how to use addsub/addsub2

Arunmoezhi Ramachandran

int i,j;
float* data;
float x,y;

data[j]   = data[i] + x;
data[i]   = data[i] - x;

data[j+1] = data[i+1] - y;
data[i+1] = data[i+1] + y;

Is there a way to use addsub/addsub2 to do the add and sub in parallel?

0 Clemens Eisserer over 12 years ago

Expert 2430 points

Addsub / Addsub2 are integer, however your data seems to be floating point.

In general, when you find an appropriate instruction, you can use it with intrinsics.

Processors forum