Re: [PATCH RESEND] btrfs: Align EOF length to block in extent_same

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, Apr 26, 2015 at 07:05:11PM +0100, Matt Robinson wrote:
> It is not currently possible to deduplicate the last block of files
> whose size is not a multiple of the block size, as the btrfs_extent_same
> ioctl returns -EINVAL if offset + size is greater than the file size or
> is not aligned to the fs block size.
> 
> For example, with the default block size of 16K and two identical
> 1,000,000 byte files, calling the extent_same ioctl with offset 0 and
> length set to 1,000,000 the call fails with -EINVAL.  The same call
> with a length of 999,424 will succeed, but the final 576 bytes can then
> not be shared.  This seems to have a larger impact on the amount of
> space actually freed by the ioctl than would be expected - in my
> testing the amount of space freed was generally reduced by 50-100% for
> files sized from a few megabytes downwards which has a significant
> negative impact on the usefulness of the extent_same ioctl in some
> circumstances.
> 
> To resolve this, this patch allows unaligned offset + length values to
> be passed to btrfs_ioctl_file_extent_same if offset + length is equal
> to the file size of both src and dest.  This is implemented in the same
> way as in btrfs_ioctl_clone.
> 
> To return to the earlier example 1,000,000 byte file - this patch would
> allow a length of 1,000,000 bytes to be passed as it is equal to the
> file lengths and would be internally extended to the end of the block
> (1,015,808), allowing one set of extents to be shared completely between
> the full length of both files.

I would love to test this, but currently I'm not using extent-same for
deduplication because of the deadlock bug.  :-/

Was there a patch for that bug that I missed?

> Signed-off-by: Matt Robinson <git@xxxxxxxxxxxxxxxxx>
> ---
>  fs/btrfs/ioctl.c | 21 ++++++++++++++-------
>  1 file changed, 14 insertions(+), 7 deletions(-)
> 
> diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
> index ca5d968..0588076 100644
> --- a/fs/btrfs/ioctl.c
> +++ b/fs/btrfs/ioctl.c
> @@ -2878,14 +2878,16 @@ static int btrfs_cmp_data(struct inode *src, u64 loff, struct inode *dst,
>  	return ret;
>  }
>  
> -static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len)
> +static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len,
> +				     u64 len_aligned)
>  {
>  	u64 bs = BTRFS_I(inode)->root->fs_info->sb->s_blocksize;
>  
>  	if (off + len > inode->i_size || off + len < off)
>  		return -EINVAL;
> +
>  	/* Check that we are block aligned - btrfs_clone() requires this */
> -	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs))
> +	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len_aligned, bs))
>  		return -EINVAL;
>  
>  	return 0;
> @@ -2895,6 +2897,8 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 len,
>  			     struct inode *dst, u64 dst_loff)
>  {
>  	int ret;
> +	u64 len_aligned = len;
> +	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
>  
>  	/*
>  	 * btrfs_clone() can't handle extents in the same file
> @@ -2909,11 +2913,15 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 len,
>  
>  	btrfs_double_lock(src, loff, dst, dst_loff, len);
>  
> -	ret = extent_same_check_offsets(src, loff, len);
> +	/* if we extend to both eofs, continue to block boundaries */
> +	if (loff + len == src->i_size && dst_loff + len == dst->i_size)
> +		len_aligned = ALIGN(src->i_size, bs) - loff;
> +
> +	ret = extent_same_check_offsets(src, loff, len, len_aligned);
>  	if (ret)
>  		goto out_unlock;
>  
> -	ret = extent_same_check_offsets(dst, dst_loff, len);
> +	ret = extent_same_check_offsets(dst, dst_loff, len, len_aligned);
>  	if (ret)
>  		goto out_unlock;
>  
> @@ -2926,7 +2934,7 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 len,
>  
>  	ret = btrfs_cmp_data(src, loff, dst, dst_loff, len);
>  	if (ret == 0)
> -		ret = btrfs_clone(src, dst, loff, len, len, dst_loff);
> +		ret = btrfs_clone(src, dst, loff, len, len_aligned, dst_loff);
>  
>  out_unlock:
>  	btrfs_double_unlock(src, loff, dst, dst_loff, len);
> @@ -3172,8 +3180,7 @@ static void clone_update_extent_map(struct inode *inode,
>   * @inode: Inode to clone to
>   * @off: Offset within source to start clone from
>   * @olen: Original length, passed by user, of range to clone
> - * @olen_aligned: Block-aligned value of olen, extent_same uses
> - *               identical values here
> + * @olen_aligned: Block-aligned value of olen
>   * @destoff: Offset within @inode to start clone
>   */
>  static int btrfs_clone(struct inode *src, struct inode *inode,
> -- 
> 2.1.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Linux Filesystem Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux