RE: [Xen-devel] [PATCH 1/3] xen/privcmd: return -ENOSYS for unimplemented IOCTLs

From: Paul Durrant
Date: Thu Feb 09 2017 - 10:32:02 EST


> -----Original Message-----
> From: Boris Ostrovsky [mailto:boris.ostrovsky@xxxxxxxxxx]
> Sent: 09 February 2017 15:26
> To: Jan Beulich <JBeulich@xxxxxxxx>; Paul Durrant
> <Paul.Durrant@xxxxxxxxxx>
> Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx; Juergen Gross <JGross@xxxxxxxx>;
> linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [Xen-devel] [PATCH 1/3] xen/privcmd: return -ENOSYS for
> unimplemented IOCTLs
>
>
>
> On 02/09/2017 09:40 AM, Jan Beulich wrote:
> >>>> On 09.02.17 at 15:17, <paul.durrant@xxxxxxxxxx> wrote:
> >> The code goes so far as to set the default return code to -ENOSYS but
> >> then overrides this to -EINVAL in the switch() statement's default
> >> case.
> >
> > If you already change this, isn't -ENOTTY the traditional way of
> > indicating unsupported ioctls?
>
> In fact, a while ago David submitted a patch to do just that:
>
> https://lists.xenproject.org/archives/html/xen-devel/2016-
> 08/msg00744.html
>
> but it never went anywhere.
>
> My question is whether anyone might be relying on current error return
> behavior.

I doubt it. It's certainly not a safe thing to do anyway. I'll change to -ENOTTY in v2 of the patch.

Paul

>
>
> -boris