From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 1/5] vfs: vfs-level fiemap interface Date: Sat, 31 May 2008 04:16:24 -0400 Message-ID: <20080531081624.GA24135@infradead.org> References: <20080525000157.GK8325@wotan.suse.de> <20080528194215.GI7263@webber.adilger.int> <20080529012435.GC12405@disturbed> <20080529130437.GD21299@infradead.org> <20080529170216.GD2985@webber.adilger.int> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Christoph Hellwig , Mark Fasheh , linux-fsdevel@vger.kernel.org, Josef Bacik To: Andreas Dilger Return-path: Received: from bombadil.infradead.org ([18.85.46.34]:59024 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751375AbYEaIQ0 (ORCPT ); Sat, 31 May 2008 04:16:26 -0400 Content-Disposition: inline In-Reply-To: <20080529170216.GD2985@webber.adilger.int> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Thu, May 29, 2008 at 11:02:16AM -0600, Andreas Dilger wrote: > > Yes. Extent merging only really makes sense for block based > > filesystems, and there should be an output flag telling the user > > that this happened. > > I've accepted that extent merging is a bad idea for the generic interface. > I think for block-mapped filesystems this pretty much has to happen. Yes, I agree and probably was a little unclear in that previous mail. It should happen for and only for block-based filesystems. And there should be a flag telling that user that it has happened.