iomap: refactor the writeback interface

Replace ->map_blocks with a new ->writeback_range, which differs in the
following ways:

 - it must also queue up the I/O for writeback, that is called into the
   slightly refactored and extended in scope iomap_add_to_ioend for
   each region
 - can handle only a part of the requested region, that is the retry
   loop for partial mappings moves to the caller
 - handles cleanup on failures as well, and thus also replaces the
   discard_folio method only implemented by XFS.

This will allow to use the iomap writeback code also for file systems
that are not block based like fuse.

Co-developed-by: Joanne Koong <joannelkoong@gmail.com>
Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/20250710133343.399917-5-hch@lst.de
Acked-by: Damien Le Moal <dlemoal@kernel.org>	# zonefs
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

1 files changed, 82 insertions, 46 deletions

diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 65485a52df3b..f6d44ab78442 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -234,6 +234,47 @@ xfs_end_bio(
 }
 
 /*
+ * We cannot cancel the ioend directly on error.  We may have already set other
+ * pages under writeback and hence we have to run I/O completion to mark the
+ * error state of the pages under writeback appropriately.
+ *
+ * If the folio has delalloc blocks on it, the caller is asking us to punch them
+ * out. If we don't, we can leave a stale delalloc mapping covered by a clean
+ * page that needs to be dirtied again before the delalloc mapping can be
+ * converted. This stale delalloc mapping can trip up a later direct I/O read
+ * operation on the same region.
+ *
+ * We prevent this by truncating away the delalloc regions on the folio. Because
+ * they are delalloc, we can do this without needing a transaction. Indeed - if
+ * we get ENOSPC errors, we have to be able to do this truncation without a
+ * transaction as there is no space left for block reservation (typically why
+ * we see a ENOSPC in writeback).
+ */
+static void
+xfs_discard_folio(
+	struct folio		*folio,
+	loff_t			pos)
+{
+	struct xfs_inode	*ip = XFS_I(folio->mapping->host);
+	struct xfs_mount	*mp = ip->i_mount;
+
+	if (xfs_is_shutdown(mp))
+		return;
+
+	xfs_alert_ratelimited(mp,
+		"page discard on page "PTR_FMT", inode 0x%llx, pos %llu.",
+			folio, ip->i_ino, pos);
+
+	/*
+	 * The end of the punch range is always the offset of the first
+	 * byte of the next folio. Hence the end offset is only dependent on the
+	 * folio itself and not the start offset that is passed in.
+	 */
+	xfs_bmap_punch_delalloc_range(ip, XFS_DATA_FORK, pos,
+				folio_pos(folio) + folio_size(folio), NULL);
+}
+
+/*
  * Fast revalidation of the cached writeback mapping. Return true if the current
  * mapping is valid, false otherwise.
  */
@@ -278,13 +319,12 @@ xfs_imap_valid(
 static int
 xfs_map_blocks(
 	struct iomap_writepage_ctx *wpc,
-	struct inode		*inode,
 	loff_t			offset,
 	unsigned int		len)
 {
-	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_inode	*ip = XFS_I(wpc->inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	ssize_t			count = i_blocksize(inode);
+	ssize_t			count = i_blocksize(wpc->inode);
 	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 	xfs_fileoff_t		end_fsb = XFS_B_TO_FSB(mp, offset + count);
 	xfs_fileoff_t		cow_fsb;
@@ -436,6 +476,24 @@ allocate_blocks:
 	return 0;
 }
 
+static ssize_t
+xfs_writeback_range(
+	struct iomap_writepage_ctx *wpc,
+	struct folio		*folio,
+	u64			offset,
+	unsigned int		len,
+	u64			end_pos)
+{
+	ssize_t			ret;
+
+	ret = xfs_map_blocks(wpc, offset, len);
+	if (!ret)
+		ret = iomap_add_to_ioend(wpc, folio, offset, end_pos, len);
+	if (ret < 0)
+		xfs_discard_folio(folio, offset);
+	return ret;
+}
+
 static bool
 xfs_ioend_needs_wq_completion(
 	struct iomap_ioend	*ioend)
@@ -488,47 +546,9 @@ xfs_submit_ioend(
 	return 0;
 }
 
-/*
- * If the folio has delalloc blocks on it, the caller is asking us to punch them
- * out. If we don't, we can leave a stale delalloc mapping covered by a clean
- * page that needs to be dirtied again before the delalloc mapping can be
- * converted. This stale delalloc mapping can trip up a later direct I/O read
- * operation on the same region.
- *
- * We prevent this by truncating away the delalloc regions on the folio. Because
- * they are delalloc, we can do this without needing a transaction. Indeed - if
- * we get ENOSPC errors, we have to be able to do this truncation without a
- * transaction as there is no space left for block reservation (typically why
- * we see a ENOSPC in writeback).
- */
-static void
-xfs_discard_folio(
-	struct folio		*folio,
-	loff_t			pos)
-{
-	struct xfs_inode	*ip = XFS_I(folio->mapping->host);
-	struct xfs_mount	*mp = ip->i_mount;
-
-	if (xfs_is_shutdown(mp))
-		return;
-
-	xfs_alert_ratelimited(mp,
-		"page discard on page "PTR_FMT", inode 0x%llx, pos %llu.",
-			folio, ip->i_ino, pos);
-
-	/*
-	 * The end of the punch range is always the offset of the first
-	 * byte of the next folio. Hence the end offset is only dependent on the
-	 * folio itself and not the start offset that is passed in.
-	 */
-	xfs_bmap_punch_delalloc_range(ip, XFS_DATA_FORK, pos,
-				folio_pos(folio) + folio_size(folio), NULL);
-}
-
 static const struct iomap_writeback_ops xfs_writeback_ops = {
-	.map_blocks		= xfs_map_blocks,
+	.writeback_range	= xfs_writeback_range,
 	.submit_ioend		= xfs_submit_ioend,
-	.discard_folio		= xfs_discard_folio,
 };
 
 struct xfs_zoned_writepage_ctx {
@@ -545,11 +565,10 @@ XFS_ZWPC(struct iomap_writepage_ctx *ctx)
 static int
 xfs_zoned_map_blocks(
 	struct iomap_writepage_ctx *wpc,
-	struct inode		*inode,
 	loff_t			offset,
 	unsigned int		len)
 {
-	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_inode	*ip = XFS_I(wpc->inode);
 	struct xfs_mount	*mp = ip->i_mount;
 	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 	xfs_fileoff_t		end_fsb = XFS_B_TO_FSB(mp, offset + len);
@@ -608,6 +627,24 @@ xfs_zoned_map_blocks(
 	return 0;
 }
 
+static ssize_t
+xfs_zoned_writeback_range(
+	struct iomap_writepage_ctx *wpc,
+	struct folio		*folio,
+	u64			offset,
+	unsigned int		len,
+	u64			end_pos)
+{
+	ssize_t			ret;
+
+	ret = xfs_zoned_map_blocks(wpc, offset, len);
+	if (!ret)
+		ret = iomap_add_to_ioend(wpc, folio, offset, end_pos, len);
+	if (ret < 0)
+		xfs_discard_folio(folio, offset);
+	return ret;
+}
+
 static int
 xfs_zoned_submit_ioend(
 	struct iomap_writepage_ctx *wpc,
@@ -621,9 +658,8 @@ xfs_zoned_submit_ioend(
 }
 
 static const struct iomap_writeback_ops xfs_zoned_writeback_ops = {
-	.map_blocks		= xfs_zoned_map_blocks,
+	.writeback_range	= xfs_zoned_writeback_range,
 	.submit_ioend		= xfs_zoned_submit_ioend,
-	.discard_folio		= xfs_discard_folio,
 };
 
 STATIC int