[Bug middle-end/111696] [11/12/13/14 Regression] Spurious -Wstringop-overflow

2024-03-07 Thread law at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111696 Jeffrey A. Law changed: What|Removed |Added CC||law at gcc dot gnu.org Prior

[Bug middle-end/111696] [11/12/13/14 Regression] Spurious -Wstringop-overflow

2023-10-05 Thread rguenth at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111696 Richard Biener changed: What|Removed |Added Target Milestone|--- |11.5 Blocks|

[Bug middle-end/111696] [11/12/13/14 Regression] Spurious -Wstringop-overflow

2023-10-04 Thread pinskia at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111696 --- Comment #2 from Andrew Pinski --- I remember seeing this similar issue before where PRE/FRE/VN decides the two addresses are the same (they are) but since waccess does not realize other passes will use the restricted address still and still

[Bug middle-end/111696] [11/12/13/14 Regression] Spurious -Wstringop-overflow

2023-10-04 Thread pinskia at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111696 --- Comment #1 from Andrew Pinski --- The vectorizer has not much to do with the warning issue except for the IR change into FRE. Before fre5: _11 = &last_14(D)->s; memcpy (_11, &cur.s, 24); After: vectp.7_19 = &accum_13(D)->sD.4831.aD.