iomap: replace folio_batch allocation with stack allocation

Zhang Yi points out that the dynamic folio_batch allocation in
iomap_fill_dirty_folios() is problematic for the ext4 on iomap work
that is under development because it doesn't sufficiently handle the
allocation failure case (by allowing a retry, for example). We've
also seen lockdep (via syzbot) complain recently about the scope of
the allocation.

The dynamic allocation was initially added for simplicity and to
help indicate whether the batch was used or not by the calling fs.
To address these issues, put the batch on the stack of
iomap_zero_range() and use a flag to control whether the batch
should be used in the iomap folio lookup path. This keeps things
simple and eliminates allocation issues with lockdep and for ext4 on
iomap.

While here, also clean up the fill helper signature to be more
consistent with the underlying filemap helper. Pass through the
return value of the filemap helper (folio count) and update the
lookup offset via an out param.

Fixes: 395ed1ef0012 ("iomap: optional zero range dirty folio processing")
Signed-off-by: Brian Foster <[e-mail]>
Link: https://patch.msgid.link/[e-mail]
Acked-by: Dave Chinner <[e-mail]>
Reviewed-by: Christoph Hellwig <[e-mail]>
Reviewed-by: Darrick J. Wong <[e-mail]>
Signed-off-by: Christian Brauner <[e-mail]>

fs/iomap/buffered-io.c

···
 	if (!mapping_large_folio_support(iter->inode->i_mapping))
 		len = min_t(size_t, len, PAGE_SIZE - offset_in_page(pos));
 
-	if (iter->fbatch) {
+	if (iter->iomap.flags & IOMAP_F_FOLIO_BATCH) {
 		struct folio *folio = folio_batch_next(iter->fbatch);
 
 		if (!folio)
···
 	 * process so return and let the caller iterate and refill the batch.
 	 */
 	if (!folio) {
-		WARN_ON_ONCE(!iter->fbatch);
+		WARN_ON_ONCE(!(iter->iomap.flags & IOMAP_F_FOLIO_BATCH));
 		return 0;
 	}
 
···
 	return status;
 }
 
-loff_t
+/**
+ * iomap_fill_dirty_folios - fill a folio batch with dirty folios
+ * @iter: Iteration structure
+ * @start: Start offset of range. Updated based on lookup progress.
+ * @end: End offset of range
+ * @iomap_flags: Flags to set on the associated iomap to track the batch.
+ *
+ * Returns the folio count directly. Also returns the associated control flag if
+ * the the batch lookup is performed and the expected offset of a subsequent
+ * lookup via out params. The caller is responsible to set the flag on the
+ * associated iomap.
+ */
+unsigned int
 iomap_fill_dirty_folios(
 	struct iomap_iter	*iter,
-	loff_t			offset,
-	loff_t			length)
+	loff_t			*start,
+	loff_t			end,
+	unsigned int		*iomap_flags)
 {
 	struct address_space	*mapping = iter->inode->i_mapping;
-	pgoff_t			start = offset >> PAGE_SHIFT;
-	pgoff_t			end = (offset + length - 1) >> PAGE_SHIFT;
+	pgoff_t			pstart = *start >> PAGE_SHIFT;
+	pgoff_t			pend = (end - 1) >> PAGE_SHIFT;
+	unsigned int		count;
 
-	iter->fbatch = kmalloc(sizeof(struct folio_batch), GFP_KERNEL);
-	if (!iter->fbatch)
-		return offset + length;
-	folio_batch_init(iter->fbatch);
+	if (!iter->fbatch) {
+		*start = end;
+		return 0;
+	}
 
-	filemap_get_folios_dirty(mapping, &start, end, iter->fbatch);
-	return (start << PAGE_SHIFT);
+	count = filemap_get_folios_dirty(mapping, &pstart, pend, iter->fbatch);
+	*start = (pstart << PAGE_SHIFT);
+	*iomap_flags |= IOMAP_F_FOLIO_BATCH;
+	return count;
 }
 EXPORT_SYMBOL_GPL(iomap_fill_dirty_folios);
 
···
 		const struct iomap_ops *ops,
 		const struct iomap_write_ops *write_ops, void *private)
 {
+	struct folio_batch fbatch;
 	struct iomap_iter iter = {
 		.inode		= inode,
 		.pos		= pos,
 		.len		= len,
 		.flags		= IOMAP_ZERO,
 		.private	= private,
+		.fbatch		= &fbatch,
 	};
 	struct address_space *mapping = inode->i_mapping;
 	int ret;
 	bool range_dirty;
+
+	folio_batch_init(&fbatch);
 
 	/*
 	 * To avoid an unconditional flush, check pagecache state and only flush
···
 	while ((ret = iomap_iter(&iter, ops)) > 0) {
 		const struct iomap *srcmap = iomap_iter_srcmap(&iter);
 
-		if (WARN_ON_ONCE(iter.fbatch &&
+		if (WARN_ON_ONCE((iter.iomap.flags & IOMAP_F_FOLIO_BATCH) &&
 				 srcmap->type != IOMAP_UNWRITTEN))
 			return -EIO;
 
-		if (!iter.fbatch &&
+		if (!(iter.iomap.flags & IOMAP_F_FOLIO_BATCH) &&
 		    (srcmap->type == IOMAP_HOLE ||
 		     srcmap->type == IOMAP_UNWRITTEN)) {
 			s64 status;

fs/iomap/iter.c

···
 
 static inline void iomap_iter_reset_iomap(struct iomap_iter *iter)
 {
-	if (iter->fbatch) {
+	if (iter->iomap.flags & IOMAP_F_FOLIO_BATCH) {
 		folio_batch_release(iter->fbatch);
-		kfree(iter->fbatch);
-		iter->fbatch = NULL;
+		folio_batch_reinit(iter->fbatch);
+		iter->iomap.flags &= ~IOMAP_F_FOLIO_BATCH;
 	}
 
 	iter->status = 0;

fs/xfs/xfs_iomap.c

···
 	 */
 	if (flags & IOMAP_ZERO) {
 		xfs_fileoff_t eof_fsb = XFS_B_TO_FSB(mp, XFS_ISIZE(ip));
-		u64 end;
 
 		if (isnullstartblock(imap.br_startblock) &&
 		    offset_fsb >= eof_fsb)
···
 		 */
 		if (imap.br_state == XFS_EXT_UNWRITTEN &&
 		    offset_fsb < eof_fsb) {
-			loff_t len = min(count,
-					 XFS_FSB_TO_B(mp, imap.br_blockcount));
+			loff_t foffset = offset, fend;
 
-			end = iomap_fill_dirty_folios(iter, offset, len);
+			fend = offset +
+			       min(count, XFS_FSB_TO_B(mp, imap.br_blockcount));
+			iomap_fill_dirty_folios(iter, &foffset, fend,
+						&iomap_flags);
 			end_fsb = min_t(xfs_fileoff_t, end_fsb,
-					XFS_B_TO_FSB(mp, end));
+					XFS_B_TO_FSB(mp, foffset));
 		}
 
 		xfs_trim_extent(&imap, offset_fsb, end_fsb - offset_fsb);

include/linux/iomap.h

···
 /*
  * Flags set by the core iomap code during operations:
  *
+ * IOMAP_F_FOLIO_BATCH indicates that the folio batch mechanism is active
+ * for this operation, set by iomap_fill_dirty_folios().
+ *
  * IOMAP_F_SIZE_CHANGED indicates to the iomap_end method that the file size
  * has changed as the result of this write operation.
  *
···
  * range it covers needs to be remapped by the high level before the operation
  * can proceed.
  */
+#define IOMAP_F_FOLIO_BATCH	(1U << 13)
 #define IOMAP_F_SIZE_CHANGED	(1U << 14)
 #define IOMAP_F_STALE		(1U << 15)
 
···
 int iomap_file_unshare(struct inode *inode, loff_t pos, loff_t len,
 		const struct iomap_ops *ops,
 		const struct iomap_write_ops *write_ops);
-loff_t iomap_fill_dirty_folios(struct iomap_iter *iter, loff_t offset,
-		loff_t length);
+unsigned int iomap_fill_dirty_folios(struct iomap_iter *iter, loff_t *start,
+		loff_t end, unsigned int *iomap_flags);
 int iomap_zero_range(struct inode *inode, loff_t pos, loff_t len,
 		bool *did_zero, const struct iomap_ops *ops,
 		const struct iomap_write_ops *write_ops, void *private);