[PATCH] virtio_blk: ioctl return value fix

[PATCH] virtio_blk: ioctl return value fix

[Christoph Hellwig]

Block driver ioctl methods must return ENOTTY and not -ENOIOCTLCMD if
they expect the block layer to handle generic ioctls.

This triggered a BLKROSET failure in xfsqa #200.


Signed-off-by: Christoph Hellwig <hch@lst.de>

Index: linux-2.6/drivers/block/virtio_blk.c
===================================================================
--- linux-2.6.orig/drivers/block/virtio_blk.c	2009-06-20 21:21:20.252930284 +0200
+++ linux-2.6/drivers/block/virtio_blk.c	2009-06-20 21:22:48.364932365 +0200
@@ -213,7 +213,7 @@ static int virtblk_ioctl(struct block_de
 	 * Only allow the generic SCSI ioctls if the host can support it.
 	 */
 	if (!virtio_has_feature(vblk->vdev, VIRTIO_BLK_F_SCSI))
-		return -ENOIOCTLCMD;
+		return -ENOTTY;
 
 	return scsi_cmd_ioctl(disk->queue, disk, mode, cmd, argp);
 }

Re: [PATCH] virtio_blk: ioctl return value fix

[Rusty Russell]

On Sun, 21 Jun 2009 04:59:41 am Christoph Hellwig wrote:
> Block driver ioctl methods must return ENOTTY and not -ENOIOCTLCMD if
> they expect the block layer to handle generic ioctls.
>
> This triggered a BLKROSET failure in xfsqa #200.

Applied.

Thanks,
Rusty.

Re: [PATCH] virtio_blk: ioctl return value fix

[Christoph Hellwig]

On Tue, Jun 23, 2009 at 04:43:10PM +0930, Rusty Russell wrote:
> On Sun, 21 Jun 2009 04:59:41 am Christoph Hellwig wrote:
> > Block driver ioctl methods must return ENOTTY and not -ENOIOCTLCMD if
> > they expect the block layer to handle generic ioctls.
> >
> > This triggered a BLKROSET failure in xfsqa #200.
> 
> Applied.

Can you send on this fix to Linus?  It fixes a regression introduced in
Linux 2.6.30.

Re: [PATCH] virtio_blk: ioctl return value fix

[Avi Kivity]

On 07/08/2009 09:12 PM, Christoph Hellwig wrote:
> On Tue, Jun 23, 2009 at 04:43:10PM +0930, Rusty Russell wrote:
>    
>> On Sun, 21 Jun 2009 04:59:41 am Christoph Hellwig wrote:
>>      
>>> Block driver ioctl methods must return ENOTTY and not -ENOIOCTLCMD if
>>> they expect the block layer to handle generic ioctls.
>>>
>>> This triggered a BLKROSET failure in xfsqa #200.
>>>        
>> Applied.
>>      
>
> Can you send on this fix to Linus?  It fixes a regression introduced in
> Linux 2.6.30.
>
>    

It should set the Cc: stable@kernel.org marker then.

-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.

Re: [PATCH] virtio_blk: ioctl return value fix

[Rusty Russell]

On Thu, 9 Jul 2009 03:42:20 am Christoph Hellwig wrote:
> On Tue, Jun 23, 2009 at 04:43:10PM +0930, Rusty Russell wrote:
> > On Sun, 21 Jun 2009 04:59:41 am Christoph Hellwig wrote:
> > > Block driver ioctl methods must return ENOTTY and not -ENOIOCTLCMD if
> > > they expect the block layer to handle generic ioctls.
> > >
> > > This triggered a BLKROSET failure in xfsqa #200.
> >
> > Applied.
>
> Can you send on this fix to Linus?  It fixes a regression introduced in
> Linux 2.6.30.

Yep, I've been away, but they'll both be pushed soon with other (minor) 
lguest/virtio fixes.  I'm holding on a final lguest regression which I need to 
track down.

Thanks,
Rusty.