mirror of
https://github.com/ollama/ollama.git
synced 2026-10-21 02:51:35 +00:00
2e036e7cdf
The MLX runner is the only Go inference runner left and is no longer experimental, so its packages leave x/. The bindings become a top-level mlx package beside the carried patches in mlx/compat, mirroring how llama/ holds the llama.cpp integration, and the runner becomes mlxrunner with the architectures nested under the package they implement. Subpackages move with their parent unless listed. x/mlxrunner/mlx mlx x/internal/mlxthread mlx/mlxthread x/internal/mlxthreadtest mlx/mlxthread/mlxthreadtest x/internal/mlxtest mlx/mlxtest x/quant mlx/quant mlx/compat/*.patch mlx/compat/mlx-c (MLX patches go in mlx/compat/mlx) x/mlxrunner mlxrunner x/models/nn mlxrunner/nn x/models/<arch> mlxrunner/model/<arch> x/mlxrunner/imports.go mlxrunner/model/architectures (new package) x/create create x/safetensors fs/safetensors x/tokenizer mlxrunner/tokenizer Every package keeps its name, so the Go changes are the import path rewrites the moves force, and the CMake, Dockerfile, CI cache keys, drift check and Darwin payload script follow the new paths. Four edits are not paths: the runner's blank architecture imports become the package mlxrunner/model/architectures, so the list to extend for a new model sits beside the architecture directories; a depguard rule keeps the two test harnesses out of non-test code, as the x/internal placement used to; the CI change filter's two entries for the long-deleted x/imagegen/mlx now name the bindings' CMake project and the carried patches, so a change to either builds the payload; and the tokenizer parity test reads its fixtures from its own testdata instead of walking out of x/. x/server and x/imagegen/manifest stay for the next two commits.
47 lines
1.4 KiB
C
47 lines
1.4 KiB
C
#ifndef MLX_DYNAMIC_H
|
|
#define MLX_DYNAMIC_H
|
|
|
|
#ifdef _WIN32
|
|
#include <windows.h>
|
|
#define DLSYM(handle, symbol) (void*)GetProcAddress((HMODULE)(handle.ctx), symbol)
|
|
#else
|
|
#include <dlfcn.h>
|
|
#define DLSYM(handle, symbol) dlsym(handle.ctx, symbol)
|
|
#endif
|
|
|
|
#include <stdint.h>
|
|
|
|
// Provide fallback typedefs for float16_t and bfloat16_t on non-ARM64
|
|
// platforms where arm_fp16.h and arm_bf16.h are not available. These are
|
|
// only used as function pointer signature placeholders since MLX requires
|
|
// Apple Silicon at runtime.
|
|
#if !defined(__aarch64__) && !defined(__ARM_FEATURE_FP16_SCALAR_ARITHMETIC)
|
|
typedef uint16_t float16_t;
|
|
#endif
|
|
|
|
#if !defined(__aarch64__) && !defined(__ARM_FEATURE_BF16)
|
|
typedef uint16_t bfloat16_t;
|
|
#endif
|
|
|
|
// Undef ERROR to avoid conflict with wingdi.h on Windows
|
|
#ifdef ERROR
|
|
#undef ERROR
|
|
#endif
|
|
#define MLX_ERROR(fmt, ...) fprintf(stderr, "%s %s - ERROR - %s:%d - " fmt "\n", __DATE__, __TIME__, __FILE__, __LINE__, ##__VA_ARGS__); return 1
|
|
#define CHECK(x) if (!(x)) { MLX_ERROR("CHECK failed: " #x); }
|
|
#define CHECK_LOAD(handle, x) *(void**)(&x##_) = DLSYM(handle, #x); CHECK(x##_)
|
|
// OPTIONAL_LOAD: load symbol if available, leave function pointer NULL otherwise
|
|
#define OPTIONAL_LOAD(handle, x) *(void**)(&x##_) = DLSYM(handle, #x)
|
|
|
|
typedef struct {
|
|
void* ctx;
|
|
} mlx_dynamic_handle;
|
|
|
|
int mlx_dynamic_load(
|
|
mlx_dynamic_handle* handle,
|
|
const char *path);
|
|
|
|
void mlx_dynamic_unload(
|
|
mlx_dynamic_handle* handle);
|
|
|
|
#endif // MLX_DYNAMIC_H
|