other/8346: libiberty test-demanagle fails on hppa64-hp-hpux11*

[dave dot anglin at nrc dot ca]

>Number:         8346
>Category:       other
>Synopsis:       libiberty test-demanagle fails on hppa64-hp-hpux11*
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Oct 24 22:16:01 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     Dave Anglin
>Release:        gcc 3.3 (experimental)
>Organization:
>Environment:
hppa64-hp-hpux11*
>Description:
cc +DA2.0W -DHAVE_CONFIG_H  -I.. -I../../../src/libiberty/testsuite/../../include  -o test-demangle \
        ../../../src/libiberty/testsuite/test-demangle.c ../libiberty.a
./test-demangle < ../../../src/libiberty/testsuite/demangle-expected
FAIL at line 791, style gnu:
in:  __cl__C11VHDLLibraryGt4PixX3Z11VHDLLibraryZ14VHDLLibraryRepZt4List1Z10VHDLEntity
out:
exp: VHDLLibrary::operator()(PixX<VHDLLibrary, VHDLLibraryRep, List<VHDLEntity> >) const
./test-demangle: 649 tests, 1 failures

The test fails irrespective of whether or not libiberty is
built with gcc or the HP compiler.  Thus, it would appear
to be a code problem in cplus-dem.c

The test did not fail prior to this patch:
2002-09-11  Zack Weinberg  <zack@codesourcery.com>

        * cplus-dem.c: Code under #ifdef MAIN moved to gcc/cp/cxxfilt.c.
        * testsuite/Makefile.in: Adjust for test-demangle.
        * testsuite/regress-demangle: Deleted.
        * testsuite/test-demangle.c: New file.
        * testsuite/demangle-expected: Change \$ to $ throughout, now that
        this file is not being read by a shell script.

I have tracked the failure to somewhere in demangle_args.

I suspect some kind of memory management issue as the specific
test does not fail if demangle-expected is edited to just
contain the failing line.

The test does not appear to fail in 32-bit environments.
>How-To-Repeat:

>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: