From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from userp1040.oracle.com ([156.151.31.81]:48974 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751500AbdBBXBd (ORCPT ); Thu, 2 Feb 2017 18:01:33 -0500 Date: Thu, 2 Feb 2017 15:01:27 -0800 From: "Darrick J. Wong" Subject: Re: [PATCH 1/4] xfs: reject all unaligned direct writes to reflinked files Message-ID: <20170202230126.GP9134@birch.djwong.org> References: <1485985359-971-1-git-send-email-hch@lst.de> <1485985359-971-2-git-send-email-hch@lst.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1485985359-971-2-git-send-email-hch@lst.de> Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: Christoph Hellwig Cc: linux-xfs@vger.kernel.org On Wed, Feb 01, 2017 at 10:42:36PM +0100, Christoph Hellwig wrote: > We currently fall back from direct to buffered writes if we detect a > remaining shared extent in the iomap_begin callback. But by the time > iomap_begin is called for the potentially unaligned end block we might > have already written most of the data to disk, which we'd now write > again using buffered I/O. To avoid this reject all writes to reflinked > files before starting I/O so that we are guaranteed to only write the > data once. > > The alternative would be to unshare the unaligned start and/or end block > before doing the I/O. I think that's doable, and will actually be > required to support reflinks on DAX file system. But it will take a > little more time and I'd rather get rid of the double write ASAP. > > Signed-off-by: Christoph Hellwig > Reviewed-by: Brian Foster > --- > fs/xfs/xfs_file.c | 7 +++++++ > fs/xfs/xfs_iomap.c | 12 +----------- > 2 files changed, 8 insertions(+), 11 deletions(-) > > diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c > index bbb9eb6..06236bf 100644 > --- a/fs/xfs/xfs_file.c > +++ b/fs/xfs/xfs_file.c > @@ -527,6 +527,13 @@ xfs_file_dio_aio_write( > if ((iocb->ki_pos & mp->m_blockmask) || > ((iocb->ki_pos + count) & mp->m_blockmask)) { > unaligned_io = 1; > + > + /* > + * We can't properly handle unaligned direct I/O to reflink > + * files yet, as we can't unshare a partial block. > + */ > + if (xfs_is_reflink_inode(ip)) > + return -EREMCHG; > iolock = XFS_IOLOCK_EXCL; > } else { > iolock = XFS_IOLOCK_SHARED; > diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c > index 767208f..d17aa3b 100644 > --- a/fs/xfs/xfs_iomap.c > +++ b/fs/xfs/xfs_iomap.c > @@ -1026,17 +1026,7 @@ xfs_file_iomap_begin( > if (error) > goto out_unlock; > > - /* > - * We're here because we're trying to do a directio write to a > - * region that isn't aligned to a filesystem block. If the > - * extent is shared, fall back to buffered mode to handle the > - * RMW. > - */ > - if (!(flags & IOMAP_REPORT) && shared) { > - trace_xfs_reflink_bounce_dio_write(ip, &imap); Can you either port the tracepoint to the new hunk or just kill it entirely? (If you opt for killing it, I'll just remove it from xfs_trace.h when I import this patch.) Otherwise, Reviewed-by: Darrick J. Wong --D > - error = -EREMCHG; > - goto out_unlock; > - } > + ASSERT((flags & IOMAP_REPORT) || !shared); > } > > if ((flags & (IOMAP_WRITE | IOMAP_ZERO)) && xfs_is_reflink_inode(ip)) { > -- > 2.1.4 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html