From f034fc50d3c7d9385c20d505ab4cf56b8fd18ac7 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 11 Apr 2022 09:17:58 +0300 Subject: perf tools: Fix misleading add event PMU debug message Fix incorrect debug message: Attempting to add event pmu 'intel_pt' with '' that may result in non-fatal errors which always appears with perf record -vv and intel_pt e.g. perf record -vv -e intel_pt//u uname The message is incorrect because there will never be non-fatal errors. Suppress the message if the PMU is 'selectable' i.e. meant to be selected directly as an event. Fixes: 4ac22b484d4c79e8 ("perf parse-events: Make add PMU verbose output clearer") Signed-off-by: Adrian Hunter Cc: Ian Rogers Cc: Jiri Olsa Link: http://lore.kernel.org/lkml/20220411061758.2458417-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/parse-events.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 24997925ae00..dd84fed698a3 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -1523,7 +1523,9 @@ int parse_events_add_pmu(struct parse_events_state *parse_state, bool use_uncore_alias; LIST_HEAD(config_terms); - if (verbose > 1) { + pmu = parse_state->fake_pmu ?: perf_pmu__find(name); + + if (verbose > 1 && !(pmu && pmu->selectable)) { fprintf(stderr, "Attempting to add event pmu '%s' with '", name); if (head_config) { @@ -1536,7 +1538,6 @@ int parse_events_add_pmu(struct parse_events_state *parse_state, fprintf(stderr, "' that may result in non-fatal errors\n"); } - pmu = parse_state->fake_pmu ?: perf_pmu__find(name); if (!pmu) { char *err_str; -- cgit v1.2.3 From d73f5d14e0cdd1f39764379250f26163913d7155 Mon Sep 17 00:00:00 2001 From: Lv Ruyi Date: Wed, 13 Apr 2022 09:33:02 +0000 Subject: perf stat: Fix error check return value of hashmap__new(), must use IS_ERR() hashmap__new() returns ERR_PTR(-ENOMEM) when it fails, so we should use IS_ERR() to check it in error handling path. Reported-by: Zeal Robot Signed-off-by: Lv Ruyi Cc: Alexander Shishkin Cc: Ian Rogers Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220413093302.2538128-1-lv.ruyi@zte.com.cn Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/stat.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c index ee6f03481215..817a2de264b4 100644 --- a/tools/perf/util/stat.c +++ b/tools/perf/util/stat.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include +#include #include #include #include @@ -311,7 +312,7 @@ static int check_per_pkg(struct evsel *counter, struct perf_counts_values *vals, if (!mask) { mask = hashmap__new(pkg_id_hash, pkg_id_equal, NULL); - if (!mask) + if (IS_ERR(mask)) return -ENOMEM; counter->per_pkg_mask = mask; -- cgit v1.2.3 From a668cc07f990d2ed19424d5c1a529521a9d1cee1 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 13 Apr 2022 14:42:32 +0300 Subject: perf tools: Fix segfault accessing sample_id xyarray perf_evsel::sample_id is an xyarray which can cause a segfault when accessed beyond its size. e.g. # perf record -e intel_pt// -C 1 sleep 1 Segmentation fault (core dumped) # That is happening because a dummy event is opened to capture text poke events accross all CPUs, however the mmap logic is allocating according to the number of user_requested_cpus. In general, perf sometimes uses the evsel cpus to open events, and sometimes the evlist user_requested_cpus. However, it is not necessary to determine which case is which because the opened event file descriptors are also in an xyarray, the size of whch can be used to correctly allocate the size of the sample_id xyarray, because there is one ID per file descriptor. Note, in the affected code path, perf_evsel fd array is subsequently used to get the file descriptor for the mmap, so it makes sense for the xyarrays to be the same size there. Fixes: d1a177595b3a824c ("libperf: Adopt perf_evlist__mmap()/munmap() from tools/perf") Fixes: 246eba8e9041c477 ("perf tools: Add support for PERF_RECORD_TEXT_POKE") Signed-off-by: Adrian Hunter Acked-by: Ian Rogers Cc: Adrian Hunter Cc: Jiri Olsa Cc: stable@vger.kernel.org # 5.5+ Link: https://lore.kernel.org/r/20220413114232.26914-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/perf/evlist.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/tools/lib/perf/evlist.c b/tools/lib/perf/evlist.c index 1b15ba13c477..a09315538a30 100644 --- a/tools/lib/perf/evlist.c +++ b/tools/lib/perf/evlist.c @@ -577,7 +577,6 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, { struct perf_evsel *evsel; const struct perf_cpu_map *cpus = evlist->user_requested_cpus; - const struct perf_thread_map *threads = evlist->threads; if (!ops || !ops->get || !ops->mmap) return -EINVAL; @@ -589,7 +588,7 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, perf_evlist__for_each_entry(evlist, evsel) { if ((evsel->attr.read_format & PERF_FORMAT_ID) && evsel->sample_id == NULL && - perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) + perf_evsel__alloc_id(evsel, evsel->fd->max_x, evsel->fd->max_y) < 0) return -ENOMEM; } -- cgit v1.2.3 From 23380e4d53305765789fd1a2cf6bddb07239cd3b Mon Sep 17 00:00:00 2001 From: Alexey Bayduraev Date: Wed, 13 Apr 2022 18:46:40 -0700 Subject: perf record: Fix per-thread option Per-thread mode doesn't have specific CPUs for events, add checks for this case. Minor fix to a pr_debug by Ian Rogers to avoid an out of bound array access. Fixes: 7954f71689f90cb2 ("perf record: Introduce thread affinity and mmap masks") Reported-by: Ian Rogers Signed-off-by: Alexey Bayduraev Signed-off-by: Ian Rogers Cc: Alexander Shishkin Cc: Alexey Bayduraev Cc: Andi Kleen Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Riccardo Mancini Cc: Stephane Eranian Link: https://lore.kernel.org/r/20220414014642.3308206-1-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index ba74fab02e62..069825c48d40 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -989,8 +989,11 @@ static int record__thread_data_init_maps(struct record_thread *thread_data, stru struct mmap *overwrite_mmap = evlist->overwrite_mmap; struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; - thread_data->nr_mmaps = bitmap_weight(thread_data->mask->maps.bits, - thread_data->mask->maps.nbits); + if (cpu_map__is_dummy(cpus)) + thread_data->nr_mmaps = nr_mmaps; + else + thread_data->nr_mmaps = bitmap_weight(thread_data->mask->maps.bits, + thread_data->mask->maps.nbits); if (mmap) { thread_data->maps = zalloc(thread_data->nr_mmaps * sizeof(struct mmap *)); if (!thread_data->maps) @@ -1007,16 +1010,17 @@ static int record__thread_data_init_maps(struct record_thread *thread_data, stru thread_data->nr_mmaps, thread_data->maps, thread_data->overwrite_maps); for (m = 0, tm = 0; m < nr_mmaps && tm < thread_data->nr_mmaps; m++) { - if (test_bit(cpus->map[m].cpu, thread_data->mask->maps.bits)) { + if (cpu_map__is_dummy(cpus) || + test_bit(cpus->map[m].cpu, thread_data->mask->maps.bits)) { if (thread_data->maps) { thread_data->maps[tm] = &mmap[m]; pr_debug2("thread_data[%p]: cpu%d: maps[%d] -> mmap[%d]\n", - thread_data, cpus->map[m].cpu, tm, m); + thread_data, perf_cpu_map__cpu(cpus, m).cpu, tm, m); } if (thread_data->overwrite_maps) { thread_data->overwrite_maps[tm] = &overwrite_mmap[m]; pr_debug2("thread_data[%p]: cpu%d: ow_maps[%d] -> ow_mmap[%d]\n", - thread_data, cpus->map[m].cpu, tm, m); + thread_data, perf_cpu_map__cpu(cpus, m).cpu, tm, m); } tm++; } @@ -3329,6 +3333,9 @@ static void record__mmap_cpu_mask_init(struct mmap_cpu_mask *mask, struct perf_c { int c; + if (cpu_map__is_dummy(cpus)) + return; + for (c = 0; c < cpus->nr; c++) set_bit(cpus->map[c].cpu, mask->bits); } @@ -3680,6 +3687,11 @@ static int record__init_thread_masks(struct record *rec) if (!record__threads_enabled(rec)) return record__init_thread_default_masks(rec, cpus); + if (cpu_map__is_dummy(cpus)) { + pr_err("--per-thread option is mutually exclusive to parallel streaming mode.\n"); + return -EINVAL; + } + switch (rec->opts.threads_spec) { case THREAD_SPEC__CPU: ret = record__init_thread_cpu_masks(rec, cpus); -- cgit v1.2.3 From 8cb7a188ac33b71ec24a9081d1521f46dcab53c4 Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Tue, 12 Apr 2022 22:10:58 +0530 Subject: perf bench: Fix numa testcase to check if CPU used to bind task is online Perf numa bench test fails with error: Testcase: ./perf bench numa mem -p 2 -t 1 -P 1024 -C 0,8 -M 1,0 -s 20 -zZq --thp 1 --no-data_rand_walk Failure snippet: <<>> Running 'numa/mem' benchmark: # Running main, "perf bench numa numa-mem -p 2 -t 1 -P 1024 -C 0,8 -M 1,0 -s 20 -zZq --thp 1 --no-data_rand_walk" perf: bench/numa.c:333: bind_to_cpumask: Assertion `!(ret)' failed. <<>> The Testcases uses CPU's 0 and 8. In function "parse_setup_cpu_list", There is check to see if cpu number is greater than max cpu's possible in the system ie via "if (bind_cpu_0 >= g->p.nr_cpus || bind_cpu_1 >= g->p.nr_cpus) {". But it could happen that system has say 48 CPU's, but only number of online CPU's is 0-7. Other CPU's are offlined. Since "g->p.nr_cpus" is 48, so function will go ahead and set bit for CPU 8 also in cpumask ( td->bind_cpumask). bind_to_cpumask function is called to set affinity using sched_setaffinity and the cpumask. Since the CPU8 is not present, set affinity will fail here with EINVAL. Fix this issue by adding a check to make sure that, CPU's provided in the input argument values are online before proceeding further and skip the test. For this, include new helper function "is_cpu_online" in "tools/perf/util/header.c". Since "BIT(x)" definition will get included from header.h, remove that from bench/numa.c Reported-by: Disha Goel Signed-off-by: Athira Jajeev Tested-by: Disha Goel Cc: Ian Rogers Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Nageswara R Sastry Cc: Srikar Dronamraju Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220412164059.42654-2-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 8 ++++++-- tools/perf/util/header.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/header.h | 1 + 3 files changed, 58 insertions(+), 2 deletions(-) diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index f2640179ada9..c838c248aa2c 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -34,6 +34,7 @@ #include #include +#include "../util/header.h" #include #include @@ -585,6 +586,11 @@ static int parse_setup_cpu_list(void) return -1; } + if (is_cpu_online(bind_cpu_0) != 1 || is_cpu_online(bind_cpu_1) != 1) { + printf("\nTest not applicable, bind_cpu_0 or bind_cpu_1 is offline\n"); + return -1; + } + BUG_ON(bind_cpu_0 < 0 || bind_cpu_1 < 0); BUG_ON(bind_cpu_0 > bind_cpu_1); @@ -752,8 +758,6 @@ static int parse_nodes_opt(const struct option *opt __maybe_unused, return parse_node_list(arg); } -#define BIT(x) (1ul << x) - static inline uint32_t lfsr_32(uint32_t lfsr) { const uint32_t taps = BIT(1) | BIT(5) | BIT(6) | BIT(31); diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index d546ff724dbe..a27132e5a5ef 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -983,6 +983,57 @@ static int write_dir_format(struct feat_fd *ff, return do_write(ff, &data->dir.version, sizeof(data->dir.version)); } +/* + * Check whether a CPU is online + * + * Returns: + * 1 -> if CPU is online + * 0 -> if CPU is offline + * -1 -> error case + */ +int is_cpu_online(unsigned int cpu) +{ + char *str; + size_t strlen; + char buf[256]; + int status = -1; + struct stat statbuf; + + snprintf(buf, sizeof(buf), + "/sys/devices/system/cpu/cpu%d", cpu); + if (stat(buf, &statbuf) != 0) + return 0; + + /* + * Check if /sys/devices/system/cpu/cpux/online file + * exists. Some cases cpu0 won't have online file since + * it is not expected to be turned off generally. + * In kernels without CONFIG_HOTPLUG_CPU, this + * file won't exist + */ + snprintf(buf, sizeof(buf), + "/sys/devices/system/cpu/cpu%d/online", cpu); + if (stat(buf, &statbuf) != 0) + return 1; + + /* + * Read online file using sysfs__read_str. + * If read or open fails, return -1. + * If read succeeds, return value from file + * which gets stored in "str" + */ + snprintf(buf, sizeof(buf), + "devices/system/cpu/cpu%d/online", cpu); + + if (sysfs__read_str(buf, &str, &strlen) < 0) + return status; + + status = atoi(str); + + free(str); + return status; +} + #ifdef HAVE_LIBBPF_SUPPORT static int write_bpf_prog_info(struct feat_fd *ff, struct evlist *evlist __maybe_unused) diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index c9e3265832d9..0eb4bc29a5a4 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -158,6 +158,7 @@ int do_write(struct feat_fd *fd, const void *buf, size_t size); int write_padded(struct feat_fd *fd, const void *bf, size_t count, size_t count_aligned); +int is_cpu_online(unsigned int cpu); /* * arch specific callback */ -- cgit v1.2.3 From f58faed7fb3f784efdbf3b401368dcf51a6e28fa Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Tue, 12 Apr 2022 22:10:59 +0530 Subject: perf bench: Fix numa bench to fix usage of affinity for machines with #CPUs > 1K The 'perf bench numa' testcase fails on systems with more than 1K CPUs. Testcase: perf bench numa mem -p 1 -t 3 -P 512 -s 100 -zZ0qcm --thp 1 Snippet of code: <<>> perf: bench/numa.c:302: bind_to_node: Assertion `!(ret)' failed. Aborted (core dumped) <<>> bind_to_node() uses "sched_getaffinity" to save the original cpumask and this call is 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 ie, use CPU_ALLOC to allocate cpumask, CPU_ALLOC_SIZE for size. Apart from fixing this for "orig_mask", apply same logic to "mask" as well which is used to setaffinity so that mask size is large enough to represent number of possible CPU's in the system. sched_getaffinity is used in one more place in perf numa bench. It is in "bind_to_cpu" function. Apply the same logic there also. Though currently no failure is reported from there, it is ideal to change getaffinity to work with such system configurations having CPU's more than default mask size supported by glibc. Also fix "sched_setaffinity" to use mask size which is large enough to represent number of possible CPU's in the system. Fixed all places where "bind_cpumask" which is part of "struct thread_data" is used such that bind_cpumask works in all configuration. Reported-by: Disha Goel Signed-off-by: Athira Jajeev Cc: Ian Rogers Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Nageswara R Sastry Cc: Srikar Dronamraju Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220412164059.42654-3-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 128 +++++++++++++++++++++++++++++++++++------------- 1 file changed, 95 insertions(+), 33 deletions(-) diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index c838c248aa2c..44e1f8a44087 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -55,7 +55,7 @@ struct thread_data { int curr_cpu; - cpu_set_t bind_cpumask; + cpu_set_t *bind_cpumask; int bind_node; u8 *process_data; int process_nr; @@ -267,71 +267,115 @@ static bool node_has_cpus(int node) return ret; } -static cpu_set_t bind_to_cpu(int target_cpu) +static cpu_set_t *bind_to_cpu(int target_cpu) { - cpu_set_t orig_mask, mask; - int ret; + int nrcpus = numa_num_possible_cpus(); + cpu_set_t *orig_mask, *mask; + size_t size; - ret = sched_getaffinity(0, sizeof(orig_mask), &orig_mask); - BUG_ON(ret); + orig_mask = CPU_ALLOC(nrcpus); + BUG_ON(!orig_mask); + size = CPU_ALLOC_SIZE(nrcpus); + CPU_ZERO_S(size, orig_mask); + + if (sched_getaffinity(0, size, orig_mask)) + goto err_out; + + mask = CPU_ALLOC(nrcpus); + if (!mask) + goto err_out; - CPU_ZERO(&mask); + CPU_ZERO_S(size, mask); if (target_cpu == -1) { int cpu; for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } else { - BUG_ON(target_cpu < 0 || target_cpu >= g->p.nr_cpus); - CPU_SET(target_cpu, &mask); + if (target_cpu < 0 || target_cpu >= g->p.nr_cpus) + goto err; + + CPU_SET_S(target_cpu, size, mask); } - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + if (sched_setaffinity(0, size, mask)) + goto err; return orig_mask; + +err: + CPU_FREE(mask); +err_out: + CPU_FREE(orig_mask); + + /* BUG_ON due to failure in allocation of orig_mask/mask */ + BUG_ON(-1); } -static cpu_set_t bind_to_node(int target_node) +static cpu_set_t *bind_to_node(int target_node) { - cpu_set_t orig_mask, mask; + int nrcpus = numa_num_possible_cpus(); + size_t size; + cpu_set_t *orig_mask, *mask; int cpu; - int ret; - ret = sched_getaffinity(0, sizeof(orig_mask), &orig_mask); - BUG_ON(ret); + orig_mask = CPU_ALLOC(nrcpus); + BUG_ON(!orig_mask); + size = CPU_ALLOC_SIZE(nrcpus); + CPU_ZERO_S(size, orig_mask); + + if (sched_getaffinity(0, size, orig_mask)) + goto err_out; + + mask = CPU_ALLOC(nrcpus); + if (!mask) + goto err_out; - CPU_ZERO(&mask); + CPU_ZERO_S(size, mask); if (target_node == NUMA_NO_NODE) { for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } else { struct bitmask *cpumask = numa_allocate_cpumask(); - BUG_ON(!cpumask); + if (!cpumask) + goto err; + if (!numa_node_to_cpus(target_node, cpumask)) { for (cpu = 0; cpu < (int)cpumask->size; cpu++) { if (numa_bitmask_isbitset(cpumask, cpu)) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } } numa_free_cpumask(cpumask); } - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + if (sched_setaffinity(0, size, mask)) + goto err; return orig_mask; + +err: + CPU_FREE(mask); +err_out: + CPU_FREE(orig_mask); + + /* BUG_ON due to failure in allocation of orig_mask/mask */ + BUG_ON(-1); } -static void bind_to_cpumask(cpu_set_t mask) +static void bind_to_cpumask(cpu_set_t *mask) { int ret; + size_t size = CPU_ALLOC_SIZE(numa_num_possible_cpus()); - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + ret = sched_setaffinity(0, size, mask); + if (ret) { + CPU_FREE(mask); + BUG_ON(ret); + } } static void mempol_restore(void) @@ -377,7 +421,7 @@ do { \ static u8 *alloc_data(ssize_t bytes0, int map_flags, int init_zero, int init_cpu0, int thp, int init_random) { - cpu_set_t orig_mask; + cpu_set_t *orig_mask = NULL; ssize_t bytes; u8 *buf; int ret; @@ -435,6 +479,7 @@ static u8 *alloc_data(ssize_t bytes0, int map_flags, /* Restore affinity: */ if (init_cpu0) { bind_to_cpumask(orig_mask); + CPU_FREE(orig_mask); mempol_restore(); } @@ -595,6 +640,7 @@ static int parse_setup_cpu_list(void) BUG_ON(bind_cpu_0 > bind_cpu_1); for (bind_cpu = bind_cpu_0; bind_cpu <= bind_cpu_1; bind_cpu += step) { + size_t size = CPU_ALLOC_SIZE(g->p.nr_cpus); int i; for (i = 0; i < mul; i++) { @@ -614,10 +660,15 @@ static int parse_setup_cpu_list(void) tprintf("%2d", bind_cpu); } - CPU_ZERO(&td->bind_cpumask); + td->bind_cpumask = CPU_ALLOC(g->p.nr_cpus); + BUG_ON(!td->bind_cpumask); + CPU_ZERO_S(size, td->bind_cpumask); for (cpu = bind_cpu; cpu < bind_cpu+bind_len; cpu++) { - BUG_ON(cpu < 0 || cpu >= g->p.nr_cpus); - CPU_SET(cpu, &td->bind_cpumask); + if (cpu < 0 || cpu >= g->p.nr_cpus) { + CPU_FREE(td->bind_cpumask); + BUG_ON(-1); + } + CPU_SET_S(cpu, size, td->bind_cpumask); } t++; } @@ -1245,7 +1296,7 @@ static void *worker_thread(void *__tdata) * by migrating to CPU#0: */ if (first_task && g->p.perturb_secs && (int)(stop.tv_sec - last_perturbance) >= g->p.perturb_secs) { - cpu_set_t orig_mask; + cpu_set_t *orig_mask; int target_cpu; int this_cpu; @@ -1269,6 +1320,7 @@ static void *worker_thread(void *__tdata) printf(" (injecting perturbalance, moved to CPU#%d)\n", target_cpu); bind_to_cpumask(orig_mask); + CPU_FREE(orig_mask); } if (details >= 3) { @@ -1402,21 +1454,31 @@ static void init_thread_data(void) for (t = 0; t < g->p.nr_tasks; t++) { struct thread_data *td = g->threads + t; + size_t cpuset_size = CPU_ALLOC_SIZE(g->p.nr_cpus); int cpu; /* Allow all nodes by default: */ td->bind_node = NUMA_NO_NODE; /* Allow all CPUs by default: */ - CPU_ZERO(&td->bind_cpumask); + td->bind_cpumask = CPU_ALLOC(g->p.nr_cpus); + BUG_ON(!td->bind_cpumask); + CPU_ZERO_S(cpuset_size, td->bind_cpumask); for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &td->bind_cpumask); + CPU_SET_S(cpu, cpuset_size, td->bind_cpumask); } } static void deinit_thread_data(void) { ssize_t size = sizeof(*g->threads)*g->p.nr_tasks; + int t; + + /* Free the bind_cpumask allocated for thread_data */ + for (t = 0; t < g->p.nr_tasks; t++) { + struct thread_data *td = g->threads + t; + CPU_FREE(td->bind_cpumask); + } free_data(g->threads, size); } -- cgit v1.2.3