linux-kernel - Re: [GIT PULL] percpu fixes for 2.6.32-rc6

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: <alpine.LFD.2.01.0911101338190.31845@localhost.localdomain>
Date:	Tue, 10 Nov 2009 13:42:51 -0800 (PST)
From:	Linus Torvalds <torvalds@...ux-foundation.org>
To:	Tejun Heo <tj@...nel.org>
cc:	Ingo Molnar <mingo@...e.hu>,
	Linux Kernel <linux-kernel@...r.kernel.org>,
	Yinghai Lu <yhlu.kernel@...il.com>
Subject: Re: [GIT PULL] percpu fixes for 2.6.32-rc6

On Wed, 11 Nov 2009, Tejun Heo wrote:
> 
> If you're talking about the three way return value, which I do agree
> to be quite ugly, I think it will be a lot safer to have three patches
> - one to fix the deadlock, another to fix the return value and the
> final one to de-uglify the function, especially as we're pretty late
> in the release cycle.

I'm certainly ok with doing it in stages if that is how you want to do it.

That said, I'm not entirely sure it's _worthwhile_, since the "return 1" 
case has apparently never ever actually worked. From a bisect standpoint, 
what's the difference between seeing

 - oh, now that we made it return the documented code and actually re-try 
   properly when dropping the lock, it turns out that the re-try code was 
   always buggy and we just hadn't noticed before because it didn't 
   trigger

or

 - oh, now that we rewrote the function to be cleaner and do the lock 
   dropping and retry more obviously, it turns out that the retry doesn't 
   actually work and leads to deadlocks.

but I don't care deeply.

I want the cleanup because I think that the code sucks from a "future 
proofing" and readability standpoint, but I really don't mind one way or 
the other whether you want to finally do that one "return 1" correctly for 
one commit, only to then fix it to not do that three-way test of a single 
function in the next one.

So whatever works - as long as the end result both looks sane and doesn't 
have the bug we clearly have now.

		Linus
--
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/