Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 13:58:46 +01:00
Code Issues Packages Projects Releases Wiki Activity

metal : use vm_allocate instead of posix_memalign on macOS (#7078)

Browse Source 
* fix: use `malloc` instead of `posix_memalign` in `ggml-metal.m` to make it not crash Electron proccesses

* fix: typo

* fix: use `vm_allocate` instead of `posix_memalign`

* fix: don't call `newBufferWithBytesNoCopy` with `NULL` when `ggml_metal_host_malloc` returns `NULL`

* fix: use `vm_allocate` only on macOS
This commit is contained in:
Gilad S 2024-05-08 22:08:10 +03:00 committed by GitHub
parent 83330d8cd6
commit 26458af1d6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 22 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
ggml-metal.m
View File

@ -265,11 +265,20 @@ static void ggml_metal_log(enum ggml_log_level level, const char * format, ...){
static void * ggml_metal_host_malloc(size_t n) { static void * ggml_metal_host_malloc(size_t n) {
void * data = NULL; void * data = NULL;
#if TARGET_OS_OSX
kern_return_t err = vm_allocate((vm_map_t) mach_task_self(), (void *) &data, n, VM_FLAGS_ANYWHERE);
if (err != KERN_SUCCESS) {
GGML_METAL_LOG_ERROR("%s: error: vm_allocate failed\n", __func__);
return NULL;
}
#else
const int result = posix_memalign((void **) &data, sysconf(_SC_PAGESIZE), n); const int result = posix_memalign((void **) &data, sysconf(_SC_PAGESIZE), n);
if (result != 0) { if (result != 0) {
GGML_METAL_LOG_ERROR("%s: error: posix_memalign failed\n", __func__); GGML_METAL_LOG_ERROR("%s: error: posix_memalign failed\n", __func__);
return NULL; return NULL;
} }
#endif
return data; return data;
} }
@ -2840,7 +2849,11 @@ GGML_CALL static void ggml_backend_metal_buffer_free_buffer(ggml_backend_buffer_
ggml_backend_metal_free_device(); ggml_backend_metal_free_device();
if (ctx->owned) { if (ctx->owned) {
#if TARGET_OS_OSX
vm_deallocate((vm_map_t)mach_task_self(), (vm_address_t)ctx->all_data, ctx->all_size);
#else
free(ctx->all_data); free(ctx->all_data);
#endif
} }
free(ctx); free(ctx);
@ -2944,14 +2957,16 @@ GGML_CALL static ggml_backend_buffer_t ggml_backend_metal_buffer_type_alloc_buff
ctx->owned = true; ctx->owned = true;
ctx->n_buffers = 1; ctx->n_buffers = 1;
if (ctx->all_data != NULL) {
ctx->buffers[0].data = ctx->all_data; ctx->buffers[0].data = ctx->all_data;
ctx->buffers[0].size = size; ctx->buffers[0].size = size;
ctx->buffers[0].metal = [device newBufferWithBytesNoCopy:ctx->all_data ctx->buffers[0].metal = [device newBufferWithBytesNoCopy:ctx->all_data
length:size_aligned length:size_aligned
options:MTLResourceStorageModeShared options:MTLResourceStorageModeShared
deallocator:nil]; deallocator:nil];
}
if (ctx->buffers[0].metal == nil) { if (ctx->all_data == NULL || ctx->buffers[0].metal == nil) {
GGML_METAL_LOG_ERROR("%s: error: failed to allocate buffer, size = %8.2f MiB\n", __func__, size_aligned / 1024.0 / 1024.0); GGML_METAL_LOG_ERROR("%s: error: failed to allocate buffer, size = %8.2f MiB\n", __func__, size_aligned / 1024.0 / 1024.0);
free(ctx); free(ctx);
ggml_backend_metal_free_device(); ggml_backend_metal_free_device();