Jigsaw test failures

[Sean Mullan]

On 1/5/11 1:03 PM, Mandy Chung wrote:
> It's a test bug that compares an original class file with a "pack-then-unpack'
> class file that is not the same as the original file as it's modified during the
> pack200 compression process.
>
> Before the fix in pack200 for (in JDK 7 b100):
> 6712743: pack200: should default to 150.7 pack format for classfiles without any
> classes.
> Changeset: http://hg.openjdk.java.net/jdk7/tl/jdk/rev/43dfa39686a1
>
> pack200 doesn't recognize class files of version 51.0 and thus doesn't do the
> compression and thus the class files match even after pack200. I missed this
> test failure when updating jigsaw repos with the recent jdk 7 builds since these
> tests failed with a different error before the merge:
> keytool error: java.security.KeyStoreException: JKS not found
>
> Here is the webrev for the test fix:
> http://cr.openjdk.java.net/~mchung/jigsaw/webrevs/test-fix.webrev.00/
>
> There are pack200 regression tests that verify if pack200 generates correct
> class files. In addition, using the jdk-module-image to build and run tests also
> verify the class files installed by jmod. I think byte-by-byte class file
> comparison isn't necessary in these tests. So ModuleFormatTest01.java is
> modified just to verify the ClassFile header (the first 10 bytes).
>
> Sean,
> Can you review the webrev?

Fix looks good.

--Sean