Date: Sun, 11 Mar 2001 22:53:21 -0600 From: Larry Rosenman <ler@lerctr.org> To: Kris Kennaway <kris@obsecurity.org> Cc: stable@FreeBSD.ORG, qa@FreeBSD.ORG Subject: Re: cputype=486 Message-ID: <20010311225320.A15729@lerami.lerctr.org> In-Reply-To: <20010311204000.A9671@lerami.lerctr.org>; from ler@lerctr.org on Sun, Mar 11, 2001 at 08:40:00PM -0600 References: <20010311120241.A15202@lerami.lerctr.org> <20010311145945.C66872@mollari.cthul.hu> <20010311193126.A6166@lerami.lerctr.org> <20010311175148.A82406@mollari.cthul.hu> <20010311195843.A7539@lerami.lerctr.org> <20010311201658.A8487@lerami.lerctr.org> <20010311204000.A9671@lerami.lerctr.org>
next in thread | previous in thread | raw e-mail | index | archive | help
* Larry Rosenman <ler@lerctr.org> [010311 20:40]: > * Larry Rosenman <ler@lerctr.org> [010311 20:16]: > > * Larry Rosenman <ler@lerctr.org> [010311 20:00]: > > > * Kris Kennaway <kris@obsecurity.org> [010311 19:52]: > > > > On Sun, Mar 11, 2001 at 07:31:26PM -0600, Larry Rosenman wrote: > > > > > * Kris Kennaway <kris@obsecurity.org> [010311 17:00]: > > > > > > On Sun, Mar 11, 2001 at 12:02:41PM -0600, Larry Rosenman wrote: > > > > > > > If I make buildworld with CPUTYPE=3Di486 (on my P-III), and t= hen=20 > > > > > > > make installworld on the 486 target, we get a signal 4 in > > > > > > > the first install of code, in strip.=20 > > > > > >=20 > > > > > > Odd. That suggests that gcc is outputting code which can't run= on the > > > > > > i486 because it uses an illegal instruction. > > > > > >=20 > > > > > > #define SIGILL 4 /* illegal instr. (not reset wh= en caught) */ > > > > > >=20 > > > > > > Does anyone else have 486 build reports with CPUTYPE, positive = or negative? > > > > > > > > > > even without CPUTYPE I'm getting the same thing.=20 > > > >=20 > > > > Well, that says it's not CPUTYPE at fault..that's good. Can you run > > > > strip outside of installworld? Perhaps it was already replaced with > > > > e.g. a pentium-optimized version by something else. > > > It works fine, as does the install from the /tmp/install.$$ directory. > > >=20 > > > SO, where to now?=20 > >=20 > > OK, the ****NEWLY BUILT**** strip does *NOT* work on the 486.... > >=20 > >=20 > > Script started on Sun Mar 11 20:10:54 2001 > > $ pwd=0D > > /usr/obj/usr/src/i386/usr/libexec/elf > > $ ls -l strui=08 =08=08 =08ip=0D > > -rwxr-xr-x 1 root wheel 430232 Mar 11 12:14 strip > > $ ./strip=0D > > Illegal instruction=20 > > $ ^D=0D > >=20 > > Script done on Sun Mar 11 20:11:09 2001 > >=20 > > So, it appears CURRENT sources don't make a valid i486 executable. > >=20 > > attached is the make buildworld output. >=20 > Here is a quick gdb... >=20 > Script started on Sun Mar 11 20:38:03 2001 > fw# gdb -c strip.core /usr/obj/usr/src/i386/usr/bin/strip=0D > GNU gdb 4.18 > Copyright 1998 Free Software Foundation, Inc. > GDB is free software, covered by the GNU General Public License, and you = are > welcome to change it and/or distribute copies of it under certain conditi= ons. > Type "show copying" to see the conditions. > There is absolutely no warranty for GDB. Type "show warranty" for detail= s. > This GDB was configured as "i386-unknown-freebsd"... > (no debugging symbols found)... > Core was generated by `strip'. > Program terminated with signal 4, Illegal instruction. > #0 0x804ef9b in __swhatbuf () > (gdb) where > #0 0x804ef9b in __swhatbuf () > #1 0x804eea2 in __smakebuf () > #2 0x80497a3 in __srefill () > #3 0x8049401 in fgets () > #4 0x804862c in getobjformat () > #5 0x80481f3 in main () > #6 0x8048135 in _start () > (gdb) fw# ^D=08=08exit >=20 > Script done on Sun Mar 11 20:38:49 2001 >=20 What else can I supply? Access is also possible to both the build box and the 486.=20 LER To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-qa" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010311225320.A15729>