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

fix race condition bug in non-inplace ggml_compute_forward_diag_mask_f32

Browse source 
memcpy needs to be synchronized across threads to avoid race conditions.
=> do it in INIT phase
This commit is contained in:
xaedes 2023-05-14 17:00:19 +02:00
parent 60f8c361ca
commit 1e6b5bf111
No known key found for this signature in database
GPG key ID: 30030EDD817EA2B1
1 changed files with 13 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

19
ggml.c
View file

@ -10450,18 +10450,25 @@ static void ggml_compute_forward_diag_mask_f32(
assert(src1->type == GGML_TYPE_I32); assert(src1->type == GGML_TYPE_I32);
assert(ggml_nelements(src1) == 2); assert(ggml_nelements(src1) == 2);
if (params->type == GGML_TASK_INIT || params->type == GGML_TASK_FINALIZE) {
return;
}
const int ith = params->ith; const int ith = params->ith;
const int nth = params->nth; const int nth = params->nth;
const int n_past = ((int32_t *) src1->data)[0]; const int n_past = ((int32_t *) src1->data)[0];
const bool inplace = (bool)((int32_t *) src1->data)[1]; const bool inplace = (bool)((int32_t *) src1->data)[1];
if (!inplace) { if (!inplace && (params->type == GGML_TASK_INIT)) {
ggml_compute_forward_dup_same_cont(params, src0, dst); // memcpy needs to be synchronized across threads to avoid race conditions.
// => do it in INIT phase
GGML_ASSERT(ggml_nelements(dst) == ggml_nelements(src0));
GGML_ASSERT(ggml_is_contiguous(dst) && ggml_is_contiguous(src0));
memcpy(
((char *) dst->data),
((char *) src0->data),
ggml_nbytes(dst));
}
if (params->type == GGML_TASK_INIT || params->type == GGML_TASK_FINALIZE) {
return;
} }
// TODO: handle transposed/permuted matrices // TODO: handle transposed/permuted matrices