linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Masami Hiramatsu (Google) <mhiramat@kernel.org>
To: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>,
	Namhyung Kim <namhyung@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>, Ian Rogers <irogers@google.com>,
	Dima Kogan <dima@secretsauce.net>,
	Alexander Lobakin <aleksander.lobakin@intel.com>,
	Przemek Kitszel <przemyslaw.kitszel@intel.com>,
	linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/4] perf-probe: Require '@' prefix for filename always
Date: Tue, 5 Nov 2024 18:36:33 +0900	[thread overview]
Message-ID: <20241105183633.28048e908539d5c638ceb10a@kernel.org> (raw)
In-Reply-To: <20241105182830.384b70727ff34391eb0ef9eb@kernel.org>

On Tue, 5 Nov 2024 18:28:30 +0900
Masami Hiramatsu (Google) <mhiramat@kernel.org> wrote:

> On Mon, 4 Nov 2024 17:10:41 -0300
> Arnaldo Carvalho de Melo <acme@kernel.org> wrote:
> 
> > On Tue, Nov 05, 2024 at 01:17:26AM +0900, Masami Hiramatsu (Google) wrote:
> > > From: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> >  
> > > Currently perf probe allows user to specify probing place without '@'
> > > prefix, for example, both `-V file:line` and `-V function:line` are
> > > allowed. But this makes a problem if a (demangled) function name is
> > > hard to be distinguished from a file name.
> >  
> > > This changes the perf-probe to require '@' prefix for filename even
> > > without function name. For example, `-V @file:line` and
> > > `-V function:line` are acceptable.
> >  
> > > With this change, users can specify filename or function correctly.
> > 
> > Well, this will break scripts that use perf probe for a given file,
> > probably the right thing not to break backwards compatibility is to
> > continue accepting and when there is a real conflict, an ambiguity that
> > makes differentiating from file to function names, then refuse it,
> > stating that it is ambiguous, probably spelling out the names of the
> > files and functions that match so and stating that to make it
> > unambiguoius, prefix file names with @.
> 
> The problem is that the ambiguous function name. For example, Go's
> main routine is `main.main`, and this is not likely to the C function
> name, so currently perf probe treats it as a filename and failed to
> find that.
> 
> I think one possible solution is to run search loop twice internally
> (search it as file name, if fails, search it as function name) if it
> looks like a file name but it does not start from `@`.
> This takes costs a bit but can keep backward compatibility.

I found another good idea, support `@*` :) 

For example, if the `main.main` is not a file, we can

 perf probe -x execfile -L 'main.main@*'

Then it is clear that `main.main` is a function name.

Thank you!

> 
> Thank you,
> 
> > 
> > - Arnaldo
> >  
> > > Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> > > ---
> > >  tools/perf/util/probe-event.c |   31 +++++++++----------------------
> > >  1 file changed, 9 insertions(+), 22 deletions(-)
> > > 
> > > diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
> > > index 665dcce482e1..913a27cbb5d9 100644
> > > --- a/tools/perf/util/probe-event.c
> > > +++ b/tools/perf/util/probe-event.c
> > > @@ -1341,7 +1341,7 @@ static bool is_c_func_name(const char *name)
> > >   * Stuff 'lr' according to the line range described by 'arg'.
> > >   * The line range syntax is described by:
> > >   *
> > > - *         SRC[:SLN[+NUM|-ELN]]
> > > + *         @SRC[:SLN[+NUM|-ELN]]
> > >   *         FNC[@SRC][:SLN[+NUM|-ELN]]
> > >   */
> > >  int parse_line_range_desc(const char *arg, struct line_range *lr)
> > > @@ -1404,16 +1404,10 @@ int parse_line_range_desc(const char *arg, struct line_range *lr)
> > >  			err = -ENOMEM;
> > >  			goto err;
> > >  		}
> > > +		if (*name != '\0')
> > > +			lr->function = name;
> > > +	} else
> > >  		lr->function = name;
> > > -	} else if (strpbrk_esc(name, "/."))
> > > -		lr->file = name;
> > > -	else if (is_c_func_name(name))/* We reuse it for checking funcname */
> > > -		lr->function = name;
> > > -	else {	/* Invalid name */
> > > -		semantic_error("'%s' is not a valid function name.\n", name);
> > > -		err = -EINVAL;
> > > -		goto err;
> > > -	}
> > >  
> > >  	return 0;
> > >  err:
> > > @@ -1463,7 +1457,7 @@ static int parse_perf_probe_point(char *arg, struct perf_probe_event *pev)
> > >  
> > >  	/*
> > >  	 * <Syntax>
> > > -	 * perf probe [GRP:][EVENT=]SRC[:LN|;PTN]
> > > +	 * perf probe [GRP:][EVENT=]@SRC[:LN|;PTN]
> > >  	 * perf probe [GRP:][EVENT=]FUNC[@SRC][+OFFS|%return|:LN|;PAT]
> > >  	 * perf probe %[GRP:]SDT_EVENT
> > >  	 */
> > > @@ -1516,19 +1510,12 @@ static int parse_perf_probe_point(char *arg, struct perf_probe_event *pev)
> > >  	/*
> > >  	 * Check arg is function or file name and copy it.
> > >  	 *
> > > -	 * We consider arg to be a file spec if and only if it satisfies
> > > -	 * all of the below criteria::
> > > -	 * - it does not include any of "+@%",
> > > -	 * - it includes one of ":;", and
> > > -	 * - it has a period '.' in the name.
> > > -	 *
> > > +	 * We consider arg to be a file spec if it starts with '@'.
> > >  	 * Otherwise, we consider arg to be a function specification.
> > >  	 */
> > > -	if (!strpbrk_esc(arg, "+@%")) {
> > > -		ptr = strpbrk_esc(arg, ";:");
> > > -		/* This is a file spec if it includes a '.' before ; or : */
> > > -		if (ptr && memchr(arg, '.', ptr - arg))
> > > -			file_spec = true;
> > > +	if (*arg == '@') {
> > > +		file_spec = true;
> > > +		arg++;
> > >  	}
> > >  
> > >  	ptr = strpbrk_esc(arg, ";:+@%");
> 
> 
> -- 
> Masami Hiramatsu (Google) <mhiramat@kernel.org>


-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>

  reply	other threads:[~2024-11-05  9:36 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-11-04 16:17 [PATCH 0/4] perf-probe: Improbe non-C language support Masami Hiramatsu (Google)
2024-11-04 16:17 ` [PATCH 1/4] perf-probe: Fix to ignore escaped characters in --lines option Masami Hiramatsu (Google)
2024-11-04 16:17 ` [PATCH 2/4] perf-probe: Require '@' prefix for filename always Masami Hiramatsu (Google)
2024-11-04 20:10   ` Arnaldo Carvalho de Melo
2024-11-05  9:28     ` Masami Hiramatsu
2024-11-05  9:36       ` Masami Hiramatsu [this message]
2024-11-04 16:17 ` [PATCH 3/4] perf-probe: Introduce quotation marks support Masami Hiramatsu (Google)
2024-11-04 20:23   ` Arnaldo Carvalho de Melo
2024-11-05  9:29     ` Masami Hiramatsu
2024-11-04 16:17 ` [PATCH 4/4] perf-probe: Replace unacceptable characters when generating event name Masami Hiramatsu (Google)
2024-11-04 20:34   ` Arnaldo Carvalho de Melo
2024-11-05  8:27     ` Masami Hiramatsu
2024-11-04 18:56 ` [PATCH 0/4] perf-probe: Improbe non-C language support Arnaldo Carvalho de Melo
2024-11-04 20:08   ` Arnaldo Carvalho de Melo
2024-11-05  9:38     ` Masami Hiramatsu
2024-11-05  9:48   ` Masami Hiramatsu

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=20241105183633.28048e908539d5c638ceb10a@kernel.org \
    --to=mhiramat@kernel.org \
    --cc=acme@kernel.org \
    --cc=aleksander.lobakin@intel.com \
    --cc=dima@secretsauce.net \
    --cc=irogers@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    --cc=przemyslaw.kitszel@intel.com \
    /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 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).