From owner-freebsd-stable Sat Mar 25 10:24:45 2000 Delivered-To: freebsd-stable@freebsd.org Received: from smtp8.teleport.com (smtp8.teleport.com [192.108.254.52]) by hub.freebsd.org (Postfix) with SMTP id 3D68437B8C5 for ; Sat, 25 Mar 2000 10:24:42 -0800 (PST) (envelope-from pl@teleport.com) Received: (qmail 397 invoked from network); 25 Mar 2000 18:24:29 -0000 Received: from user2.teleport.com (qmailr@192.108.254.12) by smtp8.teleport.com with SMTP; 25 Mar 2000 18:24:29 -0000 Received: (qmail 5406 invoked by uid 64886); 25 Mar 2000 18:24:29 -0000 Date: Sat, 25 Mar 2000 10:24:29 -0800 (PST) From: Patrick Lewis To: freebsd-stable@freebsd.org Subject: buildworld fails, hardware? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I have been trying off an on for the last couple of weeks to rebuild my system via make world & friends (I have never done this before). Currently, I have 3.4-RELEASE, and am trying to move to STABLE (RELENG_3). cvsup appears to work as expected. Basic system info: AMD k6-3, 128MB Ram, SCSI disk system. Other than this issue, the machine has worked flawlessly, albeit under light loads. Whenever I do a make buildworld, I always fail. For instance: cc -O -pipe -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/config -DFREEBSD_NATIVE -DDEFAULT_TARGET_VERSION=\"2.7.2.3\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -I/usr/obj/usr/src/tmp/usr/include -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/expr.c -o expr.o {standard input}: Assembler messages: {standard input}:14114: Error: no such 386 instruction: `mowl' *** Error code 1 Stop. ... and: cc -O -pipe -D_GNU_SOURCE -I- -I. -I/usr/src/gnu/usr.bin/binutils/ld/i386 -I/usr/src/gnu/usr.bin/binutils/ld -I/usr/src/gnu/usr.bin/binutils/ld/../libbfd/i386 -I/usr/src/gnu/usr.bin/binutils/ld/../../../../contrib/binutils/include -DDEFAULT_EMULATION=\"elf_i386\" -DTARGET=\"i386-unknown-freebsdelf\" -DSCRIPTDIR=\"/usr/obj/usr/src/tmp/usr/libdata\" -I/usr/src/gnu/usr.bin/binutils/ld/../../../../contrib/binutils/ld -I/usr/obj/usr/src/tmp/usr/include -c ldlex.c {standard input}: Assembler messages: {standard input}:4428: Error: Rest of line ignored. First ignored character valued 0xb. *** Error code 1 Stop. ... and: cc -O -pipe -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/config -DFREEBSD_NATIVE -DDEFAULT_TARGET_VERSION=\"2.7.2.3\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -I/usr/obj/usr/src/tmp/usr/include -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/combine.c -o combine.o {standard input}: Assembler messages: {standard input}:19565: Error: invalid character (0xb) in first operand *** Error code 1 Stop. To me, funky fails at differing times when compiling identical source points to a hardware problem, although it doesn't look like memory problems that are described in the Handbook (Signal 11). In the mailing list archives, I saw similar assembler type errors described with K6 chips (a K6 bug), but it looks like these were resolved by AMD a long time ago. I guess my questions are: 1) Is this a hardware issue? Could it be anything else? 2) If it is a hardware issue, any ideas what it might be? Can I narrow the options down further without buying new hardware? 3) Am I forgetting something really stupid? Thanks in advance. Patrick Lewis To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message