From: David Chinner Date: Mon, 18 Jun 2007 06:49:58 +0000 (+1000) Subject: [XFS] Map unwritten extents correctly for I/o completion processing X-Git-Tag: v2.6.23-rc1~1056^2~21 X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=effd120edb7609069cca9f3d1cb4bfae464b2f85;p=linux-2.6 [XFS] Map unwritten extents correctly for I/o completion processing If we have multiple unwritten extents within a single page, we fail to tell the I/o completion construction handlers we need a new handle for the second and subsequent blocks in the page. While we still issue the I/O correctly, we do not have the correct ranges recorded in the ioend structures and hence when we go to convert the unwritten extents we screw it up. Make sure we start a new ioend every time the mapping changes so that we convert the correct ranges on I/O completion. SGI-PV: 964647 SGI-Modid: xfs-linux-melb:xfs-kern:28797a Signed-off-by: David Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Tim Shimmin --- diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index c097e4e697..fd4105d662 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1010,6 +1010,8 @@ xfs_page_state_convert( if (buffer_unwritten(bh) || buffer_delay(bh) || ((buffer_uptodate(bh) || PageUptodate(page)) && !buffer_mapped(bh) && (unmapped || startio))) { + int new_ioend = 0; + /* * Make sure we don't use a read-only iomap */ @@ -1028,6 +1030,15 @@ xfs_page_state_convert( } if (!iomap_valid) { + /* + * if we didn't have a valid mapping then we + * need to ensure that we put the new mapping + * in a new ioend structure. This needs to be + * done to ensure that the ioends correctly + * reflect the block mappings at io completion + * for unwritten extent conversion. + */ + new_ioend = 1; if (type == IOMAP_NEW) { size = xfs_probe_cluster(inode, page, bh, head, 0); @@ -1047,7 +1058,7 @@ xfs_page_state_convert( if (startio) { xfs_add_to_ioend(inode, bh, offset, type, &ioend, - !iomap_valid); + new_ioend); } else { set_buffer_dirty(bh); unlock_buffer(bh);