On Wed, 28 Jan 2026, Siddhesh Poyarekar wrote: > On 2026-01-28 22:00, Andrew Pinski wrote: > > On Wed, Jan 28, 2026 at 6:20 PM Siddhesh Poyarekar <[email protected]> > > wrote: > >> > >> On 2026-01-28 20:37, Siddhesh Poyarekar wrote: > >>>> I think overall the idea here is solid, just the walk back I am > >>>> questioning if needed. > >>> > >>> Thanks, I was looking for assurance that I'm heading in the right > >>> direction. Thanks to your questions, I realized I should maybe try > >>> experimenting with strncpy as well; I have a feeling this will trip > >>> builtins too, not just general functions with the access attribute. I'll > >>> incorporate that into v2. > >> > >> This tunnel does indeed go deep; it's going to be a bit more involved, I > >> just need to figure out the cleanest way to do this. > > > > And unions are not the one case where the problem shows up. > > I added another simple testcase to PR 123801 which shows unions are > > not only the issue in the end :(. > > And that has been an issue since GCC 11 even. So maybe we should > > rethink waccess3 . > > Ugh, I just saw that. Let me think about this a bit more.
Doesn't the issue also equally well cause false negatives? Which would make a point of adding extra diagnoses for those aliases? IMO doing waccess this late is simply flawed. Richard. > Thanks, > Sid > > -- Richard Biener <[email protected]> SUSE Software Solutions Germany GmbH, Frankenstrasse 146, 90461 Nuernberg, Germany; GF: Jochen Jaser, Andrew McDonald, Werner Knoblich; (HRB 36809, AG Nuernberg)
