https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120086
--- Comment #19 from John David Anglin <danglin at gcc dot gnu.org> --- (In reply to Jørgen Kvalsvik from comment #18) > Ok, thanks, I pushed a fix for the gcov %zu prints, at least for gcov. Thanks. > To recap: > > * gcov-pr86536.c sporadically fails? (which has to be unrelated to these > prints) > * -fprofile-update=atomic warns (and fails the test?), is this expected on > hppa-hpux? All gcov tests pass if I define TARGET_HAVE_LIBATOMIC on hppa-hpux to true. https://gcc.gnu.org/pipermail/gcc-testresults/2025-May/846432.html It's not clear why gcov-30 passes on Linux. I searched the entire tree and couldn't find how TARGET_HAVE_LIBATOMIC gets set on Linux. Also, always setting TARGET_HAVE_LIBATOMIC to true doesn't seem 100% correct because I believe it's possible to disable building libatomic. I believe the gcov-pr86536.c were in some way related to the patch mentioned in Comment 10. gcov-pr86536.c hasn't failed when gcov-30.c isn't modified.
