mirror of https://git.ffmpeg.org/ffmpeg.git
lavc/vc1dsp: R-V V mspel_pixels
C908 X60 vc1dsp.avg_vc1_mspel_pixels_tab[0][0]_c : 14.7 13.2 vc1dsp.avg_vc1_mspel_pixels_tab[0][0]_rvv_i32 : 2.5 2.2 vc1dsp.avg_vc1_mspel_pixels_tab[1][0]_c : 3.7 3.5 vc1dsp.avg_vc1_mspel_pixels_tab[1][0]_rvv_i64 : 1.0 1.2 vc1dsp.put_vc1_mspel_pixels_tab[0][0]_c : 9.0 8.0 vc1dsp.put_vc1_mspel_pixels_tab[0][0]_rvi : 1.0 1.0 vc1dsp.put_vc1_mspel_pixels_tab[1][0]_c : 2.5 2.2 vc1dsp.put_vc1_mspel_pixels_tab[1][0]_rvi : 0.5 0.5 Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>
This commit is contained in:
sunyuechi
Rémi Denis-Courmont
parent
7b47099bc0
commit
d4083ecb7c
|
|
@ -58,6 +58,7 @@ RVV-OBJS-$(CONFIG_TAK_DECODER) += riscv/takdsp_rvv.o
|
|||
OBJS-$(CONFIG_UTVIDEO_DECODER) += riscv/utvideodsp_init.o
|
||||
RVV-OBJS-$(CONFIG_UTVIDEO_DECODER) += riscv/utvideodsp_rvv.o
|
||||
OBJS-$(CONFIG_VC1DSP) += riscv/vc1dsp_init.o
|
||||
RV-OBJS-$(CONFIG_VC1DSP) += riscv/vc1dsp_rvi.o
|
||||
RVV-OBJS-$(CONFIG_VC1DSP) += riscv/vc1dsp_rvv.o
|
||||
OBJS-$(CONFIG_VP8DSP) += riscv/vp8dsp_init.o
|
||||
RV-OBJS-$(CONFIG_VP8DSP) += riscv/vp8dsp_rvi.o
|
||||
|
|
|
|||
|
|
@ -29,19 +29,33 @@ void ff_vc1_inv_trans_8x8_dc_rvv(uint8_t *dest, ptrdiff_t stride, int16_t *block
|
|||
void ff_vc1_inv_trans_4x8_dc_rvv(uint8_t *dest, ptrdiff_t stride, int16_t *block);
|
||||
void ff_vc1_inv_trans_8x4_dc_rvv(uint8_t *dest, ptrdiff_t stride, int16_t *block);
|
||||
void ff_vc1_inv_trans_4x4_dc_rvv(uint8_t *dest, ptrdiff_t stride, int16_t *block);
|
||||
void ff_put_pixels16x16_rvi(uint8_t *dst, const uint8_t *src, ptrdiff_t line_size, int rnd);
|
||||
void ff_put_pixels8x8_rvi(uint8_t *dst, const uint8_t *src, ptrdiff_t line_size, int rnd);
|
||||
void ff_avg_pixels16x16_rvv(uint8_t *dst, const uint8_t *src, ptrdiff_t line_size, int rnd);
|
||||
void ff_avg_pixels8x8_rvv(uint8_t *dst, const uint8_t *src, ptrdiff_t line_size, int rnd);
|
||||
|
||||
av_cold void ff_vc1dsp_init_riscv(VC1DSPContext *dsp)
|
||||
{
|
||||
#if HAVE_RVV
|
||||
#if HAVE_RV
|
||||
int flags = av_get_cpu_flags();
|
||||
|
||||
# if __riscv_xlen >= 64
|
||||
if (flags & AV_CPU_FLAG_RVI) {
|
||||
dsp->put_vc1_mspel_pixels_tab[1][0] = ff_put_pixels8x8_rvi;
|
||||
dsp->put_vc1_mspel_pixels_tab[0][0] = ff_put_pixels16x16_rvi;
|
||||
}
|
||||
# endif
|
||||
#if HAVE_RVV
|
||||
if (flags & AV_CPU_FLAG_RVV_I32 && ff_rv_vlen_least(128)) {
|
||||
dsp->vc1_inv_trans_4x8_dc = ff_vc1_inv_trans_4x8_dc_rvv;
|
||||
dsp->vc1_inv_trans_4x4_dc = ff_vc1_inv_trans_4x4_dc_rvv;
|
||||
dsp->avg_vc1_mspel_pixels_tab[0][0] = ff_avg_pixels16x16_rvv;
|
||||
if (flags & AV_CPU_FLAG_RVV_I64) {
|
||||
dsp->vc1_inv_trans_8x8_dc = ff_vc1_inv_trans_8x8_dc_rvv;
|
||||
dsp->vc1_inv_trans_8x4_dc = ff_vc1_inv_trans_8x4_dc_rvv;
|
||||
dsp->avg_vc1_mspel_pixels_tab[1][0] = ff_avg_pixels8x8_rvv;
|
||||
}
|
||||
}
|
||||
#endif
|
||||
#endif
|
||||
}
|
||||
|
|
|
|||
|
|
@ -0,0 +1,47 @@
|
|||
/*
|
||||
* Copyright (c) 2024 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"
|
||||
|
||||
#if __riscv_xlen >= 64
|
||||
func ff_put_pixels8x8_rvi
|
||||
.rept 8
|
||||
ld t0, (a1)
|
||||
sd t0, (a0)
|
||||
add a1, a1, a2
|
||||
add a0, a0, a2
|
||||
.endr
|
||||
|
||||
ret
|
||||
endfunc
|
||||
|
||||
func ff_put_pixels16x16_rvi
|
||||
.rept 16
|
||||
ld t0, (a1)
|
||||
ld t1, 8(a1)
|
||||
sd t0, (a0)
|
||||
sd t1, 8(a0)
|
||||
add a1, a1, a2
|
||||
add a0, a0, a2
|
||||
.endr
|
||||
|
||||
ret
|
||||
endfunc
|
||||
#endif
|
||||
|
|
@ -111,3 +111,51 @@ func ff_vc1_inv_trans_4x4_dc_rvv, zve32x
|
|||
vsse32.v v0, (a0), a1
|
||||
ret
|
||||
endfunc
|
||||
|
||||
.macro mspel_op op pos n1 n2
|
||||
add t1, \pos, a2
|
||||
v\op\()e8.v v\n1, (\pos)
|
||||
sh1add \pos, a2, \pos
|
||||
v\op\()e8.v v\n2, (t1)
|
||||
.endm
|
||||
|
||||
.macro mspel_op_all op pos a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 a11 a12 a13 a14 a15 a16
|
||||
mspel_op \op \pos \a1 \a2
|
||||
mspel_op \op \pos \a3 \a4
|
||||
mspel_op \op \pos \a5 \a6
|
||||
mspel_op \op \pos \a7 \a8
|
||||
mspel_op \op \pos \a9 \a10
|
||||
mspel_op \op \pos \a11 \a12
|
||||
mspel_op \op \pos \a13 \a14
|
||||
mspel_op \op \pos \a15 \a16
|
||||
.endm
|
||||
|
||||
func ff_avg_pixels16x16_rvv, zve32x
|
||||
csrwi vxrm, 0
|
||||
vsetivli zero, 16, e8, m1, ta, ma
|
||||
mspel_op_all l a1 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
|
||||
mspel_op_all l a0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
|
||||
vsetvli t0, zero, e8, m8, ta, ma
|
||||
sub a0, a0, a2
|
||||
vaaddu.vv v0, v0, v16
|
||||
neg a2, a2
|
||||
vaaddu.vv v8, v8, v24
|
||||
vsetivli zero, 16, e8, m1, ta, ma
|
||||
mspel_op_all s a0 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
|
||||
|
||||
ret
|
||||
endfunc
|
||||
|
||||
func ff_avg_pixels8x8_rvv, zve64x
|
||||
csrwi vxrm, 0
|
||||
li t0, 64
|
||||
vsetivli zero, 8, e8, mf2, ta, ma
|
||||
vlse64.v v16, (a1), a2
|
||||
vlse64.v v8, (a0), a2
|
||||
vsetvli zero, t0, e8, m4, ta, ma
|
||||
vaaddu.vv v16, v16, v8
|
||||
vsetivli zero, 8, e8, mf2, ta, ma
|
||||
vsse64.v v16, (a0), a2
|
||||
|
||||
ret
|
||||
endfunc
|
||||
|
|
|
|||
Loading…
Reference in New Issue