libjava status on Tru64 UNIX V5.1

[Rainer Orth <ro at TechFak dot Uni-Bielefeld dot DE>]

I've noticed that Richard's patch

	http://gcc.gnu.org/ml/gcc-patches/2002-03/msg01043.html

fixed the bootstrap failure on Tru64 UNIX reported in PR target/5949, so at
least libjava builds again on Tru64 UNIX V5.1.  Testsuite results are here:

	http://gcc.gnu.org/ml/gcc-testresults/2002-03/msg00478.html

The libjava part doesn't look too good compared to the 3.0.5 results:

	http://gcc.gnu.org/ml/gcc-testresults/2002-03/msg00291.html

This seems to be due to the same cause as the exception handling failures
reported in PR target/5966.  E.g. Array_1 execution fails with SIGSEGV with
the following stack trace:

#0  0x0 in ?? ()
#1  0x3000038a9d0 in _ZN3gnu3gcj7convert14UnicodeToBytes17getDefaultEncoderEv
    ()
    at libjava/gnu/gcj/convert/UnicodeToBytes.java:52
#2  0x3000038a9d0 in _ZN3gnu3gcj7convert14UnicodeToBytes17getDefaultEncoderEv
    ()
    at libjava/gnu/gcj/convert/UnicodeToBytes.java:52
Error accessing memory address 0x40: Invalid argument.

I'm just running a new bootstrap on alpha-dec-osf4.0f where those EH
handling failures don't happen and libjava testsuite results should be as
good as in 3.0.5.

	Rainer