAT_CACHEREFS , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CACHEREFS , cpu_map_idx, &rsd);if (total) ratio = avg *100 / total; -if (runtime_stat_n (st,STAT_CACHEREFS , cpu, &rsd) !=0 ) +if (runtime_stat_n (st,STAT_CACHEREFS , cpu_map_idx, &rsd) !=0 )print_metric (config, ctxp,NULL ,"%8.3f %%" ,"of all cache refs" , ratio);else print_metric (config, ctxp,NULL ,NULL ,"of all cache refs" ,0 ); } else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_FRONTEND)) { -print_stalled_cycles_frontend (config, cpu, avg, out, st, &rsd); +print_stalled_cycles_frontend (config, cpu_map_idx, avg, out, st, &rsd); } else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_BACKEND)) { -print_stalled_cycles_backend (config, cpu, avg, out, st, &rsd); +print_stalled_cycles_backend (config, cpu_map_idx, avg, out, st, &rsd); } else if (evsel__match(evsel, HARDWARE, HW_CPU_CYCLES)) { - total =runtime_stat_avg (st,STAT_NSECS , cpu, &rsd); + total =runtime_stat_avg (st,STAT_NSECS , cpu_map_idx, &rsd);if (total) { ratio = avg / total;
@@ -1078,7 +1078,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp,NULL ,NULL ,"Ghz" ,0 ); } }else if (perf_stat_evsel__is (evsel,CYCLES_IN_TX )) { - total =runtime_stat_avg (st,STAT_CYCLES , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, &rsd);if (total) print_metric(config, ctxp, NULL,
@@ -1088,8 +1088,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp,NULL ,NULL ,"transactional cycles" ,0 ); }else if (perf_stat_evsel__is (evsel,CYCLES_IN_TX_CP )) { - total =runtime_stat_avg (st,STAT_CYCLES , cpu, &rsd); - total2 =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CYCLES , cpu_map_idx, &rsd); + total2 =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu_map_idx, &rsd);if (total2 < avg) total2 = avg;
@@ -1099,19 +1099,19 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,else print_metric (config, ctxp,NULL ,NULL ,"aborted cycles" ,0 ); }else if (perf_stat_evsel__is (evsel,TRANSACTION_START )) { - total =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu_map_idx, &rsd);if (avg) ratio = total / avg; -if (runtime_stat_n (st,STAT_CYCLES_IN_TX , cpu, &rsd) !=0 ) +if (runtime_stat_n (st,STAT_CYCLES_IN_TX , cpu_map_idx, &rsd) !=0 )print_metric (config, ctxp,NULL ,"%8.0f" ,"cycles / transaction" , ratio);else print_metric (config, ctxp,NULL ,NULL ,"cycles / transaction" ,0 ); }else if (perf_stat_evsel__is (evsel,ELISION_START )) { - total =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu, &rsd); + total =runtime_stat_avg (st,STAT_CYCLES_IN_TX , cpu_map_idx, &rsd);if (avg) ratio = total / avg;
@@ -1124,28 +1124,28 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,else print_metric (config, ctxp,NULL ,NULL ,"CPUs utilized" ,0 ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_FETCH_BUBBLES )) { -double fe_bound =td_fe_bound (cpu, st, &rsd); +double fe_bound =td_fe_bound (cpu_map_idx, st, &rsd);if (fe_bound >0.2 ) color =PERF_COLOR_RED ;print_metric (config, ctxp, color,"%8.1f%%" ,"frontend bound" , fe_bound *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_SLOTS_RETIRED )) { -double retiring =td_retiring (cpu, st, &rsd); +double retiring =td_retiring (cpu_map_idx, st, &rsd);if (retiring >0.7 ) color =PERF_COLOR_GREEN ;print_metric (config, ctxp, color,"%8.1f%%" ,"retiring" , retiring *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_RECOVERY_BUBBLES )) { -double bad_spec =td_bad_spec (cpu, st, &rsd); +double bad_spec =td_bad_spec (cpu_map_idx, st, &rsd);if (bad_spec >0.1 ) color =PERF_COLOR_RED ;print_metric (config, ctxp, color,"%8.1f%%" ,"bad speculation" , bad_spec *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_SLOTS_ISSUED )) { -double be_bound =td_be_bound (cpu, st, &rsd); +double be_bound =td_be_bound (cpu_map_idx, st, &rsd);const char *name ="backend bound" ;static int have_recovery_bubbles =-1 ;
@@ -1158,14 +1158,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,if (be_bound >0.2 ) color =PERF_COLOR_RED ; -if (td_total_slots (cpu, st, &rsd) >0 ) +if (td_total_slots (cpu_map_idx, st, &rsd) >0 )print_metric (config, ctxp, color,"%8.1f%%" , name, be_bound *100. );else print_metric (config, ctxp,NULL ,NULL , name,0 ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_RETIRING ) && -full_td (cpu, st, &rsd)) { -double retiring =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd)) { +double retiring =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_RETIRING , st, &rsd);if (retiring >0.7 )
@@ -1173,8 +1173,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"retiring" , retiring *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_FE_BOUND ) && -full_td (cpu, st, &rsd)) { -double fe_bound =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd)) { +double fe_bound =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_FE_BOUND , st, &rsd);if (fe_bound >0.2 )
@@ -1182,8 +1182,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"frontend bound" , fe_bound *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_BE_BOUND ) && -full_td (cpu, st, &rsd)) { -double be_bound =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd)) { +double be_bound =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_BE_BOUND , st, &rsd);if (be_bound >0.2 )
@@ -1191,8 +1191,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"backend bound" , be_bound *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_BAD_SPEC ) && -full_td (cpu, st, &rsd)) { -double bad_spec =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd)) { +double bad_spec =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_BAD_SPEC , st, &rsd);if (bad_spec >0.1 )
@@ -1200,11 +1200,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"bad speculation" , bad_spec *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_HEAVY_OPS ) && -full_td (cpu, st, &rsd) && (config->topdown_level >1 )) { -double retiring =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd) && (config->topdown_level >1 )) { +double retiring =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_RETIRING , st, &rsd); -double heavy_ops =td_metric_ratio (cpu, +double heavy_ops =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_HEAVY_OPS , st, &rsd);double light_ops = retiring - heavy_ops;
@@ -1220,11 +1220,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"light operations" , light_ops *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_BR_MISPREDICT ) && -full_td (cpu, st, &rsd) && (config->topdown_level >1 )) { -double bad_spec =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd) && (config->topdown_level >1 )) { +double bad_spec =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_BAD_SPEC , st, &rsd); -double br_mis =td_metric_ratio (cpu, +double br_mis =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_BR_MISPREDICT , st, &rsd);double m_clears = bad_spec - br_mis;
@@ -1240,11 +1240,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"machine clears" , m_clears *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_FETCH_LAT ) && -full_td (cpu, st, &rsd) && (config->topdown_level >1 )) { -double fe_bound =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd) && (config->topdown_level >1 )) { +double fe_bound =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_FE_BOUND , st, &rsd); -double fetch_lat =td_metric_ratio (cpu, +double fetch_lat =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_FETCH_LAT , st, &rsd);double fetch_bw = fe_bound - fetch_lat;
@@ -1260,11 +1260,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,print_metric (config, ctxp, color,"%8.1f%%" ,"fetch bandwidth" , fetch_bw *100. ); }else if (perf_stat_evsel__is (evsel,TOPDOWN_MEM_BOUND ) && -full_td (cpu, st, &rsd) && (config->topdown_level >1 )) { -double be_bound =td_metric_ratio (cpu, +full_td (cpu_map_idx, st, &rsd) && (config->topdown_level >1 )) { +double be_bound =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_BE_BOUND , st, &rsd); -double mem_bound =td_metric_ratio (cpu, +double mem_bound =td_metric_ratio (cpu_map_idx,STAT_TOPDOWN_MEM_BOUND , st, &rsd);double core_bound = be_bound - mem_bound;
@@ -1281,12 +1281,12 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, core_bound * 100.); }else if (evsel->metric_expr) {generic_metric (config, evsel->metric_expr, evsel->metric_events,NULL , - evsel->name, evsel->metric_name,NULL ,1 , cpu, out, st); - }else if (runtime_stat_n (st,STAT_NSECS , cpu, &rsd) !=0 ) { + evsel->name, evsel->metric_name,NULL ,1 , cpu_map_idx, out, st); + }else if (runtime_stat_n (st,STAT_NSECS , cpu_map_idx, &rsd) !=0 ) {char unit =' ' ;char unit_buf[10 ] ="/sec" ; - total =runtime_stat_avg (st,STAT_NSECS , cpu, &rsd); + total =runtime_stat_avg (st,STAT_NSECS , cpu_map_idx, &rsd);if (total) ratio =convert_unit_double (1000000000.0 * avg / total, &unit);
@@ -1294,7 +1294,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,snprintf (unit_buf,sizeof (unit_buf),"%c/sec" , unit);print_metric (config, ctxp,NULL ,"%8.3f" , unit_buf, ratio); }else if (perf_stat_evsel__is (evsel,SMI_NUM )) { -print_smi_cost (config, cpu, out, st, &rsd); +print_smi_cost (config, cpu_map_idx, out, st, &rsd); }else { num =0 ; }
@@ -1307,7 +1307,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, out->new_line (config, ctxp);generic_metric (config, mexp->metric_expr, mexp->metric_events, mexp->metric_refs, evsel->name, mexp->metric_name, - mexp->metric_unit, mexp->runtime, cpu, out, st); + mexp->metric_unit, mexp->runtime, cpu_map_idx, out, st); } }if (num ==0 )
tools/perf/util/stat.c +8 -8
@@ -356,14 +356,14 @@ static int check_per_pkg(struct evsel *counter, struct perf_counts_values *vals,static int process_counter_values (struct perf_stat_config *config,struct evsel *evsel, -int cpu,int thread, +int cpu_map_idx,int thread,struct perf_counts_values *count) {struct perf_counts_values *aggr = &evsel->counts->aggr;static struct perf_counts_values zero;bool skip = false; -if (check_per_pkg (evsel, count, cpu, &skip)) { +if (check_per_pkg (evsel, count, cpu_map_idx, &skip)) {pr_err ("failed to read per-pkg counter\n" );return -1 ; }
@@ -379,11 +379,11 @@ process_counter_values(struct perf_stat_config *config, struct evsel *evsel,case AGGR_NODE :case AGGR_NONE :if (!evsel->snapshot) -evsel__compute_deltas (evsel, cpu, thread, count); +evsel__compute_deltas (evsel, cpu_map_idx, thread, count);perf_counts_values__scale (count, config->scale,NULL );if ((config->aggr_mode ==AGGR_NONE ) && (!evsel->percore)) {perf_stat__update_shadow_stats (evsel, count->val, - cpu, &rt_stat); + cpu_map_idx, &rt_stat); }if (config->aggr_mode ==AGGR_THREAD ) {
@@ -412,15 +412,15 @@ static int process_counter_maps(struct perf_stat_config *config, {int nthreads =perf_thread_map__nr (counter->core.threads);int ncpus =evsel__nr_cpus (counter); -int cpu, thread; +int idx, thread;if (counter->core.system_wide) nthreads =1 ;for (thread =0 ; thread < nthreads; thread++) { -for (cpu =0 ; cpu < ncpus; cpu++) { -if (process_counter_values (config, counter, cpu, thread, -perf_counts (counter->counts, cpu, thread))) +for (idx =0 ; idx < ncpus; idx++) { +if (process_counter_values (config, counter, idx, thread, +perf_counts (counter->counts, idx, thread)))return -1 ; } }
tools/perf/util/stat.h +2 -2
@@ -208,7 +208,7 @@ void perf_stat__init_shadow_stats(void);void perf_stat__reset_shadow_stats (void );void perf_stat__reset_shadow_per_stat (struct runtime_stat *st);void perf_stat__update_shadow_stats (struct evsel *counter,u64 count, -int cpu,struct runtime_stat *st); +int cpu_map_idx,struct runtime_stat *st);struct perf_stat_output_ctx {void *ctx;print_metric_t print_metric;
@@ -253,5 +253,5 @@ void evlist__print_counters(struct evlist *evlist, struct perf_stat_config *conf struct target *_target, struct timespec *ts, int argc, const char **argv);struct metric_expr ; -double test_generic_metric (struct metric_expr *mexp,int cpu,struct runtime_stat *st); +double test_generic_metric (struct metric_expr *mexp,int cpu_map_idx,struct runtime_stat *st);#endif
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 43/48 libperf: Sync evsel documentation
In reply to Ian Rogerscpu was renamed cpu_map_idx, for clarity.
--- tools/lib/perf/Documentation/libperf.txt | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
tools/lib/perf/Documentation/libperf.txt +5 -5
@@ -135,16 +135,16 @@ SYNOPSIS 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); + void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu_map_idx); int perf_evsel__mmap(struct perf_evsel *evsel, int pages); void perf_evsel__munmap(struct perf_evsel *evsel); - void *perf_evsel__mmap_base(struct perf_evsel *evsel, int cpu, int thread); - int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + void *perf_evsel__mmap_base(struct perf_evsel *evsel, int cpu_map_idx, int thread); + int perf_evsel__read(struct perf_evsel *evsel, int cpu_map_idx, 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__enable_cpu(struct perf_evsel *evsel, int cpu_map_idx); int perf_evsel__disable(struct perf_evsel *evsel); - int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu_map_idx); 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);
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 44/48 perf bpf: Rename cpu to cpu_map_idx
In reply to Ian RogersSynchronize the caller in evsel with the called function. Shorten 3 lines of code in bperf_read by using perf_cpu_map__for_each_cpu.
This code is frequently using variables named cpu as cpu map indices, which doesn't matter as all CPUs are in the CPU map. It is strange in some cases the cpumap is used at all.
--- tools/perf/util/bpf_counter.c | 16 +++++++--------- tools/perf/util/bpf_counter.h | 4 ++-- 2 files changed, 9 insertions(+), 11 deletions(-)
tools/perf/util/bpf_counter.c +7 -9
@@ -265,7 +265,7 @@ static int bpf_program_profiler__read(struct evsel *evsel)return 0 ; } -static int bpf_program_profiler__install_pe (struct evsel *evsel,int cpu, +static int bpf_program_profiler__install_pe (struct evsel *evsel,int cpu_map_idx,int fd) {struct bpf_prog_profiler_bpf *skel;
@@ -277,7 +277,7 @@ static int bpf_program_profiler__install_pe(struct evsel *evsel, int cpu,assert (skel !=NULL ); ret =bpf_map_update_elem (bpf_map__fd (skel->maps.events), - &cpu, &fd,BPF_ANY ); + &cpu_map_idx, &fd,BPF_ANY );if (ret)return ret; }
@@ -566,12 +566,12 @@ static int bperf__load(struct evsel *evsel, struct target *target)return err; } -static int bperf__install_pe (struct evsel *evsel,int cpu,int fd) +static int bperf__install_pe (struct evsel *evsel,int cpu_map_idx,int fd) {struct bperf_leader_bpf *skel = evsel->leader_skel;return bpf_map_update_elem (bpf_map__fd (skel->maps.events), - &cpu, &fd,BPF_ANY ); + &cpu_map_idx, &fd,BPF_ANY ); }/*
@@ -623,9 +623,7 @@ static int bperf__read(struct evsel *evsel)case BPERF_FILTER_GLOBAL :assert (i ==0 ); - num_cpu = all_cpu_map->nr; -for (j =0 ; j < num_cpu; j++) { - cpu = all_cpu_map->map[j]; +perf_cpu_map__for_each_cpu (cpu, j, all_cpu_map) {perf_counts (evsel->counts, cpu,0 )->val = values[cpu].counter;perf_counts (evsel->counts, cpu,0 )->ena = values[cpu].enabled;perf_counts (evsel->counts, cpu,0 )->run = values[cpu].running;
@@ -757,11 +755,11 @@ static inline bool bpf_counter_skip(struct evsel *evsel) evsel->follower_skel ==NULL ; } -int bpf_counter__install_pe (struct evsel *evsel,int cpu,int fd) +int bpf_counter__install_pe (struct evsel *evsel,int cpu_map_idx,int fd) {if (bpf_counter_skip (evsel))return 0 ; -return evsel->bpf_counter_ops->install_pe (evsel, cpu, fd); +return evsel->bpf_counter_ops->install_pe (evsel, cpu_map_idx, fd); }int bpf_counter__load (struct evsel *evsel,struct target *target)
tools/perf/util/bpf_counter.h +2 -2
@@ -16,7 +16,7 @@ typedef int (*bpf_counter_evsel_op)(struct evsel *evsel);typedef int (*bpf_counter_evsel_target_op )(struct evsel *evsel,struct target *target);typedef int (*bpf_counter_evsel_install_pe_op )(struct evsel *evsel, -int cpu, +int cpu_map_idx,int fd);struct bpf_counter_ops {
@@ -40,7 +40,7 @@ int bpf_counter__enable(struct evsel *evsel);int bpf_counter__disable (struct evsel *evsel);int bpf_counter__read (struct evsel *evsel);void bpf_counter__destroy (struct evsel *evsel); -int bpf_counter__install_pe (struct evsel *evsel,int cpu,int fd); +int bpf_counter__install_pe (struct evsel *evsel,int cpu_map_idx,int fd);#else /* HAVE_BPF_SKEL */
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 45/48 perf c2c: Use more intention revealing iterator
In reply to Ian RogersUse perf_cpu_map__for_each_cpu in setup_nodes.
--- tools/perf/builtin-c2c.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
tools/perf/builtin-c2c.c +5 -5
@@ -2015,7 +2015,7 @@ static int setup_nodes(struct perf_session *session) {struct numa_node *n;unsigned long **nodes; -int node, cpu; +int node, cpu, idx;int *cpu2node;if (c2c.node_info >2 )
@@ -2057,13 +2057,13 @@ static int setup_nodes(struct perf_session *session)if (perf_cpu_map__empty (map))continue ; -for (cpu =0 ; cpu < map->nr; cpu++) { -set_bit (map->map[cpu], set); +perf_cpu_map__for_each_cpu (cpu, idx, map) { +set_bit (cpu, set); -if (WARN_ONCE (cpu2node[map->map[cpu]] !=-1 ,"node/cpu topology bug" )) +if (WARN_ONCE (cpu2node[cpu] !=-1 ,"node/cpu topology bug" ))return -EINVAL ; - cpu2node[map->map[cpu]] = node; + cpu2node[cpu] = node; } }
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 46/48 perf script: Fix flipped index and cpu
In reply to Ian Rogersperf_counts are accessed by the densely packed index.
--- tools/perf/builtin-script.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/builtin-script.c +1 -1
@@ -2129,7 +2129,7 @@ static void __process_stat(struct evsel *counter, u64 tstamp)perf_cpu_map__for_each_cpu (cpu, idx,evsel__cpus (counter)) {struct perf_counts_values *counts; - counts =perf_counts (counter->counts, cpu, thread); + counts =perf_counts (counter->counts, idx, thread); printf("%3d %8d %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %s\n", cpu,
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 47/48 perf stat: Correct first_shadow_cpu to return index
In reply to Ian Rogersperf_stat__update_shadow_stats and perf_stat__print_shadow_stats use a cpu map index rather than a CPU, but first_shadow_cpu is returning the wrong value for this. Change first_shadow_cpu to first_shadow_cpu_map_idx to make things agree.
--- tools/perf/util/stat-display.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-)
tools/perf/util/stat-display.c +7 -8
@@ -327,24 +327,23 @@ static void print_metric_header(struct perf_stat_config *config,fprintf (os->fh,"%*s " , config->metric_only_len, unit); } -static int first_shadow_cpu (struct perf_stat_config *config, -struct evsel *evsel,const struct aggr_cpu_id *id) +static int first_shadow_cpu_map_idx (struct perf_stat_config *config, +struct evsel *evsel,const struct aggr_cpu_id *id) { -struct perf_cpu_map *cpus; +struct perf_cpu_map *cpus =evsel__cpus (evsel);int cpu, idx;if (config->aggr_mode ==AGGR_NONE ) -return id->cpu; +return perf_cpu_map__idx (cpus, id->cpu);if (!config->aggr_get_id)return 0 ; - cpus =evsel__cpus (evsel);perf_cpu_map__for_each_cpu (cpu, idx, cpus) {struct aggr_cpu_id cpu_id = config->aggr_get_id (config, cpu);if (aggr_cpu_id__equal (&cpu_id, id)) -return cpu; +return idx; }return 0 ; }
@@ -503,7 +502,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int }perf_stat__print_shadow_stats (config, counter, uval, -first_shadow_cpu (config, counter, &id), +first_shadow_cpu_map_idx (config, counter, &id), &out, &config->metric_events, st);if (!config->csv_output && !config->metric_only) {print_noise (config, counter, noise);
@@ -532,7 +531,7 @@ static void aggr_update_shadow(struct perf_stat_config *config, val +=perf_counts (counter->counts, idx,0 )->val; }perf_stat__update_shadow_stats (counter, val, -first_shadow_cpu (config, counter, &id), +first_shadow_cpu_map_idx (config, counter, &id), &rt_stat); } }
Signature · 1 line
2.34.1.307.g9b7440fafd-goog
Ian Rogers
reviewerPATCH v2 48/48 perf cpumap: Give CPUs their own type.
In reply to Ian RogersA common problem is confusing CPU map indices with the CPU, by wrapping the CPU with a struct then this is avoided. This approach is similar to atomic_t.
--- tools/lib/perf/cpumap.c | 121 ++++++++++-------- tools/lib/perf/evlist.c | 4 +- tools/lib/perf/evsel.c | 9 +- tools/lib/perf/include/internal/cpumap.h | 9 +- tools/lib/perf/include/internal/evlist.h | 3 +- tools/lib/perf/include/internal/evsel.h | 4 +- tools/lib/perf/include/internal/mmap.h | 5 +- tools/lib/perf/include/perf/cpumap.h | 9 +- tools/lib/perf/mmap.c | 2 +- tools/perf/bench/epoll-ctl.c | 2 +- tools/perf/bench/epoll-wait.c | 2 +- tools/perf/bench/futex-hash.c | 2 +- tools/perf/bench/futex-lock-pi.c | 2 +- tools/perf/bench/futex-requeue.c | 2 +- tools/perf/bench/futex-wake-parallel.c | 2 +- tools/perf/bench/futex-wake.c | 2 +- tools/perf/builtin-c2c.c | 13 +- tools/perf/builtin-ftrace.c | 2 +- tools/perf/builtin-kmem.c | 2 +- tools/perf/builtin-record.c | 2 +- tools/perf/builtin-sched.c | 65 +++++----- tools/perf/builtin-script.c | 5 +- tools/perf/builtin-stat.c | 80 ++++++------ tools/perf/tests/attr.c | 6 +- tools/perf/tests/bitmap.c | 2 +- tools/perf/tests/cpumap.c | 6 +- tools/perf/tests/event_update.c | 6 +- tools/perf/tests/mem2node.c | 2 +- tools/perf/tests/mmap-basic.c | 4 +- tools/perf/tests/openat-syscall-all-cpus.c | 17 +-- tools/perf/tests/stat.c | 3 +- tools/perf/tests/topology.c | 30 +++-- tools/perf/util/affinity.c | 2 +- tools/perf/util/auxtrace.c | 12 +- tools/perf/util/auxtrace.h | 5 +- tools/perf/util/cpumap.c | 91 ++++++------- tools/perf/util/cpumap.h | 26 ++-- tools/perf/util/cputopo.c | 6 +- tools/perf/util/env.c | 29 +++-- tools/perf/util/env.h | 3 +- tools/perf/util/evlist.c | 8 +- tools/perf/util/evlist.h | 2 +- tools/perf/util/evsel.c | 6 +- tools/perf/util/expr.c | 2 +- tools/perf/util/header.c | 6 +- tools/perf/util/mmap.c | 19 +-- tools/perf/util/mmap.h | 3 +- tools/perf/util/perf_api_probe.c | 15 ++- tools/perf/util/python.c | 4 +- tools/perf/util/record.c | 11 +- .../scripting-engines/trace-event-python.c | 6 +- tools/perf/util/session.c | 10 +- tools/perf/util/stat-display.c | 34 +++-- tools/perf/util/stat.c | 2 +- tools/perf/util/stat.h | 2 +- tools/perf/util/svghelper.c | 6 +- tools/perf/util/synthetic-events.c | 12 +- tools/perf/util/synthetic-events.h | 3 +- tools/perf/util/util.h | 5 +- 59 files changed, 408 insertions(+), 347 deletions(-)
tools/lib/perf/cpumap.c +70 -51
@@ -10,16 +10,27 @@#include <ctype.h> #include <limits.h> +static struct perf_cpu_map *perf_cpu_map__alloc (int nr_cpus) +{ +struct perf_cpu_map *cpus; + + cpus =malloc (sizeof (*cpus) +sizeof (struct perf_cpu ) * nr_cpus); +if (!cpus) +return NULL ; + + cpus->nr =1 ; +refcount_set (&cpus->refcnt,1 ); +return cpus; +} +struct perf_cpu_map *perf_cpu_map__dummy_new (void ) { -struct perf_cpu_map *cpus =malloc (sizeof (*cpus) +sizeof (int )); +struct perf_cpu_map *cpus =perf_cpu_map__alloc (1 ); -if (cpus !=NULL ) { - cpus->nr =1 ; - cpus->map[0 ] =-1 ; -refcount_set (&cpus->refcnt,1 ); - } +if (!cpus) +return NULL ; + cpus->map[0 ].cpu =-1 ;return cpus; }
@@ -54,16 +65,12 @@ static struct perf_cpu_map *cpu_map__default_new(void)if (nr_cpus <0 )return NULL ; - cpus =malloc (sizeof (*cpus) + nr_cpus *sizeof (int )); -if (cpus !=NULL ) { -int i; - -for (i =0 ; i < nr_cpus; ++i) - cpus->map[i] = i; + cpus =perf_cpu_map__alloc (nr_cpus); +if (!cpus) +return NULL ; - cpus->nr = nr_cpus; -refcount_set (&cpus->refcnt,1 ); - } +for (int i =0 ; i < nr_cpus; ++i) + cpus->map[i].cpu = i;return cpus; }
@@ -73,31 +80,32 @@ struct perf_cpu_map *perf_cpu_map__default_new(void)return cpu_map__default_new (); } -static int cmp_int (const void *a,const void *b) + +static int cmp_cpu (const void *a,const void *b) { -return *(const int *)a - *(const int *)b; +const struct perf_cpu *cpu_a = a, *cpu_b = b; + +return cpu_a->cpu - cpu_b->cpu; } -static struct perf_cpu_map *cpu_map__trim_new (int nr_cpus,int *tmp_cpus) +static struct perf_cpu_map *cpu_map__trim_new (int nr_cpus,const struct perf_cpu *tmp_cpus) { -size_t payload_size = nr_cpus *sizeof (int ); -struct perf_cpu_map *cpus =malloc (sizeof (*cpus) + payload_size); +size_t payload_size = nr_cpus *sizeof (struct perf_cpu ); +struct perf_cpu_map *cpus =perf_cpu_map__alloc (nr_cpus);int i, j;if (cpus !=NULL ) {memcpy (cpus->map, tmp_cpus, payload_size); -qsort (cpus->map, nr_cpus,sizeof (int ), cmp_int); +qsort (cpus->map, nr_cpus,sizeof (struct perf_cpu ), cmp_cpu);/* Remove dups */ j =0 ;for (i =0 ; i < nr_cpus; i++) { -if (i ==0 || cpus->map[i] != cpus->map[i -1 ]) - cpus->map[j++] = cpus->map[i]; +if (i ==0 || cpus->map[i].cpu != cpus->map[i -1 ].cpu) + cpus->map[j++].cpu = cpus->map[i].cpu; } cpus->nr = j;assert (j <= nr_cpus); -refcount_set (&cpus->refcnt,1 ); } -return cpus; }
@@ -105,7 +113,7 @@ struct perf_cpu_map *perf_cpu_map__read(FILE *file) {struct perf_cpu_map *cpus =NULL ;int nr_cpus =0 ; -int *tmp_cpus =NULL , *tmp; +struct perf_cpu *tmp_cpus =NULL , *tmp;int max_entries =0 ;int n, cpu, prev;char sep;
@@ -124,24 +132,24 @@ struct perf_cpu_map *perf_cpu_map__read(FILE *file)if (new_max >= max_entries) { max_entries = new_max +MAX_NR_CPUS /2 ; - tmp =realloc (tmp_cpus, max_entries *sizeof (int )); + tmp =realloc (tmp_cpus, max_entries *sizeof (struct perf_cpu ));if (tmp ==NULL ) goto out_free_tmp; tmp_cpus = tmp; }while (++prev < cpu) - tmp_cpus[nr_cpus++] = prev; + tmp_cpus[nr_cpus++].cpu = prev; }if (nr_cpus == max_entries) { max_entries +=MAX_NR_CPUS ; - tmp =realloc (tmp_cpus, max_entries *sizeof (int )); + tmp =realloc (tmp_cpus, max_entries *sizeof (struct perf_cpu ));if (tmp ==NULL ) goto out_free_tmp; tmp_cpus = tmp; } - tmp_cpus[nr_cpus++] = cpu; + tmp_cpus[nr_cpus++].cpu = cpu;if (n ==2 && sep =='-' ) prev = cpu;else
@@ -179,7 +187,7 @@ struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list)unsigned long start_cpu, end_cpu =0 ;char *p =NULL ;int i, nr_cpus =0 ; -int *tmp_cpus =NULL , *tmp; +struct perf_cpu *tmp_cpus =NULL , *tmp;int max_entries =0 ;if (!cpu_list)
@@ -220,17 +228,17 @@ struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list)for (; start_cpu <= end_cpu; start_cpu++) {/* check for duplicates */ for (i =0 ; i < nr_cpus; i++) -if (tmp_cpus[i] == (int )start_cpu) +if (tmp_cpus[i].cpu == (int )start_cpu) goto invalid;if (nr_cpus == max_entries) { max_entries +=MAX_NR_CPUS ; - tmp =realloc (tmp_cpus, max_entries *sizeof (int )); + tmp =realloc (tmp_cpus, max_entries *sizeof (struct perf_cpu ));if (tmp ==NULL ) goto invalid; tmp_cpus = tmp; } - tmp_cpus[nr_cpus++] = (int )start_cpu; + tmp_cpus[nr_cpus++].cpu = (int )start_cpu; }if (*p) ++p;
@@ -250,12 +258,16 @@ struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list)return cpus; } -int perf_cpu_map__cpu (const struct perf_cpu_map *cpus,int idx) +struct perf_cpu perf_cpu_map__cpu (const struct perf_cpu_map *cpus,int idx) { +struct perf_cpu result = { + .cpu =-1 + }; +if (cpus && idx < cpus->nr) -return cpus->map[idx]; + result = cpus->map[idx]; -return -1 ; +return result; }int perf_cpu_map__nr (const struct perf_cpu_map *cpus)
@@ -265,10 +277,10 @@ int perf_cpu_map__nr(const struct perf_cpu_map *cpus)bool perf_cpu_map__empty (const struct perf_cpu_map *map) { -return map ? map->map[0 ] ==-1 : true; +return map ? map->map[0 ].cpu ==-1 : true; } -int perf_cpu_map__idx (const struct perf_cpu_map *cpus,int cpu) +int perf_cpu_map__idx (const struct perf_cpu_map *cpus,struct perf_cpu cpu) {int low, high;
@@ -278,13 +290,13 @@ int perf_cpu_map__idx(const struct perf_cpu_map *cpus, int cpu) low =0 ; high = cpus->nr;while (low < high) { -int idx = (low + high) /2 , - cpu_at_idx = cpus->map[idx]; +int idx = (low + high) /2 ; +struct perf_cpu cpu_at_idx = cpus->map[idx]; -if (cpu_at_idx == cpu) +if (cpu_at_idx.cpu == cpu.cpu)return idx; -if (cpu_at_idx > cpu) +if (cpu_at_idx.cpu > cpu.cpu) high = idx;else low = idx +1 ;
@@ -293,15 +305,22 @@ int perf_cpu_map__idx(const struct perf_cpu_map *cpus, int cpu)return -1 ; } -bool perf_cpu_map__has (const struct perf_cpu_map *cpus,int cpu) +bool perf_cpu_map__has (const struct perf_cpu_map *cpus,struct perf_cpu cpu) {return perf_cpu_map__idx (cpus, cpu) !=-1 ; } -int perf_cpu_map__max (struct perf_cpu_map *map) +struct perf_cpu perf_cpu_map__max (struct perf_cpu_map *map) { +struct perf_cpu result = { + .cpu =-1 + }; +// cpu_map__trim_new() qsort()s it, cpu_map__default_new() sorts it as well. -return map->nr >0 ? map->map[map->nr -1 ] :-1 ; +if (map->nr >0 ) + result = map->map[map->nr -1 ]; + +return result; }/*
@@ -315,7 +334,7 @@ int perf_cpu_map__max(struct perf_cpu_map *map)struct perf_cpu_map *perf_cpu_map__merge (struct perf_cpu_map *orig,struct perf_cpu_map *other) { -int *tmp_cpus; +struct perf_cpu *tmp_cpus;int tmp_len;int i, j, k;struct perf_cpu_map *merged;
@@ -329,19 +348,19 @@ struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig,if (!other)return orig;if (orig->nr == other->nr && - !memcmp (orig->map, other->map, orig->nr *sizeof (int ))) + !memcmp (orig->map, other->map, orig->nr *sizeof (struct perf_cpu )))return orig; tmp_len = orig->nr + other->nr; - tmp_cpus =malloc (tmp_len *sizeof (int )); + tmp_cpus =malloc (tmp_len *sizeof (struct perf_cpu ));if (!tmp_cpus)return NULL ;/* Standard merge algorithm from wikipedia */ i = j = k =0 ;while (i < orig->nr && j < other->nr) { -if (orig->map[i] <= other->map[j]) { -if (orig->map[i] == other->map[j]) +if (orig->map[i].cpu <= other->map[j].cpu) { +if (orig->map[i].cpu == other->map[j].cpu) j++; tmp_cpus[k++] = orig->map[i++]; }else
tools/lib/perf/evlist.c +2 -2
@@ -407,7 +407,7 @@ perf_evlist__mmap_cb_get(struct perf_evlist *evlist, bool overwrite, int idx)static int perf_evlist__mmap_cb_mmap (struct perf_mmap *map,struct perf_mmap_param *mp, -int output,int cpu) +int output,struct perf_cpu cpu) {return perf_mmap__mmap (map, mp, output, cpu); }
@@ -426,7 +426,7 @@ mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, int idx, struct perf_mmap_param *mp, int cpu_idx, int thread, int *_output, int *_output_overwrite) { - int evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx); + struct perf_cpu evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx);struct perf_evsel *evsel;int revent;
tools/lib/perf/evsel.c +5 -4
@@ -78,10 +78,10 @@ static int perf_evsel__alloc_mmap(struct perf_evsel *evsel, int ncpus, int nthrestatic int sys_perf_event_open (struct perf_event_attr *attr, -pid_t pid,int cpu,int group_fd, +pid_t pid,struct perf_cpu cpu,int group_fd,unsigned long flags) { -return syscall (__NR_perf_event_open, attr, pid, cpu, group_fd, flags); +return syscall (__NR_perf_event_open, attr, pid, cpu.cpu, group_fd, flags); }static int get_group_fd (struct perf_evsel *evsel,int cpu_map_idx,int thread,int *group_fd)
@@ -113,7 +113,8 @@ static int get_group_fd(struct perf_evsel *evsel, int cpu_map_idx, int thread, iint perf_evsel__open (struct perf_evsel *evsel,struct perf_cpu_map *cpus,struct perf_thread_map *threads) { -int cpu, idx, thread, err =0 ; +struct perf_cpu cpu; +int idx, thread, err =0 ;if (cpus ==NULL ) {static struct perf_cpu_map *empty_cpu_map;
@@ -252,7 +253,7 @@ int perf_evsel__mmap(struct perf_evsel *evsel, int pages)for (thread =0 ; thread <xyarray__max_y (evsel->fd); thread++) {int *fd =FD (evsel, idx, thread);struct perf_mmap *map; -int cpu =perf_cpu_map__cpu (evsel->cpus, idx); +struct perf_cpu cpu =perf_cpu_map__cpu (evsel->cpus, idx);if (fd ==NULL || *fd <0 )continue ;
tools/lib/perf/include/internal/cpumap.h +7 -2
@@ -4,6 +4,11 @@#include <linux/refcount.h> +/** A wrapper around a CPU to avoid confusion with the perf_cpu_map's map's indices. */ +struct perf_cpu { +int cpu; +}; +/** * A sized, reference counted, sorted array of integers representing CPU * numbers. This is commonly used to capture which CPUs a PMU is associated
@@ -16,13 +21,13 @@ struct perf_cpu_map {/** Length of the map array. */ int nr;/** The CPU values. */ - int map[]; + struct perf_cpu map[]; };#ifndef MAX_NR_CPUS #define MAX_NR_CPUS 2048#endif -int perf_cpu_map__idx (const struct perf_cpu_map *cpus,int cpu); +int perf_cpu_map__idx (const struct perf_cpu_map *cpus,struct perf_cpu cpu);#endif /* __LIBPERF_INTERNAL_CPUMAP_H */
tools/lib/perf/include/internal/evlist.h +2 -1
@@ -4,6 +4,7 @@#include <linux/list.h> #include <api/fd/array.h> +#include <internal/cpumap.h> #include <internal/evsel.h> #define PERF_EVLIST__HLIST_BITS 8
@@ -36,7 +37,7 @@ typedef voidtypedef struct perf_mmap * (*perf_evlist_mmap__cb_get_t )(struct perf_evlist *,bool ,int );typedef int -(*perf_evlist_mmap__cb_mmap_t )(struct perf_mmap *,struct perf_mmap_param *,int ,int ); +(*perf_evlist_mmap__cb_mmap_t )(struct perf_mmap *,struct perf_mmap_param *,int ,struct perf_cpu );struct perf_evlist_mmap_ops {perf_evlist_mmap__cb_idx_t idx;
tools/lib/perf/include/internal/evsel.h +2 -2
@@ -6,8 +6,8 @@#include <linux/perf_event.h> #include <stdbool.h> #include <sys/types.h> +#include <internal/cpumap.h> -struct perf_cpu_map ;struct perf_thread_map ;struct xyarray ;
@@ -27,7 +27,7 @@ struct perf_sample_id {* queue number. */ int idx; -int cpu; +struct perf_cpu cpu;pid_t tid;/* Holds total ID period value for PERF_SAMPLE_READ processing. */
tools/lib/perf/include/internal/mmap.h +3 -2
@@ -6,6 +6,7 @@#include <linux/refcount.h> #include <linux/types.h> #include <stdbool.h> +#include <internal/cpumap.h> /* perf sample has 16 bits size limit */ #define PERF_SAMPLE_MAX_SIZE (1 << 16)
@@ -24,7 +25,7 @@ struct perf_mmap {void *base;int mask;int fd; -int cpu; +struct perf_cpu cpu;refcount_t refcnt;u64 prev;u64 start;
@@ -46,7 +47,7 @@ size_t perf_mmap__mmap_len(struct perf_mmap *map);void perf_mmap__init (struct perf_mmap *map,struct perf_mmap *prev,bool overwrite,libperf_unmap_cb_t unmap_cb);int perf_mmap__mmap (struct perf_mmap *map,struct perf_mmap_param *mp, -int fd,int cpu); +int fd,struct perf_cpu cpu);void perf_mmap__munmap (struct perf_mmap *map);void perf_mmap__get (struct perf_mmap *map);void perf_mmap__put (struct perf_mmap *map);
tools/lib/perf/include/perf/cpumap.h +4 -5
@@ -3,11 +3,10 @@#define __LIBPERF_CPUMAP_H#include <perf/core.h> +#include <perf/cpumap.h> #include <stdio.h> #include <stdbool.h> -struct perf_cpu_map ; - LIBPERF_API struct perf_cpu_map *perf_cpu_map__dummy_new(void); LIBPERF_API struct perf_cpu_map *perf_cpu_map__default_new(void); LIBPERF_API struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list);
@@ -16,11 +15,11 @@ LIBPERF_API struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); LIBPERF_API struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig,struct perf_cpu_map *other); LIBPERF_API void perf_cpu_map__put(struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); +LIBPERF_API struct perf_cpu perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); -LIBPERF_API bool perf_cpu_map__has(const struct perf_cpu_map *map, int cpu); +LIBPERF_API struct perf_cpu perf_cpu_map__max(struct perf_cpu_map *map); +LIBPERF_API bool perf_cpu_map__has(const struct perf_cpu_map *map, struct perf_cpu cpu);#define perf_cpu_map__for_each_cpu (cpu, idx, cpus) \ for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \
tools/lib/perf/mmap.c +1 -1
@@ -32,7 +32,7 @@ size_t perf_mmap__mmap_len(struct perf_mmap *map) }int perf_mmap__mmap (struct perf_mmap *map,struct perf_mmap_param *mp, -int fd,int cpu) +int fd,struct perf_cpu cpu) { map->prev =0 ; map->mask = mp->mask;
tools/perf/bench/epoll-ctl.c +1 -1
@@ -253,7 +253,7 @@ static int do_threads(struct worker *worker, struct perf_cpu_map *cpu)if (!noaffinity) {CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset); ret =pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset);if (ret)
tools/perf/bench/epoll-wait.c +1 -1
@@ -342,7 +342,7 @@ static int do_threads(struct worker *worker, struct perf_cpu_map *cpu)if (!noaffinity) {CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset); ret =pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset);if (ret)
tools/perf/bench/futex-hash.c +1 -1
@@ -177,7 +177,7 @@ int bench_futex_hash(int argc, const char **argv) goto errmem;CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset); ret =pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset);if (ret)
tools/perf/bench/futex-lock-pi.c +1 -1
@@ -136,7 +136,7 @@ static void create_threads(struct worker *w, pthread_attr_t thread_attr, worker[i].futex = &global_futex;CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset);if (pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset))err (EXIT_FAILURE ,"pthread_attr_setaffinity_np" );
tools/perf/bench/futex-requeue.c +1 -1
@@ -131,7 +131,7 @@ static void block_threads(pthread_t *w,/* create and block all threads */ for (i =0 ; i < params.nthreads; i++) {CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset);if (pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset))err (EXIT_FAILURE ,"pthread_attr_setaffinity_np" );
tools/perf/bench/futex-wake-parallel.c +1 -1
@@ -152,7 +152,7 @@ static void block_threads(pthread_t *w, pthread_attr_t thread_attr,/* create and block all threads */ for (i =0 ; i < params.nthreads; i++) {CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset);if (pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset))err (EXIT_FAILURE ,"pthread_attr_setaffinity_np" );
tools/perf/bench/futex-wake.c +1 -1
@@ -105,7 +105,7 @@ static void block_threads(pthread_t *w,/* create and block all threads */ for (i =0 ; i < params.nthreads; i++) {CPU_ZERO (&cpuset); -CPU_SET (cpu->map[i % cpu->nr], &cpuset); +CPU_SET (perf_cpu_map__cpu (cpu, i %perf_cpu_map__nr (cpu)).cpu, &cpuset);if (pthread_attr_setaffinity_np (&thread_attr,sizeof (cpu_set_t), &cpuset))err (EXIT_FAILURE ,"pthread_attr_setaffinity_np" );
tools/perf/builtin-c2c.c +7 -6
@@ -2015,7 +2015,8 @@ static int setup_nodes(struct perf_session *session) {struct numa_node *n;unsigned long **nodes; -int node, cpu, idx; +int node, idx; +struct perf_cpu cpu;int *cpu2node;if (c2c.node_info >2 )
@@ -2038,8 +2039,8 @@ static int setup_nodes(struct perf_session *session)if (!cpu2node)return -ENOMEM ; -for (cpu =0 ; cpu < c2c.cpus_cnt; cpu++) - cpu2node[cpu] =-1 ; +for (idx =0 ; idx < c2c.cpus_cnt; idx++) + cpu2node[idx] =-1 ; c2c.cpu2node = cpu2node;
@@ -2058,12 +2059,12 @@ static int setup_nodes(struct perf_session *session)continue ;perf_cpu_map__for_each_cpu (cpu, idx, map) { -set_bit (cpu, set); +set_bit (cpu.cpu, set); -if (WARN_ONCE (cpu2node[cpu] !=-1 ,"node/cpu topology bug" )) +if (WARN_ONCE (cpu2node[cpu.cpu] !=-1 ,"node/cpu topology bug" ))return -EINVAL ; - cpu2node[cpu] = node; + cpu2node[cpu.cpu] = node; } }
tools/perf/builtin-ftrace.c +1 -1
@@ -303,7 +303,7 @@ static int set_tracing_cpumask(struct perf_cpu_map *cpumap)int ret;int last_cpu; - last_cpu =perf_cpu_map__cpu (cpumap, cpumap->nr -1 ); + last_cpu =perf_cpu_map__cpu (cpumap, cpumap->nr -1 ).cpu; mask_size = last_cpu /4 +2 ;/* one more byte for EOS */ mask_size += last_cpu /32 ;/* ',' is needed for every 32th cpus */
tools/perf/builtin-kmem.c +1 -1
@@ -192,7 +192,7 @@ static int evsel__process_alloc_node_event(struct evsel *evsel, struct perf_sampint ret =evsel__process_alloc_event (evsel, sample);if (!ret) { -int node1 =cpu__get_node (sample->cpu), +int node1 =cpu__get_node ((struct perf_cpu ){.cpu = sample->cpu}), node2 =evsel__intval (evsel, sample,"node" );if (node1 != node2)
tools/perf/builtin-record.c +1 -1
@@ -2792,7 +2792,7 @@ 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.nbits =cpu__max_cpu ().cpu; rec->affinity_mask.bits =bitmap_zalloc (rec->affinity_mask.nbits);if (!rec->affinity_mask.bits) {pr_err ("Failed to allocate thread mask for %zd cpus\n" , rec->affinity_mask.nbits);
tools/perf/builtin-sched.c +36 -29
176 lines not on this page · raw message
tools/perf/builtin-script.c +3 -2
@@ -2113,7 +2113,8 @@ static struct scripting_ops *scripting_ops;static void __process_stat (struct evsel *counter,u64 tstamp) {int nthreads =perf_thread_map__nr (counter->core.threads); -int idx, cpu, thread; +int idx, thread; +struct perf_cpu cpu;static int header_printed;if (counter->core.system_wide)
@@ -2132,7 +2133,7 @@ static void __process_stat(struct evsel *counter, u64 tstamp) counts =perf_counts (counter->counts, idx, thread); printf("%3d %8d %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %15" PRIu64 " %s\n", - cpu, + cpu.cpu, perf_thread_map__pid(counter->core.threads, thread), counts->val, counts->ena,
tools/perf/builtin-stat.c +38 -42
222 lines not on this page · raw message
tools/perf/tests/attr.c +3 -3
@@ -65,7 +65,7 @@ do { \#define WRITE_ASS (field, fmt) __WRITE_ASS(field, fmt, attr->field) -static int store_event (struct perf_event_attr *attr,pid_t pid,int cpu, +static int store_event (struct perf_event_attr *attr,pid_t pid,struct perf_cpu cpu,int fd,int group_fd,unsigned long flags) {FILE *file;
@@ -93,7 +93,7 @@ static int store_event(struct perf_event_attr *attr, pid_t pid, int cpu,/* syscall arguments */ __WRITE_ASS (fd,"d" , fd);__WRITE_ASS (group_fd,"d" , group_fd); -__WRITE_ASS (cpu,"d" , cpu); +__WRITE_ASS (cpu,"d" , cpu.cpu);__WRITE_ASS (pid,"d" , pid);__WRITE_ASS (flags,"lu" , flags);
@@ -144,7 +144,7 @@ static int store_event(struct perf_event_attr *attr, pid_t pid, int cpu,return 0 ; } -void test_attr__open (struct perf_event_attr *attr,pid_t pid,int cpu, +void test_attr__open (struct perf_event_attr *attr,pid_t pid,struct perf_cpu cpu,int fd,int group_fd,unsigned long flags) {int errno_saved = errno;
tools/perf/tests/bitmap.c +1 -1
@@ -18,7 +18,7 @@ static unsigned long *get_bitmap(const char *str, int nbits)if (map && bm) {for (i =0 ; i < map->nr; i++) -set_bit (map->map[i], bm); +set_bit (map->map[i].cpu, bm); }if (map)
tools/perf/tests/cpumap.c +3 -3
@@ -38,7 +38,7 @@ static int process_event_mask(struct perf_tool *tool __maybe_unused,TEST_ASSERT_VAL ("wrong nr" , map->nr ==20 );for (i =0 ; i <20 ; i++) { -TEST_ASSERT_VAL ("wrong cpu" , map->map[i] == i); +TEST_ASSERT_VAL ("wrong cpu" , map->map[i].cpu == i); }perf_cpu_map__put (map);
@@ -67,8 +67,8 @@ static int process_event_cpus(struct perf_tool *tool __maybe_unused, map =cpu_map__new_data (data);TEST_ASSERT_VAL ("wrong nr" , map->nr ==2 ); -TEST_ASSERT_VAL ("wrong cpu" , map->map[0 ] ==1 ); -TEST_ASSERT_VAL ("wrong cpu" , map->map[1 ] ==256 ); +TEST_ASSERT_VAL ("wrong cpu" , map->map[0 ].cpu ==1 ); +TEST_ASSERT_VAL ("wrong cpu" , map->map[1 ].cpu ==256 );TEST_ASSERT_VAL ("wrong refcnt" ,refcount_read (&map->refcnt) ==1 );perf_cpu_map__put (map);return 0 ;
tools/perf/tests/event_update.c +3 -3
@@ -76,9 +76,9 @@ static int process_event_cpus(struct perf_tool *tool __maybe_unused,TEST_ASSERT_VAL ("wrong id" , ev->id ==123 );TEST_ASSERT_VAL ("wrong type" , ev->type ==PERF_EVENT_UPDATE__CPUS );TEST_ASSERT_VAL ("wrong cpus" , map->nr ==3 ); -TEST_ASSERT_VAL ("wrong cpus" , map->map[0 ] ==1 ); -TEST_ASSERT_VAL ("wrong cpus" , map->map[1 ] ==2 ); -TEST_ASSERT_VAL ("wrong cpus" , map->map[2 ] ==3 ); +TEST_ASSERT_VAL ("wrong cpus" , map->map[0 ].cpu ==1 ); +TEST_ASSERT_VAL ("wrong cpus" , map->map[1 ].cpu ==2 ); +TEST_ASSERT_VAL ("wrong cpus" , map->map[2 ].cpu ==3 );perf_cpu_map__put (map);return 0 ; }
tools/perf/tests/mem2node.c +1 -1
@@ -31,7 +31,7 @@ static unsigned long *get_bitmap(const char *str, int nbits)if (map && bm) {for (i =0 ; i < map->nr; i++) { -set_bit (map->map[i], bm); +set_bit (map->map[i].cpu, bm); } }
tools/perf/tests/mmap-basic.c +2 -2
@@ -59,11 +59,11 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest }CPU_ZERO (&cpu_set); -CPU_SET (cpus->map[0 ], &cpu_set); +CPU_SET (cpus->map[0 ].cpu, &cpu_set);sched_setaffinity (0 ,sizeof (cpu_set), &cpu_set);if (sched_setaffinity (0 ,sizeof (cpu_set), &cpu_set) <0 ) {pr_debug ("sched_setaffinity() failed on CPU %d: %s " , - cpus->map[0 ],str_error_r (errno, sbuf,sizeof (sbuf))); + cpus->map[0 ].cpu,str_error_r (errno, sbuf,sizeof (sbuf))); goto out_free_cpus; }
tools/perf/tests/openat-syscall-all-cpus.c +9 -8
@@ -22,7 +22,8 @@static int test__openat_syscall_event_on_all_cpus (struct test_suite *test__maybe_unused ,int subtest__maybe_unused ) { -int err =-1 , fd, idx, cpu; +int err =-1 , fd, idx; +struct perf_cpu cpu;struct perf_cpu_map *cpus;struct evsel *evsel;unsigned int nr_openat_calls =111 , i;
@@ -66,15 +67,15 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb* without CPU_ALLOC. 1024 cpus in 2010 still seems * a reasonable upper limit tho :-) */ -if (cpu >=CPU_SETSIZE ) { -pr_debug ("Ignoring CPU %d\n" , cpu); +if (cpu.cpu >=CPU_SETSIZE ) { +pr_debug ("Ignoring CPU %d\n" , cpu.cpu);continue ; } -CPU_SET (cpu, &cpu_set); +CPU_SET (cpu.cpu, &cpu_set);if (sched_setaffinity (0 ,sizeof (cpu_set), &cpu_set) <0 ) {pr_debug ("sched_setaffinity() failed on CPU %d: %s " , - cpu, + cpu.cpu,str_error_r (errno, sbuf,sizeof (sbuf))); goto out_close_fd; }
@@ -82,7 +83,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb fd =openat (0 ,"/etc/passwd" ,O_RDONLY );close (fd); } -CPU_CLR (cpu, &cpu_set); +CPU_CLR (cpu.cpu, &cpu_set); } evsel->core.cpus =perf_cpu_map__get (cpus);
@@ -92,7 +93,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __maybperf_cpu_map__for_each_cpu (cpu, idx, cpus) {unsigned int expected; -if (cpu >=CPU_SETSIZE ) +if (cpu.cpu >=CPU_SETSIZE )continue ;if (evsel__read_on_cpu (evsel, idx,0 ) <0 ) {
@@ -104,7 +105,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb expected = nr_openat_calls + idx;if (perf_counts (evsel->counts, idx,0 )->val != expected) {pr_debug ("evsel__read_on_cpu: expected to intercept %d calls on cpu %d, got %" PRIu64"\n" , - expected, cpu,perf_counts (evsel->counts, idx,0 )->val); + expected, cpu.cpu,perf_counts (evsel->counts, idx,0 )->val); err =-1 ; } }
tools/perf/tests/stat.c +2 -1
@@ -87,7 +87,8 @@ static int test__synthesize_stat(struct test_suite *test __maybe_unused, int sub count.run =300 ;TEST_ASSERT_VAL ("failed to synthesize stat_config" , - !perf_event__synthesize_stat (NULL ,1 ,2 ,3 , &count, process_stat_event,NULL )); + !perf_event__synthesize_stat (NULL , (struct perf_cpu ){.cpu =1 },2 ,3 , + &count, process_stat_event,NULL ));return 0 ; }
tools/perf/tests/topology.c +16 -14
@@ -112,7 +112,9 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Session header CPU map not set" , session->header.env.cpu);for (i =0 ; i < session->header.env.nr_cpus_avail; i++) { -if (!perf_cpu_map__has (map, i)) +struct perf_cpu cpu = { .cpu = i }; + +if (!perf_cpu_map__has (map, cpu))continue ; pr_debug("CPU %d, core %d, socket %d\n", i, session->header.env.cpu[i].core_id,
@@ -122,15 +124,15 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)// Test that CPU ID contains socket, die, core and CPU for (i =0 ; i < map->nr; i++) { id =aggr_cpu_id__cpu (perf_cpu_map__cpu (map, i),NULL ); -TEST_ASSERT_VAL ("Cpu map - CPU ID doesn't match" , map->map[i] == id.cpu); +TEST_ASSERT_VAL ("Cpu map - CPU ID doesn't match" , map->map[i].cpu == id.cpu.cpu);TEST_ASSERT_VAL ("Cpu map - Core ID doesn't match" , - session->header.env.cpu[map->map[i]].core_id == id.core); + session->header.env.cpu[map->map[i].cpu].core_id == id.core);TEST_ASSERT_VAL ("Cpu map - Socket ID doesn't match" , - session->header.env.cpu[map->map[i]].socket_id == id.socket); + session->header.env.cpu[map->map[i].cpu].socket_id == id.socket);TEST_ASSERT_VAL ("Cpu map - Die ID doesn't match" , - session->header.env.cpu[map->map[i]].die_id == id.die); + session->header.env.cpu[map->map[i].cpu].die_id == id.die);TEST_ASSERT_VAL ("Cpu map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Cpu map - Thread is set" , id.thread ==-1 ); }
@@ -139,13 +141,13 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)for (i =0 ; i < map->nr; i++) { id =aggr_cpu_id__core (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Core map - Core ID doesn't match" , - session->header.env.cpu[map->map[i]].core_id == id.core); + session->header.env.cpu[map->map[i].cpu].core_id == id.core);TEST_ASSERT_VAL ("Core map - Socket ID doesn't match" , - session->header.env.cpu[map->map[i]].socket_id == id.socket); + session->header.env.cpu[map->map[i].cpu].socket_id == id.socket);TEST_ASSERT_VAL ("Core map - Die ID doesn't match" , - session->header.env.cpu[map->map[i]].die_id == id.die); + session->header.env.cpu[map->map[i].cpu].die_id == id.die);TEST_ASSERT_VAL ("Core map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Core map - Thread is set" , id.thread ==-1 ); }
@@ -154,14 +156,14 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)for (i =0 ; i < map->nr; i++) { id =aggr_cpu_id__die (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Die map - Socket ID doesn't match" , - session->header.env.cpu[map->map[i]].socket_id == id.socket); + session->header.env.cpu[map->map[i].cpu].socket_id == id.socket);TEST_ASSERT_VAL ("Die map - Die ID doesn't match" , - session->header.env.cpu[map->map[i]].die_id == id.die); + session->header.env.cpu[map->map[i].cpu].die_id == id.die);TEST_ASSERT_VAL ("Die map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Die map - Core is set" , id.core ==-1 ); -TEST_ASSERT_VAL ("Die map - CPU is set" , id.cpu ==-1 ); +TEST_ASSERT_VAL ("Die map - CPU is set" , id.cpu.cpu ==-1 );TEST_ASSERT_VAL ("Die map - Thread is set" , id.thread ==-1 ); }
@@ -169,12 +171,12 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)for (i =0 ; i < map->nr; i++) { id =aggr_cpu_id__socket (perf_cpu_map__cpu (map, i),NULL );TEST_ASSERT_VAL ("Socket map - Socket ID doesn't match" , - session->header.env.cpu[map->map[i]].socket_id == id.socket); + session->header.env.cpu[map->map[i].cpu].socket_id == id.socket);TEST_ASSERT_VAL ("Socket map - Node ID is set" , id.node ==-1 );TEST_ASSERT_VAL ("Socket map - Die ID is set" , id.die ==-1 );TEST_ASSERT_VAL ("Socket map - Core is set" , id.core ==-1 ); -TEST_ASSERT_VAL ("Socket map - CPU is set" , id.cpu ==-1 ); +TEST_ASSERT_VAL ("Socket map - CPU is set" , id.cpu.cpu ==-1 );TEST_ASSERT_VAL ("Socket map - Thread is set" , id.thread ==-1 ); }
@@ -186,7 +188,7 @@ static int check_cpu_topology(char *path, struct perf_cpu_map *map)TEST_ASSERT_VAL ("Node map - Socket is set" , id.socket ==-1 );TEST_ASSERT_VAL ("Node map - Die ID is set" , id.die ==-1 );TEST_ASSERT_VAL ("Node map - Core is set" , id.core ==-1 ); -TEST_ASSERT_VAL ("Node map - CPU is set" , id.cpu ==-1 ); +TEST_ASSERT_VAL ("Node map - CPU is set" , id.cpu.cpu ==-1 );TEST_ASSERT_VAL ("Node map - Thread is set" , id.thread ==-1 ); }perf_session__delete (session);
tools/perf/util/affinity.c +1 -1
@@ -11,7 +11,7 @@static int get_cpu_set_size (void ) { -int sz =cpu__max_cpu () +8 -1 ; +int sz =cpu__max_cpu ().cpu +8 -1 ;/* * sched_getaffinity doesn't like masks smaller than the kernel. * Hopefully that's big enough.
tools/perf/util/auxtrace.c +6 -6
@@ -123,7 +123,7 @@ int auxtrace_mmap__mmap(struct auxtrace_mmap *mm, mm->prev =0 ; mm->idx = mp->idx; mm->tid = mp->tid; - mm->cpu = mp->cpu; + mm->cpu = mp->cpu.cpu;if (!mp->len) { mm->base =NULL ;
@@ -180,7 +180,7 @@ void auxtrace_mmap_params__set_idx(struct auxtrace_mmap_params *mp,else mp->tid =-1 ; }else { - mp->cpu =-1 ; + mp->cpu.cpu =-1 ; mp->tid =perf_thread_map__pid (evlist->core.threads, idx); } }
@@ -292,7 +292,7 @@ static int auxtrace_queues__queue_buffer(struct auxtrace_queues *queues,if (!queue->set) { queue->set = true; queue->tid = buffer->tid; - queue->cpu = buffer->cpu; + queue->cpu = buffer->cpu.cpu; } buffer->buffer_nr = queues->next_buffer_nr++;
@@ -339,11 +339,11 @@ static int auxtrace_queues__split_buffer(struct auxtrace_queues *queues,return 0 ; } -static bool filter_cpu (struct perf_session *session,int cpu) +static bool filter_cpu (struct perf_session *session,struct perf_cpu cpu) {unsigned long *cpu_bitmap = session->itrace_synth_opts->cpu_bitmap; -return cpu_bitmap && cpu !=-1 && !test_bit (cpu, cpu_bitmap); +return cpu_bitmap && cpu.cpu !=-1 && !test_bit (cpu.cpu, cpu_bitmap); } static int auxtrace_queues__add_buffer(struct auxtrace_queues *queues,
@@ -399,7 +399,7 @@ int auxtrace_queues__add_event(struct auxtrace_queues *queues,struct auxtrace_buffer buffer = { .pid =-1 , .tid = event->auxtrace.tid, - .cpu = event->auxtrace.cpu, + .cpu = { event->auxtrace.cpu }, .data_offset = data_offset, .offset = event->auxtrace.offset, .reference = event->auxtrace.reference,
tools/perf/util/auxtrace.h +3 -2
@@ -15,6 +15,7 @@#include <linux/list.h> #include <linux/perf_event.h> #include <linux/types.h> +#include <internal/cpumap.h> #include <asm/bitsperlong.h> #include <asm/barrier.h>
@@ -240,7 +241,7 @@ struct auxtrace_buffer {size_t size;pid_t pid;pid_t tid; -int cpu; +struct perf_cpu cpu;void *data;off_t data_offset;void *mmap_addr;
@@ -350,7 +351,7 @@ struct auxtrace_mmap_params {int prot;int idx;pid_t tid; -int cpu; +struct perf_cpu cpu; };/**
tools/perf/util/cpumap.c +47 -44
288 lines not on this page · raw message
tools/perf/util/cpumap.h +13 -13
@@ -23,7 +23,7 @@ struct aggr_cpu_id {/** The core id as read from /sys/devices/system/cpu/cpuX/topology/core_id. */ int core;/** CPU aggregation, note there is one CPU for each SMT thread. */ -int cpu; +struct perf_cpu cpu; };/** A collection of aggr_cpu_id values, the "built" version is sorted and uniqued. */
@@ -48,28 +48,28 @@ const struct perf_cpu_map *cpu_map__online(void); /* thread unsafe */int cpu__setup_cpunode_map (void );int cpu__max_node (void ); -int cpu__max_cpu (void ); -int cpu__max_present_cpu (void ); +struct perf_cpu cpu__max_cpu (void ); +struct perf_cpu cpu__max_present_cpu (void );/** * cpu__get_node - Returns the numa node X as read from * /sys/devices/system/node/nodeX for the given CPU. */ -int cpu__get_node (int cpu); +int cpu__get_node (struct perf_cpu cpu);/** * cpu__get_socket_id - Returns the socket number as read from * /sys/devices/system/cpu/cpuX/topology/physical_package_id for the given CPU. */ -int cpu__get_socket_id (int cpu); +int cpu__get_socket_id (struct perf_cpu cpu);/** * cpu__get_die_id - Returns the die id as read from * /sys/devices/system/cpu/cpuX/topology/die_id for the given CPU. */ -int cpu__get_die_id (int cpu); +int cpu__get_die_id (struct perf_cpu cpu);/** * cpu__get_core_id - Returns the core id as read from * /sys/devices/system/cpu/cpuX/topology/core_id for the given CPU. */ -int cpu__get_core_id (int cpu); +int cpu__get_core_id (struct perf_cpu cpu);/** * cpu_aggr_map__empty_new - Create a cpu_aggr_map of size nr with every entry
@@ -77,7 +77,7 @@ int cpu__get_core_id(int cpu);*/ struct cpu_aggr_map *cpu_aggr_map__empty_new (int nr); -typedef struct aggr_cpu_id (*aggr_cpu_id_get_t )(int cpu,void *data); +typedef struct aggr_cpu_id (*aggr_cpu_id_get_t )(struct perf_cpu cpu,void *data);/** * cpu_aggr_map__new - Create a cpu_aggr_map with an aggr_cpu_id for each cpu in
@@ -98,29 +98,29 @@ struct aggr_cpu_id aggr_cpu_id__empty(void);* the socket for cpu. The function signature is compatible with * aggr_cpu_id_get_t. */ -struct aggr_cpu_id aggr_cpu_id__socket (int cpu,void *data); +struct aggr_cpu_id aggr_cpu_id__socket (struct perf_cpu cpu,void *data);/** * aggr_cpu_id__die - Create an aggr_cpu_id with the die and socket populated * with the die and socket for cpu. The function signature is compatible with * aggr_cpu_id_get_t. */ -struct aggr_cpu_id aggr_cpu_id__die (int cpu,void *data); +struct aggr_cpu_id aggr_cpu_id__die (struct perf_cpu cpu,void *data);/** * aggr_cpu_id__core - Create an aggr_cpu_id with the core, die and socket * populated with the core, die and socket for cpu. The function signature is * compatible with aggr_cpu_id_get_t. */ -struct aggr_cpu_id aggr_cpu_id__core (int cpu,void *data); +struct aggr_cpu_id aggr_cpu_id__core (struct perf_cpu cpu,void *data);/** * aggr_cpu_id__core - Create an aggr_cpu_id with the cpu, core, die and socket * populated with the cpu, core, die and socket for cpu. The function signature * is compatible with aggr_cpu_id_get_t. */ -struct aggr_cpu_id aggr_cpu_id__cpu (int cpu,void *data); +struct aggr_cpu_id aggr_cpu_id__cpu (struct perf_cpu cpu,void *data);/** * aggr_cpu_id__node - Create an aggr_cpu_id with the numa node populated for * cpu. The function signature is compatible with aggr_cpu_id_get_t. */ -struct aggr_cpu_id aggr_cpu_id__node (int cpu,void *data); +struct aggr_cpu_id aggr_cpu_id__node (struct perf_cpu cpu,void *data);#endif /* __PERF_CPUMAP_H */
tools/perf/util/cputopo.c +3 -3
@@ -187,7 +187,7 @@ struct cpu_topology *cpu_topology__new(void)struct perf_cpu_map *map;bool has_die =has_die_topology (); - ncpus =cpu__max_present_cpu (); + ncpus =cpu__max_present_cpu ().cpu;/* build online CPU map */ map =perf_cpu_map__new (NULL );
@@ -218,7 +218,7 @@ struct cpu_topology *cpu_topology__new(void) tp->core_cpus_list = addr;for (i =0 ; i < nr; i++) { -if (!perf_cpu_map__has (map, i)) +if (!perf_cpu_map__has (map, (struct perf_cpu ){ .cpu = i }))continue ; ret =build_cpu_topology (tp, i);
@@ -333,7 +333,7 @@ struct numa_topology *numa_topology__new(void) tp->nr = nr;for (i =0 ; i < nr; i++) { -if (load_numa_node (&tp->nodes[i], node_map->map[i])) { +if (load_numa_node (&tp->nodes[i], node_map->map[i].cpu)) {numa_topology__delete (tp); tp =NULL ;break ;
tools/perf/util/env.c +16 -13
@@ -285,13 +285,13 @@ int perf_env__set_cmdline(struct perf_env *env, int argc, const char *argv[])int perf_env__read_cpu_topology_map (struct perf_env *env) { -int cpu, nr_cpus; +int idx, nr_cpus;if (env->cpu !=NULL )return 0 ;if (env->nr_cpus_avail ==0 ) - env->nr_cpus_avail =cpu__max_present_cpu (); + env->nr_cpus_avail =cpu__max_present_cpu ().cpu; nr_cpus = env->nr_cpus_avail;if (nr_cpus ==-1 )
@@ -301,10 +301,12 @@ int perf_env__read_cpu_topology_map(struct perf_env *env)if (env->cpu ==NULL )return -ENOMEM ; -for (cpu =0 ; cpu < nr_cpus; ++cpu) { - env->cpu[cpu].core_id =cpu__get_core_id (cpu); - env->cpu[cpu].socket_id =cpu__get_socket_id (cpu); - env->cpu[cpu].die_id =cpu__get_die_id (cpu); +for (idx =0 ; idx < nr_cpus; ++idx) { +struct perf_cpu cpu = { .cpu = idx }; + + env->cpu[idx].core_id =cpu__get_core_id (cpu); + env->cpu[idx].socket_id =cpu__get_socket_id (cpu); + env->cpu[idx].die_id =cpu__get_die_id (cpu); } env->nr_cpus_avail = nr_cpus;
@@ -381,7 +383,7 @@ static int perf_env__read_arch(struct perf_env *env)static int perf_env__read_nr_cpus_avail (struct perf_env *env) {if (env->nr_cpus_avail ==0 ) - env->nr_cpus_avail =cpu__max_present_cpu (); + env->nr_cpus_avail =cpu__max_present_cpu ().cpu;return env->nr_cpus_avail ?0 : -ENOENT ; }
@@ -487,7 +489,7 @@ const char *perf_env__pmu_mappings(struct perf_env *env)return env->pmu_mappings; } -int perf_env__numa_node (struct perf_env *env,int cpu) +int perf_env__numa_node (struct perf_env *env,struct perf_cpu cpu) {if (!env->nr_numa_map) {struct numa_node *nn;
@@ -495,7 +497,7 @@ int perf_env__numa_node(struct perf_env *env, int cpu)for (i =0 ; i < env->nr_numa_nodes; i++) { nn = &env->numa_nodes[i]; - nr =max (nr,perf_cpu_map__max (nn->map)); + nr =max (nr,perf_cpu_map__max (nn->map).cpu); } nr++;
@@ -514,13 +516,14 @@ int perf_env__numa_node(struct perf_env *env, int cpu) env->nr_numa_map = nr;for (i =0 ; i < env->nr_numa_nodes; i++) { -int tmp, j; +struct perf_cpu tmp; +int j; nn = &env->numa_nodes[i]; -perf_cpu_map__for_each_cpu (j, tmp, nn->map) - env->numa_map[j] = i; +perf_cpu_map__for_each_cpu (tmp, j, nn->map) + env->numa_map[tmp.cpu] = i; } } -return cpu >=0 && cpu < env->nr_numa_map ? env->numa_map[cpu] :-1 ; +return cpu.cpu >=0 && cpu.cpu < env->nr_numa_map ? env->numa_map[cpu.cpu] :-1 ; }
tools/perf/util/env.h +2 -1
@@ -4,6 +4,7 @@#include <linux/types.h> #include <linux/rbtree.h> +#include "cpumap.h" #include "rwsem.h" struct perf_cpu_map ;
@@ -170,5 +171,5 @@ struct bpf_prog_info_node *perf_env__find_bpf_prog_info(struct perf_env *env,bool perf_env__insert_btf (struct perf_env *env,struct btf_node *btf_node);struct btf_node *perf_env__find_btf (struct perf_env *env,__u32 btf_id); -int perf_env__numa_node (struct perf_env *env,int cpu); +int perf_env__numa_node (struct perf_env *env,struct perf_cpu cpu);#endif /* __PERF_ENV_H */
tools/perf/util/evlist.c +4 -4
@@ -350,13 +350,13 @@ struct evlist_cpu_iterator evlist__cpu_begin(struct evlist *evlist, struct affin .cpu_map_idx = 0, .evlist_cpu_map_idx = 0, .evlist_cpu_map_nr = perf_cpu_map__nr(evlist->core.all_cpus), - .cpu = -1, + .cpu = (struct perf_cpu){ .cpu = -1}, .affinity = affinity, };if (itr.affinity) { itr.cpu =perf_cpu_map__cpu (evlist->core.all_cpus,0 ); -affinity__set (itr.affinity, itr.cpu); +affinity__set (itr.affinity, itr.cpu.cpu); itr.cpu_map_idx =perf_cpu_map__idx (itr.evsel->core.cpus, itr.cpu);/* * If this CPU isn't in the evsel's cpu map then advance through
@@ -385,7 +385,7 @@ void evlist_cpu_iterator__next(struct evlist_cpu_iterator *evlist_cpu_itr)perf_cpu_map__cpu (evlist_cpu_itr->container->core.all_cpus, evlist_cpu_itr->evlist_cpu_map_idx);if (evlist_cpu_itr->affinity) -affinity__set (evlist_cpu_itr->affinity, evlist_cpu_itr->cpu); +affinity__set (evlist_cpu_itr->affinity, evlist_cpu_itr->cpu.cpu); evlist_cpu_itr->cpu_map_idx =perf_cpu_map__idx (evlist_cpu_itr->evsel->core.cpus, evlist_cpu_itr->cpu);
@@ -819,7 +819,7 @@ perf_evlist__mmap_cb_get(struct perf_evlist *_evlist, bool overwrite, int idx)static int perf_evlist__mmap_cb_mmap (struct perf_mmap *_map,struct perf_mmap_param *_mp, -int output,int cpu) +int output,struct perf_cpu cpu) { struct mmap *map = container_of(_map, struct mmap, core); struct mmap_params *mp = container_of(_mp, struct mmap_params, core);
tools/perf/util/evlist.h +1 -1
@@ -344,7 +344,7 @@ struct evlist_cpu_iterator {/** The number of CPU map entries in evlist->core.all_cpus. */ int evlist_cpu_map_nr;/** The current CPU of the iterator. */ -int cpu; +struct perf_cpu cpu;/** If present, used to set the affinity when switching between CPUs. */ struct affinity *affinity; };
tools/perf/util/evsel.c +3 -3
@@ -1594,7 +1594,7 @@ int __evsel__read_on_cpu(struct evsel *evsel, int cpu_map_idx, int thread, boolstatic int evsel__match_other_cpu (struct evsel *evsel,struct evsel *other,int cpu_map_idx) { -int cpu; +struct perf_cpu cpu; cpu =perf_cpu_map__cpu (evsel->core.cpus, cpu_map_idx);return perf_cpu_map__idx (other->core.cpus, cpu);
@@ -2020,9 +2020,9 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,test_attr__ready ();pr_debug2_peo ("sys_perf_event_open: pid %d cpu %d group_fd %d flags %#lx" , - pid, cpus->map[idx], group_fd, evsel->open_flags); + pid, cpus->map[idx].cpu, group_fd, evsel->open_flags); - fd =sys_perf_event_open (&evsel->core.attr, pid, cpus->map[idx], + fd =sys_perf_event_open (&evsel->core.attr, pid, cpus->map[idx].cpu, group_fd, evsel->open_flags);FD (evsel, idx, thread) = fd;
tools/perf/util/expr.c +1 -1
@@ -400,7 +400,7 @@ double expr__get_literal(const char *literal)return smt_on () >0 ?1.0 :0.0 ;if (!strcmp ("#num_cpus" , literal)) -return cpu__max_present_cpu (); +return cpu__max_present_cpu ().cpu;/* * Assume that topology strings are consistent, such as CPUs "0-1"
tools/perf/util/header.c +3 -3
@@ -472,7 +472,7 @@ static int write_nrcpus(struct feat_fd *ff,u32 nrc, nra;int ret; - nrc =cpu__max_present_cpu (); + nrc =cpu__max_present_cpu ().cpu; nr =sysconf (_SC_NPROCESSORS_ONLN);if (nr <0 )
@@ -1163,7 +1163,7 @@ static int build_caches(struct cpu_cache_level caches[], u32 *cntp)u32 nr, cpu;u16 level; - nr =cpu__max_cpu (); + nr =cpu__max_cpu ().cpu;for (cpu =0 ; cpu < nr; cpu++) {for (level =0 ; level <MAX_CACHE_LVL ; level++) {
@@ -1195,7 +1195,7 @@ static int build_caches(struct cpu_cache_level caches[], u32 *cntp)static int write_cache (struct feat_fd *ff,struct evlist *evlist__maybe_unused ) { -u32 max_caches =cpu__max_cpu () *MAX_CACHE_LVL ; +u32 max_caches =cpu__max_cpu ().cpu *MAX_CACHE_LVL ;struct cpu_cache_level caches[max_caches];u32 cnt =0 , i, version =1 ;int ret;
tools/perf/util/mmap.c +10 -9
@@ -94,7 +94,7 @@ static void perf_mmap__aio_free(struct mmap *map, int idx) } } -static int perf_mmap__aio_bind (struct mmap *map,int idx,int cpu,int affinity) +static int perf_mmap__aio_bind (struct mmap *map,int idx,struct perf_cpu cpu,int affinity) {void *data;size_t mmap_len;
@@ -138,7 +138,7 @@ static void perf_mmap__aio_free(struct mmap *map, int idx) }static int perf_mmap__aio_bind (struct mmap *map__maybe_unused ,int idx__maybe_unused , -int cpu__maybe_unused ,int affinity__maybe_unused ) +struct perf_cpu cpu__maybe_unused ,int affinity__maybe_unused ) {return 0 ; }
@@ -240,7 +240,8 @@ void mmap__munmap(struct mmap *map)static void build_node_mask (int node,struct mmap_cpu_mask *mask) { -int c, cpu, nr_cpus; +int idx, nr_cpus; +struct perf_cpu cpu;const struct perf_cpu_map *cpu_map =NULL ; cpu_map =cpu_map__online ();
@@ -248,16 +249,16 @@ static void build_node_mask(int node, struct mmap_cpu_mask *mask)return ; nr_cpus =perf_cpu_map__nr (cpu_map); -for (c =0 ; c < nr_cpus; c++) { - cpu = cpu_map->map[c];/* map c index to online cpu index */ +for (idx =0 ; idx < nr_cpus; idx++) { + cpu = cpu_map->map[idx];/* map c index to online cpu index */ if (cpu__get_node (cpu) == node) -set_bit (cpu, mask->bits); +set_bit (cpu.cpu, mask->bits); } }static int perf_mmap__setup_affinity_mask (struct mmap *map,struct mmap_params *mp) { - map->affinity_mask.nbits =cpu__max_cpu (); + map->affinity_mask.nbits =cpu__max_cpu ().cpu; map->affinity_mask.bits =bitmap_zalloc (map->affinity_mask.nbits);if (!map->affinity_mask.bits)return -1 ;
@@ -265,12 +266,12 @@ static int perf_mmap__setup_affinity_mask(struct mmap *map, struct mmap_params *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 ) -set_bit (map->core.cpu, map->affinity_mask.bits); +set_bit (map->core.cpu.cpu, map->affinity_mask.bits);return 0 ; } -int mmap__mmap (struct mmap *map,struct mmap_params *mp,int fd,int cpu) +int mmap__mmap (struct mmap *map,struct mmap_params *mp,int fd,struct perf_cpu cpu) {if (perf_mmap__mmap (&map->core, &mp->core, fd, cpu)) { pr_debug2("failed to mmap perf event ring buffer, error %d\n",
tools/perf/util/mmap.h +2 -1
@@ -7,6 +7,7 @@#include <linux/types.h> #include <linux/ring_buffer.h> #include <linux/bitops.h> +#include <perf/cpumap.h> #include <stdbool.h> #include <pthread.h> // for cpu_set_t #ifdef HAVE_AIO_SUPPORT
@@ -52,7 +53,7 @@ struct mmap_params {struct auxtrace_mmap_params auxtrace_mp; }; -int mmap__mmap (struct mmap *map,struct mmap_params *mp,int fd,int cpu); +int mmap__mmap (struct mmap *map,struct mmap_params *mp,int fd,struct perf_cpu cpu);void mmap__munmap (struct mmap *map);union perf_event *perf_mmap__read_forward (struct mmap *map);
tools/perf/util/perf_api_probe.c +9 -6
@@ -11,7 +11,7 @@typedef void (*setup_probe_fn_t )(struct evsel *evsel); -static int perf_do_probe_api (setup_probe_fn_t fn,int cpu,const char *str) +static int perf_do_probe_api (setup_probe_fn_t fn,struct perf_cpu cpu,const char *str) {struct evlist *evlist;struct evsel *evsel;
@@ -29,7 +29,7 @@ static int perf_do_probe_api(setup_probe_fn_t fn, int cpu, const char *str) evsel =evlist__first (evlist);while (1 ) { - fd =sys_perf_event_open (&evsel->core.attr, pid, cpu,-1 , flags); + fd =sys_perf_event_open (&evsel->core.attr, pid, cpu.cpu,-1 , flags);if (fd <0 ) {if (pid ==-1 && errno ==EACCES ) { pid =0 ;
@@ -43,7 +43,7 @@ static int perf_do_probe_api(setup_probe_fn_t fn, int cpu, const char *str)fn (evsel); - fd =sys_perf_event_open (&evsel->core.attr, pid, cpu,-1 , flags); + fd =sys_perf_event_open (&evsel->core.attr, pid, cpu.cpu,-1 , flags);if (fd <0 ) {if (errno ==EINVAL ) err = -EINVAL ;
@@ -61,7 +61,8 @@ static bool perf_probe_api(setup_probe_fn_t fn) {const char *try[] = {"cycles:u" ,"instructions:u" ,"cpu-clock:u" ,NULL };struct perf_cpu_map *cpus; -int cpu, ret, i =0 ; +struct perf_cpu cpu; +int ret, i =0 ; cpus =perf_cpu_map__new (NULL );if (!cpus)
@@ -136,15 +137,17 @@ bool perf_can_record_cpu_wide(void) .exclude_kernel = 1, };struct perf_cpu_map *cpus; -int cpu, fd; +struct perf_cpu cpu; +int fd; cpus =perf_cpu_map__new (NULL );if (!cpus)return false; + cpu = cpus->map[0 ];perf_cpu_map__put (cpus); - fd =sys_perf_event_open (&attr,-1 , cpu,-1 ,0 ); + fd =sys_perf_event_open (&attr,-1 , cpu.cpu,-1 ,0 );if (fd <0 )return false;close (fd);
tools/perf/util/python.c +2 -2
@@ -1057,7 +1057,7 @@ static struct mmap *get_md(struct evlist *evlist, int cpu)for (i =0 ; i < evlist->core.nr_mmaps; i++) {struct mmap *md = &evlist->mmap[i]; -if (md->core.cpu_map_idx == cpu) +if (md->core.cpu == cpu)return md; }
@@ -1443,7 +1443,7 @@ PyMODINIT_FUNC PyInit_perf(void)* Dummy, to avoid dragging all the test_attr infrastructure in the python * binding. */ -void test_attr__open (struct perf_event_attr *attr,pid_t pid,int cpu, +void test_attr__open (struct perf_event_attr *attr,pid_t pid,struct perf_cpu cpu,int fd,int group_fd,unsigned long flags) { }
tools/perf/util/record.c +7 -4
@@ -106,7 +106,7 @@ void evlist__config(struct evlist *evlist, struct record_opts *opts, struct callif (opts->group)evlist__set_leader (evlist); -if (evlist->core.cpus->map[0 ] <0 ) +if (evlist->core.cpus->map[0 ].cpu <0 ) opts->no_inherit = true; use_comm_exec =perf_can_comm_exec ();
@@ -229,7 +229,8 @@ bool evlist__can_select_event(struct evlist *evlist, const char *str) {struct evlist *temp_evlist;struct evsel *evsel; -int err, fd, cpu; +int err, fd; +struct perf_cpu cpu = { .cpu =0 };bool ret = false;pid_t pid =-1 ;
@@ -246,14 +247,16 @@ bool evlist__can_select_event(struct evlist *evlist, const char *str)if (!evlist ||perf_cpu_map__empty (evlist->core.cpus)) {struct perf_cpu_map *cpus =perf_cpu_map__new (NULL ); - cpu = cpus ? cpus->map[0 ] :0 ; +if (cpus) + cpu = cpus->map[0 ]; +perf_cpu_map__put (cpus); }else { cpu = evlist->core.cpus->map[0 ]; }while (1 ) { - fd =sys_perf_event_open (&evsel->core.attr, pid, cpu,-1 , + fd =sys_perf_event_open (&evsel->core.attr, pid, cpu.cpu,-1 ,perf_event_open_cloexec_flag ());if (fd <0 ) {if (pid ==-1 && errno ==EACCES ) {
tools/perf/util/scripting-engines/trace-event-python.c +3 -3
@@ -1553,7 +1553,7 @@ static void get_handler_name(char *str, size_t size, }static void -process_stat (struct evsel *counter,int cpu,int thread,u64 tstamp, +process_stat (struct evsel *counter,struct perf_cpu cpu,int thread,u64 tstamp,struct perf_counts_values *count) {PyObject *handler, *t;
@@ -1573,7 +1573,7 @@ process_stat(struct evsel *counter, int cpu, int thread, u64 tstamp,return ; } -PyTuple_SetItem (t, n++,_PyLong_FromLong (cpu)); +PyTuple_SetItem (t, n++,_PyLong_FromLong (cpu.cpu));PyTuple_SetItem (t, n++,_PyLong_FromLong (thread));tuple_set_u64 (t, n++, tstamp);
@@ -1597,7 +1597,7 @@ static void python_process_stat(struct perf_stat_config *config,int cpu, thread;if (config->aggr_mode ==AGGR_GLOBAL ) { -process_stat (counter,-1 ,-1 , tstamp, +process_stat (counter, (struct perf_cpu ){ .cpu =-1 },-1 , tstamp, &counter->counts->aggr);return ; }
tools/perf/util/session.c +5 -5
@@ -2537,15 +2537,15 @@ int perf_session__cpu_bitmap(struct perf_session *session, }for (i =0 ; i < map->nr; i++) { -int cpu = map->map[i]; +struct perf_cpu cpu = map->map[i]; -if (cpu >= nr_cpus) { +if (cpu.cpu >= nr_cpus) {pr_err ("Requested CPU %d too large. " -"Consider raising MAX_NR_CPUS\n" , cpu); +"Consider raising MAX_NR_CPUS\n" , cpu.cpu); goto out_delete_map; } -set_bit (cpu, cpu_bitmap); +set_bit (cpu.cpu, cpu_bitmap); } err =0 ;
@@ -2597,7 +2597,7 @@ int perf_event__process_id_index(struct perf_session *session,if (!sid)return -ENOENT ; sid->idx = e->idx; - sid->cpu = e->cpu; + sid->cpu.cpu = e->cpu; sid->tid = e->tid; }return 0 ;
tools/perf/util/stat-display.c +20 -14
@@ -121,10 +121,10 @@ static void aggr_printout(struct perf_stat_config *config, id.die, config->csv_output ?0 :-3 , id.core, config->csv_sep); - }else if (id.cpu >-1 ) { + }else if (id.cpu.cpu >-1 ) {fprintf (config->output,"CPU%*d%s" , config->csv_output ?0 :-7 , - id.cpu, config->csv_sep); + id.cpu.cpu, config->csv_sep); }break ;case AGGR_THREAD :
@@ -331,7 +331,8 @@ static int first_shadow_cpu_map_idx(struct perf_stat_config *config, struct evsel *evsel, const struct aggr_cpu_id *id) { struct perf_cpu_map *cpus = evsel__cpus(evsel); -int cpu, idx; +struct perf_cpu cpu; +int idx;if (config->aggr_mode ==AGGR_NONE )return perf_cpu_map__idx (cpus, id->cpu);
@@ -513,7 +514,8 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, intstatic void aggr_update_shadow (struct perf_stat_config *config,struct evlist *evlist) { -int cpu, idx, s; +int idx, s; +struct perf_cpu cpu;struct aggr_cpu_id s2, id;u64 val;struct evsel *counter;
@@ -633,7 +635,8 @@ static void aggr_cb(struct perf_stat_config *config, struct evsel *counter, void *data, bool first) { struct aggr_data *ad = data; -int idx, cpu; +int idx; +struct perf_cpu cpu;struct perf_cpu_map *cpus;struct aggr_cpu_id s2;
@@ -666,7 +669,7 @@ static void aggr_cb(struct perf_stat_config *config,static void print_counter_aggrdata (struct perf_stat_config *config,struct evsel *counter,int s,char *prefix,bool metric_only, -bool *first,int cpu) +bool *first,struct perf_cpu cpu) {struct aggr_data ad;FILE *output = config->output;
@@ -696,7 +699,7 @@ static void print_counter_aggrdata(struct perf_stat_config *config,fprintf (output,"%s" , prefix); uval = val * counter->scale; -if (cpu !=-1 ) +if (cpu.cpu !=-1 ) id =aggr_cpu_id__cpu (cpu,/*data=*/ NULL ); printout(config, id, nr, counter, uval,
@@ -731,8 +734,8 @@ static void print_aggr(struct perf_stat_config *config, first = true;evlist__for_each_entry (evlist, counter) {print_counter_aggrdata (config, counter, s, - prefix, metric_only, - &first,/*cpu=*/ -1 ); + prefix, metric_only, + &first, (struct perf_cpu ){ .cpu =-1 }); }if (metric_only)fputc ('\n' , output);
@@ -893,7 +896,8 @@ static void print_counter(struct perf_stat_config *config,FILE *output = config->output;u64 ena, run, val;double uval; -int idx, cpu; +int idx; +struct perf_cpu cpu;struct aggr_cpu_id id;perf_cpu_map__for_each_cpu (cpu, idx,evsel__cpus (counter)) {
@@ -921,7 +925,8 @@ static void print_no_aggr_metric(struct perf_stat_config *config,struct evlist *evlist, char *prefix) { - int all_idx, cpu; + int all_idx; +struct perf_cpu cpu;perf_cpu_map__for_each_cpu (cpu, all_idx, evlist->core.cpus) {struct evsel *counter;
@@ -1211,7 +1216,8 @@ static void print_percore_thread(struct perf_stat_config *config,struct aggr_cpu_id s2, id;struct perf_cpu_map *cpus;bool first = true; -int idx, cpu; +int idx; +struct perf_cpu cpu; cpus =evsel__cpus (counter);perf_cpu_map__for_each_cpu (cpu, idx, cpus) {
@@ -1247,8 +1253,8 @@ static void print_percore(struct perf_stat_config *config,fprintf (output,"%s" , prefix);print_counter_aggrdata (config, counter, s, - prefix, metric_only, - &first,/*cpu=*/ -1 ); + prefix, metric_only, + &first, (struct perf_cpu ){ .cpu =-1 }); }if (metric_only)
tools/perf/util/stat.c +1 -1
@@ -297,7 +297,7 @@ static int check_per_pkg(struct evsel *counter, struct perf_counts_values *vals, {struct hashmap *mask = counter->per_pkg_mask;struct perf_cpu_map *cpus =evsel__cpus (counter); -int cpu =perf_cpu_map__cpu (cpus, cpu_map_idx); +struct perf_cpu cpu =perf_cpu_map__cpu (cpus, cpu_map_idx);int s, d, ret =0 ;uint64_t *key;
tools/perf/util/stat.h +1 -1
@@ -108,7 +108,7 @@ struct runtime_stat {struct rblist value_list; }; -typedef struct aggr_cpu_id (*aggr_get_id_t )(struct perf_stat_config *config,int cpu); +typedef struct aggr_cpu_id (*aggr_get_id_t )(struct perf_stat_config *config,struct perf_cpu cpu);struct perf_stat_config {enum aggr_mode aggr_mode;
tools/perf/util/svghelper.c +3 -3
@@ -728,7 +728,7 @@ static int str_to_bitmap(char *s, cpumask_t *b, int nr_cpus)int i;int ret =0 ;struct perf_cpu_map *m; -int c; +struct perf_cpu c; m =perf_cpu_map__new (s);if (!m)
@@ -736,12 +736,12 @@ static int str_to_bitmap(char *s, cpumask_t *b, int nr_cpus)for (i =0 ; i < m->nr; i++) { c = m->map[i]; -if (c >= nr_cpus) { +if (c.cpu >= nr_cpus) { ret =-1 ;break ; } -set_bit (c,cpumask_bits (b)); +set_bit (c.cpu,cpumask_bits (b)); }perf_cpu_map__put (m);
tools/perf/util/synthetic-events.c +6 -6
@@ -1191,7 +1191,7 @@ static void synthesize_cpus(struct cpu_map_entries *cpus, cpus->nr = map->nr;for (i =0 ; i < map->nr; i++) - cpus->cpu[i] = map->map[i]; + cpus->cpu[i] = map->map[i].cpu; } static void synthesize_mask(struct perf_record_record_cpu_map *mask,
@@ -1203,7 +1203,7 @@ static void synthesize_mask(struct perf_record_record_cpu_map *mask, mask->long_size =sizeof (long );for (i =0 ; i < map->nr; i++) -set_bit (map->map[i], mask->mask); +set_bit (map->map[i].cpu, mask->mask); }static size_t cpus_size (struct perf_cpu_map *map)
@@ -1219,7 +1219,7 @@ static size_t mask_size(struct perf_cpu_map *map, int *max)for (i =0 ; i < map->nr; i++) {/* bit position of the cpu is + 1 */ -int bit = map->map[i] +1 ; +int bit = map->map[i].cpu +1 ;if (bit > *max) *max = bit;
@@ -1354,7 +1354,7 @@ int perf_event__synthesize_stat_config(struct perf_tool *tool, }int perf_event__synthesize_stat (struct perf_tool *tool, -u32 cpu,u32 thread,u64 id, +struct perf_cpu cpu,u32 thread,u64 id,struct perf_counts_values *count,perf_event__handler_t process,struct machine *machine)
@@ -1366,7 +1366,7 @@ int perf_event__synthesize_stat(struct perf_tool *tool, event.header.misc =0 ; event.id = id; - event.cpu = cpu; + event.cpu = cpu.cpu; event.thread = thread; event.val = count->val; event.ena = count->ena;
@@ -1763,7 +1763,7 @@ int perf_event__synthesize_id_index(struct perf_tool *tool, perf_event__handler_ } e->idx = sid->idx; - e->cpu = sid->cpu; + e->cpu = sid->cpu.cpu; e->tid = sid->tid; } }
tools/perf/util/synthetic-events.h +2 -1
@@ -6,6 +6,7 @@#include <sys/types.h> // pid_t #include <linux/compiler.h> #include <linux/types.h> +#include <perf/cpumap.h> struct auxtrace_record ;struct dso ;
@@ -63,7 +64,7 @@ int perf_event__synthesize_sample(union perf_event *event, u64 type, u64 read_foint perf_event__synthesize_stat_config (struct perf_tool *tool,struct perf_stat_config *config,perf_event__handler_t process,struct machine *machine);int perf_event__synthesize_stat_events (struct perf_stat_config *config,struct perf_tool *tool,struct evlist *evlist,perf_event__handler_t process,bool attrs);int perf_event__synthesize_stat_round (struct perf_tool *tool,u64 time,u64 type,perf_event__handler_t process,struct machine *machine); -int perf_event__synthesize_stat (struct perf_tool *tool,u32 cpu,u32 thread,u64 id,struct perf_counts_values *count,perf_event__handler_t process,struct machine *machine); +int perf_event__synthesize_stat (struct perf_tool *tool,struct perf_cpu cpu,u32 thread,u64 id,struct perf_counts_values *count,perf_event__handler_t process,struct machine *machine);int perf_event__synthesize_thread_map2 (struct perf_tool *tool,struct perf_thread_map *threads,perf_event__handler_t process,struct machine *machine);int perf_event__synthesize_thread_map (struct perf_tool *tool,struct perf_thread_map *threads,perf_event__handler_t process,struct machine *machine,bool needs_mmap,bool mmap_data);int perf_event__synthesize_threads (struct perf_tool *tool,perf_event__handler_t process,struct machine *machine,bool needs_mmap,bool mmap_data,unsigned int nr_threads_synthesize);
tools/perf/util/util.h +4 -1
@@ -11,6 +11,9 @@#include <stddef.h> #include <linux/compiler.h> #include <sys/types.h> +#ifndef __cplusplus +#include <internal/cpumap.h> +#endif /* General helper functions */ void usage (const char *err)__noreturn ;
@@ -66,6 +69,6 @@ extern bool test_attr__enabled;void test_attr__ready (void );void test_attr__init (void );struct perf_event_attr ; -void test_attr__open (struct perf_event_attr *attr,pid_t pid,int cpu, +void test_attr__open (struct perf_event_attr *attr,pid_t pid,struct perf_cpu cpu,int fd,int group_fd,unsigned long flags);#endif /* GIT_COMPAT_UTIL_H */
Signature · 1 line
2.34.1.307.g9b7440fafd-goog