Matteo Bruni (@Mystral) commented about dlls/d3dx10_43/tests/d3dx10.c:
+ case D3D10_RESOURCE_DIMENSION_TEXTURE2D: + ID3D10Texture2D_Unmap((ID3D10Texture2D *)rb->resource, rb->sub_resource_idx); + break; + case D3D10_RESOURCE_DIMENSION_TEXTURE3D: + ID3D10Texture3D_Unmap((ID3D10Texture3D *)rb->resource, rb->sub_resource_idx); + break; + default: + trace("Unhandled resource dimension %#x.\n", rb->dimension); + break; + } + ID3D10Resource_Release(rb->resource); +} + +#define check_readback_data_u32(a, b, c) check_readback_data_u32_(__LINE__, a, b, c) +static void check_readback_data_u32_(unsigned int line, struct resource_readback *rb, + const RECT *rect, uint32_t expected) (After noticing that this dropped the `max_diff` parameter compared to the previous revision) just curious, I guess we won't need tolerance on the check even in other tests you have sitting around?
-- https://gitlab.winehq.org/wine/wine/-/merge_requests/9089#note_117964