From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964865AbXCPOaI (ORCPT ); Fri, 16 Mar 2007 10:30:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965170AbXCPOaI (ORCPT ); Fri, 16 Mar 2007 10:30:08 -0400 Received: from nf-out-0910.google.com ([64.233.182.191]:43379 "EHLO nf-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S964865AbXCPOaF (ORCPT ); Fri, 16 Mar 2007 10:30:05 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:sender:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references:x-google-sender-auth; b=jy3AjhEGPRw1hZ3zAzNvmjQekbw6U69PISwY/sEeGA2Ez2D5xMv4zSkkWGsUA03n3PrIqWCg40zusF3GVSuKZzAkue/DTQYZw+nFkuib8c6hUyudedPDLZB3SnDKivUAkMkhwn2Ot3TAUZq76AYV2vITyt04/M0tsTCK4e7bhco= Message-ID: <84144f020703160730y7cd692efx264635fe4683398b@mail.gmail.com> Date: Fri, 16 Mar 2007 16:30:04 +0200 From: "Pekka Enberg" To: "Alan Cox" Subject: Re: [PATCH 2/5] revoke: core code Cc: "Andrew Morton" , linux-kernel@vger.kernel.org, hch@infradead.org In-Reply-To: <20070316145855.5604d7d8@lxorguk.ukuu.org.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <20070315173438.efadd514.akpm@linux-foundation.org> <84144f020703152344u588da670ge14c1ec5797c2b86@mail.gmail.com> <20070316032624.c6db84fc.akpm@linux-foundation.org> <20070316042600.3e1bd424.akpm@linux-foundation.org> <20070316145855.5604d7d8@lxorguk.ukuu.org.uk> X-Google-Sender-Auth: 9061f59ad0f2201f Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On 3/16/07, Alan Cox wrote: > > I'm not sure that running do_fsync() will guarantee that all sys_write() > > callers will have finished their syscall. Probably they will have, in > > practice. But there is logic in the sync paths to deliberately bale out > > if we're competing with ongoing dirtyings, to avoid livelocking. > > For device files you really need to call into the device driver for this > (->flush etc). Sure but the do_fsync() bits are part of generic_file_revoke() which is not meant for device files at all.