Potential bug in fs/ext4/inline.c?

Potential bug in fs/ext4/inline.c?

[Theodore Ts'o]

In ext4_find_inline_entry(), there is the following:

	if (ext4_get_inline_size(dir) == EXT4_MIN_INLINE_DATA_SIZE)
		goto out;

Am I missing something, or shouldn't this be:

	if (ext4_get_inline_size(dir) <= EXT4_MIN_INLINE_DATA_SIZE)
		goto out;

?

Thanks,

						- Ted

Re: Potential bug in fs/ext4/inline.c?

[Darrick J. Wong]

On Thu, Aug 21, 2014 at 06:01:26PM -0400, Theodore Ts'o wrote:
> In ext4_find_inline_entry(), there is the following:
> 
> 	if (ext4_get_inline_size(dir) == EXT4_MIN_INLINE_DATA_SIZE)
> 		goto out;
> 
> Am I missing something, or shouldn't this be:
> 
> 	if (ext4_get_inline_size(dir) <= EXT4_MIN_INLINE_DATA_SIZE)
> 		goto out;
> 
> ?

I'm not sure it matters, since I see:

EXT4_I(inode)->i_inline_size = len + EXT4_MIN_INLINE_DATA_SIZE

scattered all over the code.  I'm not sure I like the idea that both ext4 and
e2fsprogs think that the inline data size is never <= 60, but afaict it doesn't
cause any problems.

--D

> 
> Thanks,
> 
> 						- Ted
> 
> 
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Potential bug in fs/ext4/inline.c?

[Zheng Liu]

On Thu, Aug 21, 2014 at 06:01:26PM -0400, Theodore Ts'o wrote:
> In ext4_find_inline_entry(), there is the following:
> 
> 	if (ext4_get_inline_size(dir) == EXT4_MIN_INLINE_DATA_SIZE)
> 		goto out;
> 
> Am I missing something, or shouldn't this be:
> 
> 	if (ext4_get_inline_size(dir) <= EXT4_MIN_INLINE_DATA_SIZE)
> 		goto out;

Hi Ted,

The original design idea is that the size of inline data shouldn't be
less than EXT4_MIN_INLINE_DATA_SIZE (60 bytes).  That means that the
inline data will occupy the whole i_block[] space.

If we create a new dir with inline data feature, the disk layout will
looks like below (ext4_try_create_inline_dir()):

i_block[0]: parent ino (4 bytes)
i_block+4:  dir entry (ino 0, rec_len 56 bytes)

Regards,
                                                - Zheng