From: Christoph Hellwig Subject: Re: [PATCH 13/13] ext4: Support for synchronous DAX faults Date: Thu, 24 Aug 2017 05:31:26 -0700 Message-ID: <20170824123126.GA21282@infradead.org> References: <20170817160815.30466-1-jack@suse.cz> <20170817160815.30466-14-jack@suse.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-fsdevel@vger.kernel.org, linux-nvdimm@lists.01.org, Andy Lutomirski , linux-ext4@vger.kernel.org, linux-xfs@vger.kernel.org, Christoph Hellwig , Ross Zwisler , Dan Williams , Boaz Harrosh To: Jan Kara Return-path: Received: from bombadil.infradead.org ([65.50.211.133]:39274 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751272AbdHXMb1 (ORCPT ); Thu, 24 Aug 2017 08:31:27 -0400 Content-Disposition: inline In-Reply-To: <20170817160815.30466-14-jack@suse.cz> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Thu, Aug 17, 2017 at 06:08:15PM +0200, Jan Kara wrote: > We return IOMAP_F_NEEDDSYNC flag from ext4_iomap_begin() for a > synchronous write fault when inode has some uncommitted metadata > changes. In the fault handler ext4_dax_fault() we then detect this case, > call vfs_fsync_range() to make sure all metadata is committed, and call > dax_pfn_mkwrite() to mark PTE as writeable. Note that this will also > dirty corresponding radix tree entry which is what we want - fsync(2) > will still provide data integrity guarantees for applications not using > userspace flushing. And applications using userspace flushing can avoid > calling fsync(2) and thus avoid the performance overhead. Why is this only wiered up for the huge_fault handler and not the regular?