Optimize descriptor set writes using push descriptors

We can use inline push descriptors for writing to descriptor rather than allocating a descriptor set for a one time write and freeing it as this is rather inefficient while an inline push descriptor generally ends up being a direct `memcpy` on the driver side designed for this use-case.

app/src/main/cpp/skyline/gpu/trait_manager.h

@@ -20,6 +20,7 @@ namespace skyline::gpu {
         bool supportsLogicOp{}; //!< If the device supports framebuffer logical operations during blending
         bool supportsVertexAttributeDivisor{}; //!< If the device supports a divisor for instance-rate vertex attributes (with VK_EXT_vertex_attribute_divisor)
         bool supportsVertexAttributeZeroDivisor{}; //!< If the device supports a zero divisor for instance-rate vertex attributes (with VK_EXT_vertex_attribute_divisor)
+        bool supportsPushDescriptors{}; //!< If the device supports push descriptors (with VK_KHR_push_descriptor)
         bool supportsGlobalPriority{}; //!< If the device supports global priorities for queues (with VK_EXT_global_priority)
         bool supportsMultipleViewports{}; //!< If the device supports more than one viewport
         bool supportsShaderViewportIndexLayer{}; //!< If the device supports retrieving the viewport index in shaders (with VK_EXT_shader_viewport_index_layer)