Browse Source

added kernels for cgemm, ctrmm, zgemm and ztrmm

tags/v0.2.9.rc1
wernsaar 12 years ago
parent
commit
85484a42df
5 changed files with 5657 additions and 16 deletions
  1. +21
    -16
      kernel/arm/KERNEL.ARMV7
  2. +1293
    -0
      kernel/arm/cgemm_kernel_2x2_vfpv3.S
  3. +1476
    -0
      kernel/arm/ctrmm_kernel_2x2_vfpv3.S
  4. +1329
    -0
      kernel/arm/zgemm_kernel_2x2_vfpv3.S
  5. +1538
    -0
      kernel/arm/ztrmm_kernel_2x2_vfpv3.S

+ 21
- 16
kernel/arm/KERNEL.ARMV7 View File

@@ -80,36 +80,41 @@ DGEMVTKERNEL = gemv_t.c
CGEMVTKERNEL = zgemv_t.c
ZGEMVTKERNEL = zgemv_t.c

STRMMKERNEL = ../generic/trmmkernel_2x2.c
DTRMMKERNEL = dtrmm_kernel_4x4_vfpv3.S
CTRMMKERNEL = ../generic/ztrmmkernel_2x2.c
ZTRMMKERNEL = ../generic/ztrmmkernel_2x2.c

SGEMMKERNEL = ../generic/gemmkernel_2x2.c
SGEMMONCOPY = ../generic/gemm_ncopy_2.c
SGEMMOTCOPY = ../generic/gemm_tcopy_2.c
STRMMKERNEL = strmm_kernel_4x4_vfpv3.S
DTRMMKERNEL = dtrmm_kernel_4x4_vfpv3.S
CTRMMKERNEL = ctrmm_kernel_2x2_vfpv3.S
ZTRMMKERNEL = ztrmm_kernel_2x2_vfpv3.S

#SGEMMKERNEL = ../generic/gemmkernel_2x2.c
SGEMMKERNEL = sgemm_kernel_4x4_vfpv3.S
SGEMMINCOPY =
SGEMMITCOPY =
SGEMMONCOPY = ../generic/gemm_ncopy_4.c
SGEMMOTCOPY = ../generic/gemm_tcopy_4.c
SGEMMINCOPYOBJ =
SGEMMITCOPYOBJ =
SGEMMONCOPYOBJ = sgemm_oncopy.o
SGEMMOTCOPYOBJ = sgemm_otcopy.o

#DGEMMKERNEL = ../generic/gemmkernel_2x2.c
#DGEMMKERNEL = dgemm_kernel_4x2_vfpv2.S
DGEMMKERNEL = dgemm_kernel_4x4_vfpv3.S
DGEMMINCOPY =
DGEMMITCOPY =
DGEMMONCOPY = ../generic/gemm_ncopy_4.c
DGEMMOTCOPY = ../generic/gemm_tcopy_4.c
DGEMMINCOPYOBJ =
DGEMMITCOPYOBJ =
DGEMMINCOPY = dgemm_ncopy_4_vfpv3.S
DGEMMITCOPY = ../generic/gemm_tcopy_4.c
DGEMMONCOPY = dgemm_ncopy_4_vfpv3.S
DGEMMOTCOPY = ../generic/gemm_tcopy_4.c
DGEMMINCOPYOBJ = dgemm_incopy.o
DGEMMITCOPYOBJ = dgemm_itcopy.o
DGEMMONCOPYOBJ = dgemm_oncopy.o
DGEMMOTCOPYOBJ = dgemm_otcopy.o

CGEMMKERNEL = ../generic/zgemmkernel_2x2.c
CGEMMKERNEL = cgemm_kernel_2x2_vfpv3.S
CGEMMONCOPY = ../generic/zgemm_ncopy_2.c
CGEMMOTCOPY = ../generic/zgemm_tcopy_2.c
CGEMMONCOPYOBJ = cgemm_oncopy.o
CGEMMOTCOPYOBJ = cgemm_otcopy.o

ZGEMMKERNEL = ../generic/zgemmkernel_2x2.c
ZGEMMKERNEL = zgemm_kernel_2x2_vfpv3.S
ZGEMMONCOPY = ../generic/zgemm_ncopy_2.c
ZGEMMOTCOPY = ../generic/zgemm_tcopy_2.c
ZGEMMONCOPYOBJ = zgemm_oncopy.o


+ 1293
- 0
kernel/arm/cgemm_kernel_2x2_vfpv3.S
File diff suppressed because it is too large
View File


+ 1476
- 0
kernel/arm/ctrmm_kernel_2x2_vfpv3.S
File diff suppressed because it is too large
View File


+ 1329
- 0
kernel/arm/zgemm_kernel_2x2_vfpv3.S
File diff suppressed because it is too large
View File


+ 1538
- 0
kernel/arm/ztrmm_kernel_2x2_vfpv3.S
File diff suppressed because it is too large
View File


Loading…
Cancel
Save