ggml : fix GGML_F32_VEC_FMA argument order in ggml_vec_mad1_f32 (llama/16307)

* fix GGML_F32_VEC_FMA argument order in ggml_vec_mad1_f32

* add test that fails on simd
This commit is contained in:
Sigbjørn Skjæret 2025-09-28 23:15:03 +02:00 committed by Georgi Gerganov
parent 55d45edf6d
commit 5c6e795607
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
1 changed files with 1 additions and 1 deletions

View File

@ -610,7 +610,7 @@ inline static void ggml_vec_mad1_f32(const int n, float * y, const float * x, co
for (int i = 0; i < np; i += GGML_F32_STEP) {
for (int j = 0; j < GGML_F32_ARR; j++) {
ay[j] = GGML_F32_VEC_LOAD(x + i + j*GGML_F32_EPR);
ay[j] = GGML_F32_VEC_FMA(ay[j], vs, vb);
ay[j] = GGML_F32_VEC_FMA(vb, ay[j], vs);
GGML_F32_VEC_STORE(y + i + j*GGML_F32_EPR, ay[j]);
}