Re: DP2 (I think!) crash booting from floppies

2002-11-22 Thread Udo Erdelhoff
Hi, On Fri, Nov 22, 2002 at 02:30:34PM -, local.freebsd.current wrote: All goes well until I get Extracting base into / directory then Write failure on transfer. Wrote -1 bytes of 240640 and at the bottom of the screen /mnt: write failed, filesystem is full. I had exactly the same problems

Re: make world failure...

2001-02-11 Thread Udo Erdelhoff
On Sun, Feb 11, 2001 at 10:56:18PM +0100, Robert Drehmel wrote: In 38689.981926085@critter, Poul-Henning Kamp wrote: [buildworld failure lib/libc/locale/lmessages.c] It should work with '#include stddef.h'. Yep, that seems to be enough to get past this point. I don't know if there are any

Looking for old snapshots

2001-02-06 Thread Udo Erdelhoff
Hi, I tried to install a snapshot of -current on my "new" test box. The hardware used for the box should be OK, I've managed to boot it with a recent snapshot of RELENG_4. -current is another matter. I've tried to install 20010131 and 20001204 and both die with a "Fatal Trap 18: integer divide

Re: cvs commit: src/sbin/ccdconfig ccdconfig.c

2000-08-05 Thread Udo Erdelhoff
1.17 +5 -5 src/sbin/ccdconfig/ccdconfig.c Broke world: === sbin/ccdconfig rm -f .depend mkdep -f .depend -a-I/usr/src/sbin/ccdconfig/../../sys -I/usr/obj/usr/src/i3 86/usr/include /usr/src/sbin/ccdconfig/ccdconfig.c /usr/src/sbin/ccdconfig/ccdconfig.c:725: unterminated string

Re: **HEADS UP** if you used to cvsup the crypto repo from internat !

2000-07-14 Thread Udo Erdelhoff
On Fri, Jul 14, 2000 at 01:56:31PM +0200, Sheldon Hearn wrote: Also, he clarified the last sentence for me, by saying that it's the crypto ,v files that need to be removed and not the checked out crypto files. What about the non-US cvsup mirrors? Most of them used cvsup to mirror the crypto

Re: panic in sbdrop(), propably not ppp-related

2000-07-12 Thread Udo Erdelhoff
On Wed, Jul 12, 2000 at 02:50:45PM +0100, Brian Somers wrote: I'd like to disclaim all responsibility :-I I'm almost convinced you're innocent. I've managed to restabilize my system by replacing "set device PPPoE:ed1" with "set device /dev/cuaa2". I've been pumping data through ppp for about 10

ppp-related panic in sbdrop()

2000-07-11 Thread Udo Erdelhoff
Hi, I've finally managed to capture a crashdump after a panic in sbdrop(). The machine in question uses ppp/ipfw/natd to connect a small LAN to the outside world via a DSL link. ppp started to misbehave: NS queries were sent out but didn't come back (I had tcpdumps running on both tun0 and ed1).

Re: ppp -auto gone again

2000-07-09 Thread Udo Erdelhoff
On Sun, Jul 09, 2000 at 06:04:47PM +0200, Daniel Rock wrote: - if (pri 0) + if (pri = 0) Groan. "The problem must be somewhere in ip.c, the other changes were purely cosmetic". Famous last words. Thanks, I've re-reverted to the current version, applied your patch and -auto mode

Re: Suspicious warnings in -CURRENT

2000-07-08 Thread Udo Erdelhoff
Hi, I've had the warnings, too, always after successful search operations in vi and mutt. cvs co -D '06 Jul 2000 12:00' src/lib/libc/regex/ and rebuild/reinstall of libc fixed it. It seems the bug was introduced in regcomp.c 1.20/1.121 and/or engine.c 1.8. /s/Udo (still trying to find out what

ppp -auto gone again

2000-07-07 Thread Udo Erdelhoff
Hi, ppp -auto stopped working fater I've updated my box from 06/17-Sources to yesterday's version (07/06, approx. 1500 GMT). tcpdump -ni tun0 shows the traffic but that's it. ppp.log doesn't show any obvious problems. -ddial works, sending a manual dial command (via pppctl) brings the link up

Re: ssh strangeness in -current...

2000-03-07 Thread Udo Erdelhoff
On Tue, Mar 07, 2000 at 11:26:03AM +0200, Sheldon Hearn wrote: Is it supposed to be possible to drop a "1024-bit" host key from the old ssh1 port into /etc/ssh? It works for me. I've created my host key with ssh-1.2.26 and the base system OpenSSH accepted it without any problems. To

Re: 3.4 - 4.0 upgrade problems

2000-03-07 Thread Udo Erdelhoff
On Tue, Mar 07, 2000 at 11:56:32AM +0200, Ruslan Ermilov wrote: + make buildkernel installkernel KERNEL="NAME_OF_YOUR_KERNEL" You should skip the "KERNEL=..." part. buildkernel and installkernel use KERNEL for *everything* - including the filename of the new kernel. In other words:

Re: ssh strangeness in -current...

2000-03-07 Thread Udo Erdelhoff
On Tue, Mar 07, 2000 at 06:48:56PM +0100, Brad Knowles wrote: Must be a FreeBSD vs. Solaris thing Are you using OpenSSH or the 'normal' ssh on your Solaris box? I've just tried to copy files between my FreeBSD box @home and one of 'my' Solaris boxes @work. All four possible directions work

Re: des/ssl manpages lost

2000-02-29 Thread Udo Erdelhoff
On Sun, Feb 27, 2000 at 11:43:26PM -0800, Kris Kennaway wrote: [OpenSSL 0.9.5 manpages for 'out' 0.9.4] The OpenSSL API has changed a fair bit since 0.9.4 - not substantially, but in a lot of little ways. Yikes, I'm getting old. My first reaction to this information was"Bah, whatever happend

des/ssl manpages lost

2000-02-27 Thread Udo Erdelhoff
Hi, I've noticed that the first successfull make world after the openssh import failed to update several manpages for des related functions. The functions were moved into libcrypto, the manpages were deleted alongside libdes. The missing files are: /share/man/man3/des_read_password.3.gz

Re: Am I doing this right?

2000-02-27 Thread Udo Erdelhoff
On Sun, Feb 27, 2000 at 10:37:01PM -0600, Steve Kaczkowski wrote: One source says that I should recompile my kernel, reboot into single user and do a 'make installworld'. Makes sense but I can't recompile my kernel, I get errors like: config GENERIC You're using the 3.x-version of config to

Re: des/ssl manpages lost

2000-02-27 Thread Udo Erdelhoff
On Sun, Feb 27, 2000 at 02:05:51PM -0800, Kris Kennaway wrote: [des manpages lost] We probably should revive these. OpenSSL 0.9.5 has a complete set of manpages, and I hope to import it shortly after the release, but for 4.0-REL we should keep these. What about importing the OpenSSL 0.9.5