linux-kernel - Re: [GIT PULL] block bits

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] [day] [month] [year] [list]

Date:	Wed, 27 Aug 2008 09:19:22 +0200
From:	Jens Axboe <jens.axboe@...cle.com>
To:	Jeremy Fitzhardinge <jeremy@...p.org>
Cc:	Linus Torvalds <torvalds@...ux-foundation.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [GIT PULL] block bits

On Tue, Aug 26 2008, Jeremy Fitzhardinge wrote:
> Jens Axboe wrote:
> > On Tue, Aug 26 2008, Linus Torvalds wrote:
> >   
> >> On Tue, 26 Aug 2008, Jens Axboe wrote:
> >>     
> >>> I'll move those 1-2 patches to the 2.6.28 branch and resubmit the rest
> >>> for you.
> >>>       
> >> You will *delete* the absolute crap locking patch, that has no way of 
> >> being even 2.6.28 material.
> >>     
> >
> > Yeah, I thought that was obvious, I'll bounce it back to the Xen folks.
> >   
> 
> Which, what?  Was there an objection to one of the blkfront patches?  (I
> don't remember anything locking related in there anyway.)

I see Linus' initial reply didn't make it to the list, not sure why.
This is what he said:

"it's so _incredibly_ broken that I refuse to pull any of the rest
either, because the queue is obviously utter crap.

In fact, even the "explanation" for that one is shit:

   "It shouldn't matter if an interrupt comes in whilst blkif_io_lock is
    held, as it will block on the lock [...]"

where "as it will block" is apparently shorthand for "as it will
deadlock and kill the machine"."

This is the patch in question:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commitdiff;h=b58e9034c5e2424723948a63712eaf1332a76ab8;hp=5814f8d589a1daef3b1a6f3731fd65a858dc8466

and it doesn indeed look like crap. It DOES matter if an interrupt comes
in on the local CPU while you are holding the blkif_io_lock, you'd be
dead on the spot spinning forever in the irq handler.

-- 
Jens Axboe

--
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/