sqrt should use machine instruction on x86_64
The NCG currently implements the square root MachOp with an out-of-line call. This resulted in a large performance regression in n-body (see #13570 (closed)). Fix this.
See discussion here as well.
The NCG currently implements the square root MachOp with an out-of-line call. This resulted in a large performance regression in n-body (see #13570 (closed)). Fix this.
See discussion here as well.