From mboxrd@z Thu Jan  1 00:00:00 1970
From: Thomas Hellstrom <thellstrom@vmware.com>
Subject: Re: [PATCH 2/5] drm: Break out ioctl permission check to a separate
	function
Date: Thu, 13 Mar 2014 13:28:33 +0100
Message-ID: <5321A471.2070202@vmware.com>
References: <1394708244-6565-1-git-send-email-thellstrom@vmware.com>
	<1394708244-6565-3-git-send-email-thellstrom@vmware.com>
	<CANq1E4ShMH3gend4=xkPaKUpChW8zWPC+DgVc-U+4OAm3zZrxw@mail.gmail.com>
	<5321A060.8090406@vmware.com>
	<CANq1E4T6fq9U_5ttem4qvvq9qSMSy8QY1FNOyP_GzF3-hHzN=A@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <dri-devel-bounces@lists.freedesktop.org>
Received: from smtp-outbound-1.vmware.com (smtp-outbound-1.vmware.com
	[208.91.2.12])
	by gabe.freedesktop.org (Postfix) with ESMTP id A8074FAEBA
	for <dri-devel@lists.freedesktop.org>;
	Thu, 13 Mar 2014 05:28:36 -0700 (PDT)
In-Reply-To: <CANq1E4T6fq9U_5ttem4qvvq9qSMSy8QY1FNOyP_GzF3-hHzN=A@mail.gmail.com>
List-Unsubscribe: <http://lists.freedesktop.org/mailman/options/dri-devel>,
	<mailto:dri-devel-request@lists.freedesktop.org?subject=unsubscribe>
List-Archive: <http://lists.freedesktop.org/archives/dri-devel>
List-Post: <mailto:dri-devel@lists.freedesktop.org>
List-Help: <mailto:dri-devel-request@lists.freedesktop.org?subject=help>
List-Subscribe: <http://lists.freedesktop.org/mailman/listinfo/dri-devel>,
	<mailto:dri-devel-request@lists.freedesktop.org?subject=subscribe>
Sender: dri-devel-bounces@lists.freedesktop.org
Errors-To: dri-devel-bounces@lists.freedesktop.org
To: David Herrmann <dh.herrmann@gmail.com>
Cc: Dave Airlie <airlied@redhat.com>, Thomas Hellstrom <thellstrom@vmware.com>, linux-graphics-maintainer@vmware.com, "dri-devel@lists.freedesktop.org" <dri-devel@lists.freedesktop.org>
List-Id: dri-devel@lists.freedesktop.org

On 03/13/2014 01:15 PM, David Herrmann wrote:
> Hi
>
> On Thu, Mar 13, 2014 at 1:11 PM, Thomas Hellstrom <thellstrom@vmware.com> wrote:
>> Hi.
>>
>> Thanks for reviewing. I'll incorporate your suggestions, except this
>> one, and resend.
>>
>>
>> On 03/13/2014 12:19 PM, David Herrmann wrote:
>>> Hi
>>>
>>> On Thu, Mar 13, 2014 at 11:57 AM, Thomas Hellstrom
>>> <thellstrom@vmware.com> wrote:
>> ...
>>
>> -               if (cmd & IOC_IN) {
>> -                       if (copy_from_user(kdata, (void __user *)arg,
>> -                                          usize) != 0) {
>> -                               retcode = -EFAULT;
>> +       retcode = drm_ioctl_permit(ioctl->flags, file_priv);
>> +       if (unlikely(retcode))
>>
>>> That "unlikely" seems redundant given that all error paths in
>>> drm_ioctl_permit() already are "unlikely".
>> Yes, we know that's true, but I don't think compilers in general can
>> combine branch prediction hints in that way,
>> or even have the information necessary to do it.
>> I mean even if each individual test resulting in an error is unlikely,
>> how could the compiler know that
>> all tests combined would result in an error being unlikely?
> The function is static, so the compiler can see that it returns "!=0"
> only if one of the "unlikely" branches was hit. So I think it's safe
> to assume the whole thing returns "!=0" only in unlikely conditions.
>

But a compiler can't (or shouldn't) make that assumption. Just as an
(adapted) example, imagine that
each test had a 20% probability of returning an error. The probability
of the function returning an error would
then be 68%..

> I'm no big fan of excessive likely/unlikely annotations, but I'm fine
> if you want to keep it.

Fair enough.

Thanks,
Thomas


>
> Thanks
> David
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel