X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libavfilter%2Fx86%2Faf_afir.asm;h=2cc09709a2772c39913c5ccc09b02105fe9f160a;hb=c67d2a287502845baadf986a9c63e6117a25be3f;hp=849d85e70fba83161ef73b7430537f21f5e2c6e4;hpb=f8377ffce35251bba043aeda5d81df0d411a0595;p=ffmpeg

diff --git a/libavfilter/x86/af_afir.asm b/libavfilter/x86/af_afir.asm
index 849d85e70fb..2cc09709a27 100644
--- a/libavfilter/x86/af_afir.asm
+++ b/libavfilter/x86/af_afir.asm
@@ -27,10 +27,9 @@ SECTION .text
 ; void ff_fcmul_add(float *sum, const float *t, const float *c, int len)
 ;------------------------------------------------------------------------------
 
-INIT_XMM sse3
+%macro FCMUL_ADD 0
 cglobal fcmul_add, 4,4,6, sum, t, c, len
     shl       lend, 3
-    add       lend, mmsize*2
     add         tq, lenq
     add         cq, lenq
     add       sumq, lenq
@@ -41,20 +40,30 @@ ALIGN 16
     movsldup  m3, [tq + lenq+mmsize]
     movaps    m1, [cq + lenq]
     movaps    m4, [cq + lenq+mmsize]
-    mulps     m0, m1
-    mulps     m3, m4
-    shufps    m1, m1, 0xb1
-    shufps    m4, m4, 0xb1
+    mulps     m0, m0, m1
+    mulps     m3, m3, m4
+    shufps    m1, m1, m1, 0xb1
+    shufps    m4, m4, m4, 0xb1
     movshdup  m2, [tq + lenq]
     movshdup  m5, [tq + lenq+mmsize]
-    mulps     m2, m1
-    mulps     m5, m4
-    addsubps  m0, m2
-    addsubps  m3, m5
-    addps     m0, [sumq + lenq]
-    addps     m3, [sumq + lenq+mmsize]
+    mulps     m2, m2, m1
+    mulps     m5, m5, m4
+    addsubps  m0, m0, m2
+    addsubps  m3, m3, m5
+    addps     m0, m0, [sumq + lenq]
+    addps     m3, m3, [sumq + lenq+mmsize]
     movaps    [sumq + lenq], m0
     movaps    [sumq + lenq+mmsize], m3
     add       lenq, mmsize*2
     jl .loop
-    REP_RET
+    movss xm0, [tq + lenq]
+    mulss xm0, [cq + lenq]
+    addss xm0, [sumq + lenq]
+    movss [sumq + lenq], xm0
+    RET
+%endmacro
+
+INIT_XMM sse3
+FCMUL_ADD
+INIT_YMM avx
+FCMUL_ADD