ARM jtreg results

[Pavel Tisnovsky]

Hi Andrew,

thank you for publishing test results.

You are right that javax_swing_tree_TreePath is failing on i686 and x86-64 too. The code is already fixed in OpenJDK7
but it's quite difficult to backport it to OpenJDK6 (because the Swing code was rewritten). According to Sergey who's
working for Oracle:

Sergey Malenkov wrote:
> Hi Pavel,
>
> Now I have no plan to fix the 6423060 CR in JDK 6, but it is possible
> to find and backport needed files from the finder package.
> Do you mean Open JDK 6 or JDK 6 Update Release?
>
> Thanks,
> SAM

Btw, I see many error messages in your log file which are similar to this one:

execStatus=Error. Problem deleting file: /home/aph/iced-arm/temp2/test/jdk/JTwork/scratch/.nfs000000000118d826000000b0

It could be caused by some problem with NFS (file permissions, full fs etc.)

Cheers,
Pavel

Andrew Haley wrote:
> The preliminary test results are now at
> http://aph.fedorapeople.org/test results/IcedTea6
> 1.11pre+r628e5ee78eca
> 
> I'm going through them.  Ignoring JDI tests for the moment -- the
> support is not all there -- I'm looking at others.  I'll post the
> analysis in this thread.
> 
> javax_swing_tree_TreePath.java seems simply to run out of stack space.
> Increasing the stack size causes the StackOverflowError to go away.
> However, this test doesn't work for me on x86-64 either, eventually
> failing with
> 
> Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 0
> 	at com.sun.beans.ObjectHandler.dequeueResult(ObjectHandler.java:189)
> 	at java.beans.XMLDecoder.readObject(XMLDecoder.java:218)
> 	at AbstractTest.readObject(AbstractTest.java:147)
> 	at AbstractTest.testObject(AbstractTest.java:131)
> 	at AbstractTest.test(AbstractTest.java:101)
> 	at javax_swing_tree_TreePath.main(javax_swing_tree_TreePath.java:35)
> 
> so I'm not going to worry about this one.
> 
> It may be, however, that the default stack size on ARM is too small.
> 
> Andrew.