virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-21 10:55:51 +02:00
Code Issues Releases Activity

af_afir: RISC-V V fcmul_add

Browse Source 
Segmented loads are slow, so here we use unit-strided load and narrowing shifts.

c910:
fcmul_add_c: 2179
fcmul_add_rvv_f64: 1652

c908:
fcmul_add_c: 4891.2
fcmul_add_rvv_f64: 2399.5

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>
This commit is contained in:
sunyuechi 2023-11-03 10:35:53 +08:00 committed by Rémi Denis-Courmont
parent d076517056
commit afb967b81e
4 changed files with 103 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
libavfilter/af_afirdsp.h
View File

@ -33,6 +33,7 @@ typedef struct AudioFIRDSPContext {
ptrdiff_t len);
} AudioFIRDSPContext;
void ff_afir_init_riscv(AudioFIRDSPContext *s);
void ff_afir_init_x86(AudioFIRDSPContext *s);
static void fcmul_add_c(float *sum, const float *t, const float *c, ptrdiff_t len)
@ -74,7 +75,9 @@ static av_unused void ff_afir_init(AudioFIRDSPContext *dsp)
dsp->fcmul_add = fcmul_add_c;
dsp->dcmul_add = dcmul_add_c;
#if ARCH_X86
#if ARCH_RISCV
ff_afir_init_riscv(dsp);
#elif ARCH_X86
ff_afir_init_x86(dsp);
#endif
}

2
libavfilter/riscv/Makefile Normal file
View File

@ -0,0 +1,2 @@
OBJS += riscv/af_afir_init.o
RVV-OBJS += riscv/af_afir_rvv.o

42
libavfilter/riscv/af_afir_init.c Normal file
View File

@ -0,0 +1,42 @@
/*
* Copyright (c) 2023 Institue of Software Chinese Academy of Sciences (ISCAS).
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include <stdint.h>
#include "config.h"
#include "libavutil/attributes.h"
#include "libavutil/cpu.h"
#include "libavfilter/af_afirdsp.h"
void ff_fcmul_add_rvv(float *sum, const float *t, const float *c,
ptrdiff_t len);
av_cold void ff_afir_init_riscv(AudioFIRDSPContext *s)
{
#if HAVE_RVV
int flags = av_get_cpu_flags();
if (flags & AV_CPU_FLAG_RVV_F64) {
if (flags & AV_CPU_FLAG_RVB_ADDR) {
s->fcmul_add = ff_fcmul_add_rvv;
}
}
#endif
}

55
libavfilter/riscv/af_afir_rvv.S Normal file
View File

@ -0,0 +1,55 @@
/*
* Copyright (c) 2023 Institue of Software Chinese Academy of Sciences (ISCAS).
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "libavutil/riscv/asm.S"
// void ff_fcmul_add(float *sum, const float *t, const float *c, int len)
func ff_fcmul_add_rvv, zve64f
li t1, 32
1:
vsetvli t0, a3, e32, m4, ta, ma
vle64.v v24, (a0)
sub a3, a3, t0
vnsrl.wx v16, v24, zero
vnsrl.wx v20, v24, t1
vle64.v v24, (a1)
sh3add a1, t0, a1
vnsrl.wx v0, v24, zero
vnsrl.wx v4, v24, t1
vle64.v v24, (a2)
sh3add a2, t0, a2
vnsrl.wx v8, v24, zero
vnsrl.wx v12, v24, t1
vfmacc.vv v16, v0, v8
vfmacc.vv v20, v4, v8
vfnmsac.vv v16, v4, v12
vfmacc.vv v20, v0, v12
vsseg2e32.v v16, (a0)
sh3add a0, t0, a0
bgtz a3, 1b
flw fa0, 0(a1)
flw fa1, 0(a2)
flw fa2, 0(a0)
fmadd.s fa2, fa0, fa1, fa2
fsw fa2, 0(a0)
ret
endfunc