[PATCH] xfsprogs: Don't ever try to set the device blocksize in repair

[PATCH] xfsprogs: Don't ever try to set the device blocksize in repair

[Eric Sandeen]

On 4k devices, we get this warning from repair:

# xfs_repair /dev/sdc2
xfs_repair: warning - cannot set blocksize 512 on block device /dev/sdc2: Invalid argument
Phase 1 - find and verify superblock...
...

but things proceed without trouble after that.

I'm unable to find any history or reason for setting the
device blocksize at the beginning of repair, and in any case,
things clearly work without doing so.  So, let's just remove it.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---

diff --git a/repair/init.c b/repair/init.c
index 654c406..bc10cc4 100644
--- a/repair/init.c
+++ b/repair/init.c
@@ -143,7 +143,7 @@ xfs_init(libxfs_init_t *args)
 	}
 
 	args->usebuflock = do_prefetch;
-	args->setblksize = !dangerously;
+	args->setblksize = 0;
 	args->isdirect = LIBXFS_DIRECT;
 	if (no_modify)
 		args->isreadonly = (LIBXFS_ISREADONLY | LIBXFS_ISINACTIVE);

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH] xfsprogs: Don't ever try to set the device blocksize in repair

[Alex Elder]

On Mon, 2011-02-21 at 13:06 -0600, Eric Sandeen wrote:
> On 4k devices, we get this warning from repair:
> 
> # xfs_repair /dev/sdc2
> xfs_repair: warning - cannot set blocksize 512 on block device /dev/sdc2: Invalid argument
> Phase 1 - find and verify superblock...
> ...
> 
> but things proceed without trouble after that.
> 
> I'm unable to find any history or reason for setting the
> device blocksize at the beginning of repair, and in any case,
> things clearly work without doing so.  So, let's just remove it.
> 
> Signed-off-by: Eric Sandeen <sandeen@redhat.com>
> ---

On Linux anyway, the way it's used here is pretty
pointless anyway.  It passes either 0 or 1, which
means either don't set it (if "living dangerously")
or do set it, to the minimum XFS sector size, 512
bytes.  Hence it'll screw up on > 512-byte sector
devices.

If there was a reason for setting it, it should
do so by setting it to the device's logical block
size rather than assuming 512.

But I don't see a reason to set it either, so the
change looks good to me.

Reviewed-by: Alex Elder <aelder@sgi.com>



_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs