This is the mail archive of the mailing list for the Java project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Mauve testresult status


This is an overview of the current 3.3 branch Mauve issues.

There are currently 25 unexpected Mauve failures.
- 5 Likely Mauve testsuite bugs.
- 2 Unsure about the status.
- 1 ClassLoader issue (might be simple/might be hard)
- 2 Class reflection issues (unsure how hard to solve)
- 2 Double/Float short byte value issues (pass with Kissme)
- 1 Double/Float to int conversion problem on powerpc.
- 3 Serialization issues (A bit of work but solutions do exist)
- 7 BigInteger failures, but patch exists.
- 2 DatagramSocket exception failures that seem easy to fix.

See below for the "raw data". But we are not doing that bad and I think
that by the end of the month these will all be fixed (and otherwise I
will set them to XFAIL). Two of the serialization issues seem
regressions so I will start with those.




The Raw Data:

The following failures seem to be bugs in the Mauve testsuite.
I will check the spec carfully and update Mauve.
FAIL: skip() threw StringIndexOutOfBoundsException (number 1)
FAIL: single-byte writes (number 3)
FAIL: ready() (number 1)
FAIL: getEncoding (number 1)
FAIL: available() (number 1)

The following might or might not be bugs in the testsuite (Classpath
actually has a bug report for this issue):
FAIL: skipBytes() stopped before EOF (number 1)
FAIL: skipBytes() skipped past EOF (number 1)

Classloader issue. Might get fixed by Tom his Classloader tweaks:
FAIL: test_getClassloader (number 1)

Reflection issues that I am not sure about how difficult they might be
to solve, but Kissme+Classpath passes these:
FAIL: getClasses (number 2)
FAIL: uncaught exception at "getClasses" number 3: java.lang.ArrayIndexOutOfBoundsException: 0

Double/Float issues (doesn't fail with Kissme+Classpath so in theory
could have an easy fix):
FAIL: Error: test_shortbyteValue failed - 5 (number 1)
FAIL: Error: test_shortbyteValue failed - 5 (number 1)

Double issue that only occurs on powerpc:
FAIL: Error: test_longValue returned wrong results CYGNUS: Float to int conversions - 1 (number 1)

The following are Serialization issues. The first one passes with
IK.VM.NET+Classpath, the last two pass with Kissme+Classpath but the
ObjectStream classes have diverged a bit so it is a bit of work to get
this right.
FAIL: (number 2)
FAIL: getSerialVersionUID (number 2)
FAIL: getSerialVersionUID (number 6)

There are 7 failures which are variants of:
FAIL: java.lang.ArithmeticException: not invertible (number 1)
These are all fixed by the patch submitted by Raif. So this patch or the
rewrite he is now doing with the suggestions of Per Bothner should go

The following two seem to expect a different exception then the one we
currently throw. Should be easy fix:
FAIL: Expected IOException (number 1)
FAIL: uncaught exception at "invalid_send_data" number 4: java.lang.IllegalArgumentException: Invalid length: -1

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]