Re: [PATCH v7 05/10] perf record --off-cpu: Dump off-cpu samples in BPF

[Ian Rogers <irogers@google.com>]

On Fri, Nov 8, 2024 at 12:41 PM Howard Chu <howardchu95@gmail.com> wrote:
>
> Collect tid, period, callchain, and cgroup id and dump them when off-cpu
> time threshold is reached.
>
> We don't collect the off-cpu time twice (the delta), it's either in
> direct samples, or accumulated samples that are dumped at the end of
> perf.data.
>
> Suggested-by: Namhyung Kim <namhyung@kernel.org>
> Signed-off-by: Howard Chu <howardchu95@gmail.com>
> ---
>  tools/perf/util/bpf_skel/off_cpu.bpf.c | 83 ++++++++++++++++++++++++--
>  1 file changed, 78 insertions(+), 5 deletions(-)
>
> diff --git a/tools/perf/util/bpf_skel/off_cpu.bpf.c b/tools/perf/util/bpf_skel/off_cpu.bpf.c
> index dc6acafb9353..bf652c30b1c9 100644
> --- a/tools/perf/util/bpf_skel/off_cpu.bpf.c
> +++ b/tools/perf/util/bpf_skel/off_cpu.bpf.c
> @@ -18,10 +18,18 @@
>  #define MAX_STACKS   32
>  #define MAX_ENTRIES  102400
>
> +#define MAX_CPUS  4096
> +#define MAX_OFFCPU_LEN 37
> +
> +struct stack {
> +       u64 array[MAX_STACKS];
> +};
> +
>  struct tstamp_data {
>         __u32 stack_id;
>         __u32 state;
>         __u64 timestamp;
> +       struct stack stack;
>  };
>
>  struct offcpu_key {
> @@ -39,6 +47,24 @@ struct {
>         __uint(max_entries, MAX_ENTRIES);
>  } stacks SEC(".maps");
>
> +struct offcpu_data {
> +       u64 array[MAX_OFFCPU_LEN];
> +};
> +
> +struct {
> +       __uint(type, BPF_MAP_TYPE_PERF_EVENT_ARRAY);
> +       __uint(key_size, sizeof(__u32));
> +       __uint(value_size, sizeof(__u32));
> +       __uint(max_entries, MAX_CPUS);
> +} offcpu_output SEC(".maps");

Does patch 4 build without this definition? (we're in patch 5 here). I
think this should be in patch 4.

> +
> +struct {
> +       __uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
> +       __uint(key_size, sizeof(__u32));
> +       __uint(value_size, sizeof(struct offcpu_data));
> +       __uint(max_entries, 1);
> +} offcpu_payload SEC(".maps");
> +
>  struct {
>         __uint(type, BPF_MAP_TYPE_TASK_STORAGE);
>         __uint(map_flags, BPF_F_NO_PREALLOC);
> @@ -185,6 +211,39 @@ static inline int can_record(struct task_struct *t, int state)
>         return 1;
>  }
>
> +static inline int copy_stack(struct stack *from, struct offcpu_data *to, int n)
> +{
> +       int len = 0;
> +
> +       for (int i = 0; i < MAX_STACKS && from->array[i]; ++i, ++len)
> +               to->array[n + 2 + i] = from->array[i];
> +
> +       return len;
> +}
> +

Dump is something of a generic name. Could you kernel-doc this
function to describe the behavior?

> +static int off_cpu_dump(void *ctx, struct offcpu_data *data, struct offcpu_key *key,
> +                       struct stack *stack, __u64 delta, __u64 timestamp)
> +{
> +       /* dump tid, period, callchain, and cgroup id */
> +       int n = 0, len = 0;
> +
> +       data->array[n++] = (u64)key->tgid << 32 | key->pid;
> +       data->array[n++] = delta;
> +
> +       /* data->array[n] is callchain->nr (updated later) */
> +       data->array[n + 1] = PERF_CONTEXT_USER;
> +       data->array[n + 2] = 0;
> +       len = copy_stack(stack, data, n);
> +
> +       /* update length of callchain */
> +       data->array[n] = len + 1;
> +       n += len + 2;
> +
> +       data->array[n++] = key->cgroup_id;
> +
> +       return bpf_perf_event_output(ctx, &offcpu_output, BPF_F_CURRENT_CPU, data, n * sizeof(u64));
> +}
> +
>  static int off_cpu_stat(u64 *ctx, struct task_struct *prev,
>                         struct task_struct *next, int state)
>  {
> @@ -209,6 +268,12 @@ static int off_cpu_stat(u64 *ctx, struct task_struct *prev,
>         pelem->state = state;
>         pelem->stack_id = stack_id;
>
> +       /*
> +        * If stacks are successfully collected by bpf_get_stackid(), collect them once more
> +        * in task_storage for direct off-cpu sample dumping
> +        */
> +       if (stack_id > 0 && bpf_get_stack(ctx, &pelem->stack, MAX_STACKS * sizeof(u64), BPF_F_USER_STACK)) {
> +       }

Why the empty if?

> +
>  next:
>         pelem = bpf_task_storage_get(&tstamp, next, NULL, 0);
>
> @@ -223,11 +288,19 @@ static int off_cpu_stat(u64 *ctx, struct task_struct *prev,
>                 __u64 delta = ts - pelem->timestamp;
>                 __u64 *total;
>
> -               total = bpf_map_lookup_elem(&off_cpu, &key);
> -               if (total)
> -                       *total += delta;
> -               else
> -                       bpf_map_update_elem(&off_cpu, &key, &delta, BPF_ANY);
> +               if (delta >= offcpu_thresh) {
> +                       int zero = 0;
> +                       struct offcpu_data *data = bpf_map_lookup_elem(&offcpu_payload, &zero);
> +
> +                       if (data)
> +                               off_cpu_dump(ctx, data, &key, &pelem->stack, delta, pelem->timestamp);
> +               } else {
> +                       total = bpf_map_lookup_elem(&off_cpu, &key);
> +                       if (total)
> +                               *total += delta;
> +                       else
> +                               bpf_map_update_elem(&off_cpu, &key, &delta, BPF_ANY);
> +               }

Looks good! :-)

Thanks,
Ian

>
>                 /* prevent to reuse the timestamp later */
>                 pelem->timestamp = 0;
> --
> 2.43.0
>