Date: Sat, 19 Jul 1997 02:01:40 +0200 From: Philipp Reichmuth <chokepnt@prima.ruhr.de> To: Howard Lew <hlew@www2.shoppersnet.com> Cc: David Greenman <dg@root.com>, Gary Kline <kline@tera.com>, root@counterintelligence.cdrom.com, stable@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Re: mw fails even more... Message-ID: <33D003E4.41C67EA6@prima.ruhr.de> References: <Pine.BSF.3.91.970717002443.17811D-100000@www2.shoppersnet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Howard Lew wrote: > > On Wed, 16 Jul 1997, David Greenman wrote: > > > > Now I'm even more at a loss re my upgrade to a K5 or K6 > > > chip. Perhaps someone on the Core team--like you, David-- > > > can give everybody advice on which CPU's do work flawlessly > > > with BSD and which have known or suspected woes. > > > > > > I'm not referring to occasional defective chips, but to > > > bad logic design. > > > > > > It may be simply a matter of this present rev of the K6 is > > > defective and when AMD does their next, they'll fix this. > > > > What's the cpu code on the top surface? Mine is the following: AMD K6-200-ALR (might be slashes instead) 2,9 V B9721DJAW > Hmmm... can you try running it at 166MHz and see if it works? If it does, > it may be a remarked cpu (it may still be operational if it hasn't been > already damaged by the heat). The only time I have ever seen a cpu die > was a gray market IBM 6x86 PR166+ (non L revision) due to heat death when > a customer did not plug in the fan and left the machine running over a > weekend. After that instance, we always use heatsink grease on all > metallic surface AMD and Cyrix cpus and have never had a cpu fail ever > since. (This is also something that AMD & Cyrix recommends anyway.) I don't use heatsink grease because I don't have any, but i'll fix it within two days. Anyway, my BIOS reports the CPU temperature at coredumptime to be fifty-eight degrees. Doesn't seem that bad, I think. > I think I have heard 2 people say their K6s failed to do the make world? > Have we ruled out the possibility that something has changed in the code > that may affect the way a K6 compiles it? (Did both K6s fail on > approximately the same day?) These are my last four make world attempts: ======================> begin mw 1 <================================= cc -O2 -m486 -pipe -DFREEBSD_AOUT -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=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/recog.c -o recog.o cc -O2 -m486 -pipe -DFREEBSD_AOUT -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=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/reg-stack.c -o reg-stack.o cc: Internal compiler error: program cc1 got fatal signal 11 *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. ======================> begin mw 2 <============================= cc -O2 -m486 -pipe -DFREEBSD_AOUT -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=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/print-tree.c -o print-tree.o cc -O2 -m486 -pipe -DFREEBSD_AOUT -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=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/real.c -o real.o cc: Internal compiler error: program cc1 got fatal signal 6 *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. =====================> begin mw 3 <============================== cc -O2 -m486 -pipe -DFREEBSD_AOUT -I/usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc -I/usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc/config -DFREEBSD_NATIVE -DDEFAULT_TARGET_VERSION=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc1obj/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc/c-convert.c cc -O2 -m486 -pipe -DFREEBSD_AOUT -I/usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc -I/usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc/config -DFREEBSD_NATIVE -DDEFAULT_TARGET_VERSION=\"2.7.2.1\" -DDEFAULT_TARGET_MACHINE=\"i386-unknown-freebsd\" -I/usr/obj/usr/src/gnu/usr.bin/cc/cc1obj/../cc_tools -c /usr/src/gnu/usr.bin/cc/cc1obj/../../../../contrib/gcc/c-decl.c cc: Internal compiler error: program cc1 got fatal signal 11 *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. cpp: output pipe has been closed ====================> begin mw 4 <============================== mkdep -f .depend -a -I/usr/src/gnu/usr.bin/ld/rtld/.. -I/usr/src/gnu/usr.bin/ld/rtld -I/usr/src/gnu/usr.bin/ld/rtld/../i386 -DRTLD /usr/src/gnu/usr.bin/ld/rtld/rtld.c /usr/src/gnu/usr.bin/ld/rtld/malloc.c /usr/src/gnu/usr.bin/ld/rtld/../shlib.c /usr/src/gnu/usr.bin/ld/rtld/../i386/md.c /usr/src/gnu/usr.bin/ld/rtld/../support.c /usr/src/gnu/usr.bin/ld/rtld/sbrk.c cc -O2 -m486 -pipe -I/usr/src/gnu/usr.bin/ld -I/usr/src/gnu/usr.bin/ld/i386 -c /usr/src/gnu/usr.bin/ld/ld.c cc: Internal compiler error: program cc1 got fatal signal 11 *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. ==================> end giant makeworld section <====================== Looks like the ONLY affected program is cc1. Since I usually run the machine for half an eternity and the CPU temperature was at 58° at the most (according to my board), I don't think it's a thermal problem. mw 4 was done about twenty seconds after mw3 had "completed". I can't say if there's any significance in the program getting less "far" in mw4 than in mw3. Usually, it never goes "tilt" as soon as in mw4. I NEVER had any problems with ANY other OS (including DOS & Windows 95, I'm ashamed to say) or program. My kernel is compiled with "options I586_CPU", and the _ONLY_ "special feature" is the faster 5x86 exception handler (it's a relic from my AMD 5x86 times). BTW, I don't think I'll put any more mw outputs into the mailing lists, at least not to the extents of this one. If anyone should chance to be interested in my mw outputs, feel free to mail me. Philipp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?33D003E4.41C67EA6>