More flow.c problems

[Toshiyasu Morita]

version: CVS current 3/1/98
   host: i386-linux
 target: sh-hms
   file: ftp://shell14.ba.best.com/pub.t/tm2/stress-1.8.tar.gz (g_phys.i)

I seem to have triggered another flow problem; cc1 aborts at flow.c:941 
when compiling g_phys.i:

Full invocation:

sh-hms-gcc -O2 -m4-single-only -ml -S -g g_phys.i

Result:

abort at flow.c:941

#0  make_edges (label_value_list=0x0, bb_eh_end=0xbffff410) at 
../../egcs/gcc/flow.c:941
#1  0x80d65d7 in find_basic_blocks (f=0x8278238, nregs=88, file=0x0) at 
../../egcs/gcc/flow.c:397
#2  0x80ada76 in gcse_main (f=0x8278238, file=0x0) at 
../../egcs/gcc/gcse.c:673
#3  0x804b89f in rest_of_compilation (decl=0x825c718) at 
../../egcs/gcc/toplev.c:3789
#4  0x815fd51 in finish_function (nested=0) at ../../egcs/gcc/c-decl.c:7272
#5  0x815248b in yyparse () at c-parse.y:349
#6  0x804a995 in compile_file (name=0xbffffb93 "g_phys.i") at 
../../egcs/gcc/toplev.c:2981
#7  0x804d962 in main (argc=6, argv=0xbffffa84) at 
../../egcs/gcc/toplev.c:5303
#8  0x80490fb in _start ()

Toshi