Re: [PATCH 0/6] MR5274: winevulkan: Avoid double indirection in device/queue wrappers.

[Rémi Bernon (＠rbernon)]

It does make a small difference in the vkoverhead benchmark, measurable for instance in the descriptor_buffer_offset_combined_sampler test.

Anyway I don't really mind about the last change, but I think the other ones make the code simpler, how about dropping the last commit?