llama.cpp

74d4f5b0 - vulkan: Hybrid waitForFences/getFenceStatus to reduce fence latency (#12630)

vulkan: Hybrid waitForFences/getFenceStatus to reduce fence latency (#12630) There seems to be a bubble waking up from waitForFences, which costs a few percent performance and also increased variance in performance. This change inserts an "almost_ready" fence when the graph is about 80% complete and we waitForFences for the almost_ready fence and then spin (with _mm_pauses) waiting for the final fence to be signaled.