2013-02-18 08:14:26

by Lukas Czerner

Subject: [PATCH] ext4: fix xattr block allocation/release with bigalloc

Currently when new xattr block is created or released we we would call
dquot_free_block() or dquot_alloc_block() respectively, among the else
decrementing or incrementing the number of blocks assigned to the inode
by one block.

This however does not work for bigalloc file system because we always
allocate/free the whole cluster so we have to count with that in
dquot_free_block() and dquot_alloc_block() as well.

Use the clusters-to-blocks conversion EXT4_C2B() when passing number of
blocks to the dquot_alloc/free functions to fix the problem.

The problem has been revealed by xfstests #117 (and possibly others).

Signed-off-by: Lukas Czerner <[email protected]>
---
fs/ext4/xattr.c | 7 ++++---
1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 3a91ebc..b6741be 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -549,7 +549,7 @@ ext4_xattr_release_block(handle_t *handle, struct inode *inode,
error = ext4_handle_dirty_xattr_block(handle, inode, bh);
if (IS_SYNC(inode))
ext4_handle_sync(handle);
- dquot_free_block(inode, 1);
+ dquot_free_block(inode, EXT4_C2B(EXT4_SB(inode->i_sb), 1));
ea_bdebug(bh, "refcount now=%d; releasing",
le32_to_cpu(BHDR(bh)->h_refcount));
}
@@ -832,7 +832,8 @@ inserted:
else {
/* The old block is released after updating
the inode. */
- error = dquot_alloc_block(inode, 1);
+ error = dquot_alloc_block(inode,
+ EXT4_C2B(EXT4_SB(sb), 1));
if (error)
goto cleanup;
error = ext4_journal_get_write_access(handle,
@@ -928,7 +929,7 @@ cleanup:
return error;

cleanup_dquot:
- dquot_free_block(inode, 1);
+ dquot_free_block(inode, EXT4_C2B(EXT4_SB(sb), 1));
goto cleanup;

bad_block:
--
1.7.7.6

2013-02-18 15:26:06

by Eric Sandeen

Subject: Re: [PATCH] ext4: fix xattr block allocation/release with bigalloc

On 2/18/13 2:14 AM, Lukas Czerner wrote:
> Currently when new xattr block is created or released we we would call
> dquot_free_block() or dquot_alloc_block() respectively, among the else
> decrementing or incrementing the number of blocks assigned to the inode
> by one block.
>
> This however does not work for bigalloc file system because we always
> allocate/free the whole cluster so we have to count with that in
> dquot_free_block() and dquot_alloc_block() as well.
>
> Use the clusters-to-blocks conversion EXT4_C2B() when passing number of
> blocks to the dquot_alloc/free functions to fix the problem.
>
> The problem has been revealed by xfstests #117 (and possibly others).
>
> Signed-off-by: Lukas Czerner <[email protected]>

Thanks for finding this!

Reviewed-by: Eric Sandeen <[email protected]>

> ---
> fs/ext4/xattr.c | 7 ++++---
> 1 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
> index 3a91ebc..b6741be 100644
> --- a/fs/ext4/xattr.c
> +++ b/fs/ext4/xattr.c
> @@ -549,7 +549,7 @@ ext4_xattr_release_block(handle_t *handle, struct inode *inode,
> error = ext4_handle_dirty_xattr_block(handle, inode, bh);
> if (IS_SYNC(inode))
> ext4_handle_sync(handle);
> - dquot_free_block(inode, 1);
> + dquot_free_block(inode, EXT4_C2B(EXT4_SB(inode->i_sb), 1));
> ea_bdebug(bh, "refcount now=%d; releasing",
> le32_to_cpu(BHDR(bh)->h_refcount));
> }
> @@ -832,7 +832,8 @@ inserted:
> else {
> /* The old block is released after updating
> the inode. */
> - error = dquot_alloc_block(inode, 1);
> + error = dquot_alloc_block(inode,
> + EXT4_C2B(EXT4_SB(sb), 1));
> if (error)
> goto cleanup;
> error = ext4_journal_get_write_access(handle,
> @@ -928,7 +929,7 @@ cleanup:
> return error;
>
> cleanup_dquot:
> - dquot_free_block(inode, 1);
> + dquot_free_block(inode, EXT4_C2B(EXT4_SB(sb), 1));
> goto cleanup;
>
> bad_block:
>

2013-02-18 17:25:49

by Theodore Ts'o

Subject: Re: [PATCH] ext4: fix xattr block allocation/release with bigalloc

On Mon, Feb 18, 2013 at 09:26:07AM -0600, Eric Sandeen wrote:
> On 2/18/13 2:14 AM, Lukas Czerner wrote:
> > Currently when new xattr block is created or released we we would call
> > dquot_free_block() or dquot_alloc_block() respectively, among the else
> > decrementing or incrementing the number of blocks assigned to the inode
> > by one block.
> >
> > This however does not work for bigalloc file system because we always
> > allocate/free the whole cluster so we have to count with that in
> > dquot_free_block() and dquot_alloc_block() as well.
> >
> > Use the clusters-to-blocks conversion EXT4_C2B() when passing number of
> > blocks to the dquot_alloc/free functions to fix the problem.
> >
> > The problem has been revealed by xfstests #117 (and possibly others).
> >
> > Signed-off-by: Lukas Czerner <[email protected]>
>
> Reviewed-by: Eric Sandeen <[email protected]>

Thanks, indeed. Applied.

- Ted