iomap: replace iomap_folio_ops with iomap_write_ops

The iomap_folio_ops are only used for buffered writes, including the zero
and unshare variants. Rename them to iomap_write_ops to better describe
the usage, and pass them through the call chain like the other operation
specific methods instead of through the iomap.

xfs_iomap_valid grows a IOMAP_HOLE check to keep the existing behavior
that never attached the folio_ops to a iomap representing a hole.

Signed-off-by: Christoph Hellwig <[e-mail]>
Link: https://lore.kernel.org/[e-mail]
Acked-by: Damien Le Moal <[e-mail]>
Reviewed-by: Brian Foster <[e-mail]>
Reviewed-by: Darrick J. Wong <[e-mail]>
Signed-off-by: Christian Brauner <[e-mail]>

Documentation/filesystems/iomap/design.rst

···
      struct dax_device   *dax_dev;
      void                *inline_data;
      void                *private;
-     const struct iomap_folio_ops *folio_ops;
      u64                 validity_cookie;
  };
 
···
  * ``private`` is a pointer to `filesystem-private information
    <https://lore.kernel.org/all/[e-mail]/>`_.
    This value will be passed unchanged to ``->iomap_end``.
-
- * ``folio_ops`` will be covered in the section on pagecache operations.
 
  * ``validity_cookie`` is a magic freshness value set by the filesystem
    that should be used to detect stale mappings.

Documentation/filesystems/iomap/operations.rst

···
  * ``bmap``
  * ``swap_activate``
 
-``struct iomap_folio_ops``
+``struct iomap_write_ops``
 --------------------------
-
-The ``->iomap_begin`` function for pagecache operations may set the
-``struct iomap::folio_ops`` field to an ops structure to override
-default behaviors of iomap:
 
 .. code-block:: c
 
- struct iomap_folio_ops {
+ struct iomap_write_ops {
      struct folio *(*get_folio)(struct iomap_iter *iter, loff_t pos,
                                 unsigned len);
      void (*put_folio)(struct inode *inode, loff_t pos, unsigned copied,

block/fops.c

···
 
 static ssize_t blkdev_buffered_write(struct kiocb *iocb, struct iov_iter *from)
 {
-	return iomap_file_buffered_write(iocb, from, &blkdev_iomap_ops, NULL);
+	return iomap_file_buffered_write(iocb, from, &blkdev_iomap_ops, NULL,
+			NULL);
 }
 
 /*

fs/gfs2/bmap.c

···
 gfs2_iomap_get_folio(struct iomap_iter *iter, loff_t pos, unsigned len)
 {
 	struct inode *inode = iter->inode;
+	struct gfs2_inode *ip = GFS2_I(inode);
 	unsigned int blockmask = i_blocksize(inode) - 1;
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	unsigned int blocks;
 	struct folio *folio;
 	int status;
+
+	if (!gfs2_is_jdata(ip) && !gfs2_is_stuffed(ip))
+		return iomap_get_folio(iter, pos, len);
 
 	blocks = ((pos & blockmask) + len + blockmask) >> inode->i_blkbits;
 	status = gfs2_trans_begin(sdp, RES_DINODE + blocks, 0);
···
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 
-	if (!gfs2_is_stuffed(ip))
+	if (gfs2_is_jdata(ip) && !gfs2_is_stuffed(ip))
 		gfs2_trans_add_databufs(ip->i_gl, folio,
 					offset_in_folio(folio, pos),
 					copied);
···
 	folio_unlock(folio);
 	folio_put(folio);
 
-	if (tr->tr_num_buf_new)
-		__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
-
-	gfs2_trans_end(sdp);
+	if (gfs2_is_jdata(ip) || gfs2_is_stuffed(ip)) {
+		if (tr->tr_num_buf_new)
+			__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
+		gfs2_trans_end(sdp);
+	}
 }
 
-static const struct iomap_folio_ops gfs2_iomap_folio_ops = {
+const struct iomap_write_ops gfs2_iomap_write_ops = {
 	.get_folio = gfs2_iomap_get_folio,
 	.put_folio = gfs2_iomap_put_folio,
 };
···
 		gfs2_trans_end(sdp);
 	}
 
-	if (gfs2_is_stuffed(ip) || gfs2_is_jdata(ip))
-		iomap->folio_ops = &gfs2_iomap_folio_ops;
 	return 0;
 
 out_trans_end:
···
 		return 0;
 	length = min(length, inode->i_size - from);
 	return iomap_zero_range(inode, from, length, NULL, &gfs2_iomap_ops,
-			NULL);
+			&gfs2_iomap_write_ops, NULL);
 }
 
 #define GFS2_JTRUNC_REVOKES 8192

fs/gfs2/bmap.h

···
 }
 
 extern const struct iomap_ops gfs2_iomap_ops;
+extern const struct iomap_write_ops gfs2_iomap_write_ops;
 extern const struct iomap_writeback_ops gfs2_writeback_ops;
 
 int gfs2_unstuff_dinode(struct gfs2_inode *ip);

fs/gfs2/file.c

···
 	}
 
 	pagefault_disable();
-	ret = iomap_file_buffered_write(iocb, from, &gfs2_iomap_ops, NULL);
+	ret = iomap_file_buffered_write(iocb, from, &gfs2_iomap_ops,
+			&gfs2_iomap_write_ops, NULL);
 	pagefault_enable();
 	if (ret > 0)
 		written += ret;

fs/iomap/buffered-io.c

···
 	return 0;
 }
 
-static struct folio *__iomap_get_folio(struct iomap_iter *iter, size_t len)
+static struct folio *__iomap_get_folio(struct iomap_iter *iter,
+		const struct iomap_write_ops *write_ops, size_t len)
 {
-	const struct iomap_folio_ops *folio_ops = iter->iomap.folio_ops;
 	loff_t pos = iter->pos;
 
 	if (!mapping_large_folio_support(iter->inode->i_mapping))
 		len = min_t(size_t, len, PAGE_SIZE - offset_in_page(pos));
 
-	if (folio_ops && folio_ops->get_folio)
-		return folio_ops->get_folio(iter, pos, len);
-	else
-		return iomap_get_folio(iter, pos, len);
+	if (write_ops && write_ops->get_folio)
+		return write_ops->get_folio(iter, pos, len);
+	return iomap_get_folio(iter, pos, len);
 }
 
-static void __iomap_put_folio(struct iomap_iter *iter, size_t ret,
+static void __iomap_put_folio(struct iomap_iter *iter,
+		const struct iomap_write_ops *write_ops, size_t ret,
 		struct folio *folio)
 {
-	const struct iomap_folio_ops *folio_ops = iter->iomap.folio_ops;
 	loff_t pos = iter->pos;
 
-	if (folio_ops && folio_ops->put_folio) {
-		folio_ops->put_folio(iter->inode, pos, ret, folio);
+	if (write_ops && write_ops->put_folio) {
+		write_ops->put_folio(iter->inode, pos, ret, folio);
 	} else {
 		folio_unlock(folio);
 		folio_put(folio);
···
  * offset, and length. Callers can optionally pass a max length *plen,
  * otherwise init to zero.
  */
-static int iomap_write_begin(struct iomap_iter *iter, struct folio **foliop,
+static int iomap_write_begin(struct iomap_iter *iter,
+		const struct iomap_write_ops *write_ops, struct folio **foliop,
 		size_t *poffset, u64 *plen)
 {
-	const struct iomap_folio_ops *folio_ops = iter->iomap.folio_ops;
 	const struct iomap *srcmap = iomap_iter_srcmap(iter);
 	loff_t pos = iter->pos;
 	u64 len = min_t(u64, SIZE_MAX, iomap_length(iter));
···
 	if (fatal_signal_pending(current))
 		return -EINTR;
 
-	folio = __iomap_get_folio(iter, len);
+	folio = __iomap_get_folio(iter, write_ops, len);
 	if (IS_ERR(folio))
 		return PTR_ERR(folio);
 
···
 	 * could do the wrong thing here (zero a page range incorrectly or fail
 	 * to zero) and corrupt data.
 	 */
-	if (folio_ops && folio_ops->iomap_valid) {
-		bool iomap_valid = folio_ops->iomap_valid(iter->inode,
+	if (write_ops && write_ops->iomap_valid) {
+		bool iomap_valid = write_ops->iomap_valid(iter->inode,
 							 &iter->iomap);
 		if (!iomap_valid) {
 			iter->iomap.flags |= IOMAP_F_STALE;
···
 	return 0;
 
 out_unlock:
-	__iomap_put_folio(iter, 0, folio);
-
+	__iomap_put_folio(iter, write_ops, 0, folio);
 	return status;
 }
 
···
 	return __iomap_write_end(iter->inode, pos, len, copied, folio);
 }
 
-static int iomap_write_iter(struct iomap_iter *iter, struct iov_iter *i)
+static int iomap_write_iter(struct iomap_iter *iter, struct iov_iter *i,
+		const struct iomap_write_ops *write_ops)
 {
 	ssize_t total_written = 0;
 	int status = 0;
···
 			break;
 		}
 
-		status = iomap_write_begin(iter, &folio, &offset, &bytes);
+		status = iomap_write_begin(iter, write_ops, &folio, &offset,
+				&bytes);
 		if (unlikely(status)) {
 			iomap_write_failed(iter->inode, iter->pos, bytes);
 			break;
···
 			i_size_write(iter->inode, pos + written);
 			iter->iomap.flags |= IOMAP_F_SIZE_CHANGED;
 		}
-		__iomap_put_folio(iter, written, folio);
+		__iomap_put_folio(iter, write_ops, written, folio);
 
 		if (old_size < pos)
 			pagecache_isize_extended(iter->inode, old_size, pos);
···
 
 ssize_t
 iomap_file_buffered_write(struct kiocb *iocb, struct iov_iter *i,
-		const struct iomap_ops *ops, void *private)
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private)
 {
 	struct iomap_iter iter = {
 		.inode		= iocb->ki_filp->f_mapping->host,
···
 		iter.flags |= IOMAP_DONTCACHE;
 
 	while ((ret = iomap_iter(&iter, ops)) > 0)
-		iter.status = iomap_write_iter(&iter, i);
+		iter.status = iomap_write_iter(&iter, i, write_ops);
 
 	if (unlikely(iter.pos == iocb->ki_pos))
 		return ret;
···
 }
 EXPORT_SYMBOL_GPL(iomap_write_delalloc_release);
 
-static int iomap_unshare_iter(struct iomap_iter *iter)
+static int iomap_unshare_iter(struct iomap_iter *iter,
+		const struct iomap_write_ops *write_ops)
 {
 	struct iomap *iomap = &iter->iomap;
 	u64 bytes = iomap_length(iter);
···
 		bool ret;
 
 		bytes = min_t(u64, SIZE_MAX, bytes);
-		status = iomap_write_begin(iter, &folio, &offset, &bytes);
+		status = iomap_write_begin(iter, write_ops, &folio, &offset,
+				&bytes);
 		if (unlikely(status))
 			return status;
 		if (iomap->flags & IOMAP_F_STALE)
 			break;
 
 		ret = iomap_write_end(iter, bytes, bytes, folio);
-		__iomap_put_folio(iter, bytes, folio);
+		__iomap_put_folio(iter, write_ops, bytes, folio);
 		if (WARN_ON_ONCE(!ret))
 			return -EIO;
 
···
 
 int
 iomap_file_unshare(struct inode *inode, loff_t pos, loff_t len,
-		const struct iomap_ops *ops)
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops)
 {
 	struct iomap_iter iter = {
 		.inode		= inode,
···
 
 	iter.len = min(len, size - pos);
 	while ((ret = iomap_iter(&iter, ops)) > 0)
-		iter.status = iomap_unshare_iter(&iter);
+		iter.status = iomap_unshare_iter(&iter, write_ops);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(iomap_file_unshare);
···
 	return filemap_write_and_wait_range(mapping, i->pos, end);
 }
 
-static int iomap_zero_iter(struct iomap_iter *iter, bool *did_zero)
+static int iomap_zero_iter(struct iomap_iter *iter, bool *did_zero,
+		const struct iomap_write_ops *write_ops)
 {
 	u64 bytes = iomap_length(iter);
 	int status;
···
 		bool ret;
 
 		bytes = min_t(u64, SIZE_MAX, bytes);
-		status = iomap_write_begin(iter, &folio, &offset, &bytes);
+		status = iomap_write_begin(iter, write_ops, &folio, &offset,
+				&bytes);
 		if (status)
 			return status;
 		if (iter->iomap.flags & IOMAP_F_STALE)
···
 		folio_mark_accessed(folio);
 
 		ret = iomap_write_end(iter, bytes, bytes, folio);
-		__iomap_put_folio(iter, bytes, folio);
+		__iomap_put_folio(iter, write_ops, bytes, folio);
 		if (WARN_ON_ONCE(!ret))
 			return -EIO;
 
···
 
 int
 iomap_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
-		const struct iomap_ops *ops, void *private)
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private)
 {
 	struct iomap_iter iter = {
 		.inode		= inode,
···
 	    filemap_range_needs_writeback(mapping, pos, pos + plen - 1)) {
 		iter.len = plen;
 		while ((ret = iomap_iter(&iter, ops)) > 0)
-			iter.status = iomap_zero_iter(&iter, did_zero);
+			iter.status = iomap_zero_iter(&iter, did_zero,
+					write_ops);
 
 		iter.len = len - (iter.pos - pos);
 		if (ret || !iter.len)
···
 			continue;
 		}
 
-		iter.status = iomap_zero_iter(&iter, did_zero);
+		iter.status = iomap_zero_iter(&iter, did_zero, write_ops);
 	}
 	return ret;
 }
···
 
 int
 iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops, void *private)
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private)
 {
 	unsigned int blocksize = i_blocksize(inode);
 	unsigned int off = pos & (blocksize - 1);
···
 	if (!off)
 		return 0;
 	return iomap_zero_range(inode, pos, blocksize - off, did_zero, ops,
-			private);
+			write_ops, private);
 }
 EXPORT_SYMBOL_GPL(iomap_truncate_page);
 

fs/xfs/xfs_file.c

···
 
 	trace_xfs_file_buffered_write(iocb, from);
 	ret = iomap_file_buffered_write(iocb, from,
-			&xfs_buffered_write_iomap_ops, NULL);
+			&xfs_buffered_write_iomap_ops, &xfs_iomap_write_ops,
+			NULL);
 
 	/*
 	 * If we hit a space limit, try to free up some lingering preallocated
···
 retry:
 	trace_xfs_file_buffered_write(iocb, from);
 	ret = iomap_file_buffered_write(iocb, from,
-			&xfs_buffered_write_iomap_ops, &ac);
+			&xfs_buffered_write_iomap_ops, &xfs_iomap_write_ops,
+			&ac);
 	if (ret == -ENOSPC && !cleared_space) {
 		/*
 		 * Kick off writeback to convert delalloc space and release the

fs/xfs/xfs_iomap.c

···
 {
 	struct xfs_inode	*ip = XFS_I(inode);
 
+	if (iomap->type == IOMAP_HOLE)
+		return true;
+
 	if (iomap->validity_cookie !=
 			xfs_iomap_inode_sequence(ip, iomap->flags)) {
 		trace_xfs_iomap_invalid(ip, iomap);
···
 	return true;
 }
 
-static const struct iomap_folio_ops xfs_iomap_folio_ops = {
+const struct iomap_write_ops xfs_iomap_write_ops = {
 	.iomap_valid		= xfs_iomap_valid,
 };
 
···
 		iomap->flags |= IOMAP_F_DIRTY;
 
 	iomap->validity_cookie = sequence_cookie;
-	iomap->folio_ops = &xfs_iomap_folio_ops;
 	return 0;
 }
 
···
 		return dax_zero_range(inode, pos, len, did_zero,
 				      &xfs_dax_write_iomap_ops);
 	return iomap_zero_range(inode, pos, len, did_zero,
-				&xfs_buffered_write_iomap_ops, ac);
+			&xfs_buffered_write_iomap_ops, &xfs_iomap_write_ops,
+			ac);
 }
 
 int
···
 		return dax_truncate_page(inode, pos, did_zero,
 					&xfs_dax_write_iomap_ops);
 	return iomap_truncate_page(inode, pos, did_zero,
-				   &xfs_buffered_write_iomap_ops, ac);
+			&xfs_buffered_write_iomap_ops, &xfs_iomap_write_ops,
+			ac);
 }

fs/xfs/xfs_iomap.h

···
 extern const struct iomap_ops xfs_xattr_iomap_ops;
 extern const struct iomap_ops xfs_dax_write_iomap_ops;
 extern const struct iomap_ops xfs_atomic_write_cow_iomap_ops;
+extern const struct iomap_write_ops xfs_iomap_write_ops;
 
 #endif /* __XFS_IOMAP_H__*/

fs/xfs/xfs_reflink.c

···
 				&xfs_dax_write_iomap_ops);
 	else
 		error = iomap_file_unshare(inode, offset, len,
-				&xfs_buffered_write_iomap_ops);
+				&xfs_buffered_write_iomap_ops,
+				&xfs_iomap_write_ops);
 	if (error)
 		goto out;
 

fs/zonefs/file.c

···
 	if (ret <= 0)
 		goto inode_unlock;
 
-	ret = iomap_file_buffered_write(iocb, from, &zonefs_write_iomap_ops, NULL);
+	ret = iomap_file_buffered_write(iocb, from, &zonefs_write_iomap_ops,
+			NULL, NULL);
 	if (ret == -EIO)
 		zonefs_io_error(inode, true);
 

include/linux/iomap.h

···
  */
 #define IOMAP_NULL_ADDR -1ULL	/* addr is not valid */
 
-struct iomap_folio_ops;
-
 struct iomap {
 	u64			addr; /* disk offset of mapping, bytes */
 	loff_t			offset;	/* file offset of mapping, bytes */
···
 	struct dax_device	*dax_dev; /* dax_dev for dax operations */
 	void			*inline_data;
 	void			*private; /* filesystem private */
-	const struct iomap_folio_ops *folio_ops;
 	u64			validity_cookie; /* used with .iomap_valid() */
 };
 
···
 }
 
 /*
- * When a filesystem sets folio_ops in an iomap mapping it returns, get_folio
- * and put_folio will be called for each folio written to.  This only applies
- * to buffered writes as unbuffered writes will not typically have folios
- * associated with them.
- *
  * When get_folio succeeds, put_folio will always be called to do any
  * cleanup work necessary.  put_folio is responsible for unlocking and putting
  * @folio.
  */
-struct iomap_folio_ops {
+struct iomap_write_ops {
 	struct folio *(*get_folio)(struct iomap_iter *iter, loff_t pos,
 			unsigned len);
 	void (*put_folio)(struct inode *inode, loff_t pos, unsigned copied,
···
 }
 
 ssize_t iomap_file_buffered_write(struct kiocb *iocb, struct iov_iter *from,
-		const struct iomap_ops *ops, void *private);
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private);
 int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops);
 void iomap_readahead(struct readahead_control *, const struct iomap_ops *ops);
 bool iomap_is_partially_uptodate(struct folio *, size_t from, size_t count);
···
 void iomap_invalidate_folio(struct folio *folio, size_t offset, size_t len);
 bool iomap_dirty_folio(struct address_space *mapping, struct folio *folio);
 int iomap_file_unshare(struct inode *inode, loff_t pos, loff_t len,
-		const struct iomap_ops *ops);
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops);
 int iomap_zero_range(struct inode *inode, loff_t pos, loff_t len,
-		bool *did_zero, const struct iomap_ops *ops, void *private);
+		bool *did_zero, const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private);
 int iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops, void *private);
+		const struct iomap_ops *ops,
+		const struct iomap_write_ops *write_ops, void *private);
 vm_fault_t iomap_page_mkwrite(struct vm_fault *vmf, const struct iomap_ops *ops,
 		void *private);
 typedef void (*iomap_punch_t)(struct inode *inode, loff_t offset, loff_t length,