OpenMathLib/OpenBLAS
1
0
Fork 0
mirror of https://github.com/OpenMathLib/OpenBLAS synced 2026-06-15 07:51:43 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Merge branch 'develop' of https://github.com/OpenMathLib/OpenBLAS into develop

Browse Source
This commit is contained in:
Chip Kerchner
2025-09-30 20:16:19 +00:00
parent b27a1184fd aaa5c377bc
commit dd7a1d6d4d
10 changed files with 848 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Makefile.system
View File

@@ -441,7 +441,7 @@ ifeq (x$(XCVER), x 15)
CCOMMON_OPT += -Wl,-ld_classic
FCOMMON_OPT += -Wl,-ld_classic
endif
ifeq (x$(XCVER), x 16)
ifeq ($(shell [ $(XCVER) -ge 16 ] && echo yes),yes)
override CEXTRALIB := $(filter-out(-lto_library, $(CEXTRALIB)))
endif
endif

2
kernel/power/ddot_microk_power10.c
View File

@@ -119,7 +119,7 @@ static double ddot_kernel_8 (long n, double *x, double *y)
"+b" (y) // 3
:
"m" (*(const double (*)[n]) x),
"m" (*(const double (*)[n]) y),
"m" (*(const double (*)[n]) y)
:
"cr0",
"vs32","vs33","vs34","vs35","vs36","vs37","vs38","vs39",

2
kernel/power/sasum_microk_power10.c
View File

@@ -141,7 +141,7 @@ static float sasum_kernel_32 (long n, float *x)
"=wa" (t2), // 5
"=wa" (t3) // 6
:
"m" (*(const float (*)[n]) x),
"m" (*(const float (*)[n]) x)
:
"cr0",
"vs32","vs33","vs34","vs35","vs36","vs37","vs38","vs39",

2
kernel/power/zaxpy_microk_power10.c
View File

@@ -225,7 +225,7 @@ static void zaxpy_kernel_4 (long n, double *x, double *y,
"=b" (ytmp) // 12
:
"m" (*(const double (*)[n * 2]) x),
"m" (*(const double (*)[2]) mvecp)
"m" (*(const double (*)[2]) mvecp),
"d" (alpha_r), // 15
"d" (alpha_i), // 16

18
kernel/riscv64/KERNEL.RISCV64_ZVL128B
View File

@@ -120,13 +120,21 @@ endif
DGEMMKERNEL = dgemm_kernel_$(DGEMM_UNROLL_M)x$(DGEMM_UNROLL_N)_zvl128b.c
DGEMMONCOPY = ../generic/gemm_ncopy_$(DGEMM_UNROLL_N).c
ifneq ($(filter $(DGEMM_UNROLL_N),4 8 16),)
DGEMMOTCOPY = gemm_tcopy_$(DGEMM_UNROLL_N)_rvv.c
else
DGEMMOTCOPY = ../generic/gemm_tcopy_$(DGEMM_UNROLL_N).c
endif
DGEMMONCOPYOBJ = dgemm_oncopy$(TSUFFIX).$(SUFFIX)
DGEMMOTCOPYOBJ = dgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifneq ($(DGEMM_UNROLL_M), $(DGEMM_UNROLL_N))
DGEMMINCOPY = ../generic/gemm_ncopy_$(DGEMM_UNROLL_M).c
ifneq ($(filter $(DGEMM_UNROLL_M),4 8 16),)
DGEMMITCOPY = gemm_tcopy_$(DGEMM_UNROLL_M)_rvv.c
else
DGEMMITCOPY = ../generic/gemm_tcopy_$(DGEMM_UNROLL_M).c
endif
DGEMMINCOPYOBJ = dgemm_incopy$(TSUFFIX).$(SUFFIX)
DGEMMITCOPYOBJ = dgemm_itcopy$(TSUFFIX).$(SUFFIX)
endif
@@ -258,8 +266,13 @@ endif
ifeq ($(BUILD_HFLOAT16), 1)
SHGEMMKERNEL = shgemm_kernel_$(SHGEMM_UNROLL_M)x$(SHGEMM_UNROLL_N)_zvl128b.c
ifneq ($(filter $(SHGEMM_UNROLL_N),8 16),)
SHGEMMONCOPY = gemm_ncopy_$(SHGEMM_UNROLL_N)fp_rvv.c
SHGEMMOTCOPY = gemm_tcopy_$(SHGEMM_UNROLL_N)fp_rvv.c
else
SHGEMMONCOPY = ../generic/gemm_ncopy_$(SHGEMM_UNROLL_N).c
SHGEMMOTCOPY = ../generic/gemm_tcopy_$(SHGEMM_UNROLL_N).c
endif
SHGEMMONCOPYOBJ = shgemm_oncopy$(TSUFFIX).$(SUFFIX)
SHGEMMOTCOPYOBJ = shgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifndef SHGEMM_BETA
@@ -269,8 +282,13 @@ endif
ifeq ($(BUILD_BFLOAT16), 1)
SBGEMMKERNEL = sbgemm_kernel_$(SBGEMM_UNROLL_M)x$(SBGEMM_UNROLL_N)_zvl128b.c
ifneq ($(filter $(SBGEMM_UNROLL_N),8 16),)
SBGEMMONCOPY = gemm_ncopy_$(SBGEMM_UNROLL_N)fp_rvv.c
SBGEMMOTCOPY = gemm_tcopy_$(SBGEMM_UNROLL_N)fp_rvv.c
else
SBGEMMONCOPY = ../generic/gemm_ncopy_$(SBGEMM_UNROLL_N).c
SBGEMMOTCOPY = ../generic/gemm_tcopy_$(SBGEMM_UNROLL_N).c
endif
SBGEMMONCOPYOBJ = sbgemm_oncopy$(TSUFFIX).$(SUFFIX)
SBGEMMOTCOPYOBJ = sbgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifndef SBGEMM_BETA

28
kernel/riscv64/KERNEL.RISCV64_ZVL256B
View File

@@ -119,12 +119,20 @@ endif
DGEMMKERNEL = dgemm_kernel_$(DGEMM_UNROLL_M)x$(DGEMM_UNROLL_N)_zvl256b.c
DGEMMONCOPY = ../generic/gemm_ncopy_$(DGEMM_UNROLL_N).c
ifneq ($(filter $(DGEMM_UNROLL_N),4 8 16),)
DGEMMOTCOPY = gemm_tcopy_$(DGEMM_UNROLL_N)_rvv.c
else
DGEMMOTCOPY = ../generic/gemm_tcopy_$(DGEMM_UNROLL_N).c
endif
DGEMMONCOPYOBJ = dgemm_oncopy$(TSUFFIX).$(SUFFIX)
DGEMMOTCOPYOBJ = dgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifneq ($(DGEMM_UNROLL_M), $(DGEMM_UNROLL_N))
DGEMMINCOPY = ../generic/gemm_ncopy_$(DGEMM_UNROLL_M).c
ifneq ($(filter $(DGEMM_UNROLL_M),4 8 16),)
DGEMMITCOPY = gemm_tcopy_$(DGEMM_UNROLL_M)_rvv.c
else
DGEMMITCOPY = ../generic/gemm_tcopy_$(DGEMM_UNROLL_M).c
endif
DGEMMINCOPYOBJ = dgemm_incopy$(TSUFFIX).$(SUFFIX)
DGEMMITCOPYOBJ = dgemm_itcopy$(TSUFFIX).$(SUFFIX)
endif
@@ -226,13 +234,23 @@ SOMATCOPY_CT = omatcopy_ct_rvv.c
ifeq ($(BUILD_HFLOAT16), 1)
SHGEMMKERNEL = shgemm_kernel_$(SHGEMM_UNROLL_M)x$(SHGEMM_UNROLL_N)_zvl256b.c
ifneq ($(SHGEMM_UNROLL_M), $(SHGEMM_UNROLL_N))
ifneq ($(filter $(SHGEMM_UNROLL_M),8 16),)
SHGEMMINCOPY = gemm_ncopy_$(SHGEMM_UNROLL_M)fp_rvv.c
SHGEMMITCOPY = gemm_tcopy_$(SHGEMM_UNROLL_M)fp_rvv.c
else
SHGEMMINCOPY = ../generic/gemm_ncopy_$(SHGEMM_UNROLL_M).c
SHGEMMITCOPY = ../generic/gemm_tcopy_$(SHGEMM_UNROLL_M).c
endif
SHGEMMINCOPYOBJ = shgemm_incopy$(TSUFFIX).$(SUFFIX)
SHGEMMITCOPYOBJ = shgemm_itcopy$(TSUFFIX).$(SUFFIX)
endif
ifneq ($(filter $(SHGEMM_UNROLL_N),8 16),)
SHGEMMONCOPY = gemm_ncopy_$(SHGEMM_UNROLL_N)fp_rvv.c
SHGEMMOTCOPY = gemm_tcopy_$(SHGEMM_UNROLL_N)fp_rvv.c
else
SHGEMMONCOPY = ../generic/gemm_ncopy_$(SHGEMM_UNROLL_N).c
SHGEMMOTCOPY = ../generic/gemm_tcopy_$(SHGEMM_UNROLL_N).c
endif
SHGEMMONCOPYOBJ = shgemm_oncopy$(TSUFFIX).$(SUFFIX)
SHGEMMOTCOPYOBJ = shgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifndef SHGEMM_BETA
@@ -243,13 +261,23 @@ endif
ifeq ($(BUILD_BFLOAT16), 1)
SBGEMMKERNEL = sbgemm_kernel_$(SBGEMM_UNROLL_M)x$(SBGEMM_UNROLL_N)_zvl256b.c
ifneq ($(SBGEMM_UNROLL_M), $(SBGEMM_UNROLL_N))
ifneq ($(filter $(SBGEMM_UNROLL_M),8 16),)
SBGEMMINCOPY = gemm_ncopy_$(SBGEMM_UNROLL_M)fp_rvv.c
SBGEMMITCOPY = gemm_tcopy_$(SBGEMM_UNROLL_M)fp_rvv.c
else
SBGEMMINCOPY = ../generic/gemm_ncopy_$(SBGEMM_UNROLL_M).c
SBGEMMITCOPY = ../generic/gemm_tcopy_$(SBGEMM_UNROLL_M).c
endif
SBGEMMINCOPYOBJ = sbgemm_incopy$(TSUFFIX).$(SUFFIX)
SBGEMMITCOPYOBJ = sbgemm_itcopy$(TSUFFIX).$(SUFFIX)
endif
ifneq ($(filter $(SBGEMM_UNROLL_N),8 16),)
SBGEMMONCOPY = gemm_ncopy_$(SBGEMM_UNROLL_N)fp_rvv.c
SBGEMMOTCOPY = gemm_tcopy_$(SBGEMM_UNROLL_N)fp_rvv.c
else
SBGEMMONCOPY = ../generic/gemm_ncopy_$(SBGEMM_UNROLL_N).c
SBGEMMOTCOPY = ../generic/gemm_tcopy_$(SBGEMM_UNROLL_N).c
endif
SBGEMMONCOPYOBJ = sbgemm_oncopy$(TSUFFIX).$(SUFFIX)
SBGEMMOTCOPYOBJ = sbgemm_otcopy$(TSUFFIX).$(SUFFIX)
ifndef SBGEMM_BETA

345
kernel/riscv64/gemm_ncopy_16fp_rvv.c Normal file
View File

@@ -0,0 +1,345 @@
/***************************************************************************
Copyright (c) 2025, The OpenBLAS Project
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
3. Neither the name of the OpenBLAS project nor the names of
its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*****************************************************************************/
#include "common.h"
#define VSETVL(n) __riscv_vsetvl_e16m1(n)
#define VSETVL2(n) __riscv_vsetvl_e16m2(n)
#define VSETVL4(n) __riscv_vsetvl_e16m4(n)
#define VSETVL8(n) __riscv_vsetvl_e16m8(n)
#if defined(HFLOAT16)
#define FLOAT_V_T vfloat16m1_t
#define FLOAT_V2_T vfloat16m2_t
#define FLOAT_V4_T vfloat16m4_t
#define FLOAT_V8_T vfloat16m8_t
#define FLOAT_VX2_T vfloat16m1x2_t
#define FLOAT_VX4_T vfloat16m1x4_t
#define FLOAT_VX8_T vfloat16m1x8_t
#define FLOAT_VX24_T vfloat16m4x2_t
#define FLOAT_VX42_T vfloat16m2x4_t
#define VLSEG2_FLOAT __riscv_vlse16_v_f16m2
#define VLSSEG2_FLOAT __riscv_vlsseg2e16_v_f16m1x2
#define VLSSEG4_FLOAT __riscv_vlsseg4e16_v_f16m1x4
#define VLSSEG8_FLOAT __riscv_vlsseg8e16_v_f16m1x8
#define VGET_VX2 __riscv_vget_v_f16m1x2_f16m1
#define VGET_VX4 __riscv_vget_v_f16m1x4_f16m1
#define VGET_VX8 __riscv_vget_v_f16m1x8_f16m1
#define VSET_VX2 __riscv_vset_v_f16m4_f16m4x2
#define VSET_VX4 __riscv_vset_v_f16m2_f16m2x4
#define VSET_VX8 __riscv_vset_v_f16m1_f16m1x8
#define VLEV_FLOAT __riscv_vle16_v_f16m1
#define VLEV_FLOAT2 __riscv_vle16_v_f16m2
#define VLEV_FLOAT4 __riscv_vle16_v_f16m4
#define VLEV_FLOAT8 __riscv_vle16_v_f16m8
#define VSEV_FLOAT __riscv_vse16_v_f16m1
#define VSEV_FLOAT2 __riscv_vse16_v_f16m2
#define VSEV_FLOAT8 __riscv_vse16_v_f16m8
#define VSSEG2_FLOAT __riscv_vsseg2e16_v_f16m4x2
#define VSSEG4_FLOAT __riscv_vsseg4e16_v_f16m2x4
#define VSSEG8_FLOAT __riscv_vsseg8e16_v_f16m1x8
#else
#define FLOAT_V_T vbfloat16m1_t
#define FLOAT_V2_T vbfloat16m2_t
#define FLOAT_V4_T vbfloat16m4_t
#define FLOAT_V8_T vbfloat16m8_t
#define FLOAT_VX2_T vbfloat16m1x2_t
#define FLOAT_VX4_T vbfloat16m1x4_t
#define FLOAT_VX8_T vbfloat16m1x8_t
#define FLOAT_VX24_T vbfloat16m4x2_t
#define FLOAT_VX42_T vbfloat16m2x4_t
#define VLSEG2_FLOAT __riscv_vlse16_v_bf16m2
#define VLSSEG2_FLOAT __riscv_vlsseg2e16_v_bf16m1x2
#define VLSSEG4_FLOAT __riscv_vlsseg4e16_v_bf16m1x4
#define VLSSEG8_FLOAT __riscv_vlsseg8e16_v_bf16m1x8
#define VGET_VX2 __riscv_vget_v_bf16m1x2_bf16m1
#define VGET_VX4 __riscv_vget_v_bf16m1x4_bf16m1
#define VGET_VX8 __riscv_vget_v_bf16m1x8_bf16m1
#define VSET_VX2 __riscv_vset_v_bf16m4_bf16m4x2
#define VSET_VX4 __riscv_vset_v_bf16m2_bf16m2x4
#define VSET_VX8 __riscv_vset_v_bf16m1_bf16m1x8
#define VLEV_FLOAT __riscv_vle16_v_bf16m1
#define VLEV_FLOAT2 __riscv_vle16_v_bf16m2
#define VLEV_FLOAT4 __riscv_vle16_v_bf16m4
#define VLEV_FLOAT8 __riscv_vle16_v_bf16m8
#define VSEV_FLOAT __riscv_vse16_v_bf16m1
#define VSEV_FLOAT2 __riscv_vse16_v_bf16m2
#define VSEV_FLOAT8 __riscv_vse16_v_bf16m8
#define VSSEG2_FLOAT __riscv_vsseg2e16_v_bf16m4x2
#define VSSEG4_FLOAT __riscv_vsseg4e16_v_bf16m2x4
#define VSSEG8_FLOAT __riscv_vsseg8e16_v_bf16m1x8
#endif
// Optimizes the implementation in ../generic/gemm_ncopy_16.c
int CNAME(BLASLONG m, BLASLONG n, IFLOAT *a, BLASLONG lda, IFLOAT *b)
{
BLASLONG i, j;
IFLOAT *a_offset;
IFLOAT *a_offset1, *a_offset2, *a_offset3, *a_offset4;
IFLOAT *a_offset5, *a_offset6, *a_offset7, *a_offset8;
IFLOAT *b_offset;
FLOAT_V_T v1, v2, v3, v4, v5, v6, v7, v8;
FLOAT_V_T v9, v10, v11, v12, v13, v14, v15, v16;
FLOAT_V2_T v21, v22, v23, v24;
FLOAT_V4_T v41, v42;
FLOAT_V8_T v81;
FLOAT_VX2_T vx2, vx21;
FLOAT_VX4_T vx4, vx41;
FLOAT_VX8_T vx8, vx81;
FLOAT_VX42_T vx24;
FLOAT_VX24_T vx42;
size_t vl;
//fprintf(stderr, "gemm_ncopy_16 m=%ld n=%ld lda=%ld\n", m, n, lda);
a_offset = a;
b_offset = b;
for (j = (n >> 4); j > 0; j--) {
vl = VSETVL(8);
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda * 8;
a_offset += 16 * lda;
for (i = m >> 3; i > 0; i--) {
vx8 = VLSSEG8_FLOAT(a_offset1, lda * sizeof(IFLOAT), vl);
vx81 = VLSSEG8_FLOAT(a_offset2, lda * sizeof(IFLOAT), vl);
v1 = VGET_VX8(vx8, 0);
v2 = VGET_VX8(vx8, 1);
v3 = VGET_VX8(vx8, 2);
v4 = VGET_VX8(vx8, 3);
v5 = VGET_VX8(vx8, 4);
v6 = VGET_VX8(vx8, 5);
v7 = VGET_VX8(vx8, 6);
v8 = VGET_VX8(vx8, 7);
v9 = VGET_VX8(vx81, 0);
v10 = VGET_VX8(vx81, 1);
v11 = VGET_VX8(vx81, 2);
v12 = VGET_VX8(vx81, 3);
v13 = VGET_VX8(vx81, 4);
v14 = VGET_VX8(vx81, 5);
v15 = VGET_VX8(vx81, 6);
v16 = VGET_VX8(vx81, 7);
VSEV_FLOAT(b_offset, v1, vl);
VSEV_FLOAT(b_offset + 8, v9, vl);
VSEV_FLOAT(b_offset + 16, v2, vl);
VSEV_FLOAT(b_offset + 24, v10, vl);
VSEV_FLOAT(b_offset + 32, v3, vl);
VSEV_FLOAT(b_offset + 40, v11, vl);
VSEV_FLOAT(b_offset + 48, v4, vl);
VSEV_FLOAT(b_offset + 56, v12, vl);
VSEV_FLOAT(b_offset + 64, v5, vl);
VSEV_FLOAT(b_offset + 72, v13, vl);
VSEV_FLOAT(b_offset + 80, v6, vl);
VSEV_FLOAT(b_offset + 88, v14, vl);
VSEV_FLOAT(b_offset + 96, v7, vl);
VSEV_FLOAT(b_offset + 104, v15, vl);
VSEV_FLOAT(b_offset + 112, v8, vl);
VSEV_FLOAT(b_offset + 120, v16, vl);
a_offset1 += 8;
a_offset2 += 8;
b_offset += 128;
}
if (m & 4) {
vx4 = VLSSEG4_FLOAT(a_offset1, lda * sizeof(IFLOAT), vl);
vx41 = VLSSEG4_FLOAT(a_offset2, lda * sizeof(IFLOAT), vl);
v1 = VGET_VX4(vx4, 0);
v2 = VGET_VX4(vx4, 1);
v3 = VGET_VX4(vx4, 2);
v4 = VGET_VX4(vx4, 3);
v5 = VGET_VX4(vx41, 0);
v6 = VGET_VX4(vx41, 1);
v7 = VGET_VX4(vx41, 2);
v8 = VGET_VX4(vx41, 3);
VSEV_FLOAT(b_offset, v1, vl);
VSEV_FLOAT(b_offset + 8, v5, vl);
VSEV_FLOAT(b_offset + 16, v2, vl);
VSEV_FLOAT(b_offset + 24, v6, vl);
VSEV_FLOAT(b_offset + 32, v3, vl);
VSEV_FLOAT(b_offset + 40, v7, vl);
VSEV_FLOAT(b_offset + 48, v4, vl);
VSEV_FLOAT(b_offset + 56, v8, vl);
a_offset1 += 4;
a_offset2 += 4;
b_offset += 64;
}
if (m & 2) {
vx2 = VLSSEG2_FLOAT(a_offset1, lda * sizeof(IFLOAT), vl);
vx21 = VLSSEG2_FLOAT(a_offset2, lda * sizeof(IFLOAT), vl);
v1 = VGET_VX2(vx2, 0);
v2 = VGET_VX2(vx2, 1);
v3 = VGET_VX2(vx21, 0);
v4 = VGET_VX2(vx21, 1);
VSEV_FLOAT(b_offset, v1, vl);
VSEV_FLOAT(b_offset + 8, v3, vl);
VSEV_FLOAT(b_offset + 16, v2, vl);
VSEV_FLOAT(b_offset + 24, v4, vl);
a_offset1 += 2;
a_offset2 += 2;
b_offset += 32;
}
if (m & 1) {
v21 = VLSEG2_FLOAT(a_offset1, lda * sizeof(IFLOAT), vl * 2);
VSEV_FLOAT2(b_offset, v21, vl * 2);
b_offset += 16;
}
}
if (n & 8) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset3 = a_offset2 + lda;
a_offset4 = a_offset3 + lda;
a_offset5 = a_offset4 + lda;
a_offset6 = a_offset5 + lda;
a_offset7 = a_offset6 + lda;
a_offset8 = a_offset7 + lda;
a_offset += 8 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL(i);
v1 = VLEV_FLOAT(a_offset1, vl);
v2 = VLEV_FLOAT(a_offset2, vl);
v3 = VLEV_FLOAT(a_offset3, vl);
v4 = VLEV_FLOAT(a_offset4, vl);
v5 = VLEV_FLOAT(a_offset5, vl);
v6 = VLEV_FLOAT(a_offset6, vl);
v7 = VLEV_FLOAT(a_offset7, vl);
v8 = VLEV_FLOAT(a_offset8, vl);
vx8 = VSET_VX8(vx8, 0, v1);
vx8 = VSET_VX8(vx8, 1, v2);
vx8 = VSET_VX8(vx8, 2, v3);
vx8 = VSET_VX8(vx8, 3, v4);
vx8 = VSET_VX8(vx8, 4, v5);
vx8 = VSET_VX8(vx8, 5, v6);
vx8 = VSET_VX8(vx8, 6, v7);
vx8 = VSET_VX8(vx8, 7, v8);
VSSEG8_FLOAT(b_offset, vx8, vl);
a_offset1 += vl;
a_offset2 += vl;
a_offset3 += vl;
a_offset4 += vl;
a_offset5 += vl;
a_offset6 += vl;
a_offset7 += vl;
a_offset8 += vl;
b_offset += vl*8;
}
}
if (n & 4) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset3 = a_offset2 + lda;
a_offset4 = a_offset3 + lda;
a_offset += 4 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL2(i);
v21 = VLEV_FLOAT2(a_offset1, vl);
v22 = VLEV_FLOAT2(a_offset2, vl);
v23 = VLEV_FLOAT2(a_offset3, vl);
v24 = VLEV_FLOAT2(a_offset4, vl);
vx24 = VSET_VX4(vx24, 0, v21);
vx24 = VSET_VX4(vx24, 1, v22);
vx24 = VSET_VX4(vx24, 2, v23);
vx24 = VSET_VX4(vx24, 3, v24);
VSSEG4_FLOAT(b_offset, vx24, vl);
a_offset1 += vl;
a_offset2 += vl;
a_offset3 += vl;
a_offset4 += vl;
b_offset += vl*4;
}
}
if (n & 2) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset += 2 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL4(i);
v41 = VLEV_FLOAT4(a_offset1, vl);
v42 = VLEV_FLOAT4(a_offset2, vl);
vx42 = VSET_VX2(vx42, 0, v41);
vx42 = VSET_VX2(vx42, 1, v42);
VSSEG2_FLOAT(b_offset, vx42, vl);
a_offset1 += vl;
a_offset2 += vl;
b_offset += vl*2;
}
}
if (n & 1) {
a_offset1 = a_offset;
for(i = m; i > 0; i -= vl) {
vl = VSETVL8(i);
v81 = VLEV_FLOAT8(a_offset1, vl);
VSEV_FLOAT8(b_offset, v81, vl);
a_offset1 += vl;
b_offset += vl;
}
}
return 0;
}

215
kernel/riscv64/gemm_ncopy_8fp_rvv.c Normal file
View File

@@ -0,0 +1,215 @@
/***************************************************************************
Copyright (c) 2022, The OpenBLAS Project
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
3. Neither the name of the OpenBLAS project nor the names of
its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*****************************************************************************/
#include "common.h"
#define VSETVL(n) __riscv_vsetvl_e16m1(n)
#define VSETVL2(n) __riscv_vsetvl_e16m2(n)
#define VSETVL4(n) __riscv_vsetvl_e16m4(n)
#define VSETVL8(n) __riscv_vsetvl_e16m8(n)
#if defined(HFLOAT16)
#define FLOAT_V_T vfloat16m1_t
#define FLOAT2_V_T vfloat16m2_t
#define FLOAT4_V_T vfloat16m4_t
#define FLOAT8_V_T vfloat16m8_t
#define FLOAT_VX2_T vfloat16m4x2_t
#define FLOAT_VX4_T vfloat16m2x4_t
#define FLOAT_VX8_T vfloat16m1x8_t
#define VSET_VX2 __riscv_vset_v_f16m4_f16m4x2
#define VSET_VX4 __riscv_vset_v_f16m2_f16m2x4
#define VSET_VX8 __riscv_vset_v_f16m1_f16m1x8
#define VLEV_FLOAT __riscv_vle16_v_f16m1
#define VLEV_FLOAT2 __riscv_vle16_v_f16m2
#define VLEV_FLOAT4 __riscv_vle16_v_f16m4
#define VLEV_FLOAT8 __riscv_vle16_v_f16m8
#define VSEV_FLOAT8 __riscv_vse16_v_f16m8
#define VSSEG2_FLOAT __riscv_vsseg2e16_v_f16m4x2
#define VSSEG4_FLOAT __riscv_vsseg4e16_v_f16m2x4
#define VSSEG8_FLOAT __riscv_vsseg8e16_v_f16m1x8
#else
#define FLOAT_V_T vbfloat16m1_t
#define FLOAT2_V_T vbfloat16m2_t
#define FLOAT4_V_T vbfloat16m4_t
#define FLOAT8_V_T vbfloat16m8_t
#define FLOAT_VX2_T vbfloat16m4x2_t
#define FLOAT_VX4_T vbfloat16m2x4_t
#define FLOAT_VX8_T vbfloat16m1x8_t
#define VSET_VX2 __riscv_vset_v_bf16m4_bf16m4x2
#define VSET_VX4 __riscv_vset_v_bf16m2_bf16m2x4
#define VSET_VX8 __riscv_vset_v_bf16m1_bf16m1x8
#define VLEV_FLOAT __riscv_vle16_v_bf16m1
#define VLEV_FLOAT2 __riscv_vle16_v_bf16m2
#define VLEV_FLOAT4 __riscv_vle16_v_bf16m4
#define VLEV_FLOAT8 __riscv_vle16_v_bf16m8
#define VSEV_FLOAT8 __riscv_vse16_v_bf16m8
#define VSSEG2_FLOAT __riscv_vsseg2e16_v_bf16m4x2
#define VSSEG4_FLOAT __riscv_vsseg4e16_v_bf16m2x4
#define VSSEG8_FLOAT __riscv_vsseg8e16_v_bf16m1x8
#endif
// Optimizes the implementation in ../generic/gemm_ncopy_8.c
int CNAME(BLASLONG m, BLASLONG n, IFLOAT *a, BLASLONG lda, IFLOAT *b)
{
BLASLONG i, j;
IFLOAT *a_offset;
IFLOAT *a_offset1, *a_offset2, *a_offset3, *a_offset4;
IFLOAT *a_offset5, *a_offset6, *a_offset7, *a_offset8;
IFLOAT *b_offset;
FLOAT_V_T v1, v2, v3, v4, v5, v6, v7, v8;
FLOAT2_V_T v12, v22, v32, v42;
FLOAT4_V_T v14, v24;
FLOAT8_V_T v18;
FLOAT_VX2_T vx2;
FLOAT_VX4_T vx4;
FLOAT_VX8_T vx8;
size_t vl;
//fprintf(stderr, "gemm_ncopy_8 m=%ld n=%ld lda=%ld\n", m, n, lda);
a_offset = a;
b_offset = b;
for(j = (n >> 3); j > 0; j--) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset3 = a_offset2 + lda;
a_offset4 = a_offset3 + lda;
a_offset5 = a_offset4 + lda;
a_offset6 = a_offset5 + lda;
a_offset7 = a_offset6 + lda;
a_offset8 = a_offset7 + lda;
a_offset += 8 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL(i);
v1 = VLEV_FLOAT(a_offset1, vl);
v2 = VLEV_FLOAT(a_offset2, vl);
v3 = VLEV_FLOAT(a_offset3, vl);
v4 = VLEV_FLOAT(a_offset4, vl);
v5 = VLEV_FLOAT(a_offset5, vl);
v6 = VLEV_FLOAT(a_offset6, vl);
v7 = VLEV_FLOAT(a_offset7, vl);
v8 = VLEV_FLOAT(a_offset8, vl);
vx8 = VSET_VX8(vx8, 0, v1);
vx8 = VSET_VX8(vx8, 1, v2);
vx8 = VSET_VX8(vx8, 2, v3);
vx8 = VSET_VX8(vx8, 3, v4);
vx8 = VSET_VX8(vx8, 4, v5);
vx8 = VSET_VX8(vx8, 5, v6);
vx8 = VSET_VX8(vx8, 6, v7);
vx8 = VSET_VX8(vx8, 7, v8);
VSSEG8_FLOAT(b_offset, vx8, vl);
a_offset1 += vl;
a_offset2 += vl;
a_offset3 += vl;
a_offset4 += vl;
a_offset5 += vl;
a_offset6 += vl;
a_offset7 += vl;
a_offset8 += vl;
b_offset += vl*8;
}
}
if (n & 4) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset3 = a_offset2 + lda;
a_offset4 = a_offset3 + lda;
a_offset += 4 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL2(i);
v12 = VLEV_FLOAT2(a_offset1, vl);
v22 = VLEV_FLOAT2(a_offset2, vl);
v32 = VLEV_FLOAT2(a_offset3, vl);
v42 = VLEV_FLOAT2(a_offset4, vl);
vx4 = VSET_VX4(vx4, 0, v12);
vx4 = VSET_VX4(vx4, 1, v22);
vx4 = VSET_VX4(vx4, 2, v32);
vx4 = VSET_VX4(vx4, 3, v42);
VSSEG4_FLOAT(b_offset, vx4, vl);
a_offset1 += vl;
a_offset2 += vl;
a_offset3 += vl;
a_offset4 += vl;
b_offset += vl*4;
}
}
if (n & 2) {
a_offset1 = a_offset;
a_offset2 = a_offset1 + lda;
a_offset += 2 * lda;
for(i = m; i > 0; i -= vl) {
vl = VSETVL4(i);
v14 = VLEV_FLOAT4(a_offset1, vl);
v24 = VLEV_FLOAT4(a_offset2, vl);
vx2 = VSET_VX2(vx2, 0, v14);
vx2 = VSET_VX2(vx2, 1, v24);
VSSEG2_FLOAT(b_offset, vx2, vl);
a_offset1 += vl;
a_offset2 += vl;
b_offset += vl*2;
}
}
if (n & 1) {
a_offset1 = a_offset;
for(i = m; i > 0; i -= vl) {
vl = VSETVL8(i);
v18 = VLEV_FLOAT8(a_offset1, vl);
VSEV_FLOAT8(b_offset, v18, vl);
a_offset1 += vl;
b_offset += vl;
}
}
return 0;
}

129
kernel/riscv64/gemm_tcopy_16fp_rvv.c Normal file
View File

@@ -0,0 +1,129 @@
/***************************************************************************
Copyright (c) 2025, The OpenBLAS Project
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
3. Neither the name of the OpenBLAS project nor the names of
its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*****************************************************************************/
#include "common.h"
#if defined(HFLOAT16)
#define FLOAT_V_T vfloat16m2_t
#define FLOAT_V_T_HALF vfloat16m1_t
#define FLOAT_V_T_QUARTER vfloat16m1_t
#define VLEV_FLOAT __riscv_vle16_v_f16m2
#define VLEV_FLOAT_HALF __riscv_vle16_v_f16m1
#define VLEV_FLOAT_QUARTER __riscv_vle16_v_f16m1
#define VSEV_FLOAT __riscv_vse16_v_f16m2
#define VSEV_FLOAT_HALF __riscv_vse16_v_f16m1
#define VSEV_FLOAT_QUARTER __riscv_vse16_v_f16m1
#else
#define FLOAT_V_T vbfloat16m2_t
#define FLOAT_V_T_HALF vbfloat16m1_t
#define FLOAT_V_T_QUARTER vbfloat16m1_t
#define VLEV_FLOAT __riscv_vle16_v_bf16m2
#define VLEV_FLOAT_HALF __riscv_vle16_v_bf16m1
#define VLEV_FLOAT_QUARTER __riscv_vle16_v_bf16m1
#define VSEV_FLOAT __riscv_vse16_v_bf16m2
#define VSEV_FLOAT_HALF __riscv_vse16_v_bf16m1
#define VSEV_FLOAT_QUARTER __riscv_vse16_v_bf16m1
#endif
// Optimizes the implementation in ../generic/gemm_tcopy_16.c
int CNAME(BLASLONG m, BLASLONG n, IFLOAT *a, BLASLONG lda, IFLOAT *b)
{
BLASLONG i, j;
IFLOAT *aoffset;
IFLOAT *aoffset1;
IFLOAT *boffset, *boffset1, *boffset2, *boffset3, *boffset4, *boffset5;
FLOAT_V_T v0;
FLOAT_V_T_HALF v1;
FLOAT_V_T_QUARTER v2;
// fprintf(stderr, "gemm_tcopy_16 m=%ld n=%ld lda=%ld\n", m, n, lda);
aoffset = a;
boffset = b;
boffset2 = b + m * (n & ~15);
boffset3 = b + m * (n & ~7);
boffset4 = b + m * (n & ~3);
boffset5 = b + m * (n & ~1);
for(j = m; j > 0; j--) {
aoffset1 = aoffset;
boffset1 = boffset;
aoffset += lda;
boffset += 16;
for(i = (n >> 4); i > 0; i--) {
size_t vl = 16;
v0 = VLEV_FLOAT(aoffset1, vl);
VSEV_FLOAT(boffset1, v0, vl);
aoffset1 += 16;
boffset1 += 16 * m;
}
if (n & 8) {
size_t vl = 8;
v1 = VLEV_FLOAT_HALF(aoffset1, vl);
VSEV_FLOAT_HALF(boffset2, v1, vl);
aoffset1 += 8;
boffset2 += 8;
}
if (n & 4) {
size_t vl = 4;
v2 = VLEV_FLOAT_QUARTER(aoffset1, vl);
VSEV_FLOAT_QUARTER(boffset3, v2, vl);
aoffset1 += 4;
boffset3 += 4;
}
if (n & 2) {
*(boffset4) = *(aoffset1);
*(boffset4 + 1) = *(aoffset1 + 1);
aoffset1 += 2;
boffset4 += 2;
}
if (n & 1) {
*(boffset5) = *(aoffset1);
aoffset1 ++;
boffset5 ++;
}
}
return 0;
}

109
kernel/riscv64/gemm_tcopy_8fp_rvv.c Normal file
View File

@@ -0,0 +1,109 @@
/***************************************************************************
Copyright (c) 2022, The OpenBLAS Project
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
3. Neither the name of the OpenBLAS project nor the names of
its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*****************************************************************************/
#include "common.h"
#if defined(HFLOAT16)
#define FLOAT_V_T vfloat16m1_t
#define FLOAT_V_T_HALF vfloat16m1_t
#define VLEV_FLOAT __riscv_vle16_v_f16m1
#define VLEV_FLOAT_HALF __riscv_vle16_v_f16m1
#define VSEV_FLOAT __riscv_vse16_v_f16m1
#define VSEV_FLOAT_HALF __riscv_vse16_v_f16m1
#else
#define FLOAT_V_T vbfloat16m1_t
#define FLOAT_V_T_HALF vbfloat16m1_t
#define VLEV_FLOAT __riscv_vle16_v_bf16m1
#define VLEV_FLOAT_HALF __riscv_vle16_v_bf16m1
#define VSEV_FLOAT __riscv_vse16_v_bf16m1
#define VSEV_FLOAT_HALF __riscv_vse16_v_bf16m1
#endif
int CNAME(BLASLONG m, BLASLONG n, IFLOAT *a, BLASLONG lda, IFLOAT *b)
{
BLASLONG i, j;
IFLOAT *aoffset;
IFLOAT *aoffset1;
IFLOAT *boffset, *boffset1, *boffset2, *boffset3, *boffset4;
FLOAT_V_T v0;
FLOAT_V_T_HALF v1;
// fprintf(stderr, "gemm_tcopy_8 m=%ld n=%ld lda=%ld\n", m, n, lda);
aoffset = a;
boffset = b;
boffset2 = b + m * (n & ~7);
boffset3 = b + m * (n & ~3);
boffset4 = b + m * (n & ~1);
for(j = m; j > 0; j--) {
aoffset1 = aoffset;
boffset1 = boffset;
aoffset += lda;
boffset += 8;
for(i = (n >> 3); i > 0; i--) {
size_t vl = 8;
v0 = VLEV_FLOAT(aoffset1, vl);
VSEV_FLOAT(boffset1, v0, vl);
aoffset1 += 8;
boffset1 += 8 * m;
}
if (n & 4) {
size_t vl = 4;
v1 = VLEV_FLOAT_HALF(aoffset1, vl);
VSEV_FLOAT_HALF(boffset2, v1, vl);
aoffset1 += 4;
boffset2 += 4;
}
if (n & 2) {
*(boffset3) = *(aoffset1);
*(boffset3 + 1) = *(aoffset1 + 1);
aoffset1 += 2;
boffset3 += 2;
}
if (n & 1) {
*(boffset4) = *(aoffset1);
aoffset1 ++;
boffset4 ++;
}
}
return 0;
}