java.lang/258

[Bryce McKinlay]

The following reply was made to PR java.lang/258; it has been noted by GNATS.

From: Bryce McKinlay <bryce@albatross.co.nz>
To: Warren Levy <warrenl@cygnus.com>
Cc: java-gnats@sourceware.cygnus.com, tromey@cygnus.com
Subject: Re: java.lang/258
Date: Tue, 20 Jun 2000 18:00:45 +1200

 Warren Levy wrote:
 
 > Since you closed java.lang/258, are you getting successful passes for all
 > the test cases of isAssignableFrom for the
 > libjava/testsuite/libjava.lang/Class_1.java test?
 
 Sure:
 
 [bryce@reason libjava.lang]$ javac Class_1.java
 [bryce@reason libjava.lang]$ java Class_1 > class1.out.jdk
 [bryce@reason libjava.lang]$ gcj -O2 Class_1.java -o c1 --main=Class_1
 [bryce@reason libjava.lang]$ ./c1 > class1.out.gcj
 [bryce@reason libjava.lang]$ diff -u class1.out.gcj class1.out.jdk
 [bryce@reason libjava.lang]$
 
 
 > I applied the patch and no longer get the run-time exception from before,
 > but the 3 failure test cases are still there.  Here are my results:
 >
 > --- Actual     Mon Jun 19 22:20:12 2000
 > +++ Expected   Mon Jun 19 22:20:47 2000
 > @@ -1,17 +1,17 @@
 >  Testing class `Class_1'...
 >  false
 >  true
 > -false
 > +true
 >  false
 >  true
 >  true
 >  false
 > -false
 > +true
 >  printIsAssignableFrom(C[].class, C[][].class, false); case uncompilable due to known compiler limitation.
 >  true
 >  true
 >  false
 > -false
 > +true
 >  false
 >  false
 >  true
 >
 > If you're getting no deviations from the expected values, I'll look
 > further at my build to figure out what I'm missing.
 
 These failures are different from the ones I was seeing before writing the patch. You're failing on some very
 trivial cases like "C.class, D.class" and "J.class, J.class", which seems pretty odd!
 
 I'm using gcc version 2.96 20000607 (experimental), is it possible you have a bad compiler?
 
 regards
 
   [ bryce ]