2010-01-19 21:21:53

by Eric Sandeen

Subject: [PATCH] e2fck: don't complain about i_size for known blocks past EOF

This is the userspace side of Jiaying's EOFBLOCKS patch, I think.
With Aneesh's patches for .33, Jiaying's patch, and this one,
xfstests 013/fsstress (even with direct IO enabled) has held up
through many runs.

Signed-off-by: Eric Sandeen <[email protected]>
---

diff --git a/e2fsck/pass1.c b/e2fsck/pass1.c
index 09ff1d0..0a2c6cb 100644
--- a/e2fsck/pass1.c
+++ b/e2fsck/pass1.c
@@ -1947,7 +1947,8 @@ static void check_blocks(e2fsck_t ctx, struct problem_context *pctx,
/* allow allocated blocks to end of PAGE_SIZE */
(size < (__u64)pb.last_block * fs->blocksize) &&
(pb.last_block / blkpg * blkpg != pb.last_block ||
- size < (__u64)(pb.last_block & ~(blkpg-1)) *fs->blocksize))
+ size < (__u64)(pb.last_block & ~(blkpg-1)) *fs->blocksize) &&
+ !(inode->i_flags & EXT4_EOFBLOCKS_FL))
bad_size = 3;
else if (!(extent_fs && (inode->i_flags & EXT4_EXTENTS_FL)) &&
size > ext2_max_sizes[fs->super->s_log_block_size])
diff --git a/lib/ext2fs/ext2_fs.h b/lib/ext2fs/ext2_fs.h
index 114b001..6e31c0d 100644
--- a/lib/ext2fs/ext2_fs.h
+++ b/lib/ext2fs/ext2_fs.h
@@ -276,10 +276,11 @@ struct ext2_dx_countlimit {
#define EXT2_TOPDIR_FL 0x00020000 /* Top of directory hierarchies*/
#define EXT4_HUGE_FILE_FL 0x00040000 /* Set to each huge file */
#define EXT4_EXTENTS_FL 0x00080000 /* Inode uses extents */
+#define EXT4_EOFBLOCKS_FL 0x00400000 /* Blocks allocated beyond EOF */
#define EXT2_RESERVED_FL 0x80000000 /* reserved for ext2 lib */

-#define EXT2_FL_USER_VISIBLE 0x000BDFFF /* User visible flags */
-#define EXT2_FL_USER_MODIFIABLE 0x000080FF /* User modifiable flags */
+#define EXT2_FL_USER_VISIBLE 0x004BDFFF /* User visible flags */
+#define EXT2_FL_USER_MODIFIABLE 0x004B80FF /* User modifiable flags */

/*
* ioctl commands

2010-02-24 16:25:34

by Theodore Ts'o

Subject: Re: [PATCH] e2fck: don't complain about i_size for known blocks past EOF

On Tue, Jan 19, 2010 at 03:21:50PM -0600, Eric Sandeen wrote:
> This is the userspace side of Jiaying's EOFBLOCKS patch, I think.
> With Aneesh's patches for .33, Jiaying's patch, and this one,
> xfstests 013/fsstress (even with direct IO enabled) has held up
> through many runs.
>
> Signed-off-by: Eric Sandeen <[email protected]>

Applied into the maint branch.

- Ted