stable-digest V5 #798
owner-freebsd...@freebsd.org
stable-digest Wednesday, February 19 2003 Volume 05 : Number 798
In this issue:
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Updated to today's -stable and can no longer connect to the XFree86 server
spontaneous reboot & gcc
sta103, LIVE FROM WALL STREET: VICC Test Results Are In..........
Re: spontaneous reboot & gcc
Re: spontaneous reboot & gcc
Fatal trap 12: page fault while in kernel mode
Re: spontaneous reboot & gcc
Re: spontaneous reboot & gcc
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Fatal trap 12: page fault while in kernel mode
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: make installworld fails on libperl.a
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Updated to today's -stable and can no longer connect to the XFree86 server
Re: Fatal trap 12: page fault while in kernel mode
Re: spontaneous reboot & gcc
Re: Updated to today's -stable and can no longer connect to the XFree86 server
RE: spontaneous reboot & gcc
Re: spontaneous reboot & gcc
Re: spontaneous reboot & gcc
Re: spontaneous reboot & gcc
Re: PCI oddity
PIO4 mode on ST380023A hard disk
Re: PIO4 mode on ST380023A hard disk
FTP-server
System hangs with new Intel fxp card
Re: FTP-server
Question about ipfilter logs and I may have a proof of concept patch to fix?
----------------------------------------------------------------------
Date: Tue, 18 Feb 2003 17:18:08 -0800
From: Kris Kennaway <kr...@obsecurity.org>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
- --5mZBmBd1ZkdwT1ny
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Feb 18, 2003 at 10:17:05AM -0600, Bob Willcox wrote:
> I am running XFree86 version 4.2.0 (from port XFree86-4.2.0_1,1).
>=20
> Has something recently changed that might cause this? I had previously
> updated my system on Feb 1 and wasn't having this problem with X11.
XFree86 does not listen on TCP ports by default. You have to enable
it if you want this behaviour (see man startx or man xdm).
Kris
- --5mZBmBd1ZkdwT1ny
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)
iD8DBQE+UttQWry0BWjoQKURAkr4AJ4gXR+KnrPBc/LXeJcfnTMqxg1PFgCfdeQ3
V8AthueG5l4S5jSohZf8ljE=
=lDDS
- -----END PGP SIGNATURE-----
- --5mZBmBd1ZkdwT1ny--
------------------------------
Date: Tue, 18 Feb 2003 21:38:11 -0600
From: Bob Willcox <b...@immure.com>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
On Tue, Feb 18, 2003 at 05:18:08PM -0800, Kris Kennaway wrote:
> On Tue, Feb 18, 2003 at 10:17:05AM -0600, Bob Willcox wrote:
>
> > I am running XFree86 version 4.2.0 (from port XFree86-4.2.0_1,1).
> >
> > Has something recently changed that might cause this? I had previously
> > updated my system on Feb 1 and wasn't having this problem with X11.
>
> XFree86 does not listen on TCP ports by default. You have to enable
> it if you want this behaviour (see man startx or man xdm).
Well then I am confused. I didn't change XFree86. I only updated the
system. Why would I suddenly see this new behavior?
Bob
>
> Kris
- --
Bob Willcox We seem to have forgotten the simple truth that
b...@vieo.com reason is never perfect. Only non-sense attains
Austin, TX perfection. -- Poul Henningsen [1894-1967]
------------------------------
Date: Tue, 18 Feb 2003 23:59:48 -0500 (EST)
From: Darren Henderson <dar...@nighttide.net>
Subject: spontaneous reboot & gcc
I'm seeing a spontaneous reboot in one very specific circumstance that
appears to be attributable to the use of gcc with -O2 optimization.
For some time now I've been seeing spontaneous reboots when trying to
compile pike (http://pike.ida.liu.se/) on a 1.8Ghz P4 system running
4.7-STABLE (though its been a problem for a few versions of both Pike and
FreeBSD).
This does not occur on a 700Mhz P3 or on a 133Mhz Pentium Pro system. All
the systems have recent source.
The 1.8Ghz system has 256MB of ram and appears to be rock solid in all
other respects - the reboot only occurs while attempting to compile Pike,
and always in the same location (peep.c).
Pike defaults to using -O2 optimizations. Forcing it not to do that stops
the spontaneous reboot from occurring.
There isn't any spore left behind after the reboot though I have
softupdates turned on and fsck has a bit of work to do when the box comes
back up. Turning softupdates off might reveal a bit more.
It doesn't seem reasonable that gcc could bring the box down but the fact
that the -O2 elimination keeps it from happening gives me pause to wonder
about gcc on P4 systems. I suppose not using -O2 would mean that gcc is
doing a bit less work, maybe using a bit less memory etc but the other
systems have less then half the memory of the system in question and they
don't see a problem.
Any thoughts appreciated.
______________________________________________________________________
Darren Henderson dar...@nighttide.net
Help fight junk e-mail, visit http://www.cauce.org/
------------------------------
Date: Wed, 19 Feb 2003 05:10:06 GMT
From: incomin...@cs.com
Subject: sta103, LIVE FROM WALL STREET: VICC Test Results Are In..........
sta...@txt.electronicstalk.com
<p>If you bought into our last recommendation (CIMG) early enough you had an excellent opportunity to make substantial gains (from .90 to 1.65 in just the first day). Now is your chance to do the same with our newest pick: VICC. To find out more go to <a href="http://12.148.59.67">Live From the Street</a>.</p>
<p align="center"><img border="0" src="http://12.148.59.67/bm01.gif"></p>
<p>If you no longer want to receive information from us just go to
<a href="mailto:tal...@cs.com">tal...@cs.com</a>.<p>
------------------------------
Date: Tue, 18 Feb 2003 22:05:06 -0800
From: "Doug Clements" <dscle...@linkline.com>
Subject: Re: spontaneous reboot & gcc
- ----- Original Message -----
From: "Darren Henderson" <dar...@nighttide.net>
To: "stable list" <freebsd...@FreeBSD.ORG>
Sent: Tuesday, February 18, 2003 8:59 PM
Subject: spontaneous reboot & gcc
> I'm seeing a spontaneous reboot in one very specific circumstance that
> appears to be attributable to the use of gcc with -O2 optimization.
I was getting something similar last week during a buildworld, but I chalked
it up to bad hardware (since I'm using a spiffy new P4 machine, it has no
reliable track record yet). I'll see if I can reproduce your specifics on my
machine when I get in to work tomorrow.
- --Doug
------------------------------
Date: Wed, 19 Feb 2003 17:49:27 +1100
From: Graham Menhennitt <gmenh...@optushome.com.au>
Subject: Re: spontaneous reboot & gcc
At 03:59 PM 19/02/03, Darren Henderson wrote:
>and always in the same location (peep.c).
>
>Pike defaults to using -O2 optimizations. Forcing it not to do that stops
>the spontaneous reboot from occurring.
It used to be the case (not sure if it still is) that compiling MySQL
required turning off optimisation for one file. Otherwise you would run out
of swap (unless you had _lots_ available). This may be a similar problem
since it's so reproducible.
That probably doesn't help much but you could at least check your swap
space size.
Graham
------------------------------
Date: Wed, 19 Feb 2003 15:13:21 +0100 (MET)
From: Mario Pranjic <mario....@irb.hr>
Subject: Fatal trap 12: page fault while in kernel mode
Hi!
For the last few days I'm expiriencing a strange crashes: when I try to
cvsupdate+portsdb ports tree I get the weird errors and machine crashes
(automatically rebooted)
One I got this error:
panic: ufsdirhash_lookup: bad offset in hash array.
This time I got this one:
Fatal trap 12: page fault while in kernel mode
fault virtual address = 0x3c
fault code = supervisor read, page not present
instruction pointer = 0x8:0xc0309bca
stack pointer = 0x10:0xcda58bf8
frame pointer = 0x10:0xcda58be8
code segment = base 0x0, limit 0xfffff, type 0x1b
= DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 1951 (make)
interrupt masks = none
trap number = 12
panic: page fault
The system is Athlon 700 MHz, 256 MB ECC RAM, 2 Seagate IDE disks (40 and
80 gigs), 1 SCSI Seagate disk (36 GB).
First I suspected that a power suply is malfunctioning, so I've changed
it. Didn't help. though.
I ran soime fsck on drives and it seems that disks are OK.
What could possibly be so wrong? The machine is online for about a year
and it worked fine till now.
Is there a possibility that a memory chip is dead (or should I say
deadish)?
Thanks for your help!
Mario Pranjic, dipl.ing.
sistem administrator
Knjiznica, Institut Rudjer Boskovic
- -------------------------------------
e-mail: mario....@irb.hr
ICQ: 72059629
tel: +385 1 45 60 954 (interni: 1293)
- -------------------------------------
------------------------------
Date: Wed, 19 Feb 2003 09:41:18 -0500 (EST)
From: Darren Henderson <dar...@nighttide.net>
Subject: Re: spontaneous reboot & gcc
On Tue, 18 Feb 2003, Doug Clements wrote:
> I was getting something similar last week during a buildworld, but I chalked
> it up to bad hardware (since I'm using a spiffy new P4 machine, it has no
> reliable track record yet). I'll see if I can reproduce your specifics on my
> machine when I get in to work tomorrow.
Thanks, be interested in what you see.
The box is a bit less then a year old but other then the early demise of
one hard drive its been solid. Its possible that its a hardware issue but
since this is the only issue I'm seeing it doesn't seem likely.
______________________________________________________________________
Darren Henderson dar...@nighttide.net
Help fight junk e-mail, visit http://www.cauce.org/
------------------------------
Date: Wed, 19 Feb 2003 09:46:55 -0500 (EST)
From: Darren Henderson <dar...@nighttide.net>
Subject: Re: spontaneous reboot & gcc
On Wed, 19 Feb 2003, Graham Menhennitt wrote:
> That probably doesn't help much but you could at least check your swap
> space size.
Thanks for the response. It has 256MB ram with 512MB - doesn't behave like
a resource issue, doesn't slow down and shed processes, just reboots.
Compiles ok on the 133Mhz machine which only has 96MB ram and 256MB swap.
______________________________________________________________________
Darren Henderson dar...@nighttide.net
Help fight junk e-mail, visit http://www.cauce.org/
------------------------------
Date: Wed, 19 Feb 2003 07:24:41 -0800
From: Erick Mechler <emec...@techometer.net>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
:: > XFree86 does not listen on TCP ports by default. You have to enable
:: > it if you want this behaviour (see man startx or man xdm).
::
:: Well then I am confused. I didn't change XFree86. I only updated the
:: system. Why would I suddenly see this new behavior?
For example, if you have xdm starting out of /etc/ttys and you blindly
update everything mergemaster tells you to, you might overwrite some of
your custom configs.
Cheers - Erick
------------------------------
Date: Wed, 19 Feb 2003 09:38:07 -0600
From: Bob Willcox <b...@immure.com>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
On Wed, Feb 19, 2003 at 07:24:41AM -0800, Erick Mechler wrote:
> :: > XFree86 does not listen on TCP ports by default. You have to enable
> :: > it if you want this behaviour (see man startx or man xdm).
> ::
> :: Well then I am confused. I didn't change XFree86. I only updated the
> :: system. Why would I suddenly see this new behavior?
>
> For example, if you have xdm starting out of /etc/ttys and you blindly
> update everything mergemaster tells you to, you might overwrite some of
> your custom configs.
Well, I don't use xdm. I always have (for many years now) started the
X server using startx, and I do see that startx (which hasn't been
changed on my system since last December) does explicitly specify
"-nolisten tcp" if not told otherwise. The thing that is a mystery to me
is that before upgrading to Feb 18's -stable my X server was allowing
connections. I can't explain why it used to work! :-)
Bob
>
> Cheers - Erick
- --
Bob Willcox We seem to have forgotten the simple truth that
b...@vieo.com reason is never perfect. Only non-sense attains
Austin, TX perfection. -- Poul Henningsen [1894-1967]
------------------------------
Date: Wed, 19 Feb 2003 16:40:58 +0100 (MET)
From: Mario Pranjic <mario....@irb.hr>
Subject: Re: Fatal trap 12: page fault while in kernel mode
> Is there a possibility that a memory chip is dead (or should I say
> deadish)?
As I said, the memory died. :)
The Memtest86 returned a lot of errors. I've replaced the memory module
and now it seems all right.
Mario Pranjic, dipl.ing.
sistem administrator
Knjiznica, Institut Rudjer Boskovic
- -------------------------------------
e-mail: mario....@irb.hr
ICQ: 72059629
tel: +385 1 45 60 954 (interni: 1293)
- -------------------------------------
------------------------------
Date: Wed, 19 Feb 2003 07:41:45 -0800
From: Erick Mechler <emec...@techometer.net>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
:: > For example, if you have xdm starting out of /etc/ttys and you blindly
:: > update everything mergemaster tells you to, you might overwrite some of
:: > your custom configs.
::
:: Well, I don't use xdm. I always have (for many years now) started the
:: X server using startx, and I do see that startx (which hasn't been
:: changed on my system since last December) does explicitly specify
:: "-nolisten tcp" if not told otherwise.
Perhaps you had an entry in your .xinitrc to override the command-line
settings which isn't there anymore? Eh, just grasping at straws here...
Cheers - Erick
------------------------------
Date: Wed, 19 Feb 2003 09:48:10 -0600
From: Bob Willcox <b...@immure.com>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
On Wed, Feb 19, 2003 at 07:41:45AM -0800, Erick Mechler wrote:
> :: > For example, if you have xdm starting out of /etc/ttys and you blindly
> :: > update everything mergemaster tells you to, you might overwrite some of
> :: > your custom configs.
> ::
> :: Well, I don't use xdm. I always have (for many years now) started the
> :: X server using startx, and I do see that startx (which hasn't been
> :: changed on my system since last December) does explicitly specify
> :: "-nolisten tcp" if not told otherwise.
>
> Perhaps you had an entry in your .xinitrc to override the command-line
> settings which isn't there anymore? Eh, just grasping at straws here...
I couldn't find anything...besides I haven't changed my .xinitrc in
sometime.
Bob
>
> Cheers - Erick
- --
Bob Willcox We seem to have forgotten the simple truth that
b...@vieo.com reason is never perfect. Only non-sense attains
Austin, TX perfection. -- Poul Henningsen [1894-1967]
------------------------------
Date: Wed, 19 Feb 2003 14:12:29 -0300
From: Fernan Aguero <fer...@iib.unsam.edu.ar>
Subject: Re: make installworld fails on libperl.a
Sorry for the delay in replying. Thanks Jukka and Rudolph
for pointing me to the problem.
Both of you were right. I had built perl-5.6.1 from ports in the
build host and my /etc/make.conf had some entries introduced
by the use.perl script.
I just copied the make.conf file to the install host and
everything worked fine.
Fernan
+----[ Jukka Simila <ho...@jukkis.net> (14.Feb.2003 17:57):
|
| On Fri, 2003-02-14 at 20:59, Fernan Aguero wrote:
| > I am using a FreeBSD-4.7-STABLE (Jan24) to build world and
| > kernel for an older machine (Pentium1).
| >
| > buildworld and buildkernel proceed OK on the build host.
| >
| > In the install host I do:
| >
| > mount build:/usr/src /usr/src
| > mount build:/usr/obj /usr/obj
| > make installworld
|
| maybe you have differing /etc/make.conf files on the two hosts?
|
| if the build host isn't building perl,
| you definitely get errors on the install host..
|
+----]
- --
F e r n a n A g u e r o
http://genoma.unsam.edu.ar/~fernan
------------------------------
Date: Wed, 19 Feb 2003 09:23:47 -0800
From: Nathan Kinkade <nkin...@dsl-only.net>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
- --D/26AIznG/rf8cgR
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Wed, Feb 19, 2003 at 09:48:10AM -0600, Bob Willcox wrote:
> On Wed, Feb 19, 2003 at 07:41:45AM -0800, Erick Mechler wrote:
> > :: > For example, if you have xdm starting out of /etc/ttys and you bli=
ndly=20
> > :: > update everything mergemaster tells you to, you might overwrite so=
me of=20
> > :: > your custom configs.
> > ::=20
> > :: Well, I don't use xdm. I always have (for many years now) started the
> > :: X server using startx, and I do see that startx (which hasn't been
> > :: changed on my system since last December) does explicitly specify
> > :: "-nolisten tcp" if not told otherwise.
> >=20
> > Perhaps you had an entry in your .xinitrc to override the command-line
> > settings which isn't there anymore? Eh, just grasping at straws here...
>=20
> I couldn't find anything...besides I haven't changed my .xinitrc in
> sometime.
>=20
> Bob
>=20
> >=20
> > Cheers - Erick
>=20
> --=20
> Bob Willcox We seem to have forgotten the simple truth that
> b...@vieo.com reason is never perfect. Only non-sense attains
> Austin, TX perfection. -- Poul Henningsen [1894-1967]
This is just another shot in the dark, but I see that you are trying to
connect to a machine named 'luke'. Maybe your /etc/hosts file was
overwritten and now name resolution isn't happening like it should
locally? What happens if you try to connect using the ip address like:
$ xterm -display 192.168.1.100:0
Also, when on the machine named 'luke' what is the output of `xhost`.
If you want everyone to be able to connect then it should output
something like:
$ xhost
access control disabled, clients can connect from any host
Also, can you verify 100% that X is actually listening on host 'luke' by
browsing the output of `sockstat -l4`?=20
One more thing: how did you upgrade? Are you certain that you don't
have a firewall running on the newly upgraded system?
Good luck,
Nathan
- --=20
GPG Public Key ID: 0x4250A04C
gpg --keyserver pgp.mit.edu --recv-keys 4250A04C
http://63.105.21.156/gpg_nkinkade_4250A04C.asc
- --D/26AIznG/rf8cgR
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)
iD8DBQE+U72jWZYS9EJQoEwRAo6zAKC+HhFhZs35lc0wnlUhAbUbrvUCyQCffbvt
jt39qxeBt5gIR3Pr1WhFNjY=
=NrVn
- -----END PGP SIGNATURE-----
- --D/26AIznG/rf8cgR--
------------------------------
Date: Wed, 19 Feb 2003 13:24:27 -0600
From: Bob Willcox <b...@immure.com>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
On Wed, Feb 19, 2003 at 09:23:47AM -0800, Nathan Kinkade wrote:
>
> This is just another shot in the dark, but I see that you are trying to
> connect to a machine named 'luke'. Maybe your /etc/hosts file was
> overwritten and now name resolution isn't happening like it should
> locally? What happens if you try to connect using the ip address like:
> $ xterm -display 192.168.1.100:0
I use named (DNS) to resolve all of my host names so my /etc/hosts file
is empty. Right now my X server is allowing connections (I restarted
it with "startx -listen_tcp") so I'm convinced that that was it (still
doesn't explain why it used to work).
>
> Also, when on the machine named 'luke' what is the output of `xhost`.
> If you want everyone to be able to connect then it should output
> something like:
> $ xhost
> access control disabled, clients can connect from any host
Yep.
>
> Also, can you verify 100% that X is actually listening on host 'luke' by
> browsing the output of `sockstat -l4`?
This is what I get right now:
bob@luke:pa /home/bob> sockstat -l4|grep XFree86
root XFree86 41464 1 tcp4 *:6000 *:*
but it's working now also.
>
> One more thing: how did you upgrade? Are you certain that you don't
> have a firewall running on the newly upgraded system?
I updated from source. I cvsup the cvs repo and update my /usr/src tree
from that with cvs.
>
> Good luck,
> Nathan
Thanks for you time and thoughts,
Bob
>
> --
> GPG Public Key ID: 0x4250A04C
> gpg --keyserver pgp.mit.edu --recv-keys 4250A04C
> http://63.105.21.156/gpg_nkinkade_4250A04C.asc
- --
Bob Willcox We seem to have forgotten the simple truth that
b...@vieo.com reason is never perfect. Only non-sense attains
Austin, TX perfection. -- Poul Henningsen [1894-1967]
------------------------------
Date: Wed, 19 Feb 2003 11:37:24 -0800
From: Kris Kennaway <kr...@obsecurity.org>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
- --D6z0c4W1rkZNF4Vu
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
On Wed, Feb 19, 2003 at 09:38:07AM -0600, Bob Willcox wrote:
> Well, I don't use xdm. I always have (for many years now) started the
> X server using startx, and I do see that startx (which hasn't been
> changed on my system since last December) does explicitly specify
> "-nolisten tcp" if not told otherwise. The thing that is a mystery to me
> is that before upgrading to Feb 18's -stable my X server was allowing
> connections. I can't explain why it used to work! :-)
The default was changed quite a while ago, so perhaps you just hadn't
upgraded X in a long time?
Kris
- --D6z0c4W1rkZNF4Vu
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)
iD8DBQE+U9z0Wry0BWjoQKURAuTLAKDamY7+D2b2BBblp0+POtXH+QKPyACgw8lz
cBHHIpL8EhuQgqW3J0lpt9U=
=jle1
- -----END PGP SIGNATURE-----
- --D6z0c4W1rkZNF4Vu--
------------------------------
Date: Wed, 19 Feb 2003 11:40:48 -0800
From: Kris Kennaway <kr...@obsecurity.org>
Subject: Re: Fatal trap 12: page fault while in kernel mode
- --NqNl6FRZtoRUn5bW
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Wed, Feb 19, 2003 at 04:40:58PM +0100, Mario Pranjic wrote:
> > Is there a possibility that a memory chip is dead (or should I say
> > deadish)?
>=20
> As I said, the memory died. :)
> The Memtest86 returned a lot of errors. I've replaced the memory module
> and now it seems all right.
It's so nice when problems are easily resolved :)
Kris
- --NqNl6FRZtoRUn5bW
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)
iD8DBQE+U92/Wry0BWjoQKURAneHAKDdhG+b5ps6lP3zoUKeyIgMVtC1UQCZAa6p
l2RRBxTKTlu/c0vd9wRk/bg=
=kEb6
- -----END PGP SIGNATURE-----
- --NqNl6FRZtoRUn5bW--
------------------------------
Date: Wed, 19 Feb 2003 11:42:28 -0800
From: Kris Kennaway <kr...@obsecurity.org>
Subject: Re: spontaneous reboot & gcc
- --ZrCu0B0FMx3UnjE2
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Feb 18, 2003 at 11:59:48PM -0500, Darren Henderson wrote:
>=20
> I'm seeing a spontaneous reboot in one very specific circumstance that
> appears to be attributable to the use of gcc with -O2 optimization.
Are you also compiling the kernel with -O2? If so, then "Don't do
that then". -O2 has had known serious bugs in the past.
Kris
- --ZrCu0B0FMx3UnjE2
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)
iD8DBQE+U94jWry0BWjoQKURAlM7AJ9+2KUM3FQNzfHxM2Lv3L6y5HfThQCdHUx7
Ga3TCY3/OCcVlHlmhMUBBPo=
=sDXW
- -----END PGP SIGNATURE-----
- --ZrCu0B0FMx3UnjE2--
------------------------------
Date: Wed, 19 Feb 2003 13:45:38 -0600
From: Bob Willcox <b...@immure.com>
Subject: Re: Updated to today's -stable and can no longer connect to the XFree86 server
On Wed, Feb 19, 2003 at 11:37:24AM -0800, Kris Kennaway wrote:
> On Wed, Feb 19, 2003 at 09:38:07AM -0600, Bob Willcox wrote:
>
> > Well, I don't use xdm. I always have (for many years now) started the
> > X server using startx, and I do see that startx (which hasn't been
> > changed on my system since last December) does explicitly specify
> > "-nolisten tcp" if not told otherwise. The thing that is a mystery to me
> > is that before upgrading to Feb 18's -stable my X server was allowing
> > connections. I can't explain why it used to work! :-)
>
> The default was changed quite a while ago, so perhaps you just hadn't
> upgraded X in a long time?
That's the mystery. I last updated XFree86 on Dec 23rd, and yet I have
rebooted and restarted X several times since then (with connections to
the X server working). This happened just yesterday (after upgrading the
system and rebooting.
Bob
>
> Kris
- --
Bob Willcox We seem to have forgotten the simple truth that
b...@vieo.com reason is never perfect. Only non-sense attains
Austin, TX perfection. -- Poul Henningsen [1894-1967]
------------------------------
Date: Wed, 19 Feb 2003 14:55:06 -0500
From: Don Bowman <d...@sandvine.com>
Subject: RE: spontaneous reboot & gcc
> From: Kris Kennaway [mailto:kr...@obsecurity.org]
>
> On Tue, Feb 18, 2003 at 11:59:48PM -0500, Darren Henderson wrote:
> >
> > I'm seeing a spontaneous reboot in one very specific
> circumstance that
> > appears to be attributable to the use of gcc with -O2 optimization.
>
> Are you also compiling the kernel with -O2? If so, then "Don't do
> that then". -O2 has had known serious bugs in the past.
- -O2 -pipe -malign-loops=4 -malign-jumps=4 -malign-functions=4
- -mcpu=i686 -march=i686 -fno-gcse
is what I'm using with good affect.
I found one problem for sure with global common sub-expression elimination
in the FreeBSD kernel w/ gcc 2.95.
gcc 3.X is much more reliable, but not the default compiler for 4.X.
The alignment used by default makes for fairly poor performance for
pentium-4 / xeon architectures.
- --don
------------------------------
Date: Wed, 19 Feb 2003 12:47:25 -0800
From: Doug Clements <dcle...@linkline.com>
Subject: Re: spontaneous reboot & gcc
On Wed, Feb 19, 2003 at 09:41:18AM -0500, Darren Henderson wrote:
> On Tue, 18 Feb 2003, Doug Clements wrote:
>
> > I was getting something similar last week during a buildworld, but I chalked
> > it up to bad hardware (since I'm using a spiffy new P4 machine, it has no
> > reliable track record yet). I'll see if I can reproduce your specifics on my
> > machine when I get in to work tomorrow.
>
> Thanks, be interested in what you see.
>
> The box is a bit less then a year old but other then the early demise of
> one hard drive its been solid. Its possible that its a hardware issue but
> since this is the only issue I'm seeing it doesn't seem likely.
I got Pike 7.4.10:
Pike compiled and linked ok, but the following file(s)
failed to compile with full optimization.
This may affect performance negatively.
./modules/_Crypto/lib/desQuick.o
- -bash-2.05b$
Pike7.5-20030219-182519 also seems to compile just fine.
Other than that, nothing spectacular. Sorry! Looks like I should try my buildworlds again ;)
- --Doug
------------------------------
Date: Wed, 19 Feb 2003 17:04:01 -0500 (EST)
From: Darren Henderson <dar...@nighttide.net>
Subject: Re: spontaneous reboot & gcc
On Wed, 19 Feb 2003, Kris Kennaway wrote:
> > I'm seeing a spontaneous reboot in one very specific circumstance that
> > appears to be attributable to the use of gcc with -O2 optimization.
>
> Are you also compiling the kernel with -O2? If so, then "Don't do
> that then". -O2 has had known serious bugs in the past.
Thanks for the repy. No, make.conf specs -O -pipe.
______________________________________________________________________
Darren Henderson dar...@nighttide.net
Help fight junk e-mail, visit http://www.cauce.org/
------------------------------
Date: Wed, 19 Feb 2003 17:10:10 -0500 (EST)
From: Darren Henderson <dar...@nighttide.net>
Subject: Re: spontaneous reboot & gcc
On Wed, 19 Feb 2003, Doug Clements wrote:
> Other than that, nothing spectacular. Sorry! Looks like I should try my buildworlds again ;)
Much appreciated, time to pick up another stick of ram and see if that
corrects the issue I guess. memtest indicates it checks out ok but I'm at
a loss as to what else to try - not getting Pike to complie isn't a big
issue but if I have bad memory I'd kind of like to know about it.
Thanks for trying that.
______________________________________________________________________
Darren Henderson dar...@nighttide.net
Help fight junk e-mail, visit http://www.cauce.org/
------------------------------
Date: Thu, 20 Feb 2003 01:50:14 +0200 (EET)
From: Dmitry Pryanishnikov <dmi...@atlantis.dp.ua>
Subject: Re: PCI oddity
Hello!
On Mon, 10 Feb 2003, Jason Andresen wrote:
> Are there certain cards that are known to be good about sharing interrupts?
Mine Intel D845WN motherboard is quite happy with 4 fxp interfaces (including
onboard one) on IRQ11 and 3 de interfaces on IRQ9. However, in some
combinations (e.g. trying to move some cards to the unused IRQ7 or IRQ5)
FreeBSD 4.7-RELEASE just hanged after end of the PCI device probe.
Sincerely, Dmitry
- --
Atlantis ISP, System Administrator
e-mail: dmi...@atlantis.dp.ua
nic-hdl: LYNX-RIPE
------------------------------
Date: Thu, 20 Feb 2003 01:13:20 +0100
From: Lokhorst <lokh...@fwb.eur.nl>
Subject: PIO4 mode on ST380023A hard disk
Hello,
Does anyone know the solution to the following puzzle:
Why do I get
ad0: 76319MB <ST380023A> [155061/16/63] at ata0-master PIO4
while Windows XP runs the same hard disk in UDMA5 mode (i.e., much faster)?
I use FreeBSD 4.7-STABLE from about a week ago.
Gert-Jan Lokhorst
------------------------------
Date: Thu, 20 Feb 2003 08:36:14 +0800
From: Shaun Dwyer <sdw...@arach.net.au>
Subject: Re: PIO4 mode on ST380023A hard disk
I can't tell you why it starts out in PIO mode... we have a server at work running Debian Lunix that does the same thing.
running hdparm(plus some flags) to put it into UDMA mode fixes it. Its a P3 on a Abit motherboard.
You may need to just use 'atacontrol' to set it into UDMA mode after booting (or during a boot script).
- --Shaun
Lokhorst wrote:
> Hello,
>
> Does anyone know the solution to the following puzzle:
>
> Why do I get
>
> ad0: 76319MB [155061/16/63] at ata0-master PIO4
>
> while Windows XP runs the same hard disk in UDMA5 mode (i.e., much faster)?
>
> I use FreeBSD 4.7-STABLE from about a week ago.
>
> Gert-Jan Lokhorst
>
>
> To Unsubscribe: send mail to majo...@FreeBSD.org
> with "unsubscribe freebsd-stable" in the body of the message
>
>
------------------------------
Date: Thu, 20 Feb 2003 01:43:13 +0100
From: "Vidor Demeter" <vi...@home.se>
Subject: FTP-server
Hi everybody!
I want to setup a FTP-server on my computer with FreeBSD 5.0...
Is there any help on how to begin? I tried to find some help, but I
did not find any. Could someone help me with that? How to begin?
TIA
Vidor
------------------------------
Date: Wed, 19 Feb 2003 16:56:56 -0800
From: Doug Silver <dsi...@urchin.com>
Subject: System hangs with new Intel fxp card
Hi All -
I just installed an Intel Pro card on a box that has an older 10Mb RealTe=
k=20
8029 card. Booting up a 4.7-Generic the kernel sees everything, but the=20
system completely hangs when I try to ifconfig the fxp device up. Here's=
=20
some details:
> dmesg|egrep -i "fxp|ed0"
fxp0: <Intel Pro 10/100B/100+ Ethernet> port 0xd400-0xd43f mem=20
0xeffa0000-0xeffbffff,0xeffde000-0xeffdefff irq 10 at device 9.0 on pci0
fxp0: Ethernet address 00:d0:b7:85:19:a9
ed0: <NE2000 PCI Ethernet (RealTek 8029)> port 0xd000-0xd01f irq 12 at de=
vice=20
11.0 on pci0
ed0: address 00:50:ba:a2:91:6a, type NE2000 (16 bit)
> ifconfig -a
fxp0: flags=3D8802<BROADCAST,SIMPLEX,MULTICAST> mtu 1500
ether 00:d0:b7:85:19:a9
media: Ethernet autoselect (100baseTX <full-duplex>)
status: active
ed0: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
ether 00:50:ba:a2:91:6a
The caveat to all of this is that the drive has shown some disk errors --=
=20
which will be replaced soon enough, but it seems odd that it completely h=
angs=20
the machine when I've tried to bring it online. When I ran ifconfig, it d=
id=20
not give any errors, just froze the system. I have this same card on ma=
ny=20
other machines and have never had a problem with them.
Any suggestions?
Thanks.
- -doug
------------------------------
Date: Thu, 20 Feb 2003 11:29:39 +1030
From: kibbet <kib...@kibby.au.knfpub.com>
Subject: Re: FTP-server
On Thu, Feb 20, 2003 at 01:43:13AM +0100, Vidor Demeter wrote:
> Hi everybody!
>
> I want to setup a FTP-server on my computer with FreeBSD 5.0...
> Is there any help on how to begin? I tried to find some help, but I
> did not find any. Could someone help me with that? How to begin?
> TIA
> Vidor
>
This question belongs on -questions (and then only after you have
read all the relevent documentation). Basically you need to uncomment
the following line in /etc/inetd.conf
#ftp stream tcp nowait root /usr/libexec/ftpd ftpd -l
and restart inetd;
killall -HUP inetd
Cheers,
- --
Kent Ibbetson
fl...@knfpub.com
------------------------------
Date: Wed, 19 Feb 2003 21:04:38 -0500
From: "Michael Scheidell" <sche...@secnap.net>
Subject: Question about ipfilter logs and I may have a proof of concept patch to fix?
I decided to try ipfilter instead of ipfw and had a question about the logs
and TEXT for the ICMP type rather than an icmp type code.
FBSD 4.8 4.8-PRERELEASE (today's cvs)
Feb 19 19:55:53 scanner ipmon[49]: 19:55:52.897839 fxp1 @0:16 b
207.18.92.26 -> 208.237.120.133 PR icmp len 20 84 icmp echo/0 IN
I noticed that on 'other' ipfilter logs that I have seen that log would
look like this:
Feb 19 19:55:53 scanner ipmon[49]: 19:55:52.897839 fxp1 @0:16 b
207.18.92.26 -> 208.237.120.133 PR icmp len 20 84 icmp 8/0 IN
Note this from man ipmon:
If the packet is an ICMP packet, there will be two fields
at the end, the first always being `icmp', and the next
being the ICMP message and submessage type, separated by a
slash, e.g., icmp 3/3 for a port unreachable message
started ipmon with:
ipmon_flags="-Ds"
or
ipmon_flags="-Dsp"
I suspect some 'switch' that should be regulated by the -n option is
reversed somewhere.
I have never looked at the code before today, but could it be near line 1059
of /usr/src/contrib/ipfilter/ipmon.c
I note that hostname() takes a 'res' argument (set leter depending on the
RESOLVE switches, -n an d-pP but that icmpname doesn't.
(void) sprintf(t, "%s PR icmp len %hu %hu icmp %s",
hostname(res, v, d), hl, plen,
icmpname(ic->icmp_type, ic->icmp_code));
would something like this fix it? (same would have t be done for the ipv6
part)
maybe fix/change icmpname to take an additional arg, res.
(i didn't want to go through all the source to test it)
if (res) {
(void) sprintf(t, "%s PR icmp len %hu %hu icmp %s",
hostname(res, v, d), hl, plen,
icmpname(ic->icmp_type, ic->icmp_code));
} else {
(void) sprintf(t, "%s PR icmp len %hu %hu icmp %hu/%hu",
hostname(res, v, d), hl, plen,
ic->icmp_type, ic->icmp_code);
}
possible proof of concept patch to fix?
- --- ipmon.c.orig Thu Nov 21 21:07:50 2002
+++ ipmon.c Wed Feb 19 20:58:55 2003
@@ -1054,9 +1054,15 @@
ic = (struct icmp *)((char *)ip + hl);
(void) sprintf(t, "%s -> ", hostname(res, v, s));
t += strlen(t);
+ if (res) {
(void) sprintf(t, "%s PR icmp len %hu %hu icmp %s",
hostname(res, v, d), hl, plen,
icmpname(ic->icmp_type, ic->icmp_code));
+ } else {
+ (void) sprintf(t, "%s PR icmp len %hu %hu icmp %hu/%hu",
+ hostname(res, v, d), hl, plen,
+ ic->icmp_type, ic->icmp_code);
+ }
if (ic->icmp_type == ICMP_UNREACH ||
ic->icmp_type == ICMP_SOURCEQUENCH ||
ic->icmp_type == ICMP_PARAMPROB ||
note, after patch it seems to be consistant with man, can be parsed with
'most' ipmon parse scripts.
Why should this not be there?
cd /usr/src
make most
killall ipmon
/usr/obj/usr/src/sbin/ipmon/ipmon -Dsp
tail -f /var/log/firellog:
Feb 19 20:36:38 scanner ipmon[50]: 20:36:38.155242 fxp1 @0:16 b
207.18.92.26 -> 208.237.120.133 PR icmp len 20 84 icmp echo/0 IN
Feb 19 21:02:29 scanner ipmon[573]: 21:02:29.744779 fxp1 @0:16 b
207.18.92.26 -> 208.237.120.133 PR icmp len 20 84 icmp 8/0 IN
Michael Scheidell
SECNAP Network Security, LLC
Sales: 866-SECNAPNET / (1-866-732-6276)
Main: 561-368-9561 / www.secnap.net
------------------------------
End of stable-digest V5 #798
****************************
To Unsubscribe: send mail to majo...@FreeBSD.org
with unsubscribe freebsd-stable-digest in the body of the message