Closed ax3l closed 3 days ago
Only the std:: math functions have overloads for double and float. Without the std:: prefix, we use the C functions, which will when we run ImpactX in single precision cause up and down casts to double (performance penalty).
std::
double
float
Thanks! Yes I did a line-by-line review and touch-up yesterday evening after search-replace-all.
Only the
std::
math functions have overloads fordouble
andfloat
. Without thestd::
prefix, we use the C functions, which will when we run ImpactX in single precision cause up and down casts todouble
(performance penalty).