engine: accept R32_UINT depth

gameDepthFormat now admits VK_FORMAT_R32_UINT alongside the existing
float/UNORM depth formats. The RR engine's hardwareDepth path
(DLSS.Use.HW.Depth) already handles the nonlinear interpretation;
the format check was the only gate keeping the feature on the CUDA
fallback.

Fixes Cyberpunk 2077 with RR enabled falling back to the CUDA path.

engine/game.cpp

@@ -22,7 +22,7 @@ bool gameMotionFormat(VkFormat format) {
 }
 bool gameDepthFormat(VkFormat format, VkImageAspectFlags aspect) {
     if (aspect & VK_IMAGE_ASPECT_DEPTH_BIT) return depthAspect(format);
-    return format == VK_FORMAT_R32_SFLOAT || format == VK_FORMAT_R16_SFLOAT || format == VK_FORMAT_R16_UNORM;
+    return format == VK_FORMAT_R32_SFLOAT || format == VK_FORMAT_R32_UINT || format == VK_FORMAT_R16_SFLOAT || format == VK_FORMAT_R16_UNORM;
 }
 bool gameOutputFormat(VkFormat format) { return gameColorFormat(format); }