[Mesa-stable] [PATCH v2] anv: Don't expose VK_KHX_multiview on android.

[Emil Velikov]

On 5 March 2018 at 12:07, Tapani Pälli <tapani.palli at intel.com> wrote:
>
>
> On 03/05/2018 01:46 PM, Emil Velikov wrote:
>>
>> Hi Tapani,
>>
>> On 5 March 2018 at 08:57, Tapani Pälli <tapani.palli at intel.com> wrote:
>>>
>>> Just like commit 2ffe395 does for radv.
>>>
>>> Fixes following dEQP test on i965:
>>>     dEQP-VK.api.info.android.no_unknown_extensions
>>>
>>> v2: make it !ANDROID since this extension is not about
>>>      surfaces/swapchain
>>>
>>> CC: 18.0 <mesa-stable at lists.freedesktop.org>
>>> Signed-off-by: Tapani Pälli <tapani.palli at intel.com>
>>> ---
>>>   src/intel/vulkan/anv_extensions.py | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/src/intel/vulkan/anv_extensions.py
>>> b/src/intel/vulkan/anv_extensions.py
>>> index 8d39038c43..fd0d3337d0 100644
>>> --- a/src/intel/vulkan/anv_extensions.py
>>> +++ b/src/intel/vulkan/anv_extensions.py
>>> @@ -83,7 +83,7 @@ EXTENSIONS = [
>>>       Extension('VK_KHR_wayland_surface',                   6,
>>> 'VK_USE_PLATFORM_WAYLAND_KHR'),
>>>       Extension('VK_KHR_xcb_surface',                       6,
>>> 'VK_USE_PLATFORM_XCB_KHR'),
>>>       Extension('VK_KHR_xlib_surface',                      6,
>>> 'VK_USE_PLATFORM_XLIB_KHR'),
>>> -    Extension('VK_KHX_multiview',                         1, True),
>>> +    Extension('VK_KHX_multiview',                         1,
>>> '!ANDROID'),
>>
>>
>> As reminded by Bas [0], the extensions should not be advertised in
>> released Mesa altogether.
>> Hence, I've adapted Jason's old patch for anv and radv on the 17.3 and
>> 18.0 branches [1] [2] [3] [4].
>
>
> OK cool, I missed this conversation. Then I'll drop the stable tag!
>
I might have gone overboard with the links, oopsie. Fwiw the patch is
Reviewed-by: Emil Velikov <emil.velikov at collabora.com>

> I also have a (horrible) patch to make dEQP-VK.api.info.device.extensions
> pass [1]. Problem with that test is that it hardcodes a list of 'accepted
> extensions' and everything outside that list triggers an error. To make it
> pass we actually need to disable nearly every entry in this list. Maybe a
> nicer long term solution we should have a separate EXTENSIONS list for
> Android or specific enable flag for Android ones (instead of disable)?
>
Hmm, a cleaner solution is to tweak the
VkGetDeviceProcAddr/VkGetInstanceProcAddr API to only expose 'new'
extensions... Which we seem to be doing already [A].

Does Android use version number X while hacking/allowing only a
selection of the extensions provided with said version?
Or perhaps the version we use in Mesa is wrong?

HTH
Emil

[A] See the VkGetDeviceProcAddr patch from Iago and associated code.
https://patchwork.freedesktop.org/patch/207981/