iomap: support partial page discard on writeback block mapping failure

iomap writeback mapping failure only calls into ->discard_page() if
the current page has not been added to the ioend. Accordingly, the
XFS callback assumes a full page discard and invalidation. This is
problematic for sub-page block size filesystems where some portion
of a page might have been mapped successfully before a failure to
map a delalloc block occurs. ->discard_page() is not called in that
error scenario and the bio is explicitly failed by iomap via the
error return from ->prepare_ioend(). As a result, the filesystem
leaks delalloc blocks and corrupts the filesystem block counters.

Since XFS is the only user of ->discard_page(), tweak the semantics
to invoke the callback unconditionally on mapping errors and provide
the file offset that failed to map. Update xfs_discard_page() to
discard the corresponding portion of the file and pass the range
along to iomap_invalidatepage(). The latter already properly handles
both full and sub-page scenarios by not changing any iomap or page
state on sub-page invalidations.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>

2 files changed, 16 insertions, 13 deletions

diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 8180061b9e16..e4ea1f9f94d0 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -1382,14 +1382,15 @@ iomap_writepage_map(struct iomap_writepage_ctx *wpc,
 	 * appropriately.
 	 */
 	if (unlikely(error)) {
+		/*
+		 * Let the filesystem know what portion of the current page
+		 * failed to map. If the page wasn't been added to ioend, it
+		 * won't be affected by I/O completion and we must unlock it
+		 * now.
+		 */
+		if (wpc->ops->discard_page)
+			wpc->ops->discard_page(page, file_offset);
 		if (!count) {
-			/*
-			 * If the current page hasn't been added to ioend, it
-			 * won't be affected by I/O completions and we must
-			 * discard and unlock it right here.
-			 */
-			if (wpc->ops->discard_page)
-				wpc->ops->discard_page(page);
 			ClearPageUptodate(page);
 			unlock_page(page);
 			goto done;
diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 55d126d4e096..5bf37afae5e9 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -527,13 +527,15 @@ xfs_prepare_ioend(
  */
 static void
 xfs_discard_page(
-	struct page		*page)
+	struct page		*page,
+	loff_t			fileoff)
 {
 	struct inode		*inode = page->mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	loff_t			offset = page_offset(page);
-	xfs_fileoff_t		start_fsb = XFS_B_TO_FSBT(mp, offset);
+	unsigned int		pageoff = offset_in_page(fileoff);
+	xfs_fileoff_t		start_fsb = XFS_B_TO_FSBT(mp, fileoff);
+	xfs_fileoff_t		pageoff_fsb = XFS_B_TO_FSBT(mp, pageoff);
 	int			error;
 
 	if (XFS_FORCED_SHUTDOWN(mp))
@@ -541,14 +543,14 @@ xfs_discard_page(
 
 	xfs_alert_ratelimited(mp,
 		"page discard on page "PTR_FMT", inode 0x%llx, offset %llu.",
-			page, ip->i_ino, offset);
+			page, ip->i_ino, fileoff);
 
 	error = xfs_bmap_punch_delalloc_range(ip, start_fsb,
-			i_blocks_per_page(inode, page));
+			i_blocks_per_page(inode, page) - pageoff_fsb);
 	if (error && !XFS_FORCED_SHUTDOWN(mp))
 		xfs_alert(mp, "page discard unable to remove delalloc mapping.");
 out_invalidate:
-	iomap_invalidatepage(page, 0, PAGE_SIZE);
+	iomap_invalidatepage(page, pageoff, PAGE_SIZE - pageoff);
 }
 
 static const struct iomap_writeback_ops xfs_writeback_ops = {