GNATS loosing bug reports?

[Jakub Jelinek]

On Tue, Mar 26, 2002 at 10:03:28AM +0100, Robert Schiele wrote:
> The bug report I am talking about is at
> http://gcc.gnu.org/ml/gcc-bugs/2002-03/msg00509.html.
> 
> As this bug can be reproduced in an easy way on sparcv9, I would be
> happy to see this fixed before the 3.1 release.

I've reproduced your testcase even with the 2002-03-13 patch (before that
this behaviour would be "expected"), it looks like
some instruction has incorrect length attribute in the machine description
(gcc thinks the distance from branch to the label is 29953 insn while
in reality it is 33697 insns).
Will work on a fix.

	Jakub