https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117176
--- Comment #5 from Tamar Christina <tnfchris at gcc dot gnu.org> ---
Have added more lowering to gcond handling and now we vectorize this as well:
ptrue p7.b, all
fmov z30.d, #1.0e+0
ld1d z31.d, p7/z, [sp, #1, mul vl]
fcmlt p15.d, p7/z, z31.d, z30.d
not p7.b, p7/z, p15.b
str p7, [x0, #7, mul vl]
which we didn't in GCC 14.
Running regressions and will send patch.
