This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug bootstrap/12025] combined build of i960 coff fails
- From: "wilson at tuliptree dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 22 Aug 2003 00:00:24 -0000
- Subject: [Bug bootstrap/12025] combined build of i960 coff fails
- References: <20030821220644.12025.dhazeghi@yahoo.com>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12025
------- Additional Comments From wilson at tuliptree dot org 2003-08-22 00:00 -------
Subject: Re: New: combined build of i960 coff fails
dhazeghi at yahoo dot com wrote:
> Summary: combined build of i960 coff fails
The i960 port is effectively dead. It has been unmaintained for several
years. I keep trying to kill it, but people who aren't willing to
maintain it themselves keep claiming that they need it, so it continues
on, half-dead and half-alive.
> /tmp/mainline/objdir/gcc/crt0.o(.text+0x4c): In function `_start':
> /tmp/mainline/objdir/i960-coff/libgloss/i960/../../../../combined/libgloss/i960/crt0.c:61:
> undefined reference to `_brk
>...
I don't really care, but I'm guessing that the problem is LIB_SPEC
doesn't include -lc, on the assumption that it will be provided by the
linker script. Some embedded ports will only build if you specify the
linker script. Maybe putting -Wl,-Tmon960.ld in CFLAGS_FOR_TARGET in
the toplevel Makefile would work? Or maybe we should redesign how this
stuff works.
Note that gcc/config/libgloss.h sets LIB_SPEC to a null string.