On Thu, Sep 22, 2022 at 4:55 PM Ian Rogers <irogers@xxxxxxxxxx> wrote:
>
> On Thu, Sep 22, 2022 at 1:53 PM Namhyung Kim <namhyung@xxxxxxxxxx> wrote:
> >
> > The perf_event_cgrp_id can be different on other configurations.
> > To be more portable as CO-RE, it needs to get the cgroup subsys id
> > using the bpf_core_enum_value() helper.
> >
> > Suggested-by: Ian Rogers <irogers@xxxxxxxxxx>
> > Signed-off-by: Namhyung Kim <namhyung@xxxxxxxxxx>
> > ---
> > v3 changes)
> > * check compiler features for enum value
> >
> > v2 changes)
> > * fix off_cpu.bpf.c too
> > * get perf_subsys_id only once
> >
> > tools/perf/util/bpf_skel/bperf_cgroup.bpf.c | 11 ++++++++++-
> > tools/perf/util/bpf_skel/off_cpu.bpf.c | 12 ++++++++----
> > 2 files changed, 18 insertions(+), 5 deletions(-)
> >
> > diff --git a/tools/perf/util/bpf_skel/bperf_cgroup.bpf.c b/tools/perf/util/bpf_skel/bperf_cgroup.bpf.c
> > index 292c430768b5..8e7520e273db 100644
> > --- a/tools/perf/util/bpf_skel/bperf_cgroup.bpf.c
> > +++ b/tools/perf/util/bpf_skel/bperf_cgroup.bpf.c
> > @@ -48,6 +48,7 @@ const volatile __u32 num_cpus = 1;
> >
> > int enabled = 0;
> > int use_cgroup_v2 = 0;
> > +int perf_subsys_id = -1;
> >
> > static inline int get_cgroup_v1_idx(__u32 *cgrps, int size)
> > {
> > @@ -58,7 +59,15 @@ static inline int get_cgroup_v1_idx(__u32 *cgrps, int size)
> > int level;
> > int cnt;
> >
> > - cgrp = BPF_CORE_READ(p, cgroups, subsys[perf_event_cgrp_id], cgroup);
> > + if (perf_subsys_id == -1) {
> > +#if __has_builtin(__builtin_preserve_enum_value)
> > + perf_subsys_id = bpf_core_enum_value(enum cgroup_subsys_id,
> > + perf_event_cgrp_id);
> > +#else
> > + perf_subsys_id = perf_event_cgrp_id;
> > +#endif
> > + }
> > + cgrp = BPF_CORE_READ(p, cgroups, subsys[perf_subsys_id], cgroup);
> > level = BPF_CORE_READ(cgrp, level);
> >
> > for (cnt = 0; i < MAX_LEVELS; i++) {
> > diff --git a/tools/perf/util/bpf_skel/off_cpu.bpf.c b/tools/perf/util/bpf_skel/off_cpu.bpf.c
> > index c4ba2bcf179f..e917ef7b8875 100644
> > --- a/tools/perf/util/bpf_skel/off_cpu.bpf.c
> > +++ b/tools/perf/util/bpf_skel/off_cpu.bpf.c
> > @@ -94,6 +94,8 @@ const volatile bool has_prev_state = false;
> > const volatile bool needs_cgroup = false;
> > const volatile bool uses_cgroup_v1 = false;
> >
> > +int perf_subsys_id = -1;
> > +
> > /*
> > * Old kernel used to call it task_struct->state and now it's '__state'.
> > * Use BPF CO-RE "ignored suffix rule" to deal with it like below:
> > @@ -119,11 +121,13 @@ static inline __u64 get_cgroup_id(struct task_struct *t)
> > {
> > struct cgroup *cgrp;
> >
> > - if (uses_cgroup_v1)
> > - cgrp = BPF_CORE_READ(t, cgroups, subsys[perf_event_cgrp_id], cgroup);
> > - else
> > - cgrp = BPF_CORE_READ(t, cgroups, dfl_cgrp);
> > + if (!uses_cgroup_v1)
> > + return BPF_CORE_READ(t, cgroups, dfl_cgrp, kn, id);
> > +
> > + if (perf_subsys_id == -1)
> > + perf_subsys_id = bpf_core_enum_value(enum cgroup_subsys_id, perf_event_cgrp_id);
>
> Should the "#if __has_builtin(__builtin_preserve_enum_value)" test also be here?
Ouch! Will send v4.
>
> It feels a shame that bpf_core_enum_value isn't defined something like:
>
> #if __has_builtin(__builtin_preserve_enum_value)
> #define bpf_core_enum_value(enum_type, enum_value) \
> __builtin_preserve_enum_value(*(typeof(enum_type) *)enum_value,
> BPF_ENUMVAL_VALUE)
> #else
> #define bpf_core_enum_value(enum_type, enum_value) enum_value
> #endif
Yeah, that would be much easier to use.
Thanks,
Namhyung
>
> for backward clang compatibility, but I could see why an error would
> be preferable.
>
> Thanks,
> Ian
>
> >
> > + cgrp = BPF_CORE_READ(t, cgroups, subsys[perf_subsys_id], cgroup);
> > return BPF_CORE_READ(cgrp, kn, id);
> > }
> >
> > --
> > 2.37.3.998.g577e59143f-goog
> >
