------- Comment #15 from hubicka at ucw dot cz 2010-01-16 12:30 ------- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility breaks Ada bootstrap
> No doubt about that. The problem is, this was installed without mention in > the > ChangeLog so looks quite unintentional. If this is intentional, you should > fix > the ChangeLog. Will do, the earlier version of sanity check was wrong. Actually looking into DECL_COMMON implementation, it seems that even for static variables this seems to work as intended at least on i386, so perhaps the sanity check is overactive. I am double checking that now. > > > I will have time to try to reproduce the problem only tomorrow or during > > weekend (it will be bit tricky as I know of no machine known to fail here I > > can access easily), > > Please read comment #5. I am not questioning that. I have somewhat hectic time now and did not notice situation earlier. I am finishing some papers and my thesis now. My apologize for missing this. Will finish patch today. Honza -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068