mirror of
https://github.com/amd/blis.git
synced 2026-04-20 07:38:53 +00:00
Bug fix in BF16-F32 supported AVX2 Kernels
- Bug fix in Matrix Mul post op. - Updated the config in AVX512_VNNI_BF16 context to work in AVX2 kernels Change-Id: I25980508facc38606596402dba4cfce88f4eb173
This commit is contained in:
@@ -54,6 +54,8 @@
|
||||
|
||||
#define LPGEMM_KERN_FUNC_UPD_MAP_AVX512_VNNI_BF16_TO_AVX2 \
|
||||
KMACRO(F32F32F32OF32, lpgemm_rowvar_f32f32f32of32_6x16m) \
|
||||
KMACRO(BF16BF16F32OF32, NULL) \
|
||||
KMACRO(BF16S4F32OF32, NULL) \
|
||||
|
||||
#define LPGEMM_PACKA_FUNC_MAP_AVX512_VNNI_BF16 \
|
||||
PAMACRO(U8S8S32OS32, packa_u8s8s32os32) \
|
||||
|
||||
@@ -16183,8 +16183,8 @@ POST_OPS_MATRIX_MUL_1x1F:
|
||||
// c[0:0-15]
|
||||
F32_F32_MATRIX_MUL_1COL_XMM_1ELE(xmm1,scl_fctr1,0,4);
|
||||
}
|
||||
POST_OP_LABEL_LASTK_SAFE_JUMP_WITH_NEXT_PTR
|
||||
}
|
||||
POST_OP_LABEL_LASTK_SAFE_JUMP_WITH_NEXT_PTR
|
||||
}
|
||||
POST_OPS_SWISH_1x1F:
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user