linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Milian Wolff <mail@milianw.de>
Cc: linux-perf-users@vger.kernel.org
Subject: Re: perf: remount of tracefs required
Date: Wed, 5 Aug 2015 11:35:40 -0300	[thread overview]
Message-ID: <20150805143540.GA32553@kernel.org> (raw)
In-Reply-To: <3948762.FVKqmXobpH@milian-kdab2>

Em Wed, Aug 05, 2015 at 04:17:42PM +0200, Milian Wolff escreveu:
> On Tuesday 04 August 2015 12:27:46 Arnaldo Carvalho de Melo wrote:
> >   [acme@zoo linux]$ trace usleep 1
> >   Error:	No permissions to read
> >   /sys/kernel/debug/tracing/events/raw_syscalls/sys_(enter|exit)
> >   Hint:	Try 'sudo mount -o remount,mode=755 /sys/kernel/debug/tracing'
 
> This message I do not get, not even with my perf compiled from the v4.2-rc5 
> tag. And it is exactly what I was asking for.
 
> <snip>
 
> > I guess what fixes what you report is this cset:

> >   commit ab85785aa13c36440a91a8e9f7616357de411a1f

> >     tools lib api debugfs: Check for tracefs when reporting errors

> > [acme@zoo linux]$ git describe ab85785aa13c36440a91a8e9f7616357de411a1f
> > v4.2-rc1-17-gab85785aa13c

> > So, can you try running the perf tools, built from the latest 4.2-rc
> > codebase to see if it behaves as you expect?
 
> This is not included in 4.2-rc5 as far as I can see:
 
> $ git describe
> v4.2-rc5-2-ga64d3e5
> $ git show ab85785aa13c36440a91a8e9f7616357de411a1f
> fatal: bad object ab85785aa13c36440a91a8e9f7616357de411a1f
> $ git log --grep debugfs --oneline --author=Arnaldo
> f816b3c tools lib fs debugfs: Check if debugfs is mounted when handling ENOENT
> ...

Oops, it is just in tip/perf/core, i.e. scheduled for 4.3 :-\ Wonder how
difficult would be to cherry-pick that into perf/urgent.

I.e. can you try doing something like:

  git cherry-pick ab85785aa13c36440a91a8e9f7616357de411a1f

With a checked out v4.2-rc5, or better, torvalds/master, i.e. the top of
Linus's git tree? It is possible that one or more patches just before
this one, also in perf/core, may be needed.

If it is easy, perhaps we can push it for 4.2 final.

Ah, you may want to take a look at:

https://git.kernel.org/cgit/linux/kernel/git/acme/linux.git/log/?h=perf/core

To get from this:

  [root@zoo ~]# trace -e newfstatat,unlinkat rm /tmp/foo
     0.489 ( 0.011 ms): rm/8428 newfstatat(dfd: CWD, filename: 0x9153f8, statbuf: 0x915368, flag: 256 ) = 0
     0.508 ( 0.010 ms): rm/8428 unlinkat(dfd: CWD, pathname: 0x9140c0) = 0
  [root@zoo ~]#

After doing this:

  [root@zoo ~]# perf probe 'vfs_getname=getname_flags:72 pathname=filename:string'
  Added new event:
    probe:vfs_getname    (on getname_flags:72 with pathname=filename:string)

  You can now use it in all perf tools, such as:

	perf record -e probe:vfs_getname -aR sleep 1

To this:

  [root@zoo ~]# trace -e newfstatat,unlinkat rm /tmp/foo
  rm: cannot remove ‘/tmp/foo’: No such file or directory
     0.503 ( 0.007 ms): rm/8480 newfstatat(dfd: CWD, filename: /tmp/foo, statbuf: 0x2370368, flag: 256) = -1 ENOENT No such file or directory
     0.514 ( 0.004 ms): rm/8480 unlinkat(dfd: CWD, pathname: /tmp/foo) = -1 ENOENT No such file or directory
  [root@zoo ~]#

Auto inserting that probe remains to be done, but for now can be done just once
per boot.

- Arnaldo

      reply	other threads:[~2015-08-05 14:35 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-08-04 14:44 perf: remount of tracefs required Milian Wolff
2015-08-04 15:27 ` Arnaldo Carvalho de Melo
2015-08-05 14:17   ` Milian Wolff
2015-08-05 14:35     ` Arnaldo Carvalho de Melo [this message]

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=20150805143540.GA32553@kernel.org \
    --to=acme@kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mail@milianw.de \
    /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).