From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kara Subject: Re: [PATCH 1/5] vfs: Make sync(1) writeout also block device inodes Date: Thu, 28 Jul 2011 22:13:14 +0200 Message-ID: <20110728201314.GI5044@quack.suse.cz> References: <1311719886-1130-1-git-send-email-jack@suse.cz> <1311719886-1130-2-git-send-email-jack@suse.cz> <20110727094423.GA11334@infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jan Kara , linux-fsdevel@vger.kernel.org, Curt Wohlgemuth , Al Viro To: Christoph Hellwig Return-path: Received: from cantor2.suse.de ([195.135.220.15]:58567 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755365Ab1G1UNP (ORCPT ); Thu, 28 Jul 2011 16:13:15 -0400 Content-Disposition: inline In-Reply-To: <20110727094423.GA11334@infradead.org> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Wed 27-07-11 05:44:23, Christoph Hellwig wrote: > > -static int __sync_filesystem(struct super_block *sb, int wait) > > +static void __sync_filesystem(struct super_block *sb, int wait) > > { > > - /* > > - * This should be safe, as we require bdi backing to actually > > - * write out data in the first place > > - */ > > - if (sb->s_bdi == &noop_backing_dev_info) > > - return 0; > > - > > Moving this check is not related to the block device writeback, is it? > Furthermore it gets moved deeper into the stack later on anyway, so it > gets reverted. Good point, I'll swap patches 1 & 2. > I think the series would be much cleaner if this patch gets moved > towards the end of it. > > > +static void sync_all_bdevs(int wait) > > +{ > > This function should at least have a comment explaining why we need it. Will do. Honza -- Jan Kara SUSE Labs, CR