Hi all,<br><br>I am observing an excessive use of xmm registers in the output assembly produced by x86 backend. Basically, for a code like this<br><br><pre>double test(double a, double b) {<br> double c;<br><br> c = 1.0 + sin (a + b*b);<br>
<br> return c;<br>}</pre><br>llc produced somthing like....<br><pre>movsd 16(%ebp), %xmm0<br> mulsd %xmm0, %xmm0<br> addsd 8(%ebp), %xmm0<br> movsd %xmm0, (%esp)<br> .......<br> fstpl -8(%ebp<br>
movsd -8(%ebp), %xmm0<br> addsd .LC1, %xmm0<br> movsd %xmm0, -8(%ebp)<br> fldl -8(%ebp)</pre><br>LLVM Backend is using xmms it involves a lot of register moves. llc has one option -mcpu=686, where output does not use xmm but it disables some other instructions also. Is there any fine grain optimization flag like ( -mfpmath = 387 which is present in gcc) which would just instruction backend to use fpstack instead of SSE instructions and still not limit the instructions<br>
<br>Thanks<br><br>--Kapil<br><br>.<br>