[kaffe] Mauve test results
stack at cs.utah.edu
Tue Jun 11 11:36:35 PDT 2002
> Hi all,
> I've ran the Mauve test suite (downloaded today 11-June-2002) on Kaffe and
> get the following results:
> FAIL: gnu.testlet.java.io.BufferedByteOutputStream.interrupt: single-byte
> writes (number 4)
i think this test is just plain broken, i've attached a patch that i
believe implements what they meant to do in the first place.
> FAIL: gnu.testlet.java.io.BufferedReader.boundary (number 2)
> FAIL: gnu.testlet.java.io.BufferedReader.boundary (number 3)
> FAIL: uncaught exception loading gnu.testlet.java.io.PrintStream.subclass
This is because it tries to construct a PrintStream with a null
OutputStream. Now, some jvm seem to throw an exception because of the
null and others don't...
> I'm running Kaffe release 1.0.7-rc1 on a x86/Linux 2.4.x, GCC 2.96, GNU C
> Library 2.2.2.
> Did I miss something?
which mauve is it? cvs or a tarball?
> Can anyone help me in going on further with this testing?
> Many thanks in advance.
> Guillaume AUDEON
> <mailto:audeong at thmulti.com>
> (33) (0)2 99 27 31 30
-------------- next part --------------
RCS file: /cvs/mauve/mauve/gnu/testlet/java/io/BufferedByteOutputStream/interrupt.java,v
retrieving revision 1.2
diff -u -r1.2 interrupt.java
--- interrupt.java 27 Sep 2001 15:36:17 -0000 1.2
+++ interrupt.java 11 Jun 2002 18:36:26 -0000
@@ -25,6 +25,11 @@
+ private int getCount()
+ return this.count;
public void test (TestHarness harness)
// We create an output stream that will throw an
@@ -36,7 +41,7 @@
int BUFFER = 7;
helper h = new helper (10);
- BufferedOutputStream out = new interrupt (h, BUFFER);
+ interrupt out = new interrupt (h, BUFFER);
boolean ok = false;
int i = -1;
@@ -63,7 +68,7 @@
// In theory the BufferedOutputStream should notice the
// InterruptedIOException and update its internal data structure
- harness.check (count, 4);
+ harness.check (out.getCount(), 4);
h = new helper (10);
out = new interrupt (h, BUFFER);
More information about the kaffe