https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120022
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Resolution|--- |WONTFIX
Status|UNCONFIRMED |RESOLVED
--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
CSEing inline-asm with multiple output is not going to be optimized.
Just write this in C:
A myDivMod1(U const& a, U const& b)
{
A result;
result.quo = a/b;
result.rem = a%b;
return result;
}
Provides:
myDivMod1(unsigned long const&, unsigned long const&):
.LFB5:
.cfi_startproc
mov rax, QWORD PTR [rdi]
xor edx, edx
div QWORD PTR [rsi]
ret
Which is exactly what you want.
