From c033ab8d30135dc02b09b1747c0761baefdcbb4a Mon Sep 17 00:00:00 2001
From: sunyuechi <sunyuechi@iscas.ac.cn>
Date: Sat, 20 Apr 2024 23:13:07 +0800
Subject: [PATCH 2/3] lavc/vp8dsp: R-V V loop_filter_inner
C908:
vp8_loop_filter8uv_inner_v_c: 738.2
vp8_loop_filter8uv_inner_v_rvv_i32: 455.2
vp8_loop_filter16y_inner_h_c: 685.0
vp8_loop_filter16y_inner_h_rvv_i32: 497.0
vp8_loop_filter16y_inner_v_c: 743.7
vp8_loop_filter16y_inner_v_rvv_i32: 295.7
---
libavcodec/riscv/vp8dsp_init.c | 4 ++
libavcodec/riscv/vp8dsp_rvv.S | 110 +++++++++++++++++++++++++++++++++
2 files changed, 114 insertions(+)
@@ -123,6 +123,10 @@ av_cold void ff_vp8dsp_init_riscv(VP8DSPContext *c)
c->vp8_idct_dc_add4uv = ff_vp8_idct_dc_add4uv_rvv;
}
+ c->vp8_v_loop_filter16y_inner = ff_vp8_v_loop_filter16_inner_rvv;
+ c->vp8_h_loop_filter16y_inner = ff_vp8_h_loop_filter16_inner_rvv;
+ c->vp8_v_loop_filter8uv_inner = ff_vp8_v_loop_filter8uv_inner_rvv;
+
c->vp8_v_loop_filter_simple = ff_vp8_v_loop_filter16_simple_rvv;
c->vp8_h_loop_filter_simple = ff_vp8_h_loop_filter16_simple_rvv;
}
@@ -72,6 +72,13 @@ func ff_vp8_idct_dc_add4uv_rvv, zve32x
ret
endfunc
+.macro filter_abs dst diff fI
+ vneg.v v8, \diff
+ vmax.vv \dst, v8, \diff
+ vmsleu.vx v8, \dst, \fI
+ vmand.mm v27, v27, v8
+.endm
+
.macro filter_fmin len a f1 p0f2 q0f1
.ifc \len,16
vsetvli zero, zero, e16, m2, ta, ma
@@ -101,6 +108,16 @@ endfunc
vle8.v v11, (t4)
vle8.v v17, (t1)
vle8.v v22, (\dst)
+ .if \normal
+ sub t3, t2, a6
+ sub t0, t1, a6
+ add t6, \dst, a6
+ add a7, t4, a6
+ vle8.v v2, (t3)
+ vle8.v v15, (t0)
+ vle8.v v10, (t6)
+ vle8.v v14, (a7)
+ .endif
.else
addi t1, \dst, -1
addi a6, \dst, -2
@@ -109,9 +126,28 @@ endfunc
vlse8.v v11, (t4), \stride
vlse8.v v17, (t1), \stride
vlse8.v v22, (\dst), \stride
+ .if \normal
+ addi t5, \dst, -4
+ addi t0, \dst, -3
+ addi t6, \dst, 2
+ addi a7, \dst, 3
+ vlse8.v v2, (t5), \stride
+ vlse8.v v15, (t0), \stride
+ vlse8.v v10, (t6), \stride
+ vlse8.v v14, (a7), \stride
+ .endif
.endif
vwsubu.vv v12, v1, v11 // p1-q1
vwsubu.vv v24, v22, v17 // q0-p0
+
+.if \normal
+ vwsubu.vv v30, v1, v17
+ vwsubu.vv v20, v11, v22
+ vwsubu.vv v28, v1, v15
+ vwsubu.vv v4, v2, v15
+ vwsubu.vv v6, v10, v11
+ vwsubu.vv v2, v14, v10
+.endif
vnclip.wi v23, v12, 0
.ifc \len,16
@@ -130,6 +166,26 @@ endfunc
vmacc.vx v18, a7, v8
vmsleu.vx v0, v18, \fE
+.if \normal
+ vneg.v v18, v30
+ vmax.vv v30, v18, v30
+ vmsleu.vx v27, v30, \fI
+ filter_abs v18 v28 \fI
+ filter_abs v18 v4 \fI
+ filter_abs v18 v6 \fI
+ filter_abs v18 v2 \fI
+ filter_abs v20 v20 \fI
+ vmand.mm v27, v0, v27 // vp8_simple_limit && normal
+
+ vmsgtu.vx v20, v20, \thresh // hev
+ vmsgtu.vx v3, v30, \thresh
+ vmor.mm v3, v3, v20 // v3 = hev: > thresh
+ vzext.vf2 v18, v1 // v18 = p1
+ vmand.mm v0, v27, v3 // v0 = normal && hev
+ vzext.vf2 v20, v11 // v12 = q1
+ vmnot.m v3, v3 // v3 = !hv
+.endif
+
li t5, 3
li a7, 124
li t3, 123
@@ -163,8 +219,62 @@ endfunc
vsse8.v v6, (\dst), \stride, v0.t
.endif
+.if \normal
+ vmand.mm v0, v27, v3 // vp8_normal_limit & !hv
+
+ .if \inner
+ vnclip.wi v30, v30, 0
+ filter_fmin \len v30 v24 v4 v6
+ vadd.vi v24, v24, 1
+ vsra.vi v24, v24, 1 // (f1 + 1) >> 1;
+ vadd.vv v8, v18, v24
+ vsub.vv v10, v20, v24
+ .endif
+
+ vmax.vx v8, v8, zero
+ vmax.vx v10, v10, zero
+.ifc \len,16
+ vsetvli zero, zero, e8, m1, ta, ma
+.else
+ vsetvli zero, zero, e8, mf2, ta, ma
+.endif
+ vnclipu.wi v4, v4, 0
+ vnclipu.wi v5, v6, 0
+ vnclipu.wi v6, v8, 0
+ vnclipu.wi v7, v10, 0
+ .ifc \type,v
+ vse8.v v4, (t1), v0.t
+ vse8.v v5, (\dst), v0.t
+ vse8.v v6, (t2), v0.t
+ vse8.v v7, (t4), v0.t
+ .else
+ vsse8.v v4, (t1), \stride, v0.t
+ vsse8.v v5, (\dst), \stride, v0.t
+ vsse8.v v6, (a6), \stride, v0.t
+ vsse8.v v7, (t4), \stride, v0.t
+ .endif
+.endif
.endm
+func ff_vp8_h_loop_filter16_inner_rvv, zve32x
+ vsetivli zero, 16, e8, m1, ta, ma
+ filter 16 h 1 1 a0 a1 a2 a3 a4
+ ret
+endfunc
+
+func ff_vp8_v_loop_filter16_inner_rvv, zve32x
+ vsetivli zero, 16, e8, m1, ta, ma
+ filter 16 v 1 1 a0 a1 a2 a3 a4
+ ret
+endfunc
+
+func ff_vp8_v_loop_filter8uv_inner_rvv, zve32x
+ vsetivli zero, 8, e8, mf2, ta, ma
+ filter 8 v 1 1 a0 a2 a3 a4 a5
+ filter 8 v 1 1 a1 a2 a3 a4 a5
+ ret
+endfunc
+
func ff_vp8_v_loop_filter16_simple_rvv, zve32x
vsetivli zero, 16, e8, m1, ta, ma
filter 16 v 0 0 a0 a1 a2 a3 a4
--
2.44.0