Re: New no-undefined-overflow branch

[Ian Lance Taylor]

Richard Guenther <richard.guent...@gmail.com> writes:

> I agree.  Btw, for the addition case we generate
>
>         leal    (%rsi,%rdi), %eax
>         xorl    %eax, %esi
>         xorl    %eax, %edi
>         testl   %edi, %esi
>         jns     .L2
>                 .value  0x0b0f
> .L2:
>         rep
>         ret
>
> which isn't too bad.

But I think it would normally be faster to do something like
    movl    %esi,%eax
    addl    %edi,%eax
    jo      .Ltrap
    ret
.Ltrap:
    diediedie

Ian