floraachy
/
OpenBLAS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
OpenBLAS/kernel/arm64
History
Chris Sidebottom 1361229291 Remove prefetches from SVE kernels 
This is a precursor to enabling the SVE kernels for Arm(R) Neoverse(TM)
V1 which has 256-bit SVE. Testing revealed that the SVE kernel was
actually worse in some cases than the existing kernel which seemed odd -
removing these prefetches the underlying architecture seems to do a better job
😸
 		2022-12-16 14:43:09 +00:00
..
KERNEL declare DGEMM_BETA in KERNEL.ARMV8 rather than the generic KERNEL 2019-12-20 10:11:50 +08:00
KERNEL.A64FX add sve ztrsm 2022-01-15 22:27:25 +01:00
KERNEL.ARMV8 Add workaround for NVIDIA HPC mishandling of the asm DOT kernels 2021-01-12 16:38:51 +01:00
KERNEL.ARMV8SVE update armv8sve + contributors 2022-01-18 08:28:31 +01:00
KERNEL.CORTEXA53 optimize cgemm on ARM cortex A53 & cortex A55 2021-12-12 17:22:52 +08:00
KERNEL.CORTEXA55 optimize cgemm on ARM cortex A53 & cortex A55 2021-12-12 17:22:52 +08:00
KERNEL.CORTEXA57 Add workaround for NVIDIA HPC mishandling of the asm DOT kernels 2021-01-12 16:39:35 +01:00
KERNEL.CORTEXA72 Simplifying ARMv8 build parameters 2018-11-19 16:41:49 +00:00
KERNEL.CORTEXA73 Simplifying ARMv8 build parameters 2018-11-19 16:41:49 +00:00
KERNEL.CORTEXA510 Add initial support for Phytium FT2000 series and ARMV9 Cortex 510/710/X1/X2 2022-03-27 15:29:20 +02:00
KERNEL.CORTEXA710 Add initial support for Phytium FT2000 series and ARMV9 Cortex 510/710/X1/X2 2022-03-27 15:29:20 +02:00
KERNEL.CORTEXX1 CortexX1 is ARMV8 like A7x 2022-03-28 17:28:29 +02:00
KERNEL.CORTEXX2 Add initial support for Phytium FT2000 series and ARMV9 Cortex 510/710/X1/X2 2022-03-27 15:29:20 +02:00
KERNEL.EMAG8180 Add preliminary support for EMAG8180 2020-02-19 19:00:28 +01:00
KERNEL.FALKOR Simplifying ARMv8 build parameters 2018-11-19 16:41:49 +00:00
KERNEL.FT2000 Add initial support for Phytium FT2000 series and ARMV9 Cortex 510/710/X1/X2 2022-03-27 15:29:20 +02:00
KERNEL.NEOVERSEN1 Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
KERNEL.NEOVERSEN2 Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
KERNEL.NEOVERSEV1 Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
KERNEL.THUNDERX Add workaround for NVIDIA HPC 2021-01-12 16:49:39 +01:00
KERNEL.THUNDERX2T99 Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
KERNEL.THUNDERX3T110 Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
KERNEL.TSV110 Add workaround for NVIDIA HPC 2021-01-12 16:51:35 +01:00
KERNEL.VORTEX Use Neoverse's current mix of ThunderX2 kernels for Vortex as well 2021-10-06 11:06:43 +02:00
KERNEL.generic Fix MSVC ARM64 build. Add generic kernel for ARM64 2022-06-02 16:53:54 +02:00
Makefile added experimental support for ARMV8 2013-11-24 15:47:00 +01:00
amax.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
asum.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
axpy.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
casum.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
casum_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
cgemm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
cgemm_kernel_8x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
cgemm_kernel_8x4_cortexa53.c optimize cgemm on ARM cortex A53 & cortex A55 2021-12-12 17:22:52 +08:00
cgemm_kernel_8x4_thunderx2t99.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
cgemm_kernel_sve_v1x4.S add cgemm ctrmm sve kernels 2022-01-05 09:09:18 +01:00
cgemm_ncopy_sve_v1.c sve copy functions for cgemm chemm zsymm 2022-01-05 09:12:22 +01:00
cgemm_tcopy_sve_v1.c sve copy functions for cgemm chemm zsymm 2022-01-05 09:12:22 +01:00
copy.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
copy_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
csum.S Add ARM64 implementations of ?sum 2019-03-30 22:13:36 +01:00
ctrmm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
ctrmm_kernel_8x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
ctrmm_kernel_sve_v1x4.S add cgemm ctrmm sve kernels 2022-01-05 09:09:18 +01:00
dasum_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
daxpy_thunderx.c aarch64 fix std=c18 compilation 2020-10-03 18:00:34 +03:00
daxpy_thunderx2t99.S ARM64: Improve DAXPY for ThunderX2 2020-05-07 09:22:50 -07:00
ddot_thunderx.c ARM64: Rename kernel files to have consistent naming 2017-01-24 14:53:34 +05:30
dgemm_beta.S Fix zero initialization for beta=0 case 2020-03-31 00:21:02 +02:00
dgemm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_kernel_4x4_cortexa53.c MOD: optimize normal DGEMM on ARMV8 cortex-A53 & cortex-A55 2021-11-18 21:14:43 +08:00
dgemm_kernel_4x8.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_kernel_8x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_kernel_8x4_thunderx2t99.S ARM64: Move parameters from parameter.c to param.h 2018-10-22 01:45:51 -07:00
dgemm_kernel_sve_v1x8.S some clean-up & commentary 2021-11-21 14:56:27 +01:00
dgemm_kernel_sve_v2x8.S Remove prefetches from SVE kernels 2022-12-16 14:43:09 +00:00
dgemm_ncopy_4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_ncopy_8.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_ncopy_sve_v1.c some clean-up & commentary 2021-11-21 14:56:27 +01:00
dgemm_tcopy_4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dgemm_tcopy_8.S Remove unused TEMP2 and reshuffle to leave x18 unused (reserved on OSX) 2021-09-17 09:18:25 +02:00
dgemm_tcopy_sve_v1.c some clean-up & commentary 2021-11-21 14:56:27 +01:00
dot.S ARM64: Fix utest dsdot errors 2018-02-27 10:47:55 +00:00
dot.c Wrap SVE header with __has_include check 2022-12-01 12:07:55 +00:00
dot_kernel_asimd.c Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
dot_kernel_sve.c Add SVE implementation for sdot/ddot 2022-12-01 12:07:50 +00:00
dot_thunderx.c ARM64: Rename kernel files to have consistent naming 2017-01-24 14:53:34 +05:30
dtrmm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dtrmm_kernel_4x8.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
dtrmm_kernel_8x4.S Move temp to x21 to leave x18 unused (reserved on OSX) 2021-09-17 09:24:11 +02:00
dtrmm_kernel_sve_v1x8.S some clean-up & commentary 2021-11-21 14:56:27 +01:00
dznrm2_thunderx2t99.c workaround fault with ssq=inf,scale=0 2022-07-02 23:47:17 +02:00
dznrm2_thunderx2t99_fast.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
gemv_n.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
gemv_t.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
iamax.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
iamax_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
izamax.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
izamax_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
nrm2.S Fix accidental duplication of jump instruction 2019-10-08 08:09:26 +02:00
rot.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
sasum_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
sbgemm_beta_neoversen2.c neoverse n2 sbgemm: init file 2022-06-29 10:14:21 +08:00
sbgemm_kernel_8x4_neoversen2.c Change file name to match the norm and delete useless code. 2022-10-28 17:09:39 +08:00
sbgemm_kernel_8x4_neoversen2_impl.c Change file name to match the norm and delete useless code. 2022-10-28 17:09:39 +08:00
sbgemm_ncopy_4_neoversen2.c Change file name to match the norm and delete useless code. 2022-10-28 17:09:39 +08:00
sbgemm_tcopy_8_neoversen2.c Change file name to match the norm and delete useless code. 2022-10-28 17:09:39 +08:00
scal.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
scnrm2_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
sgemm_beta.S fix initialization to zero in the NEON SGEMM_BETA kernel as well 2020-03-31 16:53:56 +02:00
sgemm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
sgemm_kernel_8x8.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
sgemm_kernel_8x8_cortexa53.S fix INIT8x4 2020-06-10 01:01:16 +08:00
sgemm_kernel_16x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
sgemm_kernel_16x4_thunderx2t99.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
sgemm_kernel_sve_v1x8.S add sgemm kernel and copy functions for sgemm and ssymm 2021-11-28 18:12:47 +01:00
sgemm_kernel_sve_v2x8.S Remove prefetches from SVE kernels 2022-12-16 14:43:09 +00:00
sgemm_ncopy_4.S change line endings from CRLF to LF 2022-11-16 22:24:01 +01:00
sgemm_ncopy_8.S sgemm copy source init 2020-06-04 02:10:45 +08:00
sgemm_ncopy_sve_v1.c add sgemm kernel and copy functions for sgemm and ssymm 2021-11-28 18:12:47 +01:00
sgemm_tcopy_8.S sgemm copy source init 2020-06-04 02:10:45 +08:00
sgemm_tcopy_16.S change line endings from CRLF to LF 2022-11-16 22:24:01 +01:00
sgemm_tcopy_sve_v1.c add sgemm kernel and copy functions for sgemm and ssymm 2021-11-28 18:12:47 +01:00
strmm_kernel_4x4.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
strmm_kernel_8x8.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
strmm_kernel_8x8_cortexa53.S use general register to speedup 2020-05-20 22:26:58 +08:00
strmm_kernel_16x4.S Move temp to x21 to leave x18 unused (reserved on OSX) 2021-09-17 09:28:19 +02:00
strmm_kernel_sve_v1x8.S strmm sve v1x8 kernel 2021-12-05 14:03:08 +01:00
sum.S Add ARM64 implementations of ?sum 2019-03-30 22:13:36 +01:00
swap.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
swap_thunderx2t99.S THUNDERX2T99: Add optimized S/D/C/Z SWAP Implementations 2017-02-03 03:55:06 -08:00
symm_lcopy_sve.c add sgemm kernel and copy functions for sgemm and ssymm 2021-11-28 18:12:47 +01:00
symm_ucopy_sve.c add sgemm kernel and copy functions for sgemm and ssymm 2021-11-28 18:12:47 +01:00
trmm_lncopy_sve_v1.c trmm sve copy fucntions for single precision 2021-11-29 21:25:05 +01:00
trmm_ltcopy_sve_v1.c trmm sve copy fucntions for single precision 2021-11-29 21:25:05 +01:00
trmm_uncopy_sve_v1.c trmm sve copy fucntions for single precision 2021-11-29 21:25:05 +01:00
trmm_utcopy_sve_v1.c trmm sve copy fucntions for single precision 2021-11-29 21:25:05 +01:00
trsm_kernel_LN_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_kernel_LT_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_kernel_RN_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_kernel_RT_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_lncopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_ltcopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_uncopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
trsm_utcopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
zamax.S Fix the functional bugs for zamax. 2020-03-09 15:36:50 +08:00
zasum.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zasum_thunderx2t99.c Fixed a few more unnecessary calls to num_cpu_avail. 2018-06-11 10:17:16 +01:00
zaxpy.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zdot.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zdot_thunderx2t99.c Eliminate uses of CREAL on left-hand side of assignments 2022-07-05 00:01:09 +02:00
zgemm_kernel_4x4.S move alpha to x19/x20 to leave x18 unused for OSX 2021-09-17 09:42:17 +02:00
zgemm_kernel_4x4_cortexa53.c MOD: add comments to a53 zgemm kernel 2021-11-25 22:48:48 +08:00
zgemm_kernel_4x4_thunderx2t99.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zgemm_kernel_sve_v1x4.S fix zgemm kernel 2021-12-29 11:42:04 +01:00
zgemm_ncopy_sve_v1.c modify sve zgemmcopy kernels 2022-01-05 09:07:28 +01:00
zgemm_tcopy_sve_v1.c modify sve zgemmcopy kernels 2022-01-05 09:07:28 +01:00
zgemv_n.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zgemv_t.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zhemm_ltcopy_sve.c combine zchemm into single file 2022-01-05 14:42:37 +01:00
zhemm_utcopy_sve.c combine zchemm into single file 2022-01-05 14:42:37 +01:00
znrm2.S Remove automatic label postfixes from macro included only once 2019-10-08 08:37:50 +02:00
zrot.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zscal.S ARM64: Convert all labels to local labels 2017-10-24 11:40:05 +00:00
zsum.S Add ARM64 implementations of ?sum 2019-03-30 22:13:36 +01:00
zsymm_lcopy_sve.c sve copy functions for cgemm chemm zsymm 2022-01-05 09:12:22 +01:00
zsymm_ucopy_sve.c sve copy functions for cgemm chemm zsymm 2022-01-05 09:12:22 +01:00
ztrmm_kernel_4x4.S Move alphaI to x22 to leave x18 unused (reserved on OSX) 2021-09-17 09:53:18 +02:00
ztrmm_kernel_sve_v1x4.S fix sve ztrmm kernel 2022-01-04 14:42:07 +01:00
ztrmm_lncopy_sve_v1.c ztrmm sve copy functions 2022-01-04 14:40:59 +01:00
ztrmm_ltcopy_sve_v1.c ztrmm sve copy functions 2022-01-04 14:40:59 +01:00
ztrmm_uncopy_sve_v1.c ztrmm sve copy functions 2022-01-04 14:40:59 +01:00
ztrmm_utcopy_sve_v1.c ztrmm sve copy functions 2022-01-04 14:40:59 +01:00
ztrsm_lncopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
ztrsm_ltcopy_sve.c fix ztrsm lt/ut copy 2022-01-16 21:39:57 +01:00
ztrsm_uncopy_sve.c add sve ztrsm 2022-01-15 22:27:25 +01:00
ztrsm_utcopy_sve.c fix ztrsm lt/ut copy 2022-01-16 21:39:57 +01:00