https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121044
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |needs-bisection
Summary|False positive |[14 Regression] False
|-Warray-bounds iwith GCC 14 |positive -Warray-bounds
|and NEON intrinsics |iwith GCC 14 and NEON
| |intrinsics
Target Milestone|--- |14.4
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I don't see any IR differences between 14 and 15 except for slightly better
range info:
15:
# RANGE [irange] size_t [33, 48] MASK 0x3f VALUE 0x0
vs 14:
# RANGE [irange] size_t [33, 48]
This is for
# i_15 = PHI <i_27(5), i_21(6)>
Which is the i used in:
for (; i < n; i++) {
r[i] = a[i] ^ b[i];
}
Where the warning is coming from.
