linux-kernel - Re: [PATCH] perf_event: mount debugfs automatically

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20091228074426.GA20039@elte.hu>
Date:	Mon, 28 Dec 2009 08:44:26 +0100
From:	Ingo Molnar <mingo@...e.hu>
To:	Xiao Guangrong <xiaoguangrong@...fujitsu.com>
Cc:	Peter Zijlstra <peterz@...radead.org>,
	Paul Mackerras <paulus@...ba.org>,
	Frederic Weisbecker <fweisbec@...il.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Clark Williams <williams@...hat.com>,
	John Kacur <jkacur@...hat.com>
Subject: Re: [PATCH] perf_event: mount debugfs automatically

* Xiao Guangrong <xiaoguangrong@...fujitsu.com> wrote:

> Mount debugfs automatically if it's not mounted, umount it
> when programme exit
> 
> Signed-off-by: Xiao Guangrong <xiaoguangrong@...fujitsu.com>
> ---
>  tools/perf/perf.c         |   10 +++++++++-
>  tools/perf/util/debugfs.c |   16 +++++++---------
>  tools/perf/util/debugfs.h |    2 +-
>  3 files changed, 17 insertions(+), 11 deletions(-)

I'm not sure that's a good idea. What happens if two perf tools are running in 
parallel:

  perf report #1 start
                          perf report #2 start
  perf report #2 exit [umount debugfs]
                          perf report #2 tries to open /debug file: kaboom
                          perf report #2 exit

But your idea is sound if we only do the first half: we should mount it under 
/sys/kernel/debug/ if it's not mounted already [and that directory is 
available], and leave it mounted there.

Furthermore please not that we must not mount it under any other path - 
mounting is always a dangerous operation because it changes the VFS namespace. 
For example some system might have some local files under /debug for whatever 
reason, we must not over-mount it and potentially destroy data because we'd 
confuse an app that writes into /debug. /sys/kernel/debug/ is a fair game to 
mount into OTOH.

(and even then we should just leave it mounted and never umount it.)

Thanks,

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/