File perf-bench-fix-futex-bench-to-correct-usage-of-affinity-for-machines-with-cpus-1k.patch of Package perf.29801
From: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Date: Wed, 6 Apr 2022 23:21:10 +0530
Subject: [PATCH] perf bench: Fix futex bench to correct usage of affinity for machines with #CPUs > 1K
Git-commit: c9c2a427dd9fe0e73eceacafb42d54f3c4535693
References: bsc#1208178
Signed-off-by: Tony Jones <tonyj@suse.de>
X-Info: IBM supplied backport.
X-Info: No 6d18804b963b7 (perf cpumap: Give CPUs their own type).
X-Info: Rework to use legacy "cpu->map" rather than perf_cpu_map
The 'perf bench futex' testcase fails on systems with more than 1K CPUs.
Testcase: perf bench futex all
Failure snippet:
<<>>Running futex/hash benchmark...
perf: pthread_create: No such file or directory
<<>>
All the futex benchmarks (ie hash, lock-api, requeue, wake,
wake-parallel), pthread_create is invoked in respective bench_futex_*
function. Though the logs shows direct failure from pthread_create,
strace logs showed that actual failure is from "sched_setaffinity"
returning EINVAL (invalid argument).
This happens because the default mask size in glibc is 1024. To overcome
this 1024 CPUs mask size limitation of cpu_set_t, change the mask size
using the CPU_*_S macros.
Patch addresses this by fixing all the futex benchmarks to use CPU_ALLOC
to allocate cpumask, CPU_ALLOC_SIZE for size, and CPU_SET_S to set the
mask.
Reported-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Athira Jajeev <atrajeev@linux.vnet.ibm.com>
Tested-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Nageswara R Sastry <rnsastry@linux.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Link: https://lore.kernel.org/r/20220406175113.87881-2-atrajeev@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
tools/perf/bench/futex-hash.c | 26 +++++++++++++++++++-------
tools/perf/bench/futex-lock-pi.c | 21 ++++++++++++++++-----
tools/perf/bench/futex-requeue.c | 21 ++++++++++++++++-----
tools/perf/bench/futex-wake-parallel.c | 21 ++++++++++++++++-----
tools/perf/bench/futex-wake.c | 22 ++++++++++++++++------
5 files changed, 83 insertions(+), 28 deletions(-)
--- a/tools/perf/bench/futex-requeue.c 2023-04-05 11:44:17.650000000 +0530
+++ b/tools/perf/bench/futex-requeue.c 2023-04-05 11:45:05.020000000 +0530
@@ -86,22 +86,33 @@ static void *workerfn(void *arg __maybe_
static void block_threads(pthread_t *w,
pthread_attr_t thread_attr, struct perf_cpu_map *cpu)
{
- cpu_set_t cpuset;
+ cpu_set_t *cpuset;
unsigned int i;
+ int nrcpus = perf_cpu_map__nr(cpu);
+ size_t size;
threads_starting = nthreads;
+ cpuset = CPU_ALLOC(nrcpus);
+ BUG_ON(!cpuset);
+ size = CPU_ALLOC_SIZE(nrcpus);
+
/* create and block all threads */
for (i = 0; i < nthreads; i++) {
- CPU_ZERO(&cpuset);
- CPU_SET(cpu->map[i % cpu->nr], &cpuset);
+ CPU_ZERO_S(size, cpuset);
+ CPU_SET_S(cpu->map[i % cpu->nr], size, cpuset);
- if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpuset))
+ if (pthread_attr_setaffinity_np(&thread_attr, size, cpuset)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+ }
- if (pthread_create(&w[i], &thread_attr, workerfn, NULL))
+ if (pthread_create(&w[i], &thread_attr, workerfn, NULL)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_create");
+ }
}
+ CPU_FREE(cpuset);
}
static void toggle_done(int sig __maybe_unused,
--- a/tools/perf/bench/futex-lock-pi.c 2023-04-05 11:44:17.650000000 +0530
+++ b/tools/perf/bench/futex-lock-pi.c 2023-04-05 11:45:05.020000000 +0530
@@ -117,11 +117,17 @@ static void *workerfn(void *arg)
static void create_threads(struct worker *w, pthread_attr_t thread_attr,
struct perf_cpu_map *cpu)
{
- cpu_set_t cpuset;
+ cpu_set_t *cpuset;
unsigned int i;
+ int nrcpus = perf_cpu_map__nr(cpu);
+ size_t size;
threads_starting = nthreads;
+ cpuset = CPU_ALLOC(nrcpus);
+ BUG_ON(!cpuset);
+ size = CPU_ALLOC_SIZE(nrcpus);
+
for (i = 0; i < nthreads; i++) {
worker[i].tid = i;
@@ -132,15 +138,20 @@ static void create_threads(struct worker
} else
worker[i].futex = &global_futex;
- CPU_ZERO(&cpuset);
- CPU_SET(cpu->map[i % cpu->nr], &cpuset);
+ CPU_ZERO_S(size, cpuset);
+ CPU_SET_S(cpu->map[i % cpu->nr], size, cpuset);
- if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpuset))
+ if (pthread_attr_setaffinity_np(&thread_attr, size, cpuset)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+ }
- if (pthread_create(&w[i].thread, &thread_attr, workerfn, &worker[i]))
+ if (pthread_create(&w[i].thread, &thread_attr, workerfn, &worker[i])) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_create");
+ }
}
+ CPU_FREE(cpuset);
}
int bench_futex_lock_pi(int argc, const char **argv)
--- a/tools/perf/bench/futex-hash.c 2023-04-05 11:44:17.650000000 +0530
+++ b/tools/perf/bench/futex-hash.c 2023-04-05 11:45:05.020000000 +0530
@@ -119,12 +119,14 @@ static void print_summary(void)
int bench_futex_hash(int argc, const char **argv)
{
int ret = 0;
- cpu_set_t cpuset;
+ cpu_set_t *cpuset;
struct sigaction act;
unsigned int i;
pthread_attr_t thread_attr;
struct worker *worker = NULL;
struct perf_cpu_map *cpu;
+ int nrcpus;
+ size_t size;
argc = parse_options(argc, argv, options, bench_futex_hash_usage, 0);
if (argc) {
@@ -162,25 +164,35 @@ int bench_futex_hash(int argc, const cha
threads_starting = nthreads;
pthread_attr_init(&thread_attr);
gettimeofday(&bench__start, NULL);
+
+ nrcpus = perf_cpu_map__nr(cpu);
+ cpuset = CPU_ALLOC(nrcpus);
+ BUG_ON(!cpuset);
+ size = CPU_ALLOC_SIZE(nrcpus);
+
for (i = 0; i < nthreads; i++) {
worker[i].tid = i;
worker[i].futex = calloc(nfutexes, sizeof(*worker[i].futex));
if (!worker[i].futex)
goto errmem;
- CPU_ZERO(&cpuset);
- CPU_SET(cpu->map[i % cpu->nr], &cpuset);
+ CPU_ZERO_S(size, cpuset);
- ret = pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpuset);
- if (ret)
+ CPU_SET_S(cpu->map[i % cpu->nr], size, cpuset);
+ ret = pthread_attr_setaffinity_np(&thread_attr, size, cpuset);
+ if (ret) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
-
+ }
ret = pthread_create(&worker[i].thread, &thread_attr, workerfn,
(void *)(struct worker *) &worker[i]);
- if (ret)
+ if (ret) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_create");
+ }
}
+ CPU_FREE(cpuset);
pthread_attr_destroy(&thread_attr);
pthread_mutex_lock(&thread_lock);
--- a/tools/perf/bench/futex-wake-parallel.c 2023-04-05 11:44:17.660000000 +0530
+++ b/tools/perf/bench/futex-wake-parallel.c 2023-04-05 11:45:05.020000000 +0530
@@ -140,22 +140,33 @@ static void *blocked_workerfn(void *arg
static void block_threads(pthread_t *w, pthread_attr_t thread_attr,
struct perf_cpu_map *cpu)
{
- cpu_set_t cpuset;
+ cpu_set_t *cpuset;
unsigned int i;
+ int nrcpus = perf_cpu_map__nr(cpu);
+ size_t size;
threads_starting = nblocked_threads;
+ cpuset = CPU_ALLOC(nrcpus);
+ BUG_ON(!cpuset);
+ size = CPU_ALLOC_SIZE(nrcpus);
+
/* create and block all threads */
for (i = 0; i < nblocked_threads; i++) {
- CPU_ZERO(&cpuset);
- CPU_SET(cpu->map[i % cpu->nr], &cpuset);
+ CPU_ZERO_S(size, cpuset);
+ CPU_SET_S(cpu->map[i % cpu->nr], size, cpuset);
- if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpuset))
+ if (pthread_attr_setaffinity_np(&thread_attr, size, cpuset)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+ }
- if (pthread_create(&w[i], &thread_attr, blocked_workerfn, NULL))
+ if (pthread_create(&w[i], &thread_attr, blocked_workerfn, NULL)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_create");
+ }
}
+ CPU_FREE(cpuset);
}
static void print_run(struct thread_data *waking_worker, unsigned int run_num)
--- a/tools/perf/bench/futex-wake.c 2023-04-05 11:44:17.660000000 +0530
+++ b/tools/perf/bench/futex-wake.c 2023-04-05 11:45:05.020000000 +0530
@@ -92,22 +92,32 @@ static void print_summary(void)
static void block_threads(pthread_t *w,
pthread_attr_t thread_attr, struct perf_cpu_map *cpu)
{
- cpu_set_t cpuset;
+ cpu_set_t *cpuset;
unsigned int i;
-
+ size_t size;
+ int nrcpus = perf_cpu_map__nr(cpu);
threads_starting = nthreads;
+ cpuset = CPU_ALLOC(nrcpus);
+ BUG_ON(!cpuset);
+ size = CPU_ALLOC_SIZE(nrcpus);
+
/* create and block all threads */
for (i = 0; i < nthreads; i++) {
- CPU_ZERO(&cpuset);
- CPU_SET(cpu->map[i % cpu->nr], &cpuset);
+ CPU_ZERO_S(size, cpuset);
+ CPU_SET_S(cpu->map[i % cpu->nr], size, cpuset);
- if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpuset))
+ if (pthread_attr_setaffinity_np(&thread_attr, size, cpuset)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+ }
- if (pthread_create(&w[i], &thread_attr, workerfn, NULL))
+ if (pthread_create(&w[i], &thread_attr, workerfn, NULL)) {
+ CPU_FREE(cpuset);
err(EXIT_FAILURE, "pthread_create");
+ }
}
+ CPU_FREE(cpuset);
}
static void toggle_done(int sig __maybe_unused,
