Okay, updated it in the reply
Rémi Denis-Courmont <r...@remlab.net> 于2024年2月13日周二 03:54写道：
> Hi,
>
> To avoid repeating the code, you can either use .repr or .irp. You can
> even
> use assembler conditionals to elide the redundant code on the last
> iteration.
>
> --
> レミ・デニ-クールモン
> http://www.remlab.net/
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
>
From d44e7ec0950ddb431aa75992cfa12d7687d39514 Mon Sep 17 00:00:00 2001
From: sunyuechi <sunyue...@iscas.ac.cn>
Date: Fri, 2 Feb 2024 12:50:07 +0800
Subject: [PATCH 2/3] 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
---
libavcodec/riscv/vp8dsp_init.c | 2 ++
libavcodec/riscv/vp8dsp_rvv.S | 16 ++++++++++++++++
2 files changed, 18 insertions(+)
diff --git a/libavcodec/riscv/vp8dsp_init.c b/libavcodec/riscv/vp8dsp_init.c
index ab020070eb..6615d3d440 100644
--- a/libavcodec/riscv/vp8dsp_init.c
+++ b/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
}
diff --git a/libavcodec/riscv/vp8dsp_rvv.S b/libavcodec/riscv/vp8dsp_rvv.S
index 8609b79937..a5a22188c1 100644
--- a/libavcodec/riscv/vp8dsp_rvv.S
+++ b/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
--
2.43.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
