[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1294038401.2016.5.camel@laptop>
Date: Mon, 03 Jan 2011 08:06:41 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: Paul Menage <menage@...gle.com>
Cc: Mike Galbraith <efault@....de>,
Miklos Vajna <vmiklos@...galware.org>,
shenghui <crosslonelyover@...il.com>,
kernel-janitors@...r.kernel.org, linux-kernel@...r.kernel.org,
mingo@...e.hu, Greg KH <greg@...ah.com>,
Paul Turner <pjt@...gle.com>,
Yong Zhang <yong.zhang0@...il.com>,
Li Zefan <lizf@...fujitsu.com>,
Balbir Singh <balbir@...ux.vnet.ibm.com>,
Srivatsa Vaddagiri <vatsa@...ibm.com>
Subject: Re: [PATCH] sched, cgroup: Use exit hook to avoid use-after-free
crash
On Sat, 2010-12-25 at 20:59 +0000, Paul Menage wrote:
> While this patch is likely fine for solving the problem, it does add
> extra work into the task exit path.
>
> Could you instead just use the pre_destroy callback to return -EBUSY
> if there are still any tasks on the cfs_rq? That way there'd only be a
> penalty on cgroup destruction, which is a much rarer operation.
No, that's broken! That would mean userspace gets the cgroup-empty
notification thing and then encounters -EBUSY, which then forces it to
go poll the state, rendering the whole notification thing pointless.
--
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/
Powered by blists - more mailing lists