Date: Tue, 9 Oct 2001 09:19:13 -0700 (PDT) From: David Wolfskill <david@catwhisker.org> To: freebsd-questions@freebsd.org Subject: Diagnosing an inconsistency? Message-ID: <200110091619.f99GJD205433@bunrab.catwhisker.org>
next in thread | raw e-mail | index | archive | help
[This is pretty long. It's a result of an apparent transient failure
during a buildworld of today's -STABLE (running yesterday's -STABLE), so
if this isn't of interest, please pass on by. Thanks....]
On my "build machine" this morning, I encountered an error during the
(daily) -STABLE buildworld. Since another machine (my laptop), which
was using very similar sources, got past that without a hiccup, I re-
started the buildworld on the build machine... and it seems to have
completed just fine... even though I didn't change anything.
So I'm wondering if anyone has suggestions for how to diagnose what might
have happened (and how to avoid it; I prefer that the machines behave in
a predictable way).
The only hint from /var/log/messages is:
Oct 9 04:38:32 freebeast /kernel: pid 79987 (cc1), uid 0: exited on signal 11 (core dumped)
The salient messages shown in the typescript of the buildworld are:
cc -pg -O -pipe -DLIBC_RCS -DSYSLIBC_RCS -I/usr/src/lib/libc_r/../libc/include -DPTHREAD_KERNEL -D_THREAD_SAFE -I/usr/src/lib/libc_r/uthread -I/usr/src/lib/libc_r/../../include -D_LOCK_DEBUG -D_PTHREADS_INVARIANTS -D__DBINTERFACE_PRIVATE -DINET6 -DPOSIX_MISTAKE -I/usr/src/lib/libc_r/../libc/locale -DBROKEN_DES -DYP -I/usr/obj/usr/src/i386/usr/include -c /usr/src/lib/libc_r/../libc/regex/regfree. c -o regfree.po
cc -fpic -DPIC -O -pipe -DLIBC_RCS -DSYSLIBC_RCS -I/usr/src/lib/libc_r/../libc/include -DPTHREAD_KERNEL -D_THREAD_SAFE -I/usr/src/lib/libc_r/uthread -I/usr/src/lib/libc_r/../../include -D_LOCK_DEBUG -D_PTHREADS_INVARIANTS -D__DBINTERFACE_PRIVATE -DINET6 -DPOSIX_MISTAKE -I/usr/src/lib/libc_r/../libc/locale -DBROKEN_DES-DYP -I/usr/obj/usr/src/i386/usr/include -c /usr/src/lib/libc_r/../libc/regex/regfree.c -o regfree.So
cc: {standard input}: Assembler messages:
{standard input}:2010: Warning: Partial line at end of file ignored
Internal compiler error: program cc1 got fatal signal 11
*** Error code 1
cc -O -pipe -DLIBC_RCS -DSYSLIBC_RCS -I/usr/src/lib/libc_r/../libc/include -DPTHREAD_KERNEL -D_THREAD_SAFE -I/usr/src/lib/libc_r/uthread -I/usr/src/lib/libc_r/../../include -D_LOCK_DEBUG -D_PTHREADS_INVARIANTS -D__DBINTERFACE_PRIVATE -DINET6 -DPOSIX_MISTAKE -I/usr/src/lib/libc_r/../libc/locale -DBROKEN_DES -DYP -I/usr/obj/usr/src/i386/usr/include -c /usr/src/lib/libc_r/../libc/stdio/_flock_stub. c -o _flock_stub.o
1 error
*** Error code 2
1 error
*** Error code 2
1 error
*** Error code 2
1 error
*** Error code 2
1 error
The command sequence I have been using for the last few months (tracking
-STABLE daily) is:
whoami && mount && cd /usr/src && uname -a && date && \
make -j8 buildworld && date && make kernel KERNCONF=FREEBEAST && \
date && make installworld && date && mergemaster -u 0022 -i && date && df -k
The non-comment lines from /etc/make.conf are:
CFLAGS= -O -pipe
COPTFLAGS= -O -pipe
COMPAT22= yes
COMPAT3X= yes
COMPAT4X= yes
PRINTERDEVICE= ps
FORCE_PKG_REGISTER= YES
WITH_PNG_MMX=YES
The machine is one I built fairly recently (mid-August): dual 866 MHz
PIIIs w/ 256 MB RAM. Complete dmesg is certainly available; this
message seemd long enough without it, for now.
Additional background that may be of interest: This is the box that I
use for mirroring the FreeBSD CVS repository; I use CVSup to do this
daily, starting at 0347 hrs. local (Pacific) time, using cvsup14 as the
CVSup server.
I use slices 1 & 2 for -STABLE, and slices 3 & 4 for -CURRENT:
freebeast[1] uname -a
FreeBSD freebeast.catwhisker.org 4.4-STABLE FreeBSD 4.4-STABLE #46: Tue Oct 9 06:04:00 PDT 2001 root@freebeast.catwhisker.org:/common/S1/obj/usr/src/sys/FREEBEAST i386
freebeast[2] df -k
Filesystem 1K-blocks Used Avail Capacity Mounted on
/dev/ad0s1a 158767 37131 108935 25% /
mfs:31 507639 1 467027 0% /tmp
/dev/ad0s1e 1873082 528816 1194420 31% /usr
/dev/ad0s2a 158767 36987 109079 25% /S2
/dev/ad0s2e 1873113 521839 1201425 30% /S2/usr
/dev/ad0s3a 158783 74429 71652 51% /S3
/dev/ad0s3e 1873153 602591 1120710 35% /S3/usr
/dev/ad0s4a 158783 74429 71652 51% /S4
/dev/ad0s4e 1872799 599227 1123749 35% /S4/usr
/dev/ad0s4h 27715371 7585046 17913096 30% /common
/dev/ad0s4g 1016303 23565 911434 3% /var
procfs 4 4 0 100% /proc
freebeast[3]
freebeast[1] uname -a
FreeBSD freebeast.catwhisker.org 5.0-CURRENT FreeBSD 5.0-CURRENT #55: Tue Oct 9 08:24:10 PDT 2001 root@freebeast.catwhisker.org:/common/S3/obj/usr/src/sys/FREEBEAST i386
freebeast[2] df -k
Filesystem 1K-blocks Used Avail Capacity Mounted on
/dev/ad0s3a 158783 74429 71652 51% /
devfs 1 1 0 100% /dev
/dev/ad0s1a 158767 37131 108935 25% /S1
/dev/ad0s1e 1873082 528816 1194420 31% /S1/usr
/dev/ad0s2a 158767 36987 109079 25% /S2
/dev/ad0s2e 1873113 521839 1201425 30% /S2/usr
/dev/ad0s3e 1873153 602591 1120710 35% /usr
/dev/ad0s4a 158783 74429 71652 51% /S4
/dev/ad0s4e 1872799 599227 1123749 35% /S4/usr
/dev/ad0s4h 27715371 7585045 17913097 30% /common
/dev/ad0s4g 1016303 23566 911433 3% /var
procfs 4 4 0 100% /proc
/dev/md10c 520140 4 478528 0% /tmp
freebeast[3]
The CVS repository is in /common, as are such directories as /usr/local
and my home directory. The /var filesystem is also common to each OS.
The machine also acts as a CVSup server for other machines at home,
especially my laptop. Just prior to the incident in question, I had:
* Re-synced freebeast's CVS repository with that of cvsup14.
* (running -STABLE, booted from slice 1) cd /usr/src && cvs update.
* Checked for CVS merge conflicts & other oddities; found none.
* Issued the above-cited "make buildworld" sequence, while running under
"script". (Disclosure: I actually have an alias for all that.)
* At 0407 hrs., the laptop synchronized its copy of the FreeBSD CVS
repository to that on freebeast.
* [Laptop also running -STABLE, booted from its slice 1) cd /usr/src &&
cvs update.]
* [Checked laptop's CVS log for merge conflicts & oddities; found none.]
* [Issued a slight modification of the "make buildworld" sequence on the
laptop -- the laptop only has a single CPU, so I use -j4; it also has
a differently-named kernel.]
...
* freebeast's cc1 fell over & died trying to compile libc/regex/regfree.c.
...
* [Laptop chugged right through...]
* I re-issued the same "make buildworld" command sequence (well, I
re-issued the alias). Salient point: At no time did I do anything
remotely similar to (say) "-DNOCLEAN" or otherwise specify odd options.
* Each one completed just fine....
* I became even more puzzled, so I posted this....
Please copy me on replies; I'm certainly willing to summarize.
Thanks,
david
--
David H. Wolfskill david@catwhisker.org
As a computing professional, I believe it would be unethical for me to
advise, recommend, or support the use (save possibly for personal
amusement) of any product that is or depends on any Microsoft product.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200110091619.f99GJD205433>