llama.cpp Release b10865 Reverts Integrated Property Commit for HIP Builds
llama.cpp released automated build b10865, which reverts pull request #24233 that restored the integrated CUDA property check for HIP builds in ggml-cuda. This routine release maintains stability for AMD GPU platform builds by rolling back a single property handler that introduced issues. It reflects the fast-paced continuous deployment cycle of the llama.cpp project. The update specifically undoes commit c7d8722922a2599dc4d77f8808d8e6c2fde5e7a2 via PR #28604 without adding new features. Updated pre-compiled binaries continue to be provided across macOS, Linux, Windows, Android, and backends including ROCm, Vulkan, and CUDA.
## BACKGROUND
llama.cpp is a widely used open-source framework for running Large Language Models efficiently on consumer hardware. AMD's HIP (Heterogeneous-Computing Interface for Portability) is a C++ runtime API that enables developers to compile CUDA code to run on AMD GPUs.