lavc/vp8dsp: R-V V vp8_idct_dc_add4y

c908:
vp8_idct_dc_add4y_c: 368.5
vp8_idct_dc_add4y_rvv_i32: 134.5

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

libavcodec/riscv/vp8dsp_init.c

@@ -26,6 +26,7 @@
 #include "libavcodec/vp8dsp.h"
 
 void ff_vp8_idct_dc_add_rvv(uint8_t *dst, int16_t block[16], ptrdiff_t stride);
+void ff_vp8_idct_dc_add4y_rvv(uint8_t *dst, int16_t block[4][16], ptrdiff_t stride);
 
 av_cold void ff_vp8dsp_init_riscv(VP8DSPContext *c)
 {
@@ -34,6 +35,7 @@ av_cold void ff_vp8dsp_init_riscv(VP8DSPContext *c)
 
     if (flags & AV_CPU_FLAG_RVV_I32 && ff_get_rv_vlenb() >= 16) {
         c->vp8_idct_dc_add = ff_vp8_idct_dc_add_rvv;
+        c->vp8_idct_dc_add4y = ff_vp8_idct_dc_add4y_rvv;
     }
 #endif
 }

libavcodec/riscv/vp8dsp_rvv.S

@@ -36,9 +36,25 @@
         vsse32.v      v0, (a0), a2
 .endm
 
+.macro vp8_idct_dc_addy
+        vp8_idct_dc_add
+        addi          a0, a0, 4
+        addi          a1, a1, 32
+.endm
+
 func ff_vp8_idct_dc_add_rvv, zve32x
         vsetivli      zero, 4, e8, mf4, ta, ma
         vp8_idct_dc_add
 
         ret
 endfunc
+
+func ff_vp8_idct_dc_add4y_rvv, zve32x
+        vsetivli      zero, 4, e8, mf4, ta, ma
+        .rept 3
+        vp8_idct_dc_addy
+        .endr
+        vp8_idct_dc_add
+
+        ret
+endfunc