[Bug tree-optimization/122286] fold (a < b) ? -1 : (a > b ? 1 : 0) into (a < b) - (a > b) or the other way around

[pinskia at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122286

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
For aarch64, cmp_double_asc and cmp_double_asc_2 are better than
cmp_double_asc_branchless.
        fcmpe   d0, d1
        cset    w0, gt
        csinv   w0, w0, wzr, ge
        ret
vs:
        fcmpe   d0, d1
        cset    w1, mi
        cset    w0, gt
        sub     w0, w1, w0
        ret

This is because forming -1/0 is easy done in csinv/cset and doing `b?-a:c` is
also easy.