From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PATCH][RFC] network splice receive Date: Tue, 5 Jun 2007 13:45:46 +0200 Message-ID: <20070605114545.GC9909@kernel.dk> References: <20070605080542.GA9909@kernel.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii To: netdev@vger.kernel.org Return-path: Received: from brick.kernel.dk ([80.160.20.94]:22264 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752786AbXFELrh (ORCPT ); Tue, 5 Jun 2007 07:47:37 -0400 Received: from nelson.home.kernel.dk (nelson.home.kernel.dk [192.168.0.33]) by kernel.dk (Postfix) with ESMTP id 0E04A2571F0 for ; Tue, 5 Jun 2007 13:47:35 +0200 (CEST) Content-Disposition: inline In-Reply-To: <20070605080542.GA9909@kernel.dk> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Tue, Jun 05 2007, Jens Axboe wrote: > Seems to work reasonably well for me, sometimes I do see small ranges > inside the output file that are not correct, but I haven't been able to > reproduce today. I think it has to do with page reuse, hence the > NET_COPY_SPLICE ifdef that you can enable to just plain copy the data > instead of referencing it. I managed to reproduce. It's segments of 68-80 bytes beyond corrupt in the middle of the out, and there might be 1-3 of such occurences in the 30mb file I tested with. The first 16 bytes of the corruption are always the same: 0000 1800 4ff3 937f e000 6381 7275 0008 Perhaps that hex pattern rings a bell with someone intimate with the networking. The remaining wrong bytes don't seem to have anything in common. Slab poisoning doesn't change the pattern, so it's not use-after-free. -- Jens Axboe