Replace SSE2 instruction in scalarproduct_float_sse() by SSE equivalent.

Fixes an AAC decoding issue with the sample from ticket #213 on machines
with SSE but without SSE2.
Based on 89411a by Reimar.

libavcodec/x86/dsputil_yasm.asm

@@ -473,7 +473,7 @@ cglobal scalarproduct_float_sse, 3,3,2, v1, v2, offset
     shufps  xmm0, xmm0, 1
     addss   xmm0, xmm1
 %if ARCH_X86_64 == 0
-    movd    r0m,  xmm0
+    movss   r0m,  xmm0
     fld     dword r0m
 %endif
     RET