Vulkan: Fix host-pinned memory for large allocations#14249

0cc4m merged 1 commit intomasterggml-org/llama.cpp:masterfrom

0cc4m/vulkan-fix-host-memory-max-sizeggml-org/llama.cpp:0cc4m/vulkan-fix-host-memory-max-sizeCopy head branch name to clipboard