Try the test case "abstract.fail.abstract2" from the mauve
"verify" module. (You will need jasmin to build this.)
This test case gets a SEGV in the unwinder:
Program received signal SIGSEGV, Segmentation fault.
uw_frame_state_for (context=0xbfff9998, fs=0xbfff98d8) at linux-unwind.h:125
125 in linux-unwind.h
Continuing results in the same SEGV endlessly.
The underlying cause of this is that a _Jv_Method has 0x0 as its
'ncode'. Instead it would be more correct to put
'throw new AbstractMethodError' there. I have a partial patch
to do this (the compiler should also do this...).
However I think the unwinder crashing is a bug regardless of whether
this other problem is fixed.
--
Summary: hang when invoking abstract method
Product: gcc
Version: 4.1.0
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: libgcj
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: tromey at gcc dot gnu dot org
CC: gcc-bugs at gcc dot gnu dot org,java-prs at gcc dot gnu
dot org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21906
