linux-kernel - Re: [PATCH] md: do not use ++ in rcu

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20100905203908.GA3228@merkur.ravnborg.org>
Date:	Sun, 5 Sep 2010 22:39:08 +0200
From:	Sam Ravnborg <sam@...nborg.org>
To:	Kulikov Vasiliy <segooon@...il.com>
Cc:	kernel-janitors@...r.kernel.org, Neil Brown <neilb@...e.de>,
	Jens Axboe <jaxboe@...ionio.com>, linux-raid@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] md: do not use ++ in rcu_dereference() argument

On Sun, Sep 05, 2010 at 11:23:35PM +0400, Kulikov Vasiliy wrote:
> On Sun, Sep 05, 2010 at 21:01 +0200, Sam Ravnborg wrote:
> > On Sun, Sep 05, 2010 at 10:32:18PM +0400, Kulikov Vasiliy wrote:
> > > From: Vasiliy Kulikov <segooon@...il.com>
> > > 
> > > rcu_dereference() is macro, so it might use its argument twice.
> > > Argument must not has side effects.
> > > 
> > > It was found by compiler warning:
> > > drivers/md/raid1.c: In function ‘read_balance’:
> > > drivers/md/raid1.c:445: warning: operation on ‘new_disk’ may be undefined
> > 
> > This change looks wrong.
> > In the original implementation new_disk is incremented and
> > then we do the array lookup.
> > With your implementation it looks like we increment it after
> > the array lookup.
> 
> No, the original code increments new_disk and then dereferences mirrors.
> 
> The full code:
> 
> 		for (rdev = rcu_dereference(conf->mirrors[new_disk].rdev);
> 		     r1_bio->bios[new_disk] == IO_BLOCKED ||
> 		     !rdev || !test_bit(In_sync, &rdev->flags)
> 			     || test_bit(WriteMostly, &rdev->flags);
> 		     rdev = rcu_dereference(conf->mirrors[++new_disk].rdev)) {
> 
> 			if (rdev && test_bit(In_sync, &rdev->flags) &&
> 				r1_bio->bios[new_disk] != IO_BLOCKED)
> 				wonly_disk = new_disk;
> 
> 			if (new_disk == conf->raid_disks - 1) {
> 				new_disk = wonly_disk;
> 				break;
> 			}
> 		}
> 
>     so,
> 
>     for (a; b; c = f(++g)) {
>        ...
>     } 

Thanks - that explains it.
This code really screams for a helper function but thats another matter.

	Sam
--
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/