Summary: double-double functions seem to not pure-optimize
           Product: D
           Version: future
          Platform: x86
        OS/Version: Windows
            Status: NEW
          Keywords: performance
          Severity: enhancement
          Priority: P2
         Component: DMD

--- Comment #0 from 2010-03-31 09:45:44 PDT ---
This small D2 program calls two times a function msin that I have tagged as
pure, but in the binary there are two sin instructions, I don't understand:

import std.c.stdio: printf;
import std.c.stdlib: atof;
import std.math: sin;
pure double msin(double x) {
    return sin(x);
void main() {
    double x = atof("0.3");
    double y = msin(x) + msin(x);
    printf("%f\n", y);

Compiled with dmd v2.042:
dmd -O -release -inline bug1.d


        fld qword ptr 4[ESP]
        ret 8

L0:     sub ESP,01Ch
        mov EAX,offset FLAT:_DATA
        push    EAX
        call    near ptr _atof
        mov ECX,offset FLAT:_DATA[4]
        fstp    qword ptr 8[ESP]
        fld qword ptr 8[ESP]
        fld qword ptr 8[ESP]
        sub ESP,8
        faddp   ST(1),ST
        fstp    qword ptr [ESP]
        push    ECX
        call    near ptr _printf
        add ESP,010h
        add ESP,01Ch
        xor EAX,EAX

Similar code is produced with this line:
double y = sin(x) + sin(x);

Configure issuemail:
------- You are receiving this mail because: -------

Reply via email to