ggml-cuda: avoid direct ROCm_Host compute on HIP integrated GPUs (port of ggml-org#25863) - #158

ggml-cuda: avoid direct ROCm_Host compute on HIP integrated GPUs (port of ggml-org#25863)#158

danielhanchen wants to merge 2 commits into

base/upstream-fcac19079unslothai/llama.cpp:base/upstream-fcac19079from

port/25863-hip-integrated-host-buftunslothai/llama.cpp:port/25863-hip-integrated-host-buftCopy head branch name to clipboard