[PATCH] ext4: let getattr report the right blocks in delalloc+bigalloc.

[PATCH] ext4: let getattr report the right blocks in delalloc+bigalloc.

[Tao Ma]

From: Tao Ma <boyu.mt@taobao.com>

In delayed allocation, i_reserved_data_blocks now indicates
clusters, not blocks. So report it in the right number.

This can be easily exposed by the following command:
echo foo > blah; du -hc blah; sync; du -hc blah

Cc: "Theodore Ts'o" <tytso@mit.edu>
Reported-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
---
 fs/ext4/inode.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index feaa82f..a5eb8c6 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4200,7 +4200,8 @@ int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
 	 * will return the blocks that include the delayed allocation
 	 * blocks for this file.
 	 */
-	delalloc_blocks = EXT4_I(inode)->i_reserved_data_blocks;
+	delalloc_blocks = EXT4_C2B(EXT4_SB(inode->i_sb),
+				EXT4_I(inode)->i_reserved_data_blocks);
 
 	stat->blocks += (delalloc_blocks << inode->i_sb->s_blocksize_bits)>>9;
 	return 0;
-- 
1.7.0.4

Re: [PATCH] ext4: let getattr report the right blocks in delalloc+bigalloc.

[Eric Sandeen]

On 1/31/12 9:58 AM, Tao Ma wrote:
> From: Tao Ma <boyu.mt@taobao.com>
> 
> In delayed allocation, i_reserved_data_blocks now indicates
> clusters, not blocks. So report it in the right number.
> 
> This can be easily exposed by the following command:
> echo foo > blah; du -hc blah; sync; du -hc blah
> 
> Cc: "Theodore Ts'o" <tytso@mit.edu>
> Reported-by: Eric Sandeen <sandeen@redhat.com>
> Signed-off-by: Tao Ma <boyu.mt@taobao.com>

Reviewed-by: Eric Sandeen <sandeen2redhat.com>

It'd be great to get this one merged.

Thanks,
-Eric

> ---
>  fs/ext4/inode.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index feaa82f..a5eb8c6 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4200,7 +4200,8 @@ int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
>  	 * will return the blocks that include the delayed allocation
>  	 * blocks for this file.
>  	 */
> -	delalloc_blocks = EXT4_I(inode)->i_reserved_data_blocks;
> +	delalloc_blocks = EXT4_C2B(EXT4_SB(inode->i_sb),
> +				EXT4_I(inode)->i_reserved_data_blocks);
>  
>  	stat->blocks += (delalloc_blocks << inode->i_sb->s_blocksize_bits)>>9;
>  	return 0;

Re: [PATCH] ext4: let getattr report the right blocks in delalloc+bigalloc.

[Ted Ts'o]

On Thu, May 31, 2012 at 12:08:50PM -0500, Eric Sandeen wrote:
> On 1/31/12 9:58 AM, Tao Ma wrote:
> > From: Tao Ma <boyu.mt@taobao.com>
> > 
> > In delayed allocation, i_reserved_data_blocks now indicates
> > clusters, not blocks. So report it in the right number.
> > 
> > This can be easily exposed by the following command:
> > echo foo > blah; du -hc blah; sync; du -hc blah
> > 
> > Cc: "Theodore Ts'o" <tytso@mit.edu>
> > Reported-by: Eric Sandeen <sandeen@redhat.com>
> > Signed-off-by: Tao Ma <boyu.mt@taobao.com>
> 
> Reviewed-by: Eric Sandeen <sandeen2redhat.com>
> 
> It'd be great to get this one merged.

I'll merge this, but delalloc and bigalloc is still pretty badly
busted.  But this will make stat return something that is closer to
being correct...

						- Ted

Re: [PATCH] ext4: let getattr report the right blocks in delalloc+bigalloc.

[Eric Sandeen]

On 5/31/12 7:43 PM, Ted Ts'o wrote:
> On Thu, May 31, 2012 at 12:08:50PM -0500, Eric Sandeen wrote:
>> On 1/31/12 9:58 AM, Tao Ma wrote:
>>> From: Tao Ma <boyu.mt@taobao.com>
>>>
>>> In delayed allocation, i_reserved_data_blocks now indicates
>>> clusters, not blocks. So report it in the right number.
>>>
>>> This can be easily exposed by the following command:
>>> echo foo > blah; du -hc blah; sync; du -hc blah
>>>
>>> Cc: "Theodore Ts'o" <tytso@mit.edu>
>>> Reported-by: Eric Sandeen <sandeen@redhat.com>
>>> Signed-off-by: Tao Ma <boyu.mt@taobao.com>
>>
>> Reviewed-by: Eric Sandeen <sandeen2redhat.com>
>>
>> It'd be great to get this one merged.
> 
> I'll merge this, but delalloc and bigalloc is still pretty badly
> busted.  But this will make stat return something that is closer to
> being correct...

Hm what else is busted?  Is there a testcase written?

How much else do you know of that's busted with bigalloc?  :)

(Is this bustedness the reason for not yet having it in the man page?)

-Eric
 
> 						- Ted