[PATCH 1/4] d3d9/tests: Test more formats in srgbtexture_test().

[Paul Gofman]

I've tested the game with AMD graphics card override under Wine
(VideoPciVendorID 0x1002, VideoPciDeviceId 0x67df) and it seems to fix
the rendering on its own. So the game expects _R5G6B5 to have SRGB reads
supported only if detects the GPU as Nvidia (at least), otherwise it
works some other way. Yet I suppose adding the SRGB read support as on
Nvidia might be a valid fix unless we want to make the support depend on
the reported video card.

On 1/21/20 16:53, Paul Gofman wrote:
> All of them are supported on Nvidia GeForce GT 650M where I tested that
> on Windows. All but _X1R5G5B5 are also supported on Intel.
> If _R5G6B5 is not supported on AMD then probably Warhammer takes some
> other path on AMD. I don't have an AMD to test that but I can try
> spoofing AMD in Wine registry keys to see if it makes any difference. In
> case it does, it might be possible to find out how it works on AMD.
>
> On 1/21/20 16:27, Henri Verbeet wrote:
>> On Mon, 13 Jan 2020 at 19:02, Paul Gofman <gofmanp at gmail.com> wrote:
>>> +        case D3DFMT_A1R5G5B5:
>>> +        case D3DFMT_X1R5G5B5:
>>> +        case D3DFMT_R5G6B5:
>> Are these generally supported on NVIDIA? They don't appear to be
>> supported on my AMD system. That's not necessarily an issue, but does
>> raise the question whether applications can really depend on these
>> formats supporting sRGB reads.
>
>