From 9aeb6aca3a9fa8caf8db263de4fa9f8077ba6634 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?R=C3=A9mi=20Denis-Courmont?= <remi@remlab.net>
Date: Mon, 26 Sep 2022 17:52:36 +0300
Subject: [PATCH] lavu/floatdsp: RISC-V V vector_fmul_reverse

---
 libavutil/riscv/float_dsp_init.c |  3 +++
 libavutil/riscv/float_dsp_rvv.S  | 21 +++++++++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/libavutil/riscv/float_dsp_init.c b/libavutil/riscv/float_dsp_init.c
index a1cd180cdc..b99e3080c9 100644
--- a/libavutil/riscv/float_dsp_init.c
+++ b/libavutil/riscv/float_dsp_init.c
@@ -33,6 +33,8 @@ void ff_vector_fmul_scalar_rvv(float *dst, const float *src, float mul,
                                 int len);
 void ff_vector_fmul_add_rvv(float *dst, const float *src0, const float *src1,
                              const float *src2, int len);
+void ff_vector_fmul_reverse_rvv(float *dst, const float *src0,
+                                 const float *src1, int len);
 void ff_butterflies_float_rvv(float *v1, float *v2, int len);
 
 void ff_vector_dmul_rvv(double *dst, const double *src0, const double *src1,
@@ -52,6 +54,7 @@ av_cold void ff_float_dsp_init_riscv(AVFloatDSPContext *fdsp)
         fdsp->vector_fmac_scalar = ff_vector_fmac_scalar_rvv;
         fdsp->vector_fmul_scalar = ff_vector_fmul_scalar_rvv;
         fdsp->vector_fmul_add = ff_vector_fmul_add_rvv;
+        fdsp->vector_fmul_reverse = ff_vector_fmul_reverse_rvv;
         fdsp->butterflies_float = ff_butterflies_float_rvv;
     }
 
diff --git a/libavutil/riscv/float_dsp_rvv.S b/libavutil/riscv/float_dsp_rvv.S
index a721c44667..fbd2777463 100644
--- a/libavutil/riscv/float_dsp_rvv.S
+++ b/libavutil/riscv/float_dsp_rvv.S
@@ -93,6 +93,27 @@ func ff_vector_fmul_add_rvv, zve32f
         ret
 endfunc
 
+// (a0) = (a1) * reverse(a2) [0..a3-1]
+func ff_vector_fmul_reverse_rvv, zve32f
+        sh2add   a2, a3, a2
+        li       t2, -4 // byte stride
+        addi     a2, a2, -4
+1:
+        vsetvli  t0, a3, e32, m1, ta, ma
+        slli     t1, t0, 2
+        vle32.v  v16, (a1)
+        sub      a3, a3, t0
+        vlse32.v v24, (a2), t2
+        add      a1, a1, t1
+        vfmul.vv v16, v16, v24
+        sub      a2, a2, t1
+        vse32.v  v16, (a0)
+        add      a0, a0, t1
+        bnez     a3, 1b
+
+        ret
+endfunc
+
 // (a0) = (a0) + (a1), (a1) = (a0) - (a1) [0..a2-1]
 func ff_butterflies_float_rvv, zve32f
 1: