From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from verein.lst.de (verein.lst.de [213.95.11.211]) by mail19.linbit.com (LINBIT Mail Daemon) with ESMTP id 9E16042039E for ; Tue, 11 Jun 2024 07:59:10 +0200 (CEST) Date: Tue, 11 Jun 2024 07:59:06 +0200 From: Christoph Hellwig To: Damien Le Moal Subject: Re: [PATCH 05/26] loop: regularize upgrading the lock size for direct I/O Message-ID: <20240611055906.GA3640@lst.de> References: <20240611051929.513387-1-hch@lst.de> <20240611051929.513387-6-hch@lst.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Cc: nvdimm@lists.linux.dev, "Michael S. Tsirkin" , Jason Wang , linux-nvme@lists.infradead.org, Song Liu , linux-mtd@lists.infradead.org, Vineeth Vijayan , linux-bcache@vger.kernel.org, Christoph Hellwig , Alasdair Kergon , drbd-dev@lists.linbit.com, linux-s390@vger.kernel.org, linux-scsi@vger.kernel.org, Richard Weinberger , Geert Uytterhoeven , Yu Kuai , dm-devel@lists.linux.dev, linux-um@lists.infradead.org, Mike Snitzer , Josef Bacik , Ming Lei , linux-raid@vger.kernel.org, linux-m68k@lists.linux-m68k.org, Mikulas Patocka , xen-devel@lists.xenproject.org, ceph-devel@vger.kernel.org, nbd@other.debian.org, Jens Axboe , linux-block@vger.kernel.org, "Martin K. Petersen" , linux-mmc@vger.kernel.org, Philipp Reisner , virtualization@lists.linux.dev, Lars Ellenberg , linuxppc-dev@lists.ozlabs.org, Roger Pau Monn?? List-Id: "*Coordination* of development, patches, contributions -- *Questions* \(even to developers\) go to drbd-user, please." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, Jun 11, 2024 at 02:56:59PM +0900, Damien Le Moal wrote: > > + if (!bsize) > > + bsize = loop_default_blocksize(lo, inode->i_sb->s_bdev); > > If bsize is specified and there is a backing dev used with direct IO, should it > be checked that bsize is a multiple of bdev_logical_block_size(backing_bdev) ? For direct I/O that check would be useful. For buffered I/O we can do read-modify-write cycles. However this series is already huge and not primarily about improving the loop driver parameter validation, so I'll defer this for now.