From 89fe17f20e7d1d10a7ec3315bf9b1816a3d47ce9 Mon Sep 17 00:00:00 2001 From: Marius Hillenbrand Date: Tue, 19 May 2020 14:56:34 +0200 Subject: [PATCH] s390x: Use new sgemm kernel also for DGEMM and DTRMM on Z14 Apply our new GEMM kernel implementation, written in C with vector intrinsics, also for DGEMM and DTRMM on Z14 and newer (i.e., architectures with FP32 SIMD instructions). As a result, we gain around 10% in performance on z15, in addition to improving maintainability. Signed-off-by: Marius Hillenbrand --- kernel/zarch/KERNEL.Z14 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/zarch/KERNEL.Z14 b/kernel/zarch/KERNEL.Z14 index 96e6745fd..3510938a7 100644 --- a/kernel/zarch/KERNEL.Z14 +++ b/kernel/zarch/KERNEL.Z14 @@ -87,7 +87,7 @@ CGEMVTKERNEL = cgemv_t_4.c ZGEMVTKERNEL = zgemv_t_4.c STRMMKERNEL = gemm_vec.c -DTRMMKERNEL = trmm8x4V.S +DTRMMKERNEL = gemm_vec.c CTRMMKERNEL = ctrmm4x4V.S ZTRMMKERNEL = ztrmm4x4V.S @@ -103,7 +103,7 @@ SGEMMOTCOPY = ../generic/gemm_tcopy_$(SGEMM_UNROLL_N).c SGEMMONCOPYOBJ = sgemm_oncopy$(TSUFFIX).$(SUFFIX) SGEMMOTCOPYOBJ = sgemm_otcopy$(TSUFFIX).$(SUFFIX) -DGEMMKERNEL = gemm8x4V.S +DGEMMKERNEL = gemm_vec.c DGEMMINCOPY = ../generic/gemm_ncopy_8.c DGEMMITCOPY = ../generic/gemm_tcopy_8.c DGEMMONCOPY = ../generic/gemm_ncopy_4.c