Hello,
On Wed, Sep 04, 2024 at 12:31:02PM +0000, Tengda Wu wrote:
> bperf has a nice ability to share PMUs, but it still does not support
> inherit events during fork(), resulting in some deviations in its stat
> results compared with perf.
>
> perf stat result:
> $ ./perf stat -e cycles,instructions -- ./perf test -w sqrtloop
>
> Performance counter stats for './perf test -w sqrtloop':
>
> 2,316,038,116 cycles
> 2,859,350,725 instructions # 1.23 insn per cycle
>
> 1.009603637 seconds time elapsed
>
> 1.004196000 seconds user
> 0.003950000 seconds sys
>
> bperf stat result:
> $ ./perf stat --bpf-counters -e cycles,instructions -- ./perf test -w sqrtloop
>
> Performance counter stats for './perf test -w sqrtloop':
>
> 18,762,093 cycles
> 23,487,766 instructions # 1.25 insn per cycle
>
> 1.008913769 seconds time elapsed
>
> 1.003248000 seconds user
> 0.004069000 seconds sys
>
> In order to support event inheritance, two new bpf programs are added
> to monitor the fork and exit of tasks respectively. When a task is
> created, add it to the filter map to enable counting, and reuse the
> `accum_key` of its parent task to count together with the parent task.
> When a task exits, remove it from the filter map to disable counting.
Right, I think we may need this for other BPF programs.
>
> After support:
> $ ./perf stat --bpf-counters -e cycles,instructions -- ./perf test -w sqrtloop
>
> Performance counter stats for './perf test -w sqrtloop':
>
> 2,316,543,537 cycles
> 2,859,677,779 instructions # 1.23 insn per cycle
>
> 1.009566332 seconds time elapsed
>
> 1.004414000 seconds user
> 0.003545000 seconds sys
>
> Signed-off-by: Tengda Wu <wutengda@xxxxxxxxxxxxxxx>
> ---
> tools/perf/util/bpf_counter.c | 9 +--
> tools/perf/util/bpf_skel/bperf_follower.bpf.c | 75 +++++++++++++++++--
> tools/perf/util/bpf_skel/bperf_u.h | 5 ++
> 3 files changed, 78 insertions(+), 11 deletions(-)
>
> diff --git a/tools/perf/util/bpf_counter.c b/tools/perf/util/bpf_counter.c
> index 7a8af60e0f51..e07ff04b934f 100644
> --- a/tools/perf/util/bpf_counter.c
> +++ b/tools/perf/util/bpf_counter.c
> @@ -529,9 +529,6 @@ static int bperf__load(struct evsel *evsel, struct target *target)
> /* set up reading map */
> bpf_map__set_max_entries(evsel->follower_skel->maps.accum_readings,
> filter_entry_cnt);
> - /* set up follower filter based on target */
> - bpf_map__set_max_entries(evsel->follower_skel->maps.filter,
> - filter_entry_cnt);
> err = bperf_follower_bpf__load(evsel->follower_skel);
> if (err) {
> pr_err("Failed to load follower skeleton\n");
> @@ -543,6 +540,7 @@ static int bperf__load(struct evsel *evsel, struct target *target)
> for (i = 0; i < filter_entry_cnt; i++) {
> int filter_map_fd;
> __u32 key;
> + struct bperf_filter_value fval = { i, 0 };
>
> if (filter_type == BPERF_FILTER_PID ||
> filter_type == BPERF_FILTER_TGID)
> @@ -553,10 +551,11 @@ static int bperf__load(struct evsel *evsel, struct target *target)
> break;
>
> filter_map_fd = bpf_map__fd(evsel->follower_skel->maps.filter);
> - bpf_map_update_elem(filter_map_fd, &key, &i, BPF_ANY);
> + bpf_map_update_elem(filter_map_fd, &key, &fval, BPF_ANY);
> }
>
> evsel->follower_skel->bss->type = filter_type;
> + evsel->follower_skel->bss->init_filter_entries = filter_entry_cnt;
Do you need this in the BPF program?
>
> err = bperf_follower_bpf__attach(evsel->follower_skel);
>
> @@ -623,7 +622,7 @@ static int bperf__read(struct evsel *evsel)
> bperf_sync_counters(evsel);
> reading_map_fd = bpf_map__fd(skel->maps.accum_readings);
>
> - for (i = 0; i < bpf_map__max_entries(skel->maps.accum_readings); i++) {
> + for (i = 0; i < skel->bss->init_filter_entries; i++) {
> struct perf_cpu entry;
> __u32 cpu;
>
> diff --git a/tools/perf/util/bpf_skel/bperf_follower.bpf.c b/tools/perf/util/bpf_skel/bperf_follower.bpf.c
> index f193998530d4..59fab421526a 100644
> --- a/tools/perf/util/bpf_skel/bperf_follower.bpf.c
> +++ b/tools/perf/util/bpf_skel/bperf_follower.bpf.c
> @@ -5,6 +5,8 @@
> #include <bpf/bpf_tracing.h>
> #include "bperf_u.h"
>
> +#define MAX_ENTRIES 102400
> +
> struct {
> __uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
> __uint(key_size, sizeof(__u32));
> @@ -22,25 +24,29 @@ struct {
> struct {
> __uint(type, BPF_MAP_TYPE_HASH);
> __uint(key_size, sizeof(__u32));
> - __uint(value_size, sizeof(__u32));
> + __uint(value_size, sizeof(struct bperf_filter_value));
> + __uint(max_entries, MAX_ENTRIES);
> + __uint(map_flags, BPF_F_NO_PREALLOC);
> } filter SEC(".maps");
>
> enum bperf_filter_type type = 0;
> int enabled = 0;
> +__u32 init_filter_entries = 0;
>
> SEC("fexit/XXX")
> int BPF_PROG(fexit_XXX)
> {
> struct bpf_perf_event_value *diff_val, *accum_val;
> __u32 filter_key, zero = 0;
> - __u32 *accum_key;
> + __u32 accum_key;
> + struct bperf_filter_value *fval;
>
> if (!enabled)
> return 0;
>
> switch (type) {
> case BPERF_FILTER_GLOBAL:
> - accum_key = &zero;
> + accum_key = zero;
> goto do_add;
> case BPERF_FILTER_CPU:
> filter_key = bpf_get_smp_processor_id();
> @@ -55,16 +61,20 @@ int BPF_PROG(fexit_XXX)
> return 0;
> }
>
> - accum_key = bpf_map_lookup_elem(&filter, &filter_key);
> - if (!accum_key)
> + fval = bpf_map_lookup_elem(&filter, &filter_key);
> + if (!fval)
> return 0;
>
> + accum_key = fval->accum_key;
> + if (fval->exited)
> + bpf_map_delete_elem(&filter, &filter_key);
> +
> do_add:
> diff_val = bpf_map_lookup_elem(&diff_readings, &zero);
> if (!diff_val)
> return 0;
>
> - accum_val = bpf_map_lookup_elem(&accum_readings, accum_key);
> + accum_val = bpf_map_lookup_elem(&accum_readings, &accum_key);
> if (!accum_val)
> return 0;
>
> @@ -75,4 +85,57 @@ int BPF_PROG(fexit_XXX)
> return 0;
> }
>
> +SEC("tp_btf/task_newtask")
> +int BPF_PROG(on_newtask, struct task_struct *task, __u64 clone_flags)
> +{
> + __u32 parent_pid, child_pid;
> + struct bperf_filter_value *parent_fval;
> + struct bperf_filter_value child_fval = { 0 };
> +
> + if (!enabled)
> + return 0;
> +
> + if (type != BPERF_FILTER_PID && type != BPERF_FILTER_TGID)
> + return 0;
I think you can load/attach this program only if the type is either
PID or TGID. Then it'd reduce the overhead.
> +
> + parent_pid = bpf_get_current_pid_tgid() >> 32;
> + child_pid = task->pid;
> +
> + /* Check if the current task is one of the target tasks to be counted */
> + parent_fval = bpf_map_lookup_elem(&filter, &parent_pid);
> + if (!parent_fval)
> + return 0;
> +
> + /* Start counting for the new task by adding it into filter map,
> + * inherit the accum key of its parent task so that they can be
> + * counted together.
> + */
> + child_fval.accum_key = parent_fval->accum_key;
> + child_fval.exited = 0;
> + bpf_map_update_elem(&filter, &child_pid, &child_fval, BPF_NOEXIST);
> +
> + return 0;
> +}
> +
> +SEC("tp_btf/sched_process_exit")
> +int BPF_PROG(on_exittask, struct task_struct *task)
> +{
> + __u32 pid;
> + struct bperf_filter_value *fval;
> +
> + if (!enabled)
> + return 0;
> +
> + if (type != BPERF_FILTER_PID && type != BPERF_FILTER_TGID)
> + return 0;
Ditto.
Thanks,
Namhyung
> +
> + /* Stop counting for this task by removing it from filter map */
> + pid = task->pid;
> + fval = bpf_map_lookup_elem(&filter, &pid);
> + if (fval)
> + fval->exited = 1;
> +
> + return 0;
> +}
> +
> char LICENSE[] SEC("license") = "Dual BSD/GPL";
> diff --git a/tools/perf/util/bpf_skel/bperf_u.h b/tools/perf/util/bpf_skel/bperf_u.h
> index 1ce0c2c905c1..4a4a753980be 100644
> --- a/tools/perf/util/bpf_skel/bperf_u.h
> +++ b/tools/perf/util/bpf_skel/bperf_u.h
> @@ -11,4 +11,9 @@ enum bperf_filter_type {
> BPERF_FILTER_TGID,
> };
>
> +struct bperf_filter_value {
> + __u32 accum_key;
> + __u8 exited;
> +};
> +
> #endif /* __BPERF_STAT_U_H */
> --
> 2.34.1
>
