llama.cpp b10001 fixes segfault in test-export-graph-ops

The llama.cpp project released version b10001, which includes a fix for a segmentation fault occurring when the `test-export-graph-ops` utility is invoked without arguments.

• The release provides binaries for macOS (Apple Silicon and Intel), iOS, Linux (Ubuntu x64, arm64, s390x with CPU, Vulkan, ROCm 7.2, OpenVINO, SYCL FP32/FP16), Android (arm64), Windows (CPU, OpenCL Adreno, CUDA 12/13, Vulkan, OpenVINO, SYCL, HIP), and openEuler.
• macOS Apple Silicon KleidiAI support is disabled in this build.
• openEuler x86 support is also disabled, while aarch64 support remains available.
• A standalone UI binary is included in the release assets.