This is a tricky bug. It is not a standards version issue, because the code compiles fine. Rather, a test case fails. (And also fails if an earlier standards version is specified.) But if the optimization is turned off, or turned down to just -O, this test is passed. So there is something being done by the optimizer that breaks the code. Not so unusual, since systems like this often have to bend the rules with pointer manipulation and such.
I could fix this present bug by simply turning optimization down to -O. But the underlying issue might break user code as well, which would be scary. So I'd rather figure out the actual problem and address it, rather than simply masking this particular manifestation. Any assistance in this regard would be quite welcome. -- Barak A. Pearlmutter <ba...@pearlmutter.net> Dept Comp Sci & Hamilton Institute, Maynooth University, Co. Kildare, Ireland http://barak.pearlmutter.net -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
