ext4: add normal write support for inline data

For a normal write case (not journalled write, not delayed
allocation), we write to the inline if the file is small and convert
it to an extent based file when the write is larger than the max
inline size.

Signed-off-by: Tao Ma <[email protected]>
Signed-off-by: "Theodore Ts'o" <[email protected]>

fs/ext4/ext4.h

@@ -2018,8 +2018,19 @@ struct buffer_head *ext4_getblk(handle_t *, struct inode *,
 						ext4_lblk_t, int, int *);
 struct buffer_head *ext4_bread(handle_t *, struct inode *,
 						ext4_lblk_t, int, int *);
+int ext4_get_block_write(struct inode *inode, sector_t iblock,
+			 struct buffer_head *bh_result, int create);
 int ext4_get_block(struct inode *inode, sector_t iblock,
 				struct buffer_head *bh_result, int create);
+int ext4_walk_page_buffers(handle_t *handle,
+			   struct buffer_head *head,
+			   unsigned from,
+			   unsigned to,
+			   int *partial,
+			   int (*fn)(handle_t *handle,
+				     struct buffer_head *bh));
+int do_journal_get_write_access(handle_t *handle,
+				struct buffer_head *bh);
 
 extern struct inode *ext4_iget(struct super_block *, unsigned long);
 extern int  ext4_write_inode(struct inode *, struct writeback_control *);

fs/ext4/extents.c

@@ -42,6 +42,7 @@
 #include <linux/fiemap.h>
 #include "ext4_jbd2.h"
 #include "ext4_extents.h"
+#include "xattr.h"
 
 #include <trace/events/ext4.h>
 
@@ -2310,7 +2311,13 @@ int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
 int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
 {
 	int index;
-	int depth = ext_depth(inode);
+	int depth;
+
+	/* If we are converting the inline data, only one is needed here. */
+	if (ext4_has_inline_data(inode))
+		return 1;
+
+	depth = ext_depth(inode);
 
 	if (chunk)
 		index = depth * 2;

fs/ext4/inline.c

@@ -14,6 +14,7 @@
 #include "ext4_jbd2.h"
 #include "ext4.h"
 #include "xattr.h"
+#include "truncate.h"
 
 #define EXT4_XATTR_SYSTEM_DATA	"data"
 #define EXT4_MIN_INLINE_DATA_SIZE	((sizeof(__le32) * EXT4_N_BLOCKS))
@@ -515,6 +516,238 @@ int ext4_readpage_inline(struct inode *inode, struct page *page)
 	return ret >= 0 ? 0 : ret;
 }
 
+static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
+					      struct inode *inode,
+					      unsigned flags)
+{
+	int ret, needed_blocks;
+	handle_t *handle = NULL;
+	int retries = 0, sem_held = 0;
+	struct page *page = NULL;
+	unsigned from, to;
+	struct ext4_iloc iloc;
+
+	if (!ext4_has_inline_data(inode)) {
+		/*
+		 * clear the flag so that no new write
+		 * will trap here again.
+		 */
+		ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA);
+		return 0;
+	}
+
+	needed_blocks = ext4_writepage_trans_blocks(inode);
+
+	ret = ext4_get_inode_loc(inode, &iloc);
+	if (ret)
+		return ret;
+
+retry:
+	handle = ext4_journal_start(inode, needed_blocks);
+	if (IS_ERR(handle)) {
+		ret = PTR_ERR(handle);
+		handle = NULL;
+		goto out;
+	}
+
+	/* We cannot recurse into the filesystem as the transaction is already
+	 * started */
+	flags |= AOP_FLAG_NOFS;
+
+	page = grab_cache_page_write_begin(mapping, 0, flags);
+	if (!page) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	down_write(&EXT4_I(inode)->xattr_sem);
+	sem_held = 1;
+	/* If some one has already done this for us, just exit. */
+	if (!ext4_has_inline_data(inode)) {
+		ret = 0;
+		goto out;
+	}
+
+	from = 0;
+	to = ext4_get_inline_size(inode);
+	if (!PageUptodate(page)) {
+		ret = ext4_read_inline_page(inode, page);
+		if (ret < 0)
+			goto out;
+	}
+
+	ret = ext4_destroy_inline_data_nolock(handle, inode);
+	if (ret)
+		goto out;
+
+	if (ext4_should_dioread_nolock(inode))
+		ret = __block_write_begin(page, from, to, ext4_get_block_write);
+	else
+		ret = __block_write_begin(page, from, to, ext4_get_block);
+
+	if (!ret && ext4_should_journal_data(inode)) {
+		ret = ext4_walk_page_buffers(handle, page_buffers(page),
+					     from, to, NULL,
+					     do_journal_get_write_access);
+	}
+
+	if (ret) {
+		unlock_page(page);
+		page_cache_release(page);
+		ext4_orphan_add(handle, inode);
+		up_write(&EXT4_I(inode)->xattr_sem);
+		sem_held = 0;
+		ext4_journal_stop(handle);
+		handle = NULL;
+		ext4_truncate_failed_write(inode);
+		/*
+		 * If truncate failed early the inode might
+		 * still be on the orphan list; we need to
+		 * make sure the inode is removed from the
+		 * orphan list in that case.
+		 */
+		if (inode->i_nlink)
+			ext4_orphan_del(NULL, inode);
+	}
+
+	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
+		goto retry;
+
+	block_commit_write(page, from, to);
+out:
+	if (page) {
+		unlock_page(page);
+		page_cache_release(page);
+	}
+	if (sem_held)
+		up_write(&EXT4_I(inode)->xattr_sem);
+	if (handle)
+		ext4_journal_stop(handle);
+	brelse(iloc.bh);
+	return ret;
+}
+
+/*
+ * Try to write data in the inode.
+ * If the inode has inline data, check whether the new write can be
+ * in the inode also. If not, create the page the handle, move the data
+ * to the page make it update and let the later codes create extent for it.
+ */
+int ext4_try_to_write_inline_data(struct address_space *mapping,
+				  struct inode *inode,
+				  loff_t pos, unsigned len,
+				  unsigned flags,
+				  struct page **pagep)
+{
+	int ret;
+	handle_t *handle;
+	struct page *page;
+	struct ext4_iloc iloc;
+
+	if (pos + len > ext4_get_max_inline_size(inode))
+		goto convert;
+
+	ret = ext4_get_inode_loc(inode, &iloc);
+	if (ret)
+		return ret;
+
+	/*
+	 * The possible write could happen in the inode,
+	 * so try to reserve the space in inode first.
+	 */
+	handle = ext4_journal_start(inode, 1);
+	if (IS_ERR(handle)) {
+		ret = PTR_ERR(handle);
+		handle = NULL;
+		goto out;
+	}
+
+	ret = ext4_prepare_inline_data(handle, inode, pos + len);
+	if (ret && ret != -ENOSPC)
+		goto out;
+
+	/* We don't have space in inline inode, so convert it to extent. */
+	if (ret == -ENOSPC) {
+		ext4_journal_stop(handle);
+		brelse(iloc.bh);
+		goto convert;
+	}
+
+	flags |= AOP_FLAG_NOFS;
+
+	page = grab_cache_page_write_begin(mapping, 0, flags);
+	if (!page) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	*pagep = page;
+	down_read(&EXT4_I(inode)->xattr_sem);
+	if (!ext4_has_inline_data(inode)) {
+		ret = 0;
+		unlock_page(page);
+		page_cache_release(page);
+		goto out_up_read;
+	}
+
+	if (!PageUptodate(page)) {
+		ret = ext4_read_inline_page(inode, page);
+		if (ret < 0)
+			goto out_up_read;
+	}
+
+	ret = 1;
+	handle = NULL;
+out_up_read:
+	up_read(&EXT4_I(inode)->xattr_sem);
+out:
+	if (handle)
+		ext4_journal_stop(handle);
+	brelse(iloc.bh);
+	return ret;
+convert:
+	return ext4_convert_inline_data_to_extent(mapping,
+						  inode, flags);
+}
+
+int ext4_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len,
+			       unsigned copied, struct page *page)
+{
+	int ret;
+	void *kaddr;
+	struct ext4_iloc iloc;
+
+	if (unlikely(copied < len)) {
+		if (!PageUptodate(page)) {
+			copied = 0;
+			goto out;
+		}
+	}
+
+	ret = ext4_get_inode_loc(inode, &iloc);
+	if (ret) {
+		ext4_std_error(inode->i_sb, ret);
+		copied = 0;
+		goto out;
+	}
+
+	down_write(&EXT4_I(inode)->xattr_sem);
+	BUG_ON(!ext4_has_inline_data(inode));
+
+	kaddr = kmap_atomic(page);
+	ext4_write_inline_data(inode, &iloc, kaddr, pos, len);
+	kunmap_atomic(kaddr);
+	SetPageUptodate(page);
+	/* clear page dirty so that writepages wouldn't work for us. */
+	ClearPageDirty(page);
+
+	up_write(&EXT4_I(inode)->xattr_sem);
+	brelse(iloc.bh);
+out:
+	return copied;
+}
+
+
 int ext4_destroy_inline_data(handle_t *handle, struct inode *inode)
 {
 	int ret;