From mboxrd@z Thu Jan 1 00:00:00 1970 From: Krzysztof Halasa Subject: Re: Strange network timeouts w/ 2.6.30.5 Date: Fri, 21 Aug 2009 11:44:35 +0200 Message-ID: References: <322550336.131250806883202.JavaMail.root@mail.holmansrus.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , linux-kernel@vger.kernel.org, netdev@vger.kernel.org To: Walt Holman Return-path: Received: from khc.piap.pl ([195.187.100.11]:55736 "EHLO khc.piap.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755081AbZHUJog (ORCPT ); Fri, 21 Aug 2009 05:44:36 -0400 In-Reply-To: <322550336.131250806883202.JavaMail.root@mail.holmansrus.com> (Walt Holman's message of "Thu\, 20 Aug 2009 17\:21\:23 -0500 \(CDT\)") Sender: netdev-owner@vger.kernel.org List-ID: Walt Holman writes: > Would something like passing a mem=xx cmdline on x86_64 be sufficient to test this? I think so, though I admit I don't remember using this personally since the introduction of e820 RAM mapping support. Dmesg will show if the memory is limited. For the swiotlb to effectively disable no RAM > 0x100000000 may be used. But this test isn't IMHO terribly important at this point - the driver makes invalid use of the DMA API and it has to change. The test could only explain _how_ exactly does it fail, we already know _why_ it does. -- Krzysztof Halasa