[Bug c/98541] warning with -Wnonnull for array parameter with bound > 0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98541 uecker at gcc dot gnu.org changed: What|Removed |Added Known to fail||11.0 Resolution|--- |FIXED Status|NEW |RESOLVED Target Milestone|--- |14.0 CC||uecker at gcc dot gnu.org --- Comment #5 from uecker at gcc dot gnu.org --- Fixed on trunk.
[Bug c/98541] warning with -Wnonnull for array parameter with bound > 0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98541 --- Comment #4 from CVS Commits --- The master branch has been updated by Martin Uecker : https://gcc.gnu.org/g:6e9ee44d96e5bda8808dd9d8ccf58d2525383f6b commit r14-5115-g6e9ee44d96e5bda8808dd9d8ccf58d2525383f6b Author: Martin Uecker Date: Thu Apr 13 19:35:15 2023 +0200 Reduce false positives for -Wnonnull for VLA parameters [PR98541] This patch limits the warning about NULL arguments to VLA parameters declared [static n]. PR c/98541 gcc/ * gimple-ssa-warn-access.cc (pass_waccess::maybe_check_access_sizes): For VLA bounds in parameters, only warn about null pointers with 'static'. gcc/testsuite: * gcc.dg/Wnonnull-4.c: Adapt test. * gcc.dg/Wstringop-overflow-40.c: Adapt test.
[Bug c/98541] warning with -Wnonnull for array parameter with bound > 0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98541 --- Comment #3 from Martin Uecker --- Having to set the length to 0 for NULL pointers is problematic also because zero-sized array types are now allowed in ISO C. On the other hand, an optional warning that warns for all parameters declared as array which are NULL (fixed length and VLA) might be useful.
[Bug c/98541] warning with -Wnonnull for array parameter with bound > 0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98541 --- Comment #2 from Martin Sebor --- (In reply to Martin Sebor from comment #1) > and necessary to prevent this use case. I mean: ...to prevent invalidating it.
[Bug c/98541] warning with -Wnonnull for array parameter with bound > 0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98541 Martin Sebor changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever confirmed|0 |1 CC||msebor at gcc dot gnu.org Last reconfirmed||2021-01-06 --- Comment #1 from Martin Sebor --- Both the warning in the VLA case and its absence in the ordinary case are by design. The VLA case gets a warning because it seems suspicious to specify a nonzero number of elements for an array and then null as the array (the VLA parameter can be null when the bound is zero). But I didn't consider the shared bound use case so I suppose the warning could be relaxed to accommodate it, and either changed to rely on the [static] for the bound as you suggest or on the VLA parameter declared with attribute nonnull. I can't think of a use case where the bound isn't shared and where setting it to nonzero and passing a null pointer would not be a bug. My rationale for not issuing a warning for ordinary arrays with constant bounds is that those are commonly treated as "optional," and so there relying on some additional notation ([static] or nonnull) seems appropriate and necessary to prevent this use case.
