🚀 go-pugleaf

I have a valid C++ program, well at least CFRONT 2.1 thinks it's
valid, that causes cc1plus in version 2.2.2 and the snapshot dated
920925 to seg fault (Sun OS 4.1.1).  The problem is virtuals is equal
to NULL in get_vtable_entry.  Here's the traceback for the 920925
snapshot from gdb 4.6:

GNU C++ version gcc-2.2.2-920925 (sparc) compiled by GNU C version gcc-2.2.2-920925.
Program received signal 11, Segmentation fault
0x5aaf0 in get_vtable_entry (virtuals=0x0, base_fndecl=0x490210)
    at cp-class.c:637
637         }
(gdb) 
(gdb) where
#0  0x5aaf0 in get_vtable_entry (virtuals=0x0, base_fndecl=0x490210)
    at cp-class.c:637
#1  0x5afbc in modify_vtable_entries (t=0x285f80, fndecl=0x6caf88, 
    base_fndecl=0x490210, pfn=0x6cb640) at cp-class.c:891
#2  0x5ec10 in finish_struct (t=0x285f80, list_of_fieldlists=0x285f80, 
    empty=1545216, warn_anon=0) at cp-class.c:2968
#3  0x10220 in yyparse () at cp-parse.y:2041
#4  0x7f4e8 in compile_file (name=0xf7fff686 "SpnObidManager.i")
    at toplev.c:1774
#5  0x82090 in main (argc=9, argv=0xf7fff56c, envp=0xf7fff594) at toplev.c:3256
(gdb) 

Jody Winston		jody@shell.com
..!{sun,psuvax1,bcm,rice,decwrl,cs.utexas.edu}!shell!jody
Consultant to:
Shell Development Company, Bellaire Research Center
P.O. Box 481, Room 4205, Houston, TX 77001	(Voice: 713 245-7574)