This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Another new egcs bootstrap failure on OSF4: ICE final.c:3555
- To: egcs-bugs at egcs dot cygnus dot com, ghazi at caip dot rutgers dot edu, rth at cygnus dot com
- Subject: Re: Another new egcs bootstrap failure on OSF4: ICE final.c:3555
- From: "Kaveh R. Ghazi" <ghazi at caip dot rutgers dot edu>
- Date: Sun, 25 Apr 1999 15:44:50 -0400 (EDT)
- Cc: egcs-patches at egcs dot cygnus dot com
> From: Richard Henderson <rth@cygnus.com>
>
> On Sat, Apr 24, 1999 at 09:23:32PM -0400, Kaveh R. Ghazi wrote:
> > > insn-emit.c: In function `gen_call_osf':
> > > insn-emit.c:4103: Internal compiler error at final.c:3555
> > > make[2]: *** [insn-emit.o] Error 1
>
> (and:DI (plus:DI (subreg:DI (reg:TI $10) 0)
> (const_int 2))
> (const_int -8))
>
> Ho hum. I should have thought to fix print_operand_address
> with that change to GO_IF_LEGITIMATE_ADDRESS. Wonder why it
> didn't show up on an ev56? It certainly should have...
> r~
>
> * alpha.h (PRINT_OPERAND_ADDRESS): Break out to ...
> * alpha.c (print_operand_address): here. Handle subregs.
Hmm, it still doesn't bootstrap. :-)
Now I get this failure in stage2:
> _divdi3
> ../../egcs-CVS19990425/gcc/libgcc2.c: In function `__divdi3':
> ../../egcs-CVS19990425/gcc/libgcc2.c:684: internal
> error--unrecognizable insn:
> (insn/i 125 124 126 (set (reg:DI 131)
> (const_int 65535 [0xffff])) -1 (nil)
> (expr_list:REG_EQUAL (const_int 65535 [0xffff])
> (nil)))
> make[2]: *** [libgcc2.a] Error 1
Regarding your ev56 system, it may help to reproduce my error
if you configure as the system I have, namely alphaev5-dec-osf4.0b.
(The ev56 will run ev5 code, right?)
--Kaveh
--
Kaveh R. Ghazi Engagement Manager / Project Services
ghazi@caip.rutgers.edu Qwest Internet Solutions