Re: [PATCH 2/2] bpf/selftests: Test skipping stacktrace

[Namhyung Kim <namhyung@kernel.org>]

Hello,

On Fri, Mar 11, 2022 at 2:23 PM Andrii Nakryiko
<andrii.nakryiko@gmail.com> wrote:
>
> On Thu, Mar 10, 2022 at 12:22 AM Namhyung Kim <namhyung@kernel.org> wrote:
> > +SEC("tracepoint/sched/sched_switch")
> > +int oncpu(struct sched_switch_args *ctx)
> > +{
> > +       __u32 max_len = TEST_STACK_DEPTH * sizeof(__u64);
> > +       __u32 key = 0, val = 0, *value_p;
> > +       __u64 *stack_p;
> > +
>
> please also add filtering by PID to avoid interference from other
> selftests when run in parallel mode

Will do!

Thanks,
Namhyung

>
> > +       value_p = bpf_map_lookup_elem(&control_map, &key);
> > +       if (value_p && *value_p)
> > +               return 0; /* skip if non-zero *value_p */
> > +
> > +       /* it should allow skipping whole buffer size entries */
> > +       key = bpf_get_stackid(ctx, &stackmap, TEST_STACK_DEPTH);
> > +       if ((int)key >= 0) {
> > +               /* The size of stackmap and stack_amap should be the same */
> > +               bpf_map_update_elem(&stackid_hmap, &key, &val, 0);
> > +               stack_p = bpf_map_lookup_elem(&stack_amap, &key);
> > +               if (stack_p) {
> > +                       bpf_get_stack(ctx, stack_p, max_len, TEST_STACK_DEPTH);
> > +                       /* it wrongly skipped all the entries and filled zero */
> > +                       if (stack_p[0] == 0)
> > +                               failed = 1;
> > +               }
> > +       } else if ((int)key == -14/*EFAULT*/) {
> > +               /* old kernel doesn't support skipping that many entries */
> > +               failed = 2;
> > +       }
> > +
> > +       return 0;
> > +}
> > +
> > +char _license[] SEC("license") = "GPL";
> > --
> > 2.35.1.723.g4982287a31-goog
> >