1 files changed, 90 insertions, 77 deletions

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 5b0d2c7d5408..8a064734e6eb 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -148,6 +148,9 @@ static int ext4_inode_is_fast_symlink(struct inode *inode)
         int ea_blocks = EXT4_I(inode)->i_file_acl ?
 		EXT4_CLUSTER_SIZE(inode->i_sb) >> 9 : 0;
 
+	if (ext4_has_inline_data(inode))
+		return 0;
+
 	return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0);
 }
 
@@ -443,7 +446,7 @@ static void ext4_map_blocks_es_recheck(handle_t *handle,
 	 * could be converted.
 	 */
 	if (!(flags & EXT4_GET_BLOCKS_NO_LOCK))
-		down_read((&EXT4_I(inode)->i_data_sem));
+		down_read(&EXT4_I(inode)->i_data_sem);
 	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		retval = ext4_ext_map_blocks(handle, inode, map, flags &
 					     EXT4_GET_BLOCKS_KEEP_SIZE);
@@ -489,8 +492,8 @@ static void ext4_map_blocks_es_recheck(handle_t *handle,
  * Otherwise, call with ext4_ind_map_blocks() to handle indirect mapping
  * based files
  *
- * On success, it returns the number of blocks being mapped or allocate.
- * if create==0 and the blocks are pre-allocated and uninitialized block,
+ * On success, it returns the number of blocks being mapped or allocated.
+ * if create==0 and the blocks are pre-allocated and unwritten block,
  * the result buffer head is unmapped. If the create ==1, it will make sure
  * the buffer head is mapped.
  *
@@ -522,6 +525,10 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
 	if (unlikely(map->m_len > INT_MAX))
 		map->m_len = INT_MAX;
 
+	/* We can handle the block number less than EXT_MAX_BLOCKS */
+	if (unlikely(map->m_lblk >= EXT_MAX_BLOCKS))
+		return -EIO;
+
 	/* Lookup extent status tree firstly */
 	if (ext4_es_lookup_extent(inode, map->m_lblk, &es)) {
 		ext4_es_lru_add(inode);
@@ -551,7 +558,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
 	 * file system block.
 	 */
 	if (!(flags & EXT4_GET_BLOCKS_NO_LOCK))
-		down_read((&EXT4_I(inode)->i_data_sem));
+		down_read(&EXT4_I(inode)->i_data_sem);
 	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		retval = ext4_ext_map_blocks(handle, inode, map, flags &
 					     EXT4_GET_BLOCKS_KEEP_SIZE);
@@ -618,12 +625,12 @@ found:
 	map->m_flags &= ~EXT4_MAP_FLAGS;
 
 	/*
-	 * New blocks allocate and/or writing to uninitialized extent
+	 * New blocks allocate and/or writing to unwritten extent
 	 * will possibly result in updating i_data, so we take
 	 * the write lock of i_data_sem, and call get_blocks()
 	 * with create == 1 flag.
 	 */
-	down_write((&EXT4_I(inode)->i_data_sem));
+	down_write(&EXT4_I(inode)->i_data_sem);
 
 	/*
 	 * if the caller is from delayed allocation writeout path
@@ -918,6 +925,7 @@ int do_journal_get_write_access(handle_t *handle,
 	 */
 	if (dirty)
 		clear_buffer_dirty(bh);
+	BUFFER_TRACE(bh, "get write access");
 	ret = ext4_journal_get_write_access(handle, bh);
 	if (!ret && dirty)
 		ret = ext4_handle_dirty_metadata(handle, NULL, bh);
@@ -1536,7 +1544,7 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock,
 		ext4_es_lru_add(inode);
 		if (ext4_es_is_hole(&es)) {
 			retval = 0;
-			down_read((&EXT4_I(inode)->i_data_sem));
+			down_read(&EXT4_I(inode)->i_data_sem);
 			goto add_delayed;
 		}
 
@@ -1573,7 +1581,7 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock,
 	 * Try to see if we can get the block without requesting a new
 	 * file system block.
 	 */
-	down_read((&EXT4_I(inode)->i_data_sem));
+	down_read(&EXT4_I(inode)->i_data_sem);
 	if (ext4_has_inline_data(inode)) {
 		/*
 		 * We will soon create blocks for this page, and let
@@ -1765,6 +1773,7 @@ static int __ext4_journalled_writepage(struct page *page,
 	BUG_ON(!ext4_handle_valid(handle));
 
 	if (inline_data) {
+		BUFFER_TRACE(inode_bh, "get write access");
 		ret = ext4_journal_get_write_access(handle, inode_bh);
 
 		err = ext4_handle_dirty_metadata(handle, inode, inode_bh);
@@ -1842,6 +1851,7 @@ static int ext4_writepage(struct page *page,
 	struct buffer_head *page_bufs = NULL;
 	struct inode *inode = page->mapping->host;
 	struct ext4_io_submit io_submit;
+	bool keep_towrite = false;
 
 	trace_ext4_writepage(page);
 	size = i_size_read(inode);
@@ -1872,6 +1882,7 @@ static int ext4_writepage(struct page *page,
 			unlock_page(page);
 			return 0;
 		}
+		keep_towrite = true;
 	}
 
 	if (PageChecked(page) && ext4_should_journal_data(inode))
@@ -1888,7 +1899,7 @@ static int ext4_writepage(struct page *page,
 		unlock_page(page);
 		return -ENOMEM;
 	}
-	ret = ext4_bio_write_page(&io_submit, page, len, wbc);
+	ret = ext4_bio_write_page(&io_submit, page, len, wbc, keep_towrite);
 	ext4_io_submit(&io_submit);
 	/* Drop io_end reference we got from init */
 	ext4_put_io_end_defer(io_submit.io_end);
@@ -1907,7 +1918,7 @@ static int mpage_submit_page(struct mpage_da_data *mpd, struct page *page)
 	else
 		len = PAGE_CACHE_SIZE;
 	clear_page_dirty_for_io(page);
-	err = ext4_bio_write_page(&mpd->io_submit, page, len, mpd->wbc);
+	err = ext4_bio_write_page(&mpd->io_submit, page, len, mpd->wbc, false);
 	if (!err)
 		mpd->wbc->nr_to_write--;
 	mpd->first_page++;
@@ -2028,7 +2039,7 @@ static int mpage_process_page_bufs(struct mpage_da_data *mpd,
  * Scan buffers corresponding to changed extent (we expect corresponding pages
  * to be already locked) and update buffer state according to new extent state.
  * We map delalloc buffers to their physical location, clear unwritten bits,
- * and mark buffers as uninit when we perform writes to uninitialized extents
+ * and mark buffers as uninit when we perform writes to unwritten extents
  * and do extent conversion after IO is finished. If the last page is not fully
  * mapped, we update @map to the next extent in the last page that needs
  * mapping. Otherwise we submit the page for IO.
@@ -2122,12 +2133,12 @@ static int mpage_map_one_extent(handle_t *handle, struct mpage_da_data *mpd)
 	struct inode *inode = mpd->inode;
 	struct ext4_map_blocks *map = &mpd->map;
 	int get_blocks_flags;
-	int err;
+	int err, dioread_nolock;
 
 	trace_ext4_da_write_pages_extent(inode, map);
 	/*
 	 * Call ext4_map_blocks() to allocate any delayed allocation blocks, or
-	 * to convert an uninitialized extent to be initialized (in the case
+	 * to convert an unwritten extent to be initialized (in the case
 	 * where we have written into one or more preallocated blocks).  It is
 	 * possible that we're going to need more metadata blocks than
 	 * previously reserved. However we must not fail because we're in
@@ -2144,7 +2155,8 @@ static int mpage_map_one_extent(handle_t *handle, struct mpage_da_data *mpd)
 	 */
 	get_blocks_flags = EXT4_GET_BLOCKS_CREATE |
 			   EXT4_GET_BLOCKS_METADATA_NOFAIL;
-	if (ext4_should_dioread_nolock(inode))
+	dioread_nolock = ext4_should_dioread_nolock(inode);
+	if (dioread_nolock)
 		get_blocks_flags |= EXT4_GET_BLOCKS_IO_CREATE_EXT;
 	if (map->m_flags & (1 << BH_Delay))
 		get_blocks_flags |= EXT4_GET_BLOCKS_DELALLOC_RESERVE;
@@ -2152,7 +2164,7 @@ static int mpage_map_one_extent(handle_t *handle, struct mpage_da_data *mpd)
 	err = ext4_map_blocks(handle, inode, map, get_blocks_flags);
 	if (err < 0)
 		return err;
-	if (map->m_flags & EXT4_MAP_UNINIT) {
+	if (dioread_nolock && (map->m_flags & EXT4_MAP_UNWRITTEN)) {
 		if (!mpd->io_submit.io_end->handle &&
 		    ext4_handle_valid(handle)) {
 			mpd->io_submit.io_end->handle = handle->h_rsv_handle;
@@ -2243,13 +2255,23 @@ static int mpage_map_and_submit_extent(handle_t *handle,
 			return err;
 	} while (map->m_len);
 
-	/* Update on-disk size after IO is submitted */
+	/*
+	 * Update on-disk size after IO is submitted.  Races with
+	 * truncate are avoided by checking i_size under i_data_sem.
+	 */
 	disksize = ((loff_t)mpd->first_page) << PAGE_CACHE_SHIFT;
 	if (disksize > EXT4_I(inode)->i_disksize) {
 		int err2;
-
-		ext4_wb_update_i_disksize(inode, disksize);
+		loff_t i_size;
+
+		down_write(&EXT4_I(inode)->i_data_sem);
+		i_size = i_size_read(inode);
+		if (disksize > i_size)
+			disksize = i_size;
+		if (disksize > EXT4_I(inode)->i_disksize)
+			EXT4_I(inode)->i_disksize = disksize;
 		err2 = ext4_mark_inode_dirty(handle, inode);
+		up_write(&EXT4_I(inode)->i_data_sem);
 		if (err2)
 			ext4_error(inode->i_sb,
 				   "Failed to mark inode %lu dirty",
@@ -3056,9 +3078,9 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
  * preallocated extents, and those write extend the file, no need to
  * fall back to buffered IO.
  *
- * For holes, we fallocate those blocks, mark them as uninitialized
+ * For holes, we fallocate those blocks, mark them as unwritten
  * If those blocks were preallocated, we mark sure they are split, but
- * still keep the range to write as uninitialized.
+ * still keep the range to write as unwritten.
  *
  * The unwritten extents will be converted to written when DIO is completed.
  * For async direct IO, since the IO may still pending when return, we
@@ -3071,13 +3093,12 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
  *
  */
 static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
-			      const struct iovec *iov, loff_t offset,
-			      unsigned long nr_segs)
+			      struct iov_iter *iter, loff_t offset)
 {
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
-	size_t count = iov_length(iov, nr_segs);
+	size_t count = iov_iter_count(iter);
 	int overwrite = 0;
 	get_block_t *get_block_func = NULL;
 	int dio_flags = 0;
@@ -3086,7 +3107,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 
 	/* Use the old path for reads and writes beyond i_size. */
 	if (rw != WRITE || final_size > inode->i_size)
-		return ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs);
+		return ext4_ind_direct_IO(rw, iocb, iter, offset);
 
 	BUG_ON(iocb->private == NULL);
 
@@ -3110,12 +3131,12 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 	 * We could direct write to holes and fallocate.
 	 *
 	 * Allocated blocks to fill the hole are marked as
-	 * uninitialized to prevent parallel buffered read to expose
+	 * unwritten to prevent parallel buffered read to expose
 	 * the stale data before DIO complete the data IO.
 	 *
 	 * As to previously fallocated extents, ext4 get_block will
 	 * just simply mark the buffer mapped but still keep the
-	 * extents uninitialized.
+	 * extents unwritten.
 	 *
 	 * For non AIO case, we will convert those unwritten extents
 	 * to written after return back from blockdev_direct_IO.
@@ -3153,8 +3174,8 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		dio_flags = DIO_LOCKING;
 	}
 	ret = __blockdev_direct_IO(rw, iocb, inode,
-				   inode->i_sb->s_bdev, iov,
-				   offset, nr_segs,
+				   inode->i_sb->s_bdev, iter,
+				   offset,
 				   get_block_func,
 				   ext4_end_io_dio,
 				   NULL,
@@ -3208,11 +3229,11 @@ retake_lock:
 }
 
 static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
-			      const struct iovec *iov, loff_t offset,
-			      unsigned long nr_segs)
+			      struct iov_iter *iter, loff_t offset)
 {
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
+	size_t count = iov_iter_count(iter);
 	ssize_t ret;
 
 	/*
@@ -3225,13 +3246,12 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
 	if (ext4_has_inline_data(inode))
 		return 0;
 
-	trace_ext4_direct_IO_enter(inode, offset, iov_length(iov, nr_segs), rw);
+	trace_ext4_direct_IO_enter(inode, offset, count, rw);
 	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
-		ret = ext4_ext_direct_IO(rw, iocb, iov, offset, nr_segs);
+		ret = ext4_ext_direct_IO(rw, iocb, iter, offset);
 	else
-		ret = ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs);
-	trace_ext4_direct_IO_exit(inode, offset,
-				iov_length(iov, nr_segs), rw, ret);
+		ret = ext4_ind_direct_IO(rw, iocb, iter, offset);
+	trace_ext4_direct_IO_exit(inode, offset, count, rw, ret);
 	return ret;
 }
 
@@ -3426,7 +3446,7 @@ unlock:
  * This required during truncate. We need to physically zero the tail end
  * of that block so it doesn't yield old data if the file is later grown.
  */
-int ext4_block_truncate_page(handle_t *handle,
+static int ext4_block_truncate_page(handle_t *handle,
 		struct address_space *mapping, loff_t from)
 {
 	unsigned offset = from & (PAGE_CACHE_SIZE-1);
@@ -3527,15 +3547,6 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length)
 	}
 
 	mutex_lock(&inode->i_mutex);
-	/* It's not possible punch hole on append only file */
-	if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) {
-		ret = -EPERM;
-		goto out_mutex;
-	}
-	if (IS_SWAPFILE(inode)) {
-		ret = -ETXTBSY;
-		goto out_mutex;
-	}
 
 	/* No need to punch hole beyond i_size */
 	if (offset >= inode->i_size)
@@ -3616,7 +3627,6 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length)
 		ret = ext4_free_hole_blocks(handle, inode, first_block,
 					    stop_block);
 
-	ext4_discard_preallocations(inode);
 	up_write(&EXT4_I(inode)->i_data_sem);
 	if (IS_SYNC(inode))
 		ext4_handle_sync(handle);
@@ -4300,12 +4310,15 @@ static int ext4_do_update_inode(handle_t *handle,
 	struct ext4_inode *raw_inode = ext4_raw_inode(iloc);
 	struct ext4_inode_info *ei = EXT4_I(inode);
 	struct buffer_head *bh = iloc->bh;
+	struct super_block *sb = inode->i_sb;
 	int err = 0, rc, block;
-	int need_datasync = 0;
+	int need_datasync = 0, set_large_file = 0;
 	uid_t i_uid;
 	gid_t i_gid;
 
-	/* For fields not not tracking in the in-memory inode,
+	spin_lock(&ei->i_raw_lock);
+
+	/* For fields not tracked in the in-memory inode,
 	 * initialise them to zero for new inodes. */
 	if (ext4_test_inode_state(inode, EXT4_STATE_NEW))
 		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
@@ -4343,8 +4356,10 @@ static int ext4_do_update_inode(handle_t *handle,
 	EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
 	EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
 
-	if (ext4_inode_blocks_set(handle, raw_inode, ei))
+	if (ext4_inode_blocks_set(handle, raw_inode, ei)) {
+		spin_unlock(&ei->i_raw_lock);
 		goto out_brelse;
+	}
 	raw_inode->i_dtime = cpu_to_le32(ei->i_dtime);
 	raw_inode->i_flags = cpu_to_le32(ei->i_flags & 0xFFFFFFFF);
 	if (likely(!test_opt2(inode->i_sb, HURD_COMPAT)))
@@ -4356,24 +4371,11 @@ static int ext4_do_update_inode(handle_t *handle,
 		need_datasync = 1;
 	}
 	if (ei->i_disksize > 0x7fffffffULL) {
-		struct super_block *sb = inode->i_sb;
 		if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
 				EXT4_FEATURE_RO_COMPAT_LARGE_FILE) ||
 				EXT4_SB(sb)->s_es->s_rev_level ==
-				cpu_to_le32(EXT4_GOOD_OLD_REV)) {
-			/* If this is the first large file
-			 * created, add a flag to the superblock.
-			 */
-			err = ext4_journal_get_write_access(handle,
-					EXT4_SB(sb)->s_sbh);
-			if (err)
-				goto out_brelse;
-			ext4_update_dynamic_rev(sb);
-			EXT4_SET_RO_COMPAT_FEATURE(sb,
-					EXT4_FEATURE_RO_COMPAT_LARGE_FILE);
-			ext4_handle_sync(handle);
-			err = ext4_handle_dirty_super(handle, sb);
-		}
+		    cpu_to_le32(EXT4_GOOD_OLD_REV))
+			set_large_file = 1;
 	}
 	raw_inode->i_generation = cpu_to_le32(inode->i_generation);
 	if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) {
@@ -4405,12 +4407,24 @@ static int ext4_do_update_inode(handle_t *handle,
 
 	ext4_inode_csum_set(inode, raw_inode, ei);
 
+	spin_unlock(&ei->i_raw_lock);
+
 	BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
 	rc = ext4_handle_dirty_metadata(handle, NULL, bh);
 	if (!err)
 		err = rc;
 	ext4_clear_inode_state(inode, EXT4_STATE_NEW);
-
+	if (set_large_file) {
+		BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access");
+		err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh);
+		if (err)
+			goto out_brelse;
+		ext4_update_dynamic_rev(sb);
+		EXT4_SET_RO_COMPAT_FEATURE(sb,
+					   EXT4_FEATURE_RO_COMPAT_LARGE_FILE);
+		ext4_handle_sync(handle);
+		err = ext4_handle_dirty_super(handle, sb);
+	}
 	ext4_update_inode_fsync_trans(handle, inode, need_datasync);
 out_brelse:
 	brelse(bh);
@@ -4423,21 +4437,20 @@ out_brelse:
  *
  * We are called from a few places:
  *
- * - Within generic_file_write() for O_SYNC files.
+ * - Within generic_file_aio_write() -> generic_write_sync() for O_SYNC files.
  *   Here, there will be no transaction running. We wait for any running
  *   transaction to commit.
  *
- * - Within sys_sync(), kupdate and such.
- *   We wait on commit, if tol to.
+ * - Within flush work (sys_sync(), kupdate and such).
+ *   We wait on commit, if told to.
  *
- * - Within prune_icache() (PF_MEMALLOC == true)
- *   Here we simply return.  We can't afford to block kswapd on the
- *   journal commit.
+ * - Within iput_final() -> write_inode_now()
+ *   We wait on commit, if told to.
  *
  * In all cases it is actually safe for us to return without doing anything,
  * because the inode has been copied into a raw inode buffer in
- * ext4_mark_inode_dirty().  This is a correctness thing for O_SYNC and for
- * knfsd.
+ * ext4_mark_inode_dirty().  This is a correctness thing for WB_SYNC_ALL
+ * writeback.
  *
  * Note that we are absolutely dependent upon all inode dirtiers doing the
  * right thing: they *must* call mark_inode_dirty() after dirtying info in
@@ -4449,15 +4462,15 @@ out_brelse:
  *	stuff();
  *	inode->i_size = expr;
  *
- * is in error because a kswapd-driven write_inode() could occur while
- * `stuff()' is running, and the new i_size will be lost.  Plus the inode
- * will no longer be on the superblock's dirty inode list.
+ * is in error because write_inode() could occur while `stuff()' is running,
+ * and the new i_size will be lost.  Plus the inode will no longer be on the
+ * superblock's dirty inode list.
  */
 int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int err;
 
-	if (current->flags & PF_MEMALLOC)
+	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC))
 		return 0;
 
 	if (EXT4_SB(inode->i_sb)->s_journal) {