Matteo Bruni (@Mystral) commented about dlls/d3dx10_43/tests/d3dx10.c:
hr = D3DX10CreateTextureFromMemory(device, dds_24bit_8_8, sizeof(dds_24bit_8_8), &load_info, NULL, &resource, &hr2); ok(hr == hr2, "Got unexpected hr2 %#lx.\n", hr2); ok(hr == S_OK, "Got unexpected hr %#lx.\n", hr); - if (img_info.Width) - { - check_image_info_values(&img_info, 8, 8, 1, 1, 4, 0, DXGI_FORMAT_R8G8B8A8_UNORM, - D3D10_RESOURCE_DIMENSION_TEXTURE2D, D3DX10_IFF_DDS, FALSE); - } + check_image_info_values(&img_info, 8, 8, 1, 1, 4, 0, DXGI_FORMAT_R8G8B8A8_UNORM, + D3D10_RESOURCE_DIMENSION_TEXTURE2D, D3DX10_IFF_DDS, FALSE);
So, this shows that the relevant behavior was already kinda covered by the tests. I'm wondering if some of the additional tests in this patch are actually redundant. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9089#note_117572