llama.cpp/ggml
hogeheer499-commits 97ae46e460 use max of hipMemGetInfo and /proc/meminfo for UMA free memory
instead of always overwriting with /proc/meminfo, take whichever
is higher. this way systems where hipMemGetInfo already reports
TTM-backed memory correctly (like Strix Halo 128GB) keep their
value, while systems where /proc/meminfo is higher still get the
full amount.

removes the HIP-specific #ifdef since the max approach works for
both CUDA and HIP.
2026-03-19 20:31:52 +01:00
..
cmake ggml: Skip backend library linking code when GGML_BACKEND_DL=ON (#15094) 2025-08-07 13:45:41 +02:00
include ggml : restore ggml_type_sizef() to aboid major version bump (ggml/1441) 2026-03-18 15:17:28 +02:00
src use max of hipMemGetInfo and /proc/meminfo for UMA free memory 2026-03-19 20:31:52 +01:00
.gitignore vulkan : cmake integration (#8119) 2024-07-13 18:12:39 +02:00
CMakeLists.txt ggml : bump version to 0.9.8 (ggml/1442) 2026-03-18 15:17:28 +02:00