如题所示，其实我把这个累加计算换成了汇编写的，所需要的cycle却变多了。（前提，DSP软件勾选了“Enable optimization”）

for(n = 0; n < 65; n++)
      {
       dctArrayOut+= a[n]*cos_11[n];                                
      }

换成了：

.section program;

.global _a_dot_c_asm;

_a_dot_c_asm:

P0 = R0;
I0 = R1;
P1 = 64;
R0 = 0;
NOP;
R1 = [P0++];
R2 = [I0++];
LSETUP (begin_loop, end_loop) LC0 = P1;

begin_loop: R1 *= R2;
             R2 = [I0++];
end_loop: R0= R0 + R1 (NS) || R1 = [P0++] || NOP;

R1 *= R2;
R0 = R0 + R1;
R0 = R0>>19;
RTS;

_a_dot_c_asm.end:

实在想不通，为什么cycle会变多呢？按理说，汇编比C要快啊？

求各位大神帮忙解答一下谢谢~