diff mbox series

[2/2] xfs: ignore extent size hints for always COW inodes

Message ID 20191011130316.13373-3-hch@lst.de (mailing list archive)
State Deferred, archived
Headers show
Series [1/2] xfs: disable xfs_ioc_space for always COW inodes | expand

Commit Message

Christoph Hellwig Oct. 11, 2019, 1:03 p.m. UTC
There is no point in applying extent size hints for always COW inodes,
as we would just have to COW any extra allocation beyond the data
actually written.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_inode.c | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Darrick J. Wong Oct. 12, 2019, 12:32 a.m. UTC | #1
On Fri, Oct 11, 2019 at 06:03:16AM -0700, Christoph Hellwig wrote:
> There is no point in applying extent size hints for always COW inodes,
> as we would just have to COW any extra allocation beyond the data
> actually written.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Looks ok, I guess?

By the way, what's the plan for always_cow inodes, seeing as it's still
only a debugging feature?

Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>

--D

> ---
>  fs/xfs/xfs_inode.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
> index 18f4b262e61c..2e94deb4610a 100644
> --- a/fs/xfs/xfs_inode.c
> +++ b/fs/xfs/xfs_inode.c
> @@ -55,6 +55,12 @@ xfs_extlen_t
>  xfs_get_extsz_hint(
>  	struct xfs_inode	*ip)
>  {
> +	/*
> +	 * No point in aligning allocations if we need to COW to actually
> +	 * write to them.
> +	 */
> +	if (xfs_is_always_cow_inode(ip))
> +		return 0;
>  	if ((ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) && ip->i_d.di_extsize)
>  		return ip->i_d.di_extsize;
>  	if (XFS_IS_REALTIME_INODE(ip))
> -- 
> 2.20.1
>
Christoph Hellwig Oct. 14, 2019, 7:19 a.m. UTC | #2
On Fri, Oct 11, 2019 at 05:32:26PM -0700, Darrick J. Wong wrote:
> On Fri, Oct 11, 2019 at 06:03:16AM -0700, Christoph Hellwig wrote:
> > There is no point in applying extent size hints for always COW inodes,
> > as we would just have to COW any extra allocation beyond the data
> > actually written.
> > 
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> 
> Looks ok, I guess?
> 
> By the way, what's the plan for always_cow inodes, seeing as it's still
> only a debugging feature?

Support for zoned devices and an O_ATOMIC-like mode that supports
data integrity safe overwrites.  I've found some time to spend on
both lately, but the former might land on the list first.
diff mbox series

Patch

diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 18f4b262e61c..2e94deb4610a 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -55,6 +55,12 @@  xfs_extlen_t
 xfs_get_extsz_hint(
 	struct xfs_inode	*ip)
 {
+	/*
+	 * No point in aligning allocations if we need to COW to actually
+	 * write to them.
+	 */
+	if (xfs_is_always_cow_inode(ip))
+		return 0;
 	if ((ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) && ip->i_d.di_extsize)
 		return ip->i_d.di_extsize;
 	if (XFS_IS_REALTIME_INODE(ip))