This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Fix wrong-code bug caused by combine_simplify_rtx()
Eric Botcazou wrote:
Let me report another problem though: very recently (3 days ago), 22 Java
failures have been introduced on SPARC (44 for a multiarch compiler):
http://gcc.gnu.org/ml/gcc-testresults/2005-04/msg00389.html
http://gcc.gnu.org/ml/gcc-testresults/2005-04/msg00508.html
on 4.0 branch.
Ugh.
I've never seen these failures, they are brand new problems. As I've never
debugged interpreted Java tests, I'm a little at a loss, but I think the Java
compiler might be miscompiled (I recompiled libjava a few times with rebuild
but not bootstrapped compilers and didn't get the failures). I'm currently
trying to pinpoint the patch.
What's your position on this?
Java's not part of the release criteria, so I don't think this is a
showstopper in-and-of itself. But, if we're miscompiling the Java
compiler, that's certainly not a good sign -- and it is itself a C
program, of course. I'd sure like to understand what's going on better
before making a decision. Have you been able to isolate the
regression-causing patch?
Will you please keep me posted?
--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304