All of lore.kernel.org
 help / color / mirror / Atom feed
From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Jiri Olsa <jolsa@kernel.org>
Cc: lkml <linux-kernel@vger.kernel.org>,
	Ingo Molnar <mingo@kernel.org>,
	Namhyung Kim <namhyung@kernel.org>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Michael Petlan <mpetlan@redhat.com>
Subject: Re: [PATCH] perf tools: Fix tracing_path_mount proper path
Date: Tue, 16 Oct 2018 12:23:22 -0300	[thread overview]
Message-ID: <20181016152322.GC3849@kernel.org> (raw)
In-Reply-To: <20181016151813.GB3849@kernel.org>

Em Tue, Oct 16, 2018 at 12:18:13PM -0300, Arnaldo Carvalho de Melo escreveu:
> Em Tue, Oct 16, 2018 at 01:48:18PM +0200, Jiri Olsa escreveu:
> > If there's no tracefs (RHEL7) support the tracing_path_mount
> > returns debugfs path which results in following fail:
> > 
> >   # perf probe sys_write
> >   kprobe_events file does not exist - please rebuild kernel with CONFIG_KPROBE_EVENTS.
> >   Error: Failed to add events.
> > 
> > In tracing_path_debugfs_mount function we need to return the
> > 'tracing' path instead of just the mount to make it work:
> > 
> >   # perf probe sys_write
> >   Added new event:
> >     probe:sys_write      (on sys_write)
> > 
> >   You can now use it in all perf tools, such as:
> > 
> >           perf record -e probe:sys_write -aR sleep 1
> > 
> > Adding the 'return tracing_path;' also to tracing_path_tracefs_mount
> > function just for consistency with tracing_path_debugfs_mount.
> > 
> > Upstream keeps working, because it has the tracefs support.
> > 
> > Link: http://lkml.kernel.org/n/tip-yiwkzexq9fk1ey1xg3gnjlw4@git.kernel.org
> > Signed-off-by: Jiri Olsa <jolsa@kernel.org>
> > ---
> >  tools/lib/api/fs/tracing_path.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/tools/lib/api/fs/tracing_path.c b/tools/lib/api/fs/tracing_path.c
> > index 120037496f77..5afb11b30fca 100644
> > --- a/tools/lib/api/fs/tracing_path.c
> > +++ b/tools/lib/api/fs/tracing_path.c
> > @@ -36,7 +36,7 @@ static const char *tracing_path_tracefs_mount(void)
> >  
> >  	__tracing_path_set("", mnt);
> >  
> > -	return mnt;
> > +	return tracing_path;
> >  }
> >  
> >  static const char *tracing_path_debugfs_mount(void)
> > @@ -49,7 +49,7 @@ static const char *tracing_path_debugfs_mount(void)
> >  
> >  	__tracing_path_set("tracing/", mnt);
> >  
> > -	return mnt;
> > +	return tracing_path;
> 
> Humm, colour me confused, by the name of the function I expected
> tracing_path_debugfs_mount() to return the debugfs mount.
> 
> tracing_path_tracefs_mount() would return the tracefs mount point, so,
> there, returning tracing_path would be ok.
> 
> Is there some other way to figure out that tracefs isn't available and
> thus we should use debugfs_mount + "/tracing/" for the "tracefs" mount?
> 
> This may well be just the naming confusion... lemme read this a bit
> more...

Indeed, naming confusion, these are static things, used only by:

const char *tracing_path_mount(void)
{
        const char *mnt;

        mnt = tracing_path_tracefs_mount();
        if (mnt)
                return mnt;

        mnt = tracing_path_debugfs_mount();

        return mnt;
}

This one could as well just return tracing_path and the return be used
just to figure if it managed to read the mount point for tracefs or
debugfs...

Something like:

static int tracing_path_tracefs_mount(void)
{
	const char *mnt = tracefs__mount();
	if (!mnt)
		return -1;

	__tracing_path_set("", mnt);
	return 0;
}

static const char *tracing_path_debugfs_mount(void)
{
	const char *mnt = debugfs__mount();
	if (!mnt)
		return -1;

	__tracing_path_set("tracing/", mnt);
	return 0;
}

const char *tracing_path_mount(void)
{
	if (tracing_path_tracefs_mount() < 0 &&
	    tracing_path_debugfs_mount() < 0)
		return NULL;

	return tracing_path;
}

But anyway, your patch fixes things, applying...

- Arnaldo

  reply	other threads:[~2018-10-16 15:23 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-10-16 11:48 [PATCH] perf tools: Fix tracing_path_mount proper path Jiri Olsa
2018-10-16 15:18 ` Arnaldo Carvalho de Melo
2018-10-16 15:23   ` Arnaldo Carvalho de Melo [this message]
2018-10-18  6:19 ` [tip:perf/urgent] " tip-bot for Jiri Olsa

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20181016152322.GC3849@kernel.org \
    --to=acme@kernel.org \
    --cc=a.p.zijlstra@chello.nl \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=mpetlan@redhat.com \
    --cc=namhyung@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.