From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH 1/1] 3c59x: Acquire vortex lock instead of disabling irq Date: Wed, 17 Mar 2010 11:47:17 -0700 (PDT) Message-ID: <20100317.114717.30960972.davem@davemloft.net> References: <40ec3ea41003171120x26d2e8b4q6b59c992ab28e454@mail.gmail.com> <20100317.112706.143765408.davem@davemloft.net> <40ec3ea41003171144s6acbf2fdl4434f2147b8244a0@mail.gmail.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org To: chase.douglas@canonical.com Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:52423 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754132Ab0CQSq4 convert rfc822-to-8bit (ORCPT ); Wed, 17 Mar 2010 14:46:56 -0400 In-Reply-To: <40ec3ea41003171144s6acbf2fdl4434f2147b8244a0@mail.gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: =46rom: Chase Douglas Date: Wed, 17 Mar 2010 14:44:30 -0400 > On Wed, Mar 17, 2010 at 2:27 PM, David Miller w= rote: >> From: Chase Douglas >> Date: Wed, 17 Mar 2010 14:20:45 -0400 >> >>> If that's the case, what's the solution? It's not safe to call >>> disable_irq* in softirq context anymore. >> >> That's a huge problem, because such a restriction has broken >> several drivers. =A03c59x is not the only one which uses this >> technique for this reason. =A0The 8390 one does too. >=20 > So basically the answer right now is: it's broken and needs to be > reworked, and a switch to disabling irqs is deemed inadequate. Is tha= t > accurate? Yep. The bug is whatever caused disable_irq*() to stop working where these drivers have been using it for what feels like a century :-)