public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Rusty Russell <rusty@rustcorp.com.au>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>,
	LKML <linux-kernel@vger.kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Brandon Philips <brandon@ifup.org>,
	Jon Masters <jonathan@jonmasters.org>,
	Tejun Heo <htejun@gmail.com>,
	Masami Hiramatsu <mhiramat@redhat.com>
Subject: Re: [Regression] Crash in load_module() while freeing args
Date: Wed, 26 May 2010 21:27:24 +0930	[thread overview]
Message-ID: <201005262127.26235.rusty@rustcorp.com.au> (raw)
In-Reply-To: <201005261730.59058.rusty@rustcorp.com.au>

On Wed, 26 May 2010 05:30:58 pm Rusty Russell wrote:
> On Wed, 26 May 2010 09:17:32 am Linus Torvalds wrote:
> > 
> > On Wed, 26 May 2010, Rafael J. Wysocki wrote:
> > >
> > > I'm not able to reproduce the issue with the following commit reverted:
> > > 
> > > commit 480b02df3aa9f07d1c7df0cd8be7a5ca73893455
> > > Author: Rusty Russell <rusty@rustcorp.com.au>
> > > Date:   Wed May 19 17:33:39 2010 -0600
> > > 
> > >     module: drop the lock while waiting for module to complete initialization.
> > 
> > Hmm. That does seem to be buggy. We can't just drop and re-take the lock: 
> > that may make sense _internally_ as far as resolve_symbol() itself is 
> > concerned, but the caller will its own local variables, and some of those 
> > will no longer be valid if the lock was dropped. 
> 
> Well, yes, obviously I missed something :(  I'll look at it tonight after
> Arabella is asleep.

See if you can spot it (I acked the patch, so I can't point fingers):

 free_core:
	module_free(mod, mod->module_core);
	/* mod will be freed with core. Don't access it beyond this line! */
 free_percpu:
	percpu_modfree(mod);

Only a year after Masami fixed that and added the comment, too :(

I suspect that the increased parallelism enabled by this patch uncovered this
bug.  Does this fix it?

(Side note: the locking should be simplified.  No code before simplify_symbols
actually needs the lock, so we should grab it just for that, then again at the
end.  We use kobjects to protect us from multiple loads as a side-effect, but
we should move that registration to the end).

Subject: module: fix reference to mod->percpu after freeing module.

The comment about the mod being freed is self-explanatory, but neither
Tejun nor I read it.  This bug was introduced in 259354deaa, after it
had previously been fixed in 6e2b75740b.  How embarrassing.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Tejun Heo <tj@kernel.org>
Cc: Masami Hiramatsu <mhiramat@redhat.com>

diff --git a/kernel/module.c b/kernel/module.c
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -2031,6 +2031,7 @@ static noinline struct module *load_modu
 	long err = 0;
 	void *ptr = NULL; /* Stops spurious gcc warning */
 	unsigned long symoffs, stroffs, *strmap;
+	void __percpu *percpu;
 
 	mm_segment_t old_fs;
 
@@ -2175,6 +2176,8 @@ static noinline struct module *load_modu
 			goto free_mod;
 		sechdrs[pcpuindex].sh_flags &= ~(unsigned long)SHF_ALLOC;
 	}
+	/* Keep this around for failure path. */
+	percpu = mod_percpu(mod);
 
 	/* Determine total sizes, and put offsets in sh_entsize.  For now
 	   this is done generically; there doesn't appear to be any
@@ -2480,7 +2483,7 @@ static noinline struct module *load_modu
 	module_free(mod, mod->module_core);
 	/* mod will be freed with core. Don't access it beyond this line! */
  free_percpu:
-	percpu_modfree(mod);
+	free_percpu(percpu);
  free_mod:
 	kfree(args);
 	kfree(strmap);

  reply	other threads:[~2010-05-26 11:57 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-05-25 21:00 [Regression] Crash in load_module() while freeing args Rafael J. Wysocki
2010-05-25 22:54 ` Rafael J. Wysocki
2010-05-25 23:47   ` Linus Torvalds
2010-05-26  8:00     ` Rusty Russell
2010-05-26 11:57       ` Rusty Russell [this message]
2010-05-26 22:56         ` Rafael J. Wysocki
2010-05-26 23:07           ` Linus Torvalds
2010-05-27  5:26           ` Rusty Russell
2010-05-27 18:46             ` Brandon Philips
2010-05-31  9:40               ` Rusty Russell
2010-05-31 12:00                 ` [PATCH 0/2] kernel/module.c locking changes Rusty Russell
2010-05-31 12:01                   ` [PATCH 1/2] module: make locking more fine-grained Rusty Russell
2010-05-31 12:02                     ` [PATCH 2/2] module: fix bne2 "gave up waiting for init of module libcrc32c" Rusty Russell
2010-05-31 16:48                       ` Andrew Morton
2010-05-31 18:19                         ` Linus Torvalds
2010-05-31 20:15                           ` Linus Torvalds
2010-05-31 20:16                             ` [PATCH 1/2] Make the module 'usage' lists be two-way Linus Torvalds
2010-05-31 20:17                               ` [PATCH 2/2] module: wait for other modules after dropping the module_mutex Linus Torvalds
2010-06-01  1:37                               ` [PATCH 1/2] Make the module 'usage' lists be two-way Rusty Russell
2010-06-01  3:42                                 ` Rusty Russell
2010-06-01  4:00                                   ` Linus Torvalds
2010-06-01  4:05                                     ` Linus Torvalds
2010-06-01  2:44                               ` Américo Wang
2010-06-01  3:51                                 ` Linus Torvalds
2010-06-01  1:57                             ` [PATCH 2/2] module: fix bne2 "gave up waiting for init of module libcrc32c" Rusty Russell
2010-06-01  3:40                               ` Linus Torvalds
2010-06-01  4:27                                 ` Linus Torvalds
2010-06-01  5:19                                 ` Rusty Russell
2010-06-02  3:15                                   ` Rusty Russell
2010-06-01  1:21                           ` Rusty Russell
2010-06-01  3:24                             ` Linus Torvalds
2010-06-01  5:22                               ` Rusty Russell
2010-06-01 14:58                                 ` Linus Torvalds
2010-06-01 17:53                                   ` Linus Torvalds
2010-06-01 23:24                                     ` Brandon Philips
2010-06-01 23:51                                       ` Linus Torvalds
2010-06-02  2:10                                         ` Brandon Philips
2010-06-02  3:03                                           ` Rusty Russell
2010-06-02  4:35                                           ` Linus Torvalds
2010-06-02  4:44                                             ` Linus Torvalds
2010-06-02  6:35                                               ` Rusty Russell
2010-06-02  7:45                                                 ` Linus Torvalds
2010-06-02  8:12                                                   ` Linus Torvalds
2010-06-02  9:07                                                     ` Rusty Russell
2010-06-02  5:52                                             ` Rusty Russell
2010-06-02  7:21                                               ` Linus Torvalds
2010-06-02 14:06                                                 ` Rusty Russell
2010-06-02 14:50                                                   ` Linus Torvalds
2010-06-03 13:06                                                     ` Rusty Russell
2010-06-02 16:53                                                   ` Brandon Philips
2010-06-02 18:01                                                   ` Linus Torvalds
2010-06-03  5:20                                                     ` Rusty Russell
2010-06-03 16:24                                                       ` Linus Torvalds
2010-06-04  1:02                                                         ` Rusty Russell
2010-06-04  1:55                                                           ` Linus Torvalds
2010-06-04  5:20                                                             ` Rusty Russell
2010-06-04 22:48                                                               ` Linus Torvalds
2010-06-05  1:49                                                                 ` Rusty Russell
2010-06-02  3:09                                   ` Rusty Russell
2010-06-02  4:32                                     ` Linus Torvalds
2010-06-02  4:56                                     ` Linus Torvalds
2010-06-02  5:52                                       ` Rusty Russell
2010-06-02  6:59                                         ` Linus Torvalds
2010-06-01  1:04                         ` Rusty Russell
2010-06-01  5:38                     ` [PATCH 1/2] module: make locking more fine-grained Américo Wang
2010-06-01  5:55                       ` Rusty Russell
2010-05-27 21:57             ` [Regression] Crash in load_module() while freeing args Rafael J. Wysocki
2010-05-31  7:54               ` Rusty Russell
2010-05-31 10:23               ` [PATCH] module: fix reference to mod->percpu after freeing module Rusty Russell
2010-05-31 10:25                 ` Tejun Heo
2010-05-26 15:41       ` [Regression] Crash in load_module() while freeing args Linus Torvalds

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=201005262127.26235.rusty@rustcorp.com.au \
    --to=rusty@rustcorp.com.au \
    --cc=akpm@linux-foundation.org \
    --cc=brandon@ifup.org \
    --cc=htejun@gmail.com \
    --cc=jonathan@jonmasters.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mhiramat@redhat.com \
    --cc=rjw@sisk.pl \
    --cc=torvalds@linux-foundation.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox