From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from verein.lst.de ([213.95.11.211]:54899 "EHLO newverein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754900AbeDWRYI (ORCPT ); Mon, 23 Apr 2018 13:24:08 -0400 Date: Mon, 23 Apr 2018 19:25:31 +0200 From: Christoph Hellwig Subject: Re: [PATCH 06/13] xfs: fetch dquots directly during quotacheck Message-ID: <20180423172531.GF834@lst.de> References: <152440954198.29601.14390250048915099607.stgit@magnolia> <152440959026.29601.18114675159046762818.stgit@magnolia> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <152440959026.29601.18114675159046762818.stgit@magnolia> Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: "Darrick J. Wong" Cc: linux-xfs@vger.kernel.org, hch@lst.de On Sun, Apr 22, 2018 at 08:06:30AM -0700, Darrick J. Wong wrote: > From: Darrick J. Wong > > Quotacheck only runs during mount, which means that there are no other > processes in the system that could be doing chown or chproj. Therefore > there's no potential for racing to attach dquots to the inode so we can > drop all the ILOCK and race detection bits from quotacheck. Looks good: Reviewed-by: Christoph Hellwig Btw, one thing I've wondered for a while is if there is any point in keeping the dquot pointers in the inode at all. With the radix tree the dquots should be a few indirections and thus very few cache lines away at any point in time, while saving us all the setup overhead, as well as shrinking the inode.