@@ -73,14 +73,12 @@ static void off_cpu_start(void *arg)
struct evlist *evlist = arg;
/* update task filter for the given workload */
- if (!skel->bss->has_cpu && !skel->bss->has_task &&
+ if (skel->rodata->has_task && skel->rodata->uses_tgid &&
perf_thread_map__pid(evlist->core.threads, 0) != -1) {
int fd;
u32 pid;
u8 val = 1;
- skel->bss->has_task = 1;
- skel->bss->uses_tgid = 1;
fd = bpf_map__fd(skel->maps.task_filter);
pid = perf_thread_map__pid(evlist->core.threads, 0);
bpf_map_update_elem(fd, &pid, &val, BPF_ANY);
@@ -148,6 +146,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
if (target->cpu_list) {
ncpus = perf_cpu_map__nr(evlist->core.user_requested_cpus);
bpf_map__set_max_entries(skel->maps.cpu_filter, ncpus);
+ skel->rodata->has_cpu = 1;
}
if (target->pid) {
@@ -173,11 +172,16 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
ntasks = MAX_PROC;
bpf_map__set_max_entries(skel->maps.task_filter, ntasks);
+ skel->rodata->has_task = 1;
+ skel->rodata->uses_tgid = 1;
} else if (target__has_task(target)) {
ntasks = perf_thread_map__nr(evlist->core.threads);
bpf_map__set_max_entries(skel->maps.task_filter, ntasks);
+ skel->rodata->has_task = 1;
} else if (target__none(target)) {
bpf_map__set_max_entries(skel->maps.task_filter, MAX_PROC);
+ skel->rodata->has_task = 1;
+ skel->rodata->uses_tgid = 1;
}
if (evlist__first(evlist)->cgrp) {
@@ -186,6 +190,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
if (!cgroup_is_v2("perf_event"))
skel->rodata->uses_cgroup_v1 = true;
+ skel->rodata->has_cgroup = 1;
}
if (opts->record_cgroup) {
@@ -208,7 +213,6 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
u32 cpu;
u8 val = 1;
- skel->bss->has_cpu = 1;
fd = bpf_map__fd(skel->maps.cpu_filter);
for (i = 0; i < ncpus; i++) {
@@ -220,8 +224,6 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
if (target->pid) {
u8 val = 1;
- skel->bss->has_task = 1;
- skel->bss->uses_tgid = 1;
fd = bpf_map__fd(skel->maps.task_filter);
strlist__for_each_entry(pos, pid_slist) {
@@ -240,7 +242,6 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
u32 pid;
u8 val = 1;
- skel->bss->has_task = 1;
fd = bpf_map__fd(skel->maps.task_filter);
for (i = 0; i < ntasks; i++) {
@@ -253,7 +254,6 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,
struct evsel *evsel;
u8 val = 1;
- skel->bss->has_cgroup = 1;
fd = bpf_map__fd(skel->maps.cgroup_filter);
evlist__for_each_entry(evlist, evsel) {
@@ -85,10 +85,11 @@ struct task_struct___old {
} __attribute__((preserve_access_index));
int enabled = 0;
-int has_cpu = 0;
-int has_task = 0;
-int has_cgroup = 0;
-int uses_tgid = 0;
+
+const volatile int has_cpu = 0;
+const volatile int has_task = 0;
+const volatile int has_cgroup = 0;
+const volatile int uses_tgid = 0;
const volatile bool has_prev_state = false;
const volatile bool needs_cgroup = false;
The control knobs set before loading BPF programs should be declared as 'const volatile' so that it can be optimized by the BPF core. Signed-off-by: Namhyung Kim <namhyung@kernel.org> --- tools/perf/util/bpf_off_cpu.c | 16 ++++++++-------- tools/perf/util/bpf_skel/off_cpu.bpf.c | 9 +++++---- 2 files changed, 13 insertions(+), 12 deletions(-)