PATCH in Java constant propagation
Tom Tromey
tromey@redhat.com
Wed Sep 11 12:52:00 GMT 2002
>>>>> "Per" == Per Bothner <per@bothner.com> writes:
Per> I believe this fixes the bug that martin.kahlert@infineon.com
Per> reported (subject: Is this a bug?). I *think* it doesn't break
Per> anything else, but I did get a lot of testsuite errors, so I'm
Per> not sure. I looks like I will be inaccessible for a the rest of
Per> the week, so I'd appreciate it if someone could test it, and if
Per> it works, check it in.
I applied and tested this patch. I rebuilt libgcj from scratch with
it, and I ran the test suite including Mauve. I get just the ordinary
expected problems, and no regressions (against a compiler without the
patch -- some of the currently-expected failures seem to represent
regressions against earlier compilers).
I'm checking this in.
Tom
More information about the Java
mailing list