From mboxrd@z Thu Jan 1 00:00:00 1970 From: Steve Wise Subject: Re: [PATCH] RDMA/cxgb4: don't declare wr_wait objects on the stack. Date: Mon, 16 May 2011 10:56:45 -0500 Message-ID: <4DD1493D.3060401@opengridcomputing.com> References: <20110513183727.32157.8873.stgit@build.ogc.int> <4DD14036.9050704@opengridcomputing.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Roland Dreier Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: linux-rdma@vger.kernel.org On 05/16/2011 10:53 AM, Roland Dreier wrote: > On Mon, May 16, 2011 at 8:18 AM, Steve Wise wrote: >> Roland, I need to recall this patch. It appears to have some problem. > Good thing I didn't apply it yet. > > I did think it should be possible to declare wait objects on the stack... the > origin of completions was to handle exactly that issue; the older semaphore > structure was vulnerable to the wakeup after free, but completions handle > that. But I didn't look at the cxgb4 code at all... I'm pretty sure the race exists as described in my commit comment. And I don't see how the wait object code could avoid this issue. Unless I'm just all wrong. :) -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html