[FFmpeg-cvslog] avcodec/x86/audiodsp: add scalarproduct avx2

[Paul B Mahol]

ffmpeg | branch: master | Paul B Mahol <onemda at gmail.com> | Mon Sep 12 18:53:31 2022 +0200| [37a503ac879ca7677beb7423c33a6c5d24dd6396] | committer: Paul B Mahol

avcodec/x86/audiodsp: add scalarproduct avx2

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=37a503ac879ca7677beb7423c33a6c5d24dd6396
---

 libavcodec/x86/audiodsp.asm    | 18 ++++++++++++++++++
 libavcodec/x86/audiodsp_init.c |  6 ++++++
 2 files changed, 24 insertions(+)

diff --git a/libavcodec/x86/audiodsp.asm b/libavcodec/x86/audiodsp.asm
index b604b0443c..f64077cb13 100644
--- a/libavcodec/x86/audiodsp.asm
+++ b/libavcodec/x86/audiodsp.asm
@@ -44,6 +44,24 @@ cglobal scalarproduct_int16, 3,3,3, v1, v2, order
     movd   eax, m2
     RET
 
+%if HAVE_AVX2_EXTERNAL
+INIT_YMM avx2
+cglobal scalarproduct_int16, 3,3,2, v1, v2, order
+    add orderd, orderd
+    add v1q, orderq
+    add v2q, orderq
+    neg orderq
+    pxor    m1, m1
+.loop:
+    movu    m0, [v1q + orderq]
+    pmaddwd m0, [v2q + orderq]
+    paddd   m1, m0
+    add     orderq, mmsize
+    jl .loop
+    HADDD   m1, m0
+    movd   eax, xm1
+    RET
+%endif
 
 ;-----------------------------------------------------------------------------
 ; void ff_vector_clip_int32(int32_t *dst, const int32_t *src, int32_t min,
diff --git a/libavcodec/x86/audiodsp_init.c b/libavcodec/x86/audiodsp_init.c
index aa5e43e570..68aa3b2129 100644
--- a/libavcodec/x86/audiodsp_init.c
+++ b/libavcodec/x86/audiodsp_init.c
@@ -24,6 +24,9 @@
 #include "libavutil/x86/cpu.h"
 #include "libavcodec/audiodsp.h"
 
+int32_t ff_scalarproduct_int16_avx2(const int16_t *v1, const int16_t *v2,
+                                    int order);
+
 int32_t ff_scalarproduct_int16_sse2(const int16_t *v1, const int16_t *v2,
                                     int order);
 
@@ -53,4 +56,7 @@ av_cold void ff_audiodsp_init_x86(AudioDSPContext *c)
 
     if (EXTERNAL_SSE4(cpu_flags))
         c->vector_clip_int32 = ff_vector_clip_int32_sse4;
+
+    if (EXTERNAL_AVX2_FAST(cpu_flags))
+        c->scalarproduct_int16 = ff_scalarproduct_int16_avx2;
 }