Hi Paolo,

On Sun, Feb 05, 2012 at 11:44:57PM +0100, Paolo Bonzini wrote:

On 02/05/2012 11:11 PM, Willy Tarreau wrote:

quoted

Signed-off-by: Ben Hutchings<redacted>
Signed-off-by: Greg Kroah-Hartman<gregkh@linuxfoundation.org>
[wt: no scsi_cmd_blk_ioctl in 2.6.27, change callers instead. cciss is OK,
 ub,virtio_blk,ide-floppy,sd need fixing, cdrom&st can be ignored ]

NACK, you aren't fixing virtio-blk to call scsi_verify_blk_ioctl, are you?

You need to backport the patch that introduced scsi_cmd_blk_ioctl, 
which, in fact was in the same patch series as this one.

Thanks for this report, I missed this patch. I thought the reason for the lack
of scsi_cmd_blk_ioctl() it was one of the differences between 2.6.27 and 2.6.32,
so I have adapted the callers I identified to perform the same test as what's
in scsi_cmd_blk_ioctl(). I *think* I did it correctly, but it would be safer
to apply the original patch, of course. Now I see the patch (618 in my queue).
 I'll merge it and it will simplify the backport.

Thank you,
Willy