vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

metal : enable ggml-alloc (#2627)

Browse source 
* metal: enable ggml-alloc

Make ggml-alloc work with concurrently dispatch.

* style-fix

Co-authored-by: slaren <slarengh@gmail.com>

---------

Co-authored-by: slaren <slarengh@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Shouzheng Liu 2023-08-16 16:08:28 -04:00 • committed by GitHub
parent bf83bff674
commit fc8ef549e5
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
5 changed files with 61 additions and 26 deletions
Download patch file Download diff file Expand all files Collapse all files

15
ggml-metal.m
View file

@ -236,11 +236,12 @@ void ggml_metal_set_n_cb(struct ggml_metal_context * ctx, int n_cb) {
ctx->n_cb = n_cb;
}
bool ggml_metal_if_optimized(struct ggml_metal_context * ctx) {
if (ctx->concur_list_len) {
return true;
}
return false;
int ggml_metal_if_optimized(struct ggml_metal_context * ctx) {
return ctx->concur_list_len;
}
int * ggml_metal_get_concur_list(struct ggml_metal_context * ctx) {
return ctx->concur_list;
}
// finds the Metal buffer that contains the tensor data on the GPU device
@ -383,7 +384,7 @@ void ggml_metal_get_tensor(
void ggml_metal_graph_find_concurrency(
struct ggml_metal_context * ctx,
struct ggml_cgraph * gf) {
struct ggml_cgraph * gf, bool check_mem) {
int search_depth = gf->n_nodes; //we only find concurrency in this range to avoid wasting too much time
int nodes_unused[GGML_MAX_CONCUR];
@ -430,7 +431,7 @@ void ggml_metal_graph_find_concurrency(
}
}
}
if (exe_flag) {
if (exe_flag && check_mem) {
// check if nodes[i]'s data will be overwritten by a node before nodes[i].
// if node[5] and node[3] write to the same memory region, then we can't issue node[5] before node[3]
int64_t data_start = (int64_t) gf->nodes[i]->data;