Skip to content

Commit 643a0b5

Browse files
authored
Allow VortexM4 on the direct_SME fast path only for clang-based compilers
1 parent d7b0fcc commit 643a0b5

File tree

1 file changed

+5
-1
lines changed

1 file changed

+5
-1
lines changed

interface/gemm.c

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -555,7 +555,11 @@ void CNAME(enum CBLAS_ORDER order, enum CBLAS_TRANSPOSE TransA, enum CBLAS_TRANS
555555
#endif
556556
#if defined(ARCH_ARM64) && (defined(USE_SGEMM_KERNEL_DIRECT)||defined(DYNAMIC_ARCH))
557557
#if defined(DYNAMIC_ARCH)
558-
if (strcmp(gotoblas_corename(), "armv9sme") == 0 || strcmp(gotoblas_corename(), "vortexm4") == 0)
558+
if (strcmp(gotoblas_corename(), "armv9sme") == 0
559+
#if defined(__clang__)
560+
|| strcmp(gotoblas_corename(), "vortexm4") == 0
561+
#endif
562+
)
559563
// if (support_sme1())
560564
#endif
561565
if (order == CblasRowMajor && m==lda && n ==ldb && k==ldc && beta == 0 && alpha == 1.0 && TransA == CblasNoTrans && TransB == CblasNoTrans&& SGEMM_DIRECT_PERFORMANT(m,n,k)) {

0 commit comments

Comments
 (0)