From mboxrd@z Thu Jan 1 00:00:00 1970 From: Manfred Spraul Subject: Re: [git patches] net driver updates Date: Sun, 21 May 2006 16:56:14 +0200 Message-ID: <44707F8E.8010506@colorfullife.com> References: <20060520042856.GA7218@havoc.gtf.org> <20060520105547.220f2bea.akpm@osdl.org> <200605210015.15847.ak@suse.de> <447012B2.9050207@colorfullife.com> <4579880.1148217864672.SLOX.WebMail.wwwrun@imap-dhs.suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Cc: Andrew Morton , Linus Torvalds , jeff@garzik.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Ayaz Abdulla Return-path: Received: from dbl.q-ag.de ([213.172.117.3]:6049 "EHLO dbl.q-ag.de") by vger.kernel.org with ESMTP id S964881AbWEUO4q (ORCPT ); Sun, 21 May 2006 10:56:46 -0400 To: Andreas Kleen In-Reply-To: <4579880.1148217864672.SLOX.WebMail.wwwrun@imap-dhs.suse.de> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Andreas Kleen wrote: >>No idea, but unlikely. The fix removes a duplicate request_irq call. >>Is >>it possible that the both instances run concurrently? >> >> > >The system has two Forcedeth ports, but only one has a cable connected. >I don't think there is any parallelism. Just one connection with a lot >of data. It didn't happen with 2.6.16. > > > You misunderstood me: Due to a broken patch, the driver did request_irq(irqnum, handler,...) twice. Thus the irq handler was registered twice. The rx synchronization assumes that the irq handler doesn't run concurrently. I'm not sure what happens if the irq handler is registered twice: is it possible that it runs twice at the same time, i.e. is the synchronization in the irq subsystem irq number or registration call based? >If you don't have any other good ideas I will try to track it down. > > > I don't have any good ideas, please try to figure out what's wrong. Is there a debug switch for the network layer that forces the network layer to verify the CHECKSUM_UNNECESSARY blocks? -- Manfred