The Character.unicode test in mauve gives lots of false failures and has been added to the mauve-libgcj ignore list. Should be tested again when Character.unicode has been rewritten to support Unicode 3.0 (which our Character implementation implements). Release: 3.1 How-To-Repeat: In your configured and compiled mauve obj dir do: echo gnu.testlet.java.lang.Character.unicode | ./SimpleTestHarness
Hello, would it be possible for you to also check whether this bug is still present in a current version of gcc, and if so, send a reduced testcase? Thanks, Dara
Hello, scratch the my previous message. I can reproduce the failures here on gcc 3.3: FAIL: gnu.testlet.java.lang.Character.unicode: Character a05e:UNDEFINED incorectly reported as javaindetifierstart (number 1) FAIL: gnu.testlet.java.lang.Character.unicode: Character a05e:UNDEFINED incorectly reported as javaidentifierpart (number 1) ... Thanks, Dara
See Dara's comment.
This appears to be working fine now. Once PR 17587 is fixed, I will check in the appropriate mauve-libgcj patch.
Subject: Bug 6182 CVSROOT: /cvs/gcc Module name: gcc Changes by: tromey@gcc.gnu.org 2004-09-23 00:06:10 Modified files: libjava : ChangeLog mauve-libgcj Log message: PR libgcj/6182: * mauve-libgcj: Enable java.lang.Character tests. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libjava/ChangeLog.diff?cvsroot=gcc&r1=1.3066&r2=1.3067 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libjava/mauve-libgcj.diff?cvsroot=gcc&r1=1.28&r2=1.29
Tests re-enabled.