Re: Jikes RVM & mauve tests
> "Eric" == Eric Blake <[EMAIL PROTECTED]> writes: Eric> I stand by the Classpath version of Character. I updated it to Eric> correctly obey the same Unicode version as JDK 1.4, and verified Eric> everything against the JDK. However, I never found the time to go and Eric> fix Mauve to do the same. So the Mauve tests for Character are Eric> currently broken. Yeah. They were always a little bogus, I'm afraid. Back when they were written the Character documentation was notably less clear... Eric> Back in Feb 02 or so, I posted a patch to gcj to "merge" as much of Eric> Character and String as possible between Classpath and libgcj, but it Eric> was never applied (back in the 3.1 days), so I would practically have Eric> to start the merge again to catch everything that has happened in both Eric> trees in the meantime. First, I'm completely responsible for the String merge not going forward. I'm really sorry about that. Second, I thought the Character changes did go in? Tom ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Re: Jikes RVM & mauve tests
Thanks for the information. It's useful to know that the test and libraries are out of synch.It only takes about 20 seconds to run all of the java.lang.Character.unicode tests on Jikes RVM, so we'll probably just xfail the ones that are currently failing but go ahead and run the rest of the test anyways. I'll probably set something up next week to run the mauve tests as part of our daily Jikes RVM regression tests. The whole thing only takes a few minutes to run. Seems like nightly checks of classpath and Jikes RVM CVS heads are a worthwhile thing to be doing. --dave ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Re: Jikes RVM & mauve tests
Eric Blake <[EMAIL PROTECTED]> writes: > I stand by the Classpath version of Character. I updated it to > correctly obey the same Unicode version as JDK 1.4, and verified > everything against the JDK. However, I never found the time to go and > fix Mauve to do the same. So the Mauve tests for Character are > currently broken. > > Back in Feb 02 or so, I posted a patch to gcj to "merge" as much of > Character and String as possible between Classpath and libgcj, but it > was never applied (back in the 3.1 days), so I would practically have > to start the merge again to catch everything that has happened in both > trees in the meantime. I've checked and it's been touched a few times to add or remove methods, mostly by Mark. Hopefully none of those were impropoer gcj/classpath merges as you noted back in April last year. Brian -- Brian Jones <[EMAIL PROTECTED]> ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Re: Jikes RVM & mauve tests
Brian Jones wrote: David P Grove <[EMAIL PROTECTED]> writes: Hi, I've been experimenting a little bit with running Mauve tests on top of Jikes RVM and GNU Classpath. This generally does pretty well, except that there are massive failures on java.lang.Character.unicode ("make check KEYS=!java. java.lang.Character" yields 136460 of 3603803 tests failed). Before looking into this further, I just wanted to verify that the classpath libraries are expected to pass these tests. There may be Character problems... I tend not to run it because it (a) takes a while and (b) inflates positive results out of proportion with other class tests. I believe this is another area where gcj/classpath differ as well. I stand by the Classpath version of Character. I updated it to correctly obey the same Unicode version as JDK 1.4, and verified everything against the JDK. However, I never found the time to go and fix Mauve to do the same. So the Mauve tests for Character are currently broken. Back in Feb 02 or so, I posted a patch to gcj to "merge" as much of Character and String as possible between Classpath and libgcj, but it was never applied (back in the 3.1 days), so I would practically have to start the merge again to catch everything that has happened in both trees in the meantime. -- This signature intentionally left boring. Eric Blake [EMAIL PROTECTED] BYU student, free software programmer ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Re: Jikes RVM & mauve tests
Hi, On Fri, 2003-01-31 at 23:57, David P Grove wrote: > I've been experimenting a little bit with running Mauve tests on top > of Jikes RVM and GNU Classpath. This generally does pretty well, except > that there are massive failures on java.lang.Character.unicode ("make check > KEYS=!java. java.lang.Character" yields 136460 of 3603803 tests failed). > Before looking into this further, I just wanted to verify that the > classpath libraries are expected to pass these tests. Both Kissme and gcj disable this test by default. If I remember correctly they are based on an old unicode spec. > On the bright side, "make check KEYS=!java.net > !java.lang.Character.unicode" yields 127 of 2017 tests failed. Roughly > half of these failures are because Jikes RVM has never bothered to > implement thread priorities (easily fixed). Note that there are still deficiencies in Classpath that can explain some of these failures. With kissme and the above KEYS I get: 74 of 2037 tests failed Brian has results posted on: http://www.haphazard.org/~cbj/classpath/automauve/summary.html And I reqularly post results on gcc-testresults for gcj. See for the latests: http://gcc.gnu.org/ml/gcc-testresults/2003-01/msg01418.html (Note that gcj uses a exclude file and an xfail file.) > I excluded java.net because of the following javac compilation > problem...is this expected? > public class TestSocketImplFactory >^ > gnu/testlet/java/net/Socket/SocketTest.java:203: Socket() has protected > access in java.net.Socket > sock = new Socket (); // unconnected socket > The no argument constructor is public in Classpath (specnote says since 1.4). I never noticed this since I use either gcj or jikes with the Classpath libraries to compile Mauve. Maybe we could move that part of the test into a new Socket14.java testlet that explicitly sets Tags: JDK1.4. Cheers, Mark ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Re: Jikes RVM & mauve tests
David P Grove <[EMAIL PROTECTED]> writes: > Hi, > > I've been experimenting a little bit with running Mauve tests > on top of Jikes RVM and GNU Classpath. This generally does pretty > well, except that there are massive failures on > java.lang.Character.unicode ("make check > KEYS=!java. java.lang.Character" yields 136460 of 3603803 tests > failed). Before looking into this further, I just wanted to verify > that the classpath libraries are expected to pass these tests. There may be Character problems... I tend not to run it because it (a) takes a while and (b) inflates positive results out of proportion with other class tests. I believe this is another area where gcj/classpath differ as well. > On the bright side, "make check KEYS=!java.net > !java.lang.Character.unicode" yields 127 of 2017 tests failed. > Roughly half of these failures are because Jikes RVM has never > bothered to implement thread priorities (easily fixed). > > I excluded java.net because of the following javac compilation > problem...is this expected? > public class TestSocketImplFactory >^ > gnu/testlet/java/net/Socket/SocketTest.java:203: Socket() has protected > access in java.net.Socket > sock = new Socket (); // unconnected socket > This last one is because I was looking at more recent docs which indicate it is a public constructor but in JDK 1.1 it was protected. I need to move this particular test to another file with appropriate tags entry. On the bright side it should have compiled and worked for you using JDK 1.4 (you mention javac) or Classpath + Jikes. Thanks, Brian -- Brian Jones <[EMAIL PROTECTED]> ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
Jikes RVM & mauve tests
Hi, I've been experimenting a little bit with running Mauve tests on top of Jikes RVM and GNU Classpath. This generally does pretty well, except that there are massive failures on java.lang.Character.unicode ("make check KEYS=!java. java.lang.Character" yields 136460 of 3603803 tests failed). Before looking into this further, I just wanted to verify that the classpath libraries are expected to pass these tests. On the bright side, "make check KEYS=!java.net !java.lang.Character.unicode" yields 127 of 2017 tests failed. Roughly half of these failures are because Jikes RVM has never bothered to implement thread priorities (easily fixed). I excluded java.net because of the following javac compilation problem...is this expected? public class TestSocketImplFactory ^ gnu/testlet/java/net/Socket/SocketTest.java:203: Socket() has protected access in java.net.Socket sock = new Socket (); // unconnected socket thanks, --dave ___ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath