ext4: avoid potential hang in mpage_submit_io() when blocksize < pagesize
Yongqiang Yang [Wed, 14 Dec 2011 02:51:55 +0000 (21:51 -0500)]
commit 13a79a4741d37fda2fbafb953f0f301dc007928f upstream.

If there is an unwritten but clean buffer in a page and there is a
dirty buffer after the buffer, then mpage_submit_io does not write the
dirty buffer out.  As a result, da_writepages loops forever.

This patch fixes the problem by checking dirty flag.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

fs/ext4/inode.c

index d33c737..5dbdb6b 100644 (file)
@@ -1261,8 +1261,11 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd,
                                        clear_buffer_unwritten(bh);
                                }
 
-                               /* skip page if block allocation undone */
-                               if (buffer_delay(bh) || buffer_unwritten(bh))
+                               /*
+                                * skip page if block allocation undone and
+                                * block is dirty
+                                */
+                               if (ext4_bh_delay_or_unwritten(NULL, bh))
                                        skip_page = 1;
                                bh = bh->b_this_page;
                                block_start += bh->b_size;