Re: [RFC v3] drm/i915: Select engines via class and instance in execbuffer2

From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
To: Emil Velikov <emil.l.velikov@gmail.com>
Cc: Ben Widawsky <ben@bwidawsk.net>,
	intel-vaapi-media@lists.01.org,
	"intel-gfx@lists.freedesktop.org"
	<Intel-gfx@lists.freedesktop.org>,
	ML mesa-dev <mesa-dev@lists.freedesktop.org>,
	Daniel Vetter <daniel.vetter@intel.com>
Subject: Re: [RFC v3] drm/i915: Select engines via class and instance in execbuffer2
Date: Thu, 18 May 2017 15:55:27 +0100	[thread overview]
Message-ID: <ff61310a-6802-739f-7abc-b68395038ee8@linux.intel.com> (raw)
In-Reply-To: <CACvgo52-sV=zc737nBU_1HQLf8s0i6-7cxmU-74ZzpUO55Ncgw@mail.gmail.com>

On 18/05/2017 15:10, Emil Velikov wrote:
> Hi Tvrtko,
>
> On 18 May 2017 at 14:06, Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> wrote:
>
>> struct drm_i915_engine_info {
>>         /** Engine instance number. */
>>         __u32   instance;
>>         __u32   rsvd;
>>
>>         /** Engine specific features and info. */
>> #define DRM_I915_ENGINE_HAS_HEVC        BIT(0)
>>         __u8 features;
>>         __u8 rsvd;
>>
>>         __32 info;
>> };
>>
> To spare yourself from writing a compat ioctl, you want to have the
> members at the same offset on both 32 and 64bit builds.
> At the same times the whole struct size should be multiple of 64bit.
>
> This and a few others are covered in Daniel Vetter's "Botching up ioctls" [1]

Yes, thanks, I failed to add up to 64. :)

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx