[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <ZrpAxUEnCUNBcwCS@x1>
Date: Mon, 12 Aug 2024 14:05:09 -0300
From: Arnaldo Carvalho de Melo <acme@...nel.org>
To: Namhyung Kim <namhyung@...nel.org>
Cc: Ian Rogers <irogers@...gle.com>, Kan Liang <kan.liang@...ux.intel.com>,
Jiri Olsa <jolsa@...nel.org>,
Adrian Hunter <adrian.hunter@...el.com>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...nel.org>, LKML <linux-kernel@...r.kernel.org>,
linux-perf-users@...r.kernel.org,
Dan Carpenter <dan.carpenter@...aro.org>
Subject: Re: [PATCH] perf lock contention: Change stack_id type to s32
On Mon, Aug 12, 2024 at 09:57:27AM -0700, Namhyung Kim wrote:
> On Mon, Aug 12, 2024 at 9:52 AM Arnaldo Carvalho de Melo
> <acme@...nel.org> wrote:
> >
> > On Mon, Aug 12, 2024 at 01:09:40PM -0300, Arnaldo Carvalho de Melo wrote:
> > > On Sat, Aug 10, 2024 at 12:17:04PM -0700, Namhyung Kim wrote:
> > > > The bpf_get_stackid() helper returns a signed type to check whether it
> > > > failed to get a stacktrace or not. But it saved the result in u32 and
> > > > checked if the value is negative.
> > > >
> > > > 376 if (needs_callstack) {
> > > > 377 pelem->stack_id = bpf_get_stackid(ctx, &stacks,
> > > > 378 BPF_F_FAST_STACK_CMP | stack_skip);
> > > > --> 379 if (pelem->stack_id < 0)
> > > >
> > > > ./tools/perf/util/bpf_skel/lock_contention.bpf.c:379 contention_begin()
> > > > warn: unsigned 'pelem->stack_id' is never less than zero.
> > > >
> > > > Let's change the type to s32 instead.
> > > >
> > > > Fixes: 6d499a6b3d90 ("perf lock: Print the number of lost entries for BPF")
> > >
> > > Thanks, applied to perf-tools-next,
> >
> > I'll try to fix this later, but now it fails the first 'make -C
> > tools/perf build-test' target, that you can run directly as:
> >
> > ⬢[acme@...lbox perf-tools-next]$ tools/perf/tests/perf-targz-src-pkg tools/perf
> > <SNIP>
> > CLANG /tmp/tmp.FRZLVEwqdz/perf-6.11.0-rc2/tools/perf/util/bpf_skel/.tmp/augmented_raw_syscalls.bpf.o
> > In file included from util/bpf_skel/lock_contention.bpf.c:9:
> > util/bpf_skel/lock_data.h:10:2: error: unknown type name 's32'; did you mean 'u32'?
> > 10 | s32 stack_id;
> > | ^~~
> > | u32
> > util/bpf_skel/vmlinux.h:17:15: note: 'u32' declared here
> > 17 | typedef __u32 u32;
> > | ^
>
> Oops, sorry about this. There was a kernel test robot report.
Cool, I didn't see it, but its good its doing this work.
> It seems we need 'typedef __s32 s32;' too.
Please send a v2 then,
Thanks,
- Arnaldo
> Thanks,
> Namhyung
>
>
> > In file included from util/bpf_skel/lock_contention.bpf.c:9:
> > util/bpf_skel/lock_data.h:14:2: error: unknown type name 's32'; did you mean 'u32'?
> > 14 | s32 stack_id;
> > | ^~~
> > | u32
> > util/bpf_skel/vmlinux.h:17:15: note: 'u32' declared here
> > 17 | typedef __u32 u32;
> > | ^
> > 2 errors generated.
> > make[2]: *** [Makefile.perf:1247: /tmp/tmp.FRZLVEwqdz/perf-6.11.0-rc2/tools/perf/util/bpf_skel/.tmp/lock_contention.bpf.o] Error 1
> > make[2]: *** Waiting for unfinished jobs....
> > make[1]: *** [Makefile.perf:292: sub-make] Error 2
> > make: *** [Makefile:76: all] Error 2
> > make: Leaving directory '/tmp/tmp.FRZLVEwqdz/perf-6.11.0-rc2/tools/perf'
> > ⬢[acme@...lbox perf-tools-next]$
Powered by blists - more mailing lists