added optimized blas level1 dot kernels for single and double precision

This commit is contained in:
wernsaar
2013-11-07 17:22:03 +01:00
parent 8fa93be06e
commit c8f1aeb154
3 changed files with 597 additions and 17 deletions
+12 -17
View File
@@ -45,13 +45,13 @@ DAXPYKERNEL = ../arm/axpy.c
CAXPYKERNEL = ../arm/zaxpy.c
ZAXPYKERNEL = ../arm/zaxpy.c
SCOPYKERNEL = ../arm/copy.c
DCOPYKERNEL = ../arm/copy.c
CCOPYKERNEL = ../arm/zcopy.c
ZCOPYKERNEL = ../arm/zcopy.c
SCOPYKERNEL = scopy_vfpv3.S
DCOPYKERNEL = dcopy_vfpv3.S
CCOPYKERNEL = ccopy_vfpv3.S
ZCOPYKERNEL = zcopy_vfpv3.S
SDOTKERNEL = ../arm/dot.c
DDOTKERNEL = ../arm/dot.c
SDOTKERNEL = sdot_vfpv3.S
DDOTKERNEL = ddot_vfpv3.S
CDOTKERNEL = ../arm/zdot.c
ZDOTKERNEL = ../arm/zdot.c
@@ -108,15 +108,15 @@ DGEMMOTCOPYOBJ = dgemm_otcopy.o
CGEMMKERNEL = cgemm_kernel_2x2_vfpv3.S
CGEMMONCOPY = cgemm_ncopy_2_vfpv3.S
CGEMMOTCOPY = ../generic/zgemm_tcopy_2.c
CGEMMONCOPYOBJ = cgemm_oncopy.o
CGEMMOTCOPYOBJ = cgemm_otcopy.o
CGEMMOTCOPY = cgemm_tcopy_2_vfpv3.S
CGEMMONCOPYOBJ = cgemm_oncopy.o
CGEMMOTCOPYOBJ = cgemm_otcopy.o
ZGEMMKERNEL = zgemm_kernel_2x2_vfpv3.S
ZGEMMONCOPY = zgemm_ncopy_2_vfpv3.S
ZGEMMOTCOPY = ../generic/zgemm_tcopy_2.c
ZGEMMONCOPYOBJ = zgemm_oncopy.o
ZGEMMOTCOPYOBJ = zgemm_otcopy.o
ZGEMMOTCOPY = zgemm_tcopy_2_vfpv3.S
ZGEMMONCOPYOBJ = zgemm_oncopy.o
ZGEMMOTCOPYOBJ = zgemm_otcopy.o
STRSMKERNEL_LN = ../generic/trsm_kernel_LN.c
STRSMKERNEL_LT = ../generic/trsm_kernel_LT.c
@@ -138,9 +138,4 @@ ZTRSMKERNEL_LT = ../generic/trsm_kernel_LT.c
ZTRSMKERNEL_RN = ../generic/trsm_kernel_RN.c
ZTRSMKERNEL_RT = ../generic/trsm_kernel_RT.c
CGEMM3MKERNEL = zgemm3m_kernel_8x4_core2.S
ZGEMM3MKERNEL = zgemm3m_kernel_4x4_core2.S