| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <20190802201959.184992-1-meyerk@stormcage.eag.rdlabs.hpecorp.net>
Date: Fri, 2 Aug 2019 15:19:59 -0500
From: Kyle Meyer <meyerk@....com>
To: unlisted-recipients:; (no To-header on input)
Cc: Kyle Meyer <meyerk@....com>, Kyle Meyer <kyle.meyer@....com>,
Russ Anderson <russ.anderson@....com>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
linux-kernel@...r.kernel.org
Subject: [PATCH] perf tools: Replace MAX_NR_CPUS with nr_cpus_onln
The variables nr_cpus_onln and max_caches are dynamic alternatives for
MAX_NR_CPUS and MAX_CACHES as they are initialized at runtime. MAX_NR_CPUS
is still used by DECLARE_BITMAP() at compile time, however, nr_cpus_onln
replaces it elsewhere throughout perf.
This patch was tested using "perf record -a -g" on both an eight socket
(288 CPUs) system and a single socket (36 CPUs) system. Each system was then
rebooted single socket (36 CPUs) / eight socket (288 CPUs) and "perf
report" used to read the perf.data file. "perf report --header" was used to
confirm that each perf.data had information on 288 CPUs / 36 CPUs.
Signed-off-by: Kyle Meyer <kyle.meyer@....com>
Cc: Russ Anderson <russ.anderson@....com>
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Ingo Molnar <mingo@...hat.com>
Cc: Arnaldo Carvalho de Melo <acme@...nel.org>
Cc: Alexander Shishkin <alexander.shishkin@...ux.intel.com>
Cc: Jiri Olsa <jolsa@...hat.com>
Cc: Namhyung Kim <namhyung@...nel.org>
Cc: linux-kernel@...r.kernel.org
---
tools/perf/perf.c | 10 ++++++++++
tools/perf/perf.h | 1 +
tools/perf/util/header.c | 7 +++----
tools/perf/util/machine.c | 11 +++++------
tools/perf/util/stat.c | 4 ++--
tools/perf/util/svghelper.c | 10 +++++-----
6 files changed, 26 insertions(+), 17 deletions(-)
diff --git a/tools/perf/perf.c b/tools/perf/perf.c
index 97e2628ea5dd..658bf8501bb0 100644
--- a/tools/perf/perf.c
+++ b/tools/perf/perf.c
@@ -428,6 +428,16 @@ int main(int argc, const char **argv)
const char *cmd;
char sbuf[STRERR_BUFSIZE];
+ nr_cpus_onln = sysconf(_SC_NPROCESSORS_ONLN);
+ if (nr_cpus_onln < 0) {
+ fprintf(stderr, "Cannot determine the number of CPUs currently online.\n");
+ goto out;
+ }
+ if (nr_cpus_onln > MAX_NR_CPUS) {
+ fprintf(stderr, "The number of CPUs currently online is too large, consider raising MAX_NR_CPUS.\n");
+ nr_cpus_onln = MAX_NR_CPUS;
+ }
+
/* libsubcmd init */
exec_cmd_init("perf", PREFIX, PERF_EXEC_PATH, EXEC_PATH_ENVIRONMENT);
pager_init(PERF_PAGER_ENVIRONMENT);
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 74d0124d38f3..603391cac85b 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -29,6 +29,7 @@ static inline unsigned long long rdclock(void)
#define MAX_NR_CPUS 2048
#endif
+int nr_cpus_onln;
extern const char *input_name;
extern bool perf_host, perf_guest;
extern const char perf_version_string[];
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index b04c2b6b28b3..8b0cb20a770c 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -1121,16 +1121,15 @@ static int build_caches(struct cpu_cache_level caches[], u32 size, u32 *cntp)
return 0;
}
-#define MAX_CACHES (MAX_NR_CPUS * 4)
-
static int write_cache(struct feat_fd *ff,
struct evlist *evlist __maybe_unused)
{
- struct cpu_cache_level caches[MAX_CACHES];
+ u32 max_caches = (nr_cpus_onln * 4);
+ struct cpu_cache_level caches[max_caches];
u32 cnt = 0, i, version = 1;
int ret;
- ret = build_caches(caches, MAX_CACHES, &cnt);
+ ret = build_caches(caches, max_caches, &cnt);
if (ret)
goto out;
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index f6ee7fbad3e4..3ad77d5e8aab 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -2615,7 +2615,7 @@ int __machine__synthesize_threads(struct machine *machine, struct perf_tool *too
pid_t machine__get_current_tid(struct machine *machine, int cpu)
{
- if (cpu < 0 || cpu >= MAX_NR_CPUS || !machine->current_tid)
+ if (cpu < 0 || cpu >= nr_cpus_onln || !machine->current_tid)
return -1;
return machine->current_tid[cpu];
@@ -2632,16 +2632,15 @@ int machine__set_current_tid(struct machine *machine, int cpu, pid_t pid,
if (!machine->current_tid) {
int i;
- machine->current_tid = calloc(MAX_NR_CPUS, sizeof(pid_t));
+ machine->current_tid = calloc(nr_cpus_onln, sizeof(pid_t));
if (!machine->current_tid)
return -ENOMEM;
- for (i = 0; i < MAX_NR_CPUS; i++)
+ for (i = 0; i < nr_cpus_onln; i++)
machine->current_tid[i] = -1;
}
- if (cpu >= MAX_NR_CPUS) {
- pr_err("Requested CPU %d too large. ", cpu);
- pr_err("Consider raising MAX_NR_CPUS\n");
+ if (cpu >= nr_cpus_onln) {
+ pr_err("Requested CPU %d too large, there are %d CPUs currently online.\n", cpu, nr_cpus_onln);
return -EINVAL;
}
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
index e4e4e3bf8b2b..42dddbd2f23c 100644
--- a/tools/perf/util/stat.c
+++ b/tools/perf/util/stat.c
@@ -208,7 +208,7 @@ void perf_evlist__reset_stats(struct evlist *evlist)
static void zero_per_pkg(struct evsel *counter)
{
if (counter->per_pkg_mask)
- memset(counter->per_pkg_mask, 0, MAX_NR_CPUS);
+ memset(counter->per_pkg_mask, 0, nr_cpus_onln);
}
static int check_per_pkg(struct evsel *counter,
@@ -227,7 +227,7 @@ static int check_per_pkg(struct evsel *counter,
return 0;
if (!mask) {
- mask = zalloc(MAX_NR_CPUS);
+ mask = zalloc(nr_cpus_onln);
if (!mask)
return -ENOMEM;
diff --git a/tools/perf/util/svghelper.c b/tools/perf/util/svghelper.c
index ae6a534a7a80..0404bd87812a 100644
--- a/tools/perf/util/svghelper.c
+++ b/tools/perf/util/svghelper.c
@@ -706,7 +706,7 @@ static void scan_thread_topology(int *map, struct topology *t, int cpu, int *pos
for_each_set_bit(thr,
cpumask_bits(&t->sib_thr[i]),
- MAX_NR_CPUS)
+ nr_cpus_onln)
if (map[thr] == -1)
map[thr] = (*pos)++;
}
@@ -721,7 +721,7 @@ static void scan_core_topology(int *map, struct topology *t)
for (i = 0; i < t->sib_core_nr; i++)
for_each_set_bit(cpu,
cpumask_bits(&t->sib_core[i]),
- MAX_NR_CPUS)
+ nr_cpus_onln)
scan_thread_topology(map, t, cpu, &pos);
}
@@ -738,7 +738,7 @@ static int str_to_bitmap(char *s, cpumask_t *b)
for (i = 0; i < m->nr; i++) {
c = m->map[i];
- if (c >= MAX_NR_CPUS) {
+ if (c >= nr_cpus_onln) {
ret = -1;
break;
}
@@ -785,13 +785,13 @@ int svg_build_topology_map(char *sib_core, int sib_core_nr,
sib_thr += strlen(sib_thr) + 1;
}
- topology_map = malloc(sizeof(int) * MAX_NR_CPUS);
+ topology_map = malloc(sizeof(int) * nr_cpus_onln);
if (!topology_map) {
fprintf(stderr, "topology: no memory\n");
goto exit;
}
- for (i = 0; i < MAX_NR_CPUS; i++)
+ for (i = 0; i < nr_cpus_onln; i++)
topology_map[i] = -1;
scan_core_topology(topology_map, &t);
--
2.12.3
Powered by blists - more mailing lists