Re: [PATCH 0/6] MR6360: d3dx9: Beginning of pixel format handling changes.

[Matteo Bruni (＠Mystral)]

Matteo Bruni (@Mystral) commented about dlls/d3dx9_36/tests/surface.c:
> -
>    check_pixel_float4(&lockrect, 1, 1,  pixdata_a32b32g32r32f[12], pixdata_a32b32g32r32f[13],
> -
>            pixdata_a32b32g32r32f[14], pixdata_a32b32g32r32f[15], FALSE);
> -
>    hr = IDirect3DSurface9_UnlockRect(surf);
> -
>    ok(hr == D3D_OK, "Failed to unlock surface, hr %#lx.\n", hr);
> -
> -
>    /* Signed normalized value to full range float. */
> -
>    hr = D3DXLoadSurfaceFromMemory(surf, NULL, NULL, pixdata_q8w8v8u8, D3DFMT_Q8W8V8U8, 8, NULL, &rect,
> -
>            D3DX_FILTER_NONE, 0);
> -
>    todo_wine ok(hr == D3D_OK, "Got unexpected hr %#lx.\n", hr);
> -
> -
>    hr = IDirect3DSurface9_LockRect(surf, &lockrect, NULL, D3DLOCK_READONLY);
> -
>    ok(hr == D3D_OK, "Failed to lock surface, hr %#lx.\n", hr);
> -
>    check_pixel_float4(&lockrect, 0, 0,  0.0f,       0.125984f,  0.251969f,  0.377953f, TRUE);
> -
>    check_pixel_float4(&lockrect, 1, 0,  0.503937f,  0.629921f,  0.755906f,  1.0f,      TRUE);
> -
>    check_pixel_float4(&lockrect, 0, 1, -1.0f,      -1.0f,      -0.755906f, -0.629921f, TRUE);
> -
>    check_pixel_float4(&lockrect, 1, 1, -0.503937f, -0.377953f, -0.251969f, -0.007874f, TRUE);

Good test, showing the two encodings for -1.0f and exact 0.0f for SNORM.