| Message ID | 20221006160044.3397237-1-sumanthk@linux.ibm.com (mailing list archive) |
|---|---|
| State | RFC |
| Delegated to: | BPF |
| Headers | show |
| Series | Re: [tip: perf/core] perf: Use sample_flags for raw_data | expand |
| Context | Check | Description |
|---|---|---|
| netdev/tree_selection | success | Guessing tree name failed - patch did not apply |
| bpf/vmtest-bpf-PR | fail | merge-conflict |
Hello, On Thu, Oct 6, 2022 at 9:01 AM Sumanth Korikkar <sumanthk@linux.ibm.com> wrote: > > Hi, > > This causes segfaults. > > Steps to recreate: > * Run ./samples/bpf/trace_output > BUG pid 9 cookie 1001000000004 sized 4 > BUG pid 9 cookie 1001000000004 sized 4 > BUG pid 9 cookie 1001000000004 sized 4 > Segmentation fault (core dumped) > > Problem: > * The following commit sets data->raw to NULL, when the raw data is not filled > by PMU driver. This leads to stale data. > > * raw data could also be filled by bpf_perf_event_output(), bpf_event_output() > ... > 686 perf_sample_data_init(sd, 0, 0); > 687 sd->raw = &raw; > 688 > 689 err = __bpf_perf_event_output(regs, map, flags, sd); > ... > > * The below patch eliminates segfaults. However, contradicts with > the description mentioned in this commit (Filled by only PMU driver). Thank you for the fix. Don't worry about the description - it said it's usually filled by PMU drivers and it should be fine as long as you set the sample flags after filling the raw data. Acked-by: Namhyung Kim <namhyung@kernel.org> Thanks, Namhyung > > diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c > index 49fb9ec8366d..1ed08967fb97 100644 > --- a/kernel/trace/bpf_trace.c > +++ b/kernel/trace/bpf_trace.c > @@ -687,6 +687,7 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, > > perf_sample_data_init(sd, 0, 0); > sd->raw = &raw; > + sd->sample_flags |= PERF_SAMPLE_RAW; > > err = __bpf_perf_event_output(regs, map, flags, sd); > > @@ -745,6 +746,7 @@ u64 bpf_event_output(struct bpf_map *map, u64 flags, void *meta, u64 meta_size, > perf_fetch_caller_regs(regs); > perf_sample_data_init(sd, 0, 0); > sd->raw = &raw; > + sd->sample_flags |= PERF_SAMPLE_RAW; > > ret = __bpf_perf_event_output(regs, map, flags, sd); > out: > > -- > Thanks, > Sumanth
On Thu, Oct 06, 2022 at 06:00:44PM +0200, Sumanth Korikkar wrote: > Hi, > > This causes segfaults. > > Steps to recreate: > * Run ./samples/bpf/trace_output > BUG pid 9 cookie 1001000000004 sized 4 > BUG pid 9 cookie 1001000000004 sized 4 > BUG pid 9 cookie 1001000000004 sized 4 > Segmentation fault (core dumped) > > Problem: > * The following commit sets data->raw to NULL, when the raw data is not filled > by PMU driver. This leads to stale data. > > * raw data could also be filled by bpf_perf_event_output(), bpf_event_output() > ... > 686 perf_sample_data_init(sd, 0, 0); > 687 sd->raw = &raw; > 688 > 689 err = __bpf_perf_event_output(regs, map, flags, sd); > ... > > * The below patch eliminates segfaults. However, contradicts with > the description mentioned in this commit (Filled by only PMU driver). hi, could you please resend the patch with formal changelog and Fixes tag? thanks, jirka > > diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c > index 49fb9ec8366d..1ed08967fb97 100644 > --- a/kernel/trace/bpf_trace.c > +++ b/kernel/trace/bpf_trace.c > @@ -687,6 +687,7 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, > > perf_sample_data_init(sd, 0, 0); > sd->raw = &raw; > + sd->sample_flags |= PERF_SAMPLE_RAW; > > err = __bpf_perf_event_output(regs, map, flags, sd); > > @@ -745,6 +746,7 @@ u64 bpf_event_output(struct bpf_map *map, u64 flags, void *meta, u64 meta_size, > perf_fetch_caller_regs(regs); > perf_sample_data_init(sd, 0, 0); > sd->raw = &raw; > + sd->sample_flags |= PERF_SAMPLE_RAW; > > ret = __bpf_perf_event_output(regs, map, flags, sd); > out: > > -- > Thanks, > Sumanth
diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 49fb9ec8366d..1ed08967fb97 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -687,6 +687,7 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, perf_sample_data_init(sd, 0, 0); sd->raw = &raw; + sd->sample_flags |= PERF_SAMPLE_RAW; err = __bpf_perf_event_output(regs, map, flags, sd); @@ -745,6 +746,7 @@ u64 bpf_event_output(struct bpf_map *map, u64 flags, void *meta, u64 meta_size, perf_fetch_caller_regs(regs); perf_sample_data_init(sd, 0, 0); sd->raw = &raw; + sd->sample_flags |= PERF_SAMPLE_RAW; ret = __bpf_perf_event_output(regs, map, flags, sd); out: