libgcj/7578: gcc-3.1.1 libjava regression: tons of new FAILS on check

[nirkko@dkf6.unibe.ch]

>Number:         7578
>Category:       libgcj
>Synopsis:       gcc-3.1.1 libjava regression: tons of new FAILS on check
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Aug 12 08:06:01 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     nirkko@dkf6.unibe.ch
>Release:        gcc-3.1.1
>Organization:
>Environment:
Sun SPARC Solaris 2.6 (and Linux, and others...)
>Description:
after gcc bootstrap:
whereas in gcc-3.1, gmake -k check results
for libjava were in the 'usual' range, now
in gcc-3.1.1 there are tons of unexpected FAILS,
which seems a regression to me. Looking at your
"successful builds" webpage, this seems so also
on other platforms (for example, Linux-i86).
Do tests FAIL because of assembler warnings,
and newer assemblers (gas-2.13) issue more of them?
or was there a major change in the testsuite or
in libjava code?
>How-To-Repeat:
take gcc-3.1 and gcc-3.1.1
'gmake bootstrap'  and  'gmake -k check'  each of them.
compare results.
>Fix:

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