[Top] [All Lists]

Re: Segfaults while linking

Subject: Re: Segfaults while linking
From: Richard van den Berg <>
Date: Sun, 20 Jun 1999 23:57:01 +0200 (MET DST)
In-reply-to: <>
Hello Karsten,

> I have a strange problem on my Maxine and would like to ask if somebody
> could try to reproduce this. When compiling the inn-2.2-package from RH60,
> I always get the same error:

That fast I haven't reproduced the errors you encounter, but I have one
to add on MAXine building glibc-2.0.7-20:

make[1]: Leaving directory `/usr/src/redhat/BUILD/glibc-2.0.7/wcsmbs'
make  -C time subdir_lib
make: *** [time/subdir_lib] Segmentation fault (core dumped)

and on 5000/133: 
make[1]: Leaving directory `/usr/src/redhat/BUILD/glibc-2.0.7/ctype'
make  -C db others
make[1]: Entering directory `/usr/src/redhat/BUILD/glibc-2.0.7/db'
gcc -g  -shared -o  -Wl,-dynamic-linker=/lib/ -B../csu/ 
-Wl,   -L.. -L../elf -L../nss -L../nis -L. -L../resolv 
-Wl,-rpath-link=..:../elf:../nss:../nis:.:../resolv -Wl,--whole-archive 
libdb_pic.a ../ ../ -W
collect2: ld terminated with signal 6 [Aborted], core dumped
make[1]: *** [] Error 1
make[1]: Leaving directory `/usr/src/redhat/BUILD/glibc-2.0.7/db'
make: *** [db/others] Error 2

> The system is based on the declinuxroot-990128.tgz; egcs-1.0.2-9,
> glibc-devel-2.0.7-20 and perl-5.004-6 are from and
> make-3.77-6, gdbm-1.7.3-19 and python-1.5.1-4 are RH60 packages which I
> have compiled myself (they are on The python-package
> required some small changes before compiling on a glibc20 system, but that
> should not cause the linker to segfault, in particular as python is not
> used while building inn. I have tried replacing perl-5.004-6 by
> perl-5.00503 from RH6 but that did not change the behaviour.

Which binutils? I'm using 2.8.1 with R3k patch.

> Any ideas?

I would like to now how to start a kernel without compiled in serial
support, is it possible that a kernel boots without an initial console? I
build such a kernel, but it didn't completly started, I guess it paniced
not having a console. As Vladimir pointed out the instability on R3K
DEC's is outside common R3k code, with that in mind I'm curious what the
MAXine does with its native console supported. That's what I'm currently
trying to do: building a kernel with the framebuffer supported.


<Prev in Thread] Current Thread [Next in Thread>