Files
ports/misc/llama.cpp/patches/patch-tests_test-opt_cpp
sthen bfebc5a1d1 update llama.cpp and use new devel/libggml port
rename package llama-cpp -> llama.cpp to match the common name and
port dir name

tweak/ok volker
2025-10-01 19:44:07 +00:00

15 lines
498 B
Plaintext

fix build on ILP32
Index: tests/test-opt.cpp
--- tests/test-opt.cpp.orig
+++ tests/test-opt.cpp
@@ -98,7 +98,7 @@ static helper_ctx_data helper_get_ctx_data(
struct ggml_context * ctx_compute;
{
struct ggml_init_params params = {
- /*.mem_size =*/ (2*ndata + 2)*ggml_tensor_overhead(),
+ /*.mem_size =*/ static_cast<size_t>((2*ndata + 2)*ggml_tensor_overhead()),
/*.mem_buffer =*/ nullptr,
/*.no_alloc =*/ true,
};