FYI: JamVM: there is something odd going on with the langtools tests

[Dr Andrew John Hughes]

On 23:54 Wed 02 Mar     , Xerxes Rånby wrote:
> On 2011-02-28 01:44, Xerxes Rånby wrote:
> > On 2011-02-28 00:37, Xerxes Rånby wrote:
> >> On 2011-02-28 00:01, Mark Wielaard wrote:
> >>  > BTW. If you look at the buildbot test results there is something
> >>  > odd going on with the langtools tests. There are some failures that
> >>  > we should take a look at. But the odd thing is that they seem to
> >>  > just abort after T6397044. I can reproduce this locally.
> >>  > The logs stop at T6397044, but I have a (passing) T6397286.jtr
> >>  > file. Which seems to mean the actual failure/crash/abort is in
> >>  > that one or in processing the results from it. I haven't debugged
> >>  > it further.
> >>  >
> >> Yes this is odd.
> >>
> >> I have attached a patch for icedtea6 that disables a lot more
> >> Unrecognized options. I belive this will help us to locate tests that
> >> take-down JamVM in a unclean way. By finding these failing tests we will
> >> have a easier time to fix remaining bugs in JamVM.
> > Updated the patch once more.
> > > Ok for TRUNK ?
> 
> FYI: I have committed this patch with a little sorting to icedtea6 head.
> http://icedtea.classpath.org/hg/icedtea6/rev/6c809c02b993
> 

You posted this as RFC.  Why did you not then wait for a review?

This seems like the wrong approach.  Surely it should default to ignoring an
unrecognised option instead of having this whitelist of HotSpot options.
I'm pretty sure CACAO does that rather than this mess.

> It will allow the builders to better spot if we introduce any regressions while fixing the remaining list of failures.
> 
> >
> > Now whem running make check-hotspot the list of remaining FAILED and Error tests have been reduced to a handfull.
> >
> > make check-hotspot
> >
> > FAILED: compiler/6603011/Test.java
> >     Unexpected exit from test [exit code: 136]
> >     * Needs investigation!
> >
> > FAILED: compiler/6800154/Test6800154.java
> >     Unexpected exit from test [exit code: 136]
> >     * Needs investigation!
> >
> > Error:  compiler/6866651/Test.java
> >     elapsed time (seconds): 120.004 (timed out)
> >
> > Error:  compiler/6901572/Test.java
> >     elapsed time (seconds): 120.001 (timed out)
> >
> > FAILED: gc/6581734/Test6581734.java
> >     Exception: No matching memory pools found: test with -XX:+UseConcMarkSweepGC
> >     * Test are unsuitable for JamVM
> >     it should be ignored if possible when using JamVM.
> >
> > FAILED: runtime/6626217/Test6626217.sh
> >     Execution failed: exit code 1
> >     * Needs investigation!
> >
> > Error:  runtime/6925573/SortMethodsTest.java
> >     elapsed time (seconds): 120.001 (timed out)   
> >
> > FAILED: runtime/6929067/Test6929067.sh
> >     Execution failed: exit code 127
> >     * Needs investigation!
> >
> > Test results: passed: 91; failed: 5; error: 3
> 
> Cheers
> Xerxes
> 

-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

Support Free Java!
Contribute to GNU Classpath and IcedTea
http://www.gnu.org/software/classpath
http://icedtea.classpath.org
PGP Key: F5862A37 (https://keys.indymedia.org/)
Fingerprint = EA30 D855 D50F 90CD F54D  0698 0713 C3ED F586 2A37