Re: [PATCH v3 2/2] xen/vm-events: Move parts of monitor_domctl code to common-side.

From: "Jan Beulich" <JBeulich@suse.com>
To: Corneliu ZUZU <czuzu@bitdefender.com>
Cc: Tamas K Lengyel <tamas@tklengyel.com>, Keir Fraser <keir@xen.org>,
	Ian Campbell <ian.campbell@citrix.com>,
	Razvan Cojocaru <rcojocaru@bitdefender.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	xen-devel@lists.xen.org,
	Stefano Stabellini <stefano.stabellini@citrix.com>
Subject: Re: [PATCH v3 2/2] xen/vm-events: Move parts of monitor_domctl code to common-side.
Date: Mon, 15 Feb 2016 09:44:53 -0700	[thread overview]
Message-ID: <56C20E9502000078000D24B6@prv-mh.provo.novell.com> (raw)
In-Reply-To: <56C1FCBB.5070004@bitdefender.com>

>>> On 15.02.16 at 17:28, <czuzu@bitdefender.com> wrote:
> On 2/15/2016 4:08 PM, Jan Beulich wrote:
>>
>>> After changing 1 to 1U though, I don't understand why we should also
>>> range-check mop->event.
>>> I'm imagining when (mop->event > 31):
>>> * (1U << mop->event) = 0 or >= (0x1 + 0xFFFFFFFF) (?)
>> No, it's plain undefined.
> 
> Weirdo C, didn't know that!
> I've just read http://www.danielvik.com/2010/05/c-language-quirks.html .
> That's crazy, I can't believe such 'quirks' exist and that I never knew 
> of them.
> So then, would this do:
> 
> /* sanity check - avoid '<<' operator undefined behavior */
> if ( unlikely(mop->event > 31) )
>      return -EINVAL;
> if ( unlikely(!(arch_monitor_get_capabilities(d) & (1U << mop->event))) )
>      return -EOPNOTSUPP;

I'd say -EOPNOTSUPP in both cases, but if the maintainers like
-EINVAL better I wouldn't insist on my preference.

Jan