linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file
@ 2025-11-07 17:31 Thomas Falcon
  2025-11-11 22:07 ` Ian Rogers
  2025-11-11 22:38 ` Namhyung Kim
  0 siblings, 2 replies; 5+ messages in thread
From: Thomas Falcon @ 2025-11-07 17:31 UTC (permalink / raw)
  To: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
	Namhyung Kim, Mark Rutland, Alexander Shishkin, Jiri Olsa,
	Ian Rogers, Adrian Hunter
  Cc: linux-perf-users, linux-kernel, Thomas Falcon

With commit f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF
info"), the write_bpf_( prog_info() | btf() ) functions exit
without writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.

process_bpf_( prog_info() | btf() ), however, still expect a "count"
value to exist in the data file. If btf information is empty, for
example, process_bpf_btf will read garbage or some other data as the
number of btf nodes in the data file. As a result, the data file will
not be processed correctly.

Instead, write the count to the data file and exit if it is zero.

Fixes: f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF info")
Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
---
 tools/perf/util/header.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index db2ad19fa50d..54968881481c 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -1022,12 +1022,9 @@ static int write_bpf_prog_info(struct feat_fd *ff,
 
 	down_read(&env->bpf_progs.lock);
 
-	if (env->bpf_progs.infos_cnt == 0)
-		goto out;
-
 	ret = do_write(ff, &env->bpf_progs.infos_cnt,
 		       sizeof(env->bpf_progs.infos_cnt));
-	if (ret < 0)
+	if (ret < 0 || env->bpf_progs.infos_cnt == 0)
 		goto out;
 
 	root = &env->bpf_progs.infos;
@@ -1067,13 +1064,10 @@ static int write_bpf_btf(struct feat_fd *ff,
 
 	down_read(&env->bpf_progs.lock);
 
-	if (env->bpf_progs.btfs_cnt == 0)
-		goto out;
-
 	ret = do_write(ff, &env->bpf_progs.btfs_cnt,
 		       sizeof(env->bpf_progs.btfs_cnt));
 
-	if (ret < 0)
+	if (ret < 0 || env->bpf_progs.btfs_cnt == 0)
 		goto out;
 
 	root = &env->bpf_progs.btfs;
-- 
2.47.3


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file
  2025-11-07 17:31 [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file Thomas Falcon
@ 2025-11-11 22:07 ` Ian Rogers
  2025-11-11 22:38 ` Namhyung Kim
  1 sibling, 0 replies; 5+ messages in thread
From: Ian Rogers @ 2025-11-11 22:07 UTC (permalink / raw)
  To: Thomas Falcon
  Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
	Namhyung Kim, Mark Rutland, Alexander Shishkin, Jiri Olsa,
	Adrian Hunter, linux-perf-users, linux-kernel

On Fri, Nov 7, 2025 at 9:32 AM Thomas Falcon <thomas.falcon@intel.com> wrote:
>
> With commit f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF
> info"), the write_bpf_( prog_info() | btf() ) functions exit
> without writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.
>
> process_bpf_( prog_info() | btf() ), however, still expect a "count"
> value to exist in the data file. If btf information is empty, for
> example, process_bpf_btf will read garbage or some other data as the
> number of btf nodes in the data file. As a result, the data file will
> not be processed correctly.
>
> Instead, write the count to the data file and exit if it is zero.
>
> Fixes: f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF info")
> Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>

Reviewed-by: Ian Rogers <irogers@google.com>

Thanks!
Ian

> ---
>  tools/perf/util/header.c | 10 ++--------
>  1 file changed, 2 insertions(+), 8 deletions(-)
>
> diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
> index db2ad19fa50d..54968881481c 100644
> --- a/tools/perf/util/header.c
> +++ b/tools/perf/util/header.c
> @@ -1022,12 +1022,9 @@ static int write_bpf_prog_info(struct feat_fd *ff,
>
>         down_read(&env->bpf_progs.lock);
>
> -       if (env->bpf_progs.infos_cnt == 0)
> -               goto out;
> -
>         ret = do_write(ff, &env->bpf_progs.infos_cnt,
>                        sizeof(env->bpf_progs.infos_cnt));
> -       if (ret < 0)
> +       if (ret < 0 || env->bpf_progs.infos_cnt == 0)
>                 goto out;
>
>         root = &env->bpf_progs.infos;
> @@ -1067,13 +1064,10 @@ static int write_bpf_btf(struct feat_fd *ff,
>
>         down_read(&env->bpf_progs.lock);
>
> -       if (env->bpf_progs.btfs_cnt == 0)
> -               goto out;
> -
>         ret = do_write(ff, &env->bpf_progs.btfs_cnt,
>                        sizeof(env->bpf_progs.btfs_cnt));
>
> -       if (ret < 0)
> +       if (ret < 0 || env->bpf_progs.btfs_cnt == 0)
>                 goto out;
>
>         root = &env->bpf_progs.btfs;
> --
> 2.47.3
>

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file
  2025-11-07 17:31 [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file Thomas Falcon
  2025-11-11 22:07 ` Ian Rogers
@ 2025-11-11 22:38 ` Namhyung Kim
  2025-11-13 12:52   ` Arnaldo Carvalho de Melo
  1 sibling, 1 reply; 5+ messages in thread
From: Namhyung Kim @ 2025-11-11 22:38 UTC (permalink / raw)
  To: Thomas Falcon, Arnaldo Carvalho de Melo
  Cc: Peter Zijlstra, Ingo Molnar, Mark Rutland, Alexander Shishkin,
	Jiri Olsa, Ian Rogers, Adrian Hunter, linux-perf-users,
	linux-kernel

On Fri, Nov 07, 2025 at 11:31:50AM -0600, Thomas Falcon wrote:
> With commit f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF
> info"), the write_bpf_( prog_info() | btf() ) functions exit
> without writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.
> 
> process_bpf_( prog_info() | btf() ), however, still expect a "count"
> value to exist in the data file. If btf information is empty, for
> example, process_bpf_btf will read garbage or some other data as the
> number of btf nodes in the data file. As a result, the data file will
> not be processed correctly.
> 
> Instead, write the count to the data file and exit if it is zero.

Oh, I'm afraid it'd create a compatibility problem.  But I think this is
a right behavior and it's should be fine if it goes to the stable soon.

Arnaldo, can you please take this into perf-tools tree for v6.18?

Thanks,
Namhyung

> 
> Fixes: f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF info")
> Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
> ---
>  tools/perf/util/header.c | 10 ++--------
>  1 file changed, 2 insertions(+), 8 deletions(-)
> 
> diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
> index db2ad19fa50d..54968881481c 100644
> --- a/tools/perf/util/header.c
> +++ b/tools/perf/util/header.c
> @@ -1022,12 +1022,9 @@ static int write_bpf_prog_info(struct feat_fd *ff,
>  
>  	down_read(&env->bpf_progs.lock);
>  
> -	if (env->bpf_progs.infos_cnt == 0)
> -		goto out;
> -
>  	ret = do_write(ff, &env->bpf_progs.infos_cnt,
>  		       sizeof(env->bpf_progs.infos_cnt));
> -	if (ret < 0)
> +	if (ret < 0 || env->bpf_progs.infos_cnt == 0)
>  		goto out;
>  
>  	root = &env->bpf_progs.infos;
> @@ -1067,13 +1064,10 @@ static int write_bpf_btf(struct feat_fd *ff,
>  
>  	down_read(&env->bpf_progs.lock);
>  
> -	if (env->bpf_progs.btfs_cnt == 0)
> -		goto out;
> -
>  	ret = do_write(ff, &env->bpf_progs.btfs_cnt,
>  		       sizeof(env->bpf_progs.btfs_cnt));
>  
> -	if (ret < 0)
> +	if (ret < 0 || env->bpf_progs.btfs_cnt == 0)
>  		goto out;
>  
>  	root = &env->bpf_progs.btfs;
> -- 
> 2.47.3
> 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file
  2025-11-11 22:38 ` Namhyung Kim
@ 2025-11-13 12:52   ` Arnaldo Carvalho de Melo
  2025-11-14  7:21     ` Namhyung Kim
  0 siblings, 1 reply; 5+ messages in thread
From: Arnaldo Carvalho de Melo @ 2025-11-13 12:52 UTC (permalink / raw)
  To: Namhyung Kim
  Cc: Thomas Falcon, Peter Zijlstra, Ingo Molnar, Mark Rutland,
	Alexander Shishkin, Jiri Olsa, Ian Rogers, Adrian Hunter,
	linux-perf-users, linux-kernel

On Tue, Nov 11, 2025 at 02:38:07PM -0800, Namhyung Kim wrote:
> On Fri, Nov 07, 2025 at 11:31:50AM -0600, Thomas Falcon wrote:
> > With commit f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF
> > info"), the write_bpf_( prog_info() | btf() ) functions exit
> > without writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.
> > 
> > process_bpf_( prog_info() | btf() ), however, still expect a "count"
> > value to exist in the data file. If btf information is empty, for
> > example, process_bpf_btf will read garbage or some other data as the
> > number of btf nodes in the data file. As a result, the data file will
> > not be processed correctly.
> > 
> > Instead, write the count to the data file and exit if it is zero.
> 
> Oh, I'm afraid it'd create a compatibility problem.  But I think this is
> a right behavior and it's should be fine if it goes to the stable soon.
> 
> Arnaldo, can you please take this into perf-tools tree for v6.18?

Ok, agreed, I'm taking your request as an Acked-by, ok?

- Arnaldo
 
> Thanks,
> Namhyung
> 
> > 
> > Fixes: f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF info")
> > Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
> > ---
> >  tools/perf/util/header.c | 10 ++--------
> >  1 file changed, 2 insertions(+), 8 deletions(-)
> > 
> > diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
> > index db2ad19fa50d..54968881481c 100644
> > --- a/tools/perf/util/header.c
> > +++ b/tools/perf/util/header.c
> > @@ -1022,12 +1022,9 @@ static int write_bpf_prog_info(struct feat_fd *ff,
> >  
> >  	down_read(&env->bpf_progs.lock);
> >  
> > -	if (env->bpf_progs.infos_cnt == 0)
> > -		goto out;
> > -
> >  	ret = do_write(ff, &env->bpf_progs.infos_cnt,
> >  		       sizeof(env->bpf_progs.infos_cnt));
> > -	if (ret < 0)
> > +	if (ret < 0 || env->bpf_progs.infos_cnt == 0)
> >  		goto out;
> >  
> >  	root = &env->bpf_progs.infos;
> > @@ -1067,13 +1064,10 @@ static int write_bpf_btf(struct feat_fd *ff,
> >  
> >  	down_read(&env->bpf_progs.lock);
> >  
> > -	if (env->bpf_progs.btfs_cnt == 0)
> > -		goto out;
> > -
> >  	ret = do_write(ff, &env->bpf_progs.btfs_cnt,
> >  		       sizeof(env->bpf_progs.btfs_cnt));
> >  
> > -	if (ret < 0)
> > +	if (ret < 0 || env->bpf_progs.btfs_cnt == 0)
> >  		goto out;
> >  
> >  	root = &env->bpf_progs.btfs;
> > -- 
> > 2.47.3
> > 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file
  2025-11-13 12:52   ` Arnaldo Carvalho de Melo
@ 2025-11-14  7:21     ` Namhyung Kim
  0 siblings, 0 replies; 5+ messages in thread
From: Namhyung Kim @ 2025-11-14  7:21 UTC (permalink / raw)
  To: Arnaldo Carvalho de Melo
  Cc: Thomas Falcon, Peter Zijlstra, Ingo Molnar, Mark Rutland,
	Alexander Shishkin, Jiri Olsa, Ian Rogers, Adrian Hunter,
	linux-perf-users, linux-kernel

On Thu, Nov 13, 2025 at 09:52:40AM -0300, Arnaldo Carvalho de Melo wrote:
> On Tue, Nov 11, 2025 at 02:38:07PM -0800, Namhyung Kim wrote:
> > On Fri, Nov 07, 2025 at 11:31:50AM -0600, Thomas Falcon wrote:
> > > With commit f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF
> > > info"), the write_bpf_( prog_info() | btf() ) functions exit
> > > without writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.
> > > 
> > > process_bpf_( prog_info() | btf() ), however, still expect a "count"
> > > value to exist in the data file. If btf information is empty, for
> > > example, process_bpf_btf will read garbage or some other data as the
> > > number of btf nodes in the data file. As a result, the data file will
> > > not be processed correctly.
> > > 
> > > Instead, write the count to the data file and exit if it is zero.
> > 
> > Oh, I'm afraid it'd create a compatibility problem.  But I think this is
> > a right behavior and it's should be fine if it goes to the stable soon.
> > 
> > Arnaldo, can you please take this into perf-tools tree for v6.18?
> 
> Ok, agreed, I'm taking your request as an Acked-by, ok?

Sure, thanks!
Namhyung

> > 
> > > 
> > > Fixes: f0d0f978f3f58 ("perf header: Don't write empty BPF/BTF info")
> > > Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
> > > ---
> > >  tools/perf/util/header.c | 10 ++--------
> > >  1 file changed, 2 insertions(+), 8 deletions(-)
> > > 
> > > diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
> > > index db2ad19fa50d..54968881481c 100644
> > > --- a/tools/perf/util/header.c
> > > +++ b/tools/perf/util/header.c
> > > @@ -1022,12 +1022,9 @@ static int write_bpf_prog_info(struct feat_fd *ff,
> > >  
> > >  	down_read(&env->bpf_progs.lock);
> > >  
> > > -	if (env->bpf_progs.infos_cnt == 0)
> > > -		goto out;
> > > -
> > >  	ret = do_write(ff, &env->bpf_progs.infos_cnt,
> > >  		       sizeof(env->bpf_progs.infos_cnt));
> > > -	if (ret < 0)
> > > +	if (ret < 0 || env->bpf_progs.infos_cnt == 0)
> > >  		goto out;
> > >  
> > >  	root = &env->bpf_progs.infos;
> > > @@ -1067,13 +1064,10 @@ static int write_bpf_btf(struct feat_fd *ff,
> > >  
> > >  	down_read(&env->bpf_progs.lock);
> > >  
> > > -	if (env->bpf_progs.btfs_cnt == 0)
> > > -		goto out;
> > > -
> > >  	ret = do_write(ff, &env->bpf_progs.btfs_cnt,
> > >  		       sizeof(env->bpf_progs.btfs_cnt));
> > >  
> > > -	if (ret < 0)
> > > +	if (ret < 0 || env->bpf_progs.btfs_cnt == 0)
> > >  		goto out;
> > >  
> > >  	root = &env->bpf_progs.btfs;
> > > -- 
> > > 2.47.3
> > > 

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2025-11-14  7:21 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-11-07 17:31 [PATCH] perf: write bpf_prog (infos|btfs)_cnt to data file Thomas Falcon
2025-11-11 22:07 ` Ian Rogers
2025-11-11 22:38 ` Namhyung Kim
2025-11-13 12:52   ` Arnaldo Carvalho de Melo
2025-11-14  7:21     ` Namhyung Kim

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).