[Bug target/82580] Optimize comparisons for __int128 on x86-64

[ubizjak at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82580

--- Comment #7 from Uroš Bizjak <ubizjak at gmail dot com> ---
(In reply to Jakub Jelinek from comment #6)

> i.e. for the flags we first clear %eax and then setX %al, but for f2
>         cmpq    %rdi, %rdx
>         sbbq    %rsi, %rcx
>         setb    %al
>         movzbl  %al, %eax
> Seems this is done in
> ;; Convert setcc + movzbl to xor + setcc if operands don't overlap.
> peephole2s, do we need a peephole for this as well, or some more general
> peephole2 or other means of optimizing this?
No, this would break carry flag propagation from cmp to sbb.