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

ggml : reading the runtime sve config of the cpu (#8709)

Browse source 
* ggml : reading the runtime sve config of the cpu

* change to one time init to prevent performance drop

* prefix variable to avoid possible conflicts

* revert xxhash fix and add brackets

---------

Co-authored-by: domke <673751-domke@users.noreply.gitlab.com>
This commit is contained in:
jdomke 2024-08-04 01:34:41 +09:00 • committed by GitHub
parent b72c20b85c
commit 76614f352e
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 30 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

9
ggml/src/ggml.c
View file

@ -37,6 +37,9 @@
#include <unistd.h>
#endif
#if defined(__ARM_FEATURE_SVE)
int ggml_sve_cnt_b = 0;
#endif
#if defined(__ARM_FEATURE_SVE) || defined(__ARM_FEATURE_MATMUL_INT8)
#undef GGML_USE_LLAMAFILE
#endif
@ -3558,6 +3561,12 @@ struct ggml_context * ggml_init(struct ggml_init_params params) {
GGML_ASSERT_ALIGNED(ctx->mem_buffer);
#if defined(__ARM_FEATURE_SVE)
if (!ggml_sve_cnt_b) {
ggml_sve_cnt_b = PR_SVE_VL_LEN_MASK & prctl(PR_SVE_GET_VL);
}
#endif
GGML_PRINT_DEBUG("%s: context initialized\n", __func__);
ggml_critical_section_end();