From 052b0f6eaf8b1f02669884a177bc3ce463133a42 Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Fri, 24 Apr 2015 10:00:48 -0700 Subject: perf bench futex: Fix hung wakeup tasks after requeueing The futex-requeue benchmark can hang because of missing wakeups once the benchmark is done, ie: [Run 1]: Requeued 1024 of 1024 threads in 0.3290 ms perf: couldn't wakeup all tasks (135/1024) This bug, while perhaps suggesting missing wakeups in kernel futex code, is merely a consequence of the crappy FUTEX_CMP_REQUEUE man page, incorrectly mentioning that the number of requeued tasks is in fact returned, not the wakeups. This patch acknowledges this and updates the corresponding futex_wake code around it. Signed-off-by: Davidlohr Bueso Cc: Mel Gorman Link: http://lkml.kernel.org/r/1429894848.10273.44.camel@stgolabs.net Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/futex-requeue.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'tools/perf/bench') diff --git a/tools/perf/bench/futex-requeue.c b/tools/perf/bench/futex-requeue.c index bedff6b5b3cf..ad0d9b5342fb 100644 --- a/tools/perf/bench/futex-requeue.c +++ b/tools/perf/bench/futex-requeue.c @@ -132,6 +132,9 @@ int bench_futex_requeue(int argc, const char **argv, if (!fshared) futex_flag = FUTEX_PRIVATE_FLAG; + if (nrequeue > nthreads) + nrequeue = nthreads; + printf("Run summary [PID %d]: Requeuing %d threads (from [%s] %p to %p), " "%d at a time.\n\n", getpid(), nthreads, fshared ? "shared":"private", &futex1, &futex2, nrequeue); @@ -161,20 +164,18 @@ int bench_futex_requeue(int argc, const char **argv, /* Ok, all threads are patiently blocked, start requeueing */ gettimeofday(&start, NULL); - for (nrequeued = 0; nrequeued < nthreads; nrequeued += nrequeue) { + while (nrequeued < nthreads) { /* * Do not wakeup any tasks blocked on futex1, allowing * us to really measure futex_wait functionality. */ - futex_cmp_requeue(&futex1, 0, &futex2, 0, - nrequeue, futex_flag); + nrequeued += futex_cmp_requeue(&futex1, 0, &futex2, 0, + nrequeue, futex_flag); } + gettimeofday(&end, NULL); timersub(&end, &start, &runtime); - if (nrequeued > nthreads) - nrequeued = nthreads; - update_stats(&requeued_stats, nrequeued); update_stats(&requeuetime_stats, runtime.tv_usec); @@ -184,7 +185,7 @@ int bench_futex_requeue(int argc, const char **argv, } /* everybody should be blocked on futex2, wake'em up */ - nrequeued = futex_wake(&futex2, nthreads, futex_flag); + nrequeued = futex_wake(&futex2, nrequeued, futex_flag); if (nthreads != nrequeued) warnx("couldn't wakeup all tasks (%d/%d)", nrequeued, nthreads); -- cgit v1.2.3 From 24f1ced167e5e011040b4c3aae75aee45a79eed5 Mon Sep 17 00:00:00 2001 From: Petr Holasek Date: Thu, 16 Apr 2015 17:38:17 +0200 Subject: perf bench numa: Fixes of --quiet argument Corrected description and fixed function of --quiet argument. Signed-off-by: Petr Holasek Reviewed-by: Ingo Molnar Cc: Jiri Olsa Link: http://lkml.kernel.org/r/1429198699-25039-2-git-send-email-pholasek@redhat.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'tools/perf/bench') diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index ebfa163b80b5..cd872e9c3a9c 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -180,7 +180,7 @@ static const struct option options[] = { OPT_INTEGER('H', "thp" , &p0.thp, "MADV_NOHUGEPAGE < 0 < MADV_HUGEPAGE"), OPT_BOOLEAN('c', "show_convergence", &p0.show_convergence, "show convergence details"), OPT_BOOLEAN('m', "measure_convergence", &p0.measure_convergence, "measure convergence latency"), - OPT_BOOLEAN('q', "quiet" , &p0.show_quiet, "bzero the initial allocations"), + OPT_BOOLEAN('q', "quiet" , &p0.show_quiet, "quiet mode"), OPT_BOOLEAN('S', "serialize-startup", &p0.serialize_startup,"serialize thread startup"), /* Special option string parsing callbacks: */ @@ -1395,7 +1395,7 @@ static void print_res(const char *name, double val, if (!name) name = "main,"; - if (g->p.show_quiet) + if (!g->p.show_quiet) printf(" %-30s %15.3f, %-15s %s\n", name, val, txt_unit, txt_short); else printf(" %14.3f %s\n", val, txt_long); -- cgit v1.2.3 From 1d90a685eb75a56648d7dd22c704a1a6da516de9 Mon Sep 17 00:00:00 2001 From: Petr Holasek Date: Thu, 16 Apr 2015 17:38:19 +0200 Subject: perf bench numa: Fix immediate meeting of convergence condition This patch fixes the race in the beginning of benchmark run when some threads hasn't got assigned curr_cpu yet so they don't occur in nodes-of-process stats and benchmark concludes that all remaining threads are converged already. The race can be reproduced with small amount of threads and some bigger amount of shared process memory, e.g. one process, two threads and 5GB of process memory. Signed-off-by: Petr Holasek Reviewed-by: Ingo Molnar Cc: Jiri Olsa Link: http://lkml.kernel.org/r/1429198699-25039-4-git-send-email-pholasek@redhat.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'tools/perf/bench') diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index cd872e9c3a9c..ba5efa4710b5 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -828,6 +828,9 @@ static int count_process_nodes(int process_nr) td = g->threads + task_nr; node = numa_node_of_cpu(td->curr_cpu); + if (node < 0) /* curr_cpu was likely still -1 */ + return 0; + node_present[node] = 1; } @@ -882,6 +885,11 @@ static void calc_convergence_compression(int *strong) for (p = 0; p < g->p.nr_proc; p++) { unsigned int nodes = count_process_nodes(p); + if (!nodes) { + *strong = 0; + return; + } + nodes_min = min(nodes, nodes_min); nodes_max = max(nodes, nodes_max); } -- cgit v1.2.3