Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [aspectj-dev] AspectJ 1.1rc1 Package-and-Test dry-run report

It does look good!  But ajcTestsFailing should pass for a final candidate.
(Replies inline below.)


Jim.Hugunin@xxxxxxxx wrote:
> 
> Yay!  This looks really good.  I hope that you and Wes can work out a good way of merging this detailed script into the build documentation.

I'll reply to Adrian's email on point.

> I have a few small comments.
> 
> > * ran harness -release -logFail ajcTests.xml , saved results to
> > ajcTests-result-1.1rc1.txt
> >       - **2 failures**
> >       (1) source locations within expressions  [anticipated, as discussed
> > on aspectj-dev]
> 
> I'm going to defer tracking this down until after rc1 when we have some "leisure" time.  Hopefully Erik and Wes will be able to chime in about the configurations of their machines and the behavior of this test that they observe.

I get the same failures, arguably bugs in the tests.

> >       (2) assert and pertarget crashes compiler - this test passes on a
> > 1.4
> > JDK, but fails under 1.3.1
> >       as used for our testing (see Environment section)
> 
> This test requires a 1.4 vm to run on because it uses assertions.  I've modified the entry in the test suite to specify 'vm="1.4"' which I believe will fix this by not running the compiled code unless a 1.4 VM is available.  I haven't verified that the fix works.
> 
> > * ran harness -release -logFail ajcTestsFailing.xl, saved results to
> > ajcTestsFailing-result-1.1rc1.txt
> >       - all 5 fail (expected?? - but deviation from release notes in
> > readme-build-and-test-aspectj.html
> 
> I see where this is mentioned in the build instructions, but I don't understand it.  Tests in ajcTestsFailing are expected to fail, and I don't see any reason to run them as part of the (already long enough) release process.  Hopefully Wes can help us understand this part of the instructions?

ajcTestsFailing.xml contains tests for bugs that the compiler has never passed,
so these are the bugs to fix or invalidate before a final release. (The bad 
tests to fix are in ajcTestsBroken.xml.)

> Even though your testing didn't find any bugs for me, I have two fairly high-priority bug reports that were submitted last night that I'll try to fix today for tommorrow's final build.

also fyi, I crashed the compiler a few times in incremental mode while 
resolving types, etc. I'll try to verify, reproduce and submit after 
last night's fixes.

Wes


Back to the top