https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83348
Bug ID: 83348
Summary: Missed optimization in math expression: can be used
std::pow function
Product: gcc
Version: 8.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: zamazan4ik at tut dot by
Target Milestone: ---
gcc(trunk) with optimization flags '--std=c++17 -O3 -march=native -ffast-math'
for this code:
double f(double a)
{
return a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*
a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*
a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*
a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*
a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*
a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a*a;
}
generates this assembly:
f(double):
vmulsd xmm1, xmm0, xmm0
vmulsd xmm1, xmm1, xmm1
vmulsd xmm1, xmm1, xmm1
vmulsd xmm1, xmm1, xmm1
vmulsd xmm0, xmm1, xmm0
vmulsd xmm1, xmm0, xmm0
vmulsd xmm0, xmm1, xmm0
vmulsd xmm0, xmm0, xmm0
ret
But here it can be simplified by using std::pow function. If you will increase
length of this multiply chain, gcc just will add more and more multiplications.
