1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-23 12:43:46 +02:00
FFmpeg/libavcodec/riscv/mpegvideoencdsp_rvv.S
Rémi Denis-Courmont d8fb44c0aa lavc/mpegvideoencdsp: R-V V add_8x8basis
T-Head C908:
add_8x8basis_c:      440.6
add_8x8basis_rvv_i32: 70.3

SpacemiT X60:
add_8x8basis_c:      436.3
add_8x8basis_rvv_i32: 40.5
2024-08-19 22:41:13 +03:00

128 lines
3.6 KiB
ArmAsm

/*
* Copyright © 2024 Rémi Denis-Courmont.
*
* 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"
.equ BASIS_SHIFT, 16
.equ RECON_SHIFT, 6
func ff_try_8x8basis_rvv, zve32x, b
li t1, 64
csrwi vxrm, 0
vsetvli t0, t1, e32, m8, ta, ma
vmv.v.x v24, zero
vmv.s.x v1, zero
1:
vsetvli zero, zero, e16, m4, ta, ma
vle16.v v4, (a2)
sub t1, t1, t0
vwmul.vx v16, v4, a3
sh1add a2, t0, a2
vle16.v v8, (a0)
sh1add a0, t0, a0
vnclip.wi v4, v16, BASIS_SHIFT - RECON_SHIFT
vle16.v v12, (a1)
sh1add a1, t0, a1
vadd.vv v4, v8, v4
vsra.vi v4, v4, RECON_SHIFT
vwmul.vv v16, v12, v4
vsetvli zero, zero, e32, m8, ta, ma
vmul.vv v16, v16, v16
vsra.vi v16, v16, 4
vadd.vv v24, v24, v16
bnez t1, 1b
vredsum.vs v1, v24, v1
vmv.x.s a0, v1
srai a0, a0, 2
ret
endfunc
func ff_add_8x8basis_rvv, zve32x, b
li t1, 64
csrwi vxrm, 0
1:
vsetvli t0, t1, e16, m4, ta, ma
vle16.v v4, (a1)
sub t1, t1, t0
vwmul.vx v16, v4, a2
sh1add a1, t0, a1
vle16.v v8, (a0)
vnclip.wi v4, v16, BASIS_SHIFT - RECON_SHIFT
vadd.vv v4, v8, v4
vse16.v v4, (a0)
sh1add a0, t0, a0
bnez t1, 1b
ret
endfunc
func ff_pix_sum_rvv, zve64x, b
lpad 0
vsetivli t0, 16, e16, m1, ta, ma
slli t1, t0, 3
vlsseg2e64.v v8, (a0), a1 # poor man's vlse128.v
li t2, 16
vmv.s.x v1, zero
vsetvli zero, t1, e8, m4, ta, ma
vwaddu.vv v16, v8, v12
beq t0, t2, 2f # VLEN >= 256
1:
sh3add a0, a1, a0
vsetivli zero, 8, e16, m1, ta, ma
vlsseg2e64.v v8, (a0), a1
vsetvli zero, t1, e8, m4, ta, ma
vwaddu.wv v16, v16, v8
vwaddu.wv v16, v16, v12
2:
vsetvli zero, zero, e16, m8, ta, ma
vredsum.vs v1, v16, v1
vmv.x.s a0, v1
zext.h a0, a0
ret
endfunc
func ff_pix_norm1_rvv, zve64x
lpad 0
li t1, 16
vsetvli t0, zero, e32, m8, ta, ma
vmv.v.x v16, zero
vmv.s.x v1, zero
1:
vsetvli t0, t1, e64, m2, ta, ma
vlsseg2e64.v v8, (a0), a1
slli t2, t0, 3
vsetvli zero, t2, e8, m2, ta, ma
vwmulu.vv v24, v8, v8
sub t1, t1, t0
vwmulu.vv v28, v10, v10
mul t3, a1, t0
vsetvli zero, zero, e16, m4, tu, ma
vwaddu.wv v16, v16, v24
add a0, t3, a0
vwaddu.wv v16, v16, v28
bnez t1, 1b
vsetvli zero, zero, e32, m8, ta, ma
vredsum.vs v1, v16, v1
vmv.x.s a0, v1
ret
endfunc