Re: Travel router part 2

[D'Arcy Cain]

On 2018-09-05 08:03 AM, Roy Marples wrote:
>> and have a named configured to use the forwarders in
>> /etc/namedb/forwarders.  Whatever the ISP dhcp gives me is stuffed into
>> the forwarders and used as last resort.  This has been a robust solution
>> for many open wireless access points.
> 
> Since NetBSD-6, dhclient-script has shipped with resolvconf(8) support
> that will do that for you.

Do you run dhclient if you have PPPOE set up?  Once the interface is up
I already have an IP address so what does dhclient do?

Note that the router is also the DHCP server so it also has static IPs
on the internal interfaces.  At some point I will add a second wifi card
to connect to campground wifi or tether to my phone.  At that point
dhclient will probably be set up to talk to that second wifi since it
will replace the PPPOE connection.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Travel router part 2

[D'Arcy Cain]

On 2018-09-02 01:08 PM, D'Arcy Cain wrote:
> On 2018-09-02 12:47 PM, Robert Swindells wrote:
>> % man pppoe
>>
>> Follow the instructions on what to put in /etc/ifconfig.pppoe0.
> 
> I guess so.  I have usually tried to keep everything in rc.conf but I
> suppose that that's better than manual intervention.
> 
> Thanks again.

So here is mu ifconfig.pppoe0 file

-- ifconfig.pppoe0 
create
! /sbin/ifconfig re0 up
! pppoectl -e re0 $int
! pppoectl $int myauthproto=pap 'xxx' 'myauthsecret=' hisauthproto=none
inet 0.0.0.0 netmask 0x
#! pppoectl $int max-auth-failure=1
up


When I restart the network I ger this.

  ifconfig: SIOCGIFAFLAG_IN: Can't assign requested address

but it works.  The man page actually says that the fifth line should be
this.

  inet 0.0.0.0 0.0.0.1 netmask 0x

but then this is what I get and it doesn't work.

[  404.125896] pppoe: connected to CMLDON7502W
[  404.295963] pppoe0: ipcp illegal conf-req in state initial
[  404.315971] pppoe0: ipcp illegal conf-req in state initial
[  404.335979] pppoe0: ipcp illegal conf-req in state initial
...

The same thing happens if I leave the line out altogether.  I guess I
will live with the error in the logs unless someone knows why it is
doing that.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Travel router part 2

[D'Arcy Cain]

On 2018-09-02 12:47 PM, Robert Swindells wrote:
> % man pppoe
> 
> Follow the instructions on what to put in /etc/ifconfig.pppoe0.

I guess so.  I have usually tried to keep everything in rc.conf but I
suppose that that's better than manual intervention.

Thanks again.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Travel router part 2

[D'Arcy Cain]

On 2018-09-02 12:10 PM, Robert Swindells wrote:
> 
> D'Arcy Cain  wrote:
>> On 2018-09-02 10:03 AM, Robert Swindells wrote:
>>>> The idea is to use re0 as my DSL connection and later to use athn1 in
>>>> the campgrounds as I travel.  Here is the relevant portion of my rc.conf.
>>>
>>> Maybe use wm0 as the external interface.
>>
>> What would that give me?
> 
> You trimmed the paragraph of my reply that described what it would give
> you.

Sorry, I thought that it was unrelated.  Are you saying that using wm0
would allow me to set the MTU but re0 doesn't?

> If you don't have an MTU of 1500 on pppoe0 you will have to deal with
> MSS-clamping for your internal networks.

# ifconfig pppoe0 mtu 1508
ifconfig: SIOCSIFMTU: Invalid argument

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Travel router part 2A - NAT

[D'Arcy Cain]

On 2018-09-02 10:11 AM, Robert Swindells wrote:
> 
> D'Arcy Cain  wrote:
>> On 2018-09-02 09:34 AM, D'Arcy Cain wrote:
>>> Some of you may remember that I was trying to set up a NetBSD machine to
>>> do all of my routing in my trailer while we travel.  I hit a snag when
>>
>> So, basic on my new setup is mostly working but I can't seem to get NAT
>> working at all.  I am trying to use NPF which I am not too familiar with
>> but the docs seem pretty clear.  Here is my npf.conf.  Can some NPF
>> experts check this out please.  Ta.
> 
> Try reloading the ruleset:
> 
> # npfctl reload
> 
> Before doing this, you could check on what rules it was using by
> doing:
> 
> # npfctl show
> 
> You may find that the npf startup script has been run before the
> pppoe(4) negociation had finished. This can mean that your $ext_if
> rule variable just contains 0.0.0.0 so the rules won't work.

That wasn't quite it but it did lead me to the fix.  I was mapping to
the re0 interface.  I needed to map to the pppoe0 interface.  When I did
"npfctl show" I saw "map re0 dynamic any -> 0.0.0.0 ...etc".  As soon as
I changed the ext_if variable to pppoe0 it worked.

Thanks.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Travel router part 2

[D'Arcy Cain]

On 2018-09-02 10:03 AM, Robert Swindells wrote:
>> The idea is to use re0 as my DSL connection and later to use athn1 in
>> the campgrounds as I travel.  Here is the relevant portion of my rc.conf.
> 
> Maybe use wm0 as the external interface.

What would that give me?  Or is it a "try and see if it works"
suggestion?  In any case, re0 works fine.  See below.

> In /etc/ppp/ip-up I have:
> 
> #! /bin/sh
> /sbin/route add default $5
> 
> /etc/ppp/ip-down is:
> 
> #! /bin/sh
> /sbin/route delete default $5

I did that and it worked.  Thanks.

Any thoughts on picking up the DNS servers?  It's not too bad because my
DHCP server can be modified as needed so it is only one location and in
any case I always include Google's public servers.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Travel router part 2A - NAT

[D'Arcy Cain]

On 2018-09-02 09:34 AM, D'Arcy Cain wrote:
> Some of you may remember that I was trying to set up a NetBSD machine to
> do all of my routing in my trailer while we travel.  I hit a snag when

So, basic on my new setup is mostly working but I can't seem to get NAT
working at all.  I am trying to use NPF which I am not too familiar with
but the docs seem pretty clear.  Here is my npf.conf.  Can some NPF
experts check this out please.  Ta.

-- npf.conf ---
$ext_if = inet4(re0)
$int_if = inet4(wm0)
$wifi_if = inet4(athn0)

table  type hash file "/etc/enemies.list"
table  type tree dynamic

$services_tcp = { http, https, smtp, domain, 6000, 9022 }
$services_udp = { domain, ntp, 6000 }
$localnet = { 192.168.215.0/24 }
$localwifi = { 192.168.225.0/24 }

alg "icmp"

map $ext_if dynamic $localnet -> $ext_if
map $wifi_if dynamic $localwifi -> $ext_if

group "external" on $ext_if {
pass stateful out final all

block in final from 
pass stateful in final family inet4 proto tcp to $ext_if port ssh
pass stateful in final proto tcp to $ext_if port $services_tcp
pass stateful in final proto udp to $ext_if port $services_udp
pass stateful in final proto udp to $ext_if port 33434-33600 #
Traceroute
}

group "internal" on $int_if {
block in all
block in final from 

# Ingress filtering as per BCP 38 / RFC 2827.
pass in final from $localnet
pass out final all
}

group "wifi" on $wifi_if {
block in all
block in final from 

# Ingress filtering as per BCP 38 / RFC 2827.
pass in final from $localwifi
pass out final all
}

group default {
pass in final on lo0 all
pass stateful out final to any
block in all
}
-

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: LVM tutorial and snapshots

[D'Arcy Cain]

On 2018-04-30 11:14 AM, Michael van Elst wrote:
>> "Write snapshot driver -only skeleton was written yet."
> 
> That's the problem. While a few bugs have been fixed, edges have
> been smoothed and userland tools have been updated, the mirror target
> (base for mirrors, snapshots and moving of LVs) is still missing.

So is it not ready for a production system then?  Is it solid except for
those items?

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

LVM tutorial and snapshots

[D'Arcy Cain]

https://wiki.netbsd.org/tutorials/how_to_use_lvm_on_netbsd/

I was looking at the above page and I wonder how out of date it is.
Other than changing CRLF to LF it hasn't been touched since it was
imported in 2011.

In particular, I wonder how accurate the following line is.

"Write snapshot driver -only skeleton was written yet."

It's the only TODO not marked "DONE" yet.

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Building tools on current

[D'Arcy Cain]

On 01/29/2018 04:05 PM, D'Arcy Cain wrote:
> So if I have rebuilt kernel, userland and pkgsrc multiple times since
> switching to LLVM can I safely remove those files?  I think I can see
> why one wouldn't want to blindly remove them but I am pretty sure that I
> don't have anything left that uses them.

Nope.  Removing them made all sorts of things fail.

FYI, I have gone back to a vanilla GCC system and tools, kernel,
distribution and all my packages built perfectly.  I think I will stay
away from LLVM for now.

-- 
D'Arcy J.M. Cain <da...@netbsd.org>
http://www.NetBSD.org/ IM:da...@vex.net

Re: Building tools on current

[D'Arcy Cain]

On 01/29/2018 03:57 PM, Joerg Sonnenberger wrote:
> On Mon, Jan 29, 2018 at 03:44:17PM -0600, D'Arcy Cain wrote:
>> So why are all those files still there?  If they have been deprecated or
>> moved, shouldn't postinstall have removed them?  What would happen if I
>> removed those as well as other old files in /bin, /usr/lib and /lib?  In
>> particular, why do I need gcc if I have converted to an LLVM system?
> 
> gcc is not marked obsolete just because yhou are building with LLVM.
> That is pretty intentional, actually.

So if I have rebuilt kernel, userland and pkgsrc multiple times since
switching to LLVM can I safely remove those files?  I think I can see
why one wouldn't want to blindly remove them but I am pretty sure that I
don't have anything left that uses them.

-- 
D'Arcy J.M. Cain <da...@netbsd.org>
http://www.NetBSD.org/ IM:da...@vex.net

Building tools on current

[D'Arcy Cain]

This started failing for me recently.  The culprit seems to be a missing
header.  I was able to build tools by commenting out the line "#include
" in /usr/include/gcc-5/x86intrin.h.  However, I am
running a system completely LLVM.  I think I have the same question here
as in another message on the pkgsrc list.  I will repeat it here.

I last built my world on Jan 24.  A tail of "ls -lt" in /usr/bin looks
like this:

-r-xr-xr-x  1 root  wheel  53976 Jan 24 14:36 dns-sd
-r-xr-xr-x  2 root  wheel1348568 Jun 24  2017 cc
-r-xr-xr-x  2 root  wheel1348568 Jun 24  2017 gcc
-r-xr-xr-x  2 root  wheel 898016 Jun 24  2017 c++
-r-xr-xr-x  2 root  wheel1348616 Jun 24  2017 cpp
-r-xr-xr-x  2 root  wheel 898016 Jun 24  2017 g++
-r-xr-xr-x  2 root  wheel1348616 Jun 24  2017 gcpp
-r-xr-xr-x  1 root  wheel 424904 Jun 24  2017 gcov
-r-xr-xr-x  1 root  wheel 41 Oct  2  2016 c11
-r-xr-xr-x  1 root  wheel 41 Jan 13  2016 c89
-r-xr-xr-x  1 root  wheel 41 Jan 13  2016 c99
-r-xr-xr-x  1 root  wheel   3197 Jan 13  2016 rcsfreeze
-r-xr-xr-x  1 root  wheel  28392 Jan 13  2016 texi2dvi
-r-xr-xr-x  3 root  wheel   2972 Oct 10  2015 zegrep
-r-xr-xr-x  3 root  wheel   2972 Oct 10  2015 zfgrep
-r-xr-xr-x  3 root  wheel   2972 Oct 10  2015 zgrep
-r-xr-xr-x  1 root  wheel   4187 Jan 28  2014 rump_wmd
-r-xr-xr-x  1 root  wheel   5011 Jan  6  2014 genassym
-r-xr-xr-x  2 root  wheel   2003 Dec  6  2013 zless
-r-xr-xr-x  2 root  wheel   2003 Dec  6  2013 zmore
-r-xr-xr-x  1 root  wheel   1248 Dec 25  2011 cleantags
-r-xr-xr-x  2 root  wheel   3186 Apr 14  2010 zcmp
-r-xr-xr-x  2 root  wheel   3186 Apr 14  2010 zdiff
-r-xr-xr-x  1 root  wheel   3293 Apr 27  2008 znew
-r-xr-xr-x  1 root  wheel   3946 Jan 14  2007 lorder
-r-xr-xr-x  1 root  wheel   2984 Jul  6  2005 spell
-r-xr-xr-x  1 root  wheel   2285 Jun 29  2005 shar
-r-xr-xr-x  1 root  wheel   4009 May  1  2004 gzexe
-r-xr-xr-x  1 root  wheel   1377 Dec 28  2003 zforce
-r-xr-xr-x  1 root  wheel   3962 Aug 18  2003 vgrind
-r-xr-xr-x  1 root  wheel869 Jul 30  2003 unifdefall
-r-xr-xr-x  1 root  wheel   2520 Mar 28  2003 lp
-r-xr-xr-x  1 root  wheel   1515 Jul 27  2000 skeyaudit
-r-xr-xr-x  1 root  wheel   1984 Mar 22  2000 pagesize
-r-xr-xr-x  1 root  wheel 33 Jul 30  1999 machine
-r-xr-xr-x  1 root  wheel 18 Dec 29  1994 false
-r-xr-xr-x  1 root  wheel 18 Dec 29  1994 true
-r-xr-xr-x  1 root  wheel   1922 Dec  7  1994 clear

So why are all those files still there?  If they have been deprecated or
moved, shouldn't postinstall have removed them?  What would happen if I
removed those as well as other old files in /bin, /usr/lib and /lib?  In
particular, why do I need gcc if I have converted to an LLVM system?

-- 
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: -current binary packages

[D'Arcy Cain]

On 06/28/17 15:23, Riccardo Mottola wrote:
I want install netbsd-current, but I cannot locate „-current“ pkgin 
and binary packages. I have „-current“ pkgsrc source packages, but no 
binary. Where are locate it?


Maybe RelEng ?

http://nycdn.netbsd.org/pub/NetBSD-daily/


Isn't that for base?  Radim wants packages.

--
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: -current binary packages

[D'Arcy Cain]

On 06/28/17 06:25, Radim Horáček wrote:
I want install netbsd-current, but I cannot locate „-current“ pkgin and 
binary packages. I have „-current“ pkgsrc source packages, but no 
binary. Where are locate it?


ftp://ftp.netbsd.org/pub/pkgsrc/packages/NetBSD/

However the last build seems to be for over a month ago, at least for 
amd64.  They should work with -current but may give warnings that it 
wasn't built for your platform.


I cannot get everything to build on my -current system.  Too many 
packages die due to segfault.  Perhaps that is why no new builds are 
being done.


Has anyone been able to build the latest pkgsrc on the latest -current? 
In particular GTK based packages such as xfce4 and libreoffice.  One of 
the failing dependencies for me is sysutils/libnotify.  It builds fine 
on 7.1.0_PATCH.


--
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Seg Faults building devel/gobject-inspection with latest 8.0_BETA

[D'Arcy Cain]

On 06/22/17 17:19, Joerg Sonnenberger wrote:

On Wed, Jun 21, 2017 at 06:25:26PM -0400, D'Arcy Cain wrote:

On 06/21/17 17:02, D'Arcy Cain wrote:

On 06/21/17 11:04, D'Arcy Cain wrote:

#0  0x70e05a605663 in ?? ()
#1  0x70e05a200585 in ?? () from /usr/pkg/lib/libgthread-2.0.so.0
#2  0x70e0665e9b60 in ?? ()
#3  0x70e05a200669 in _fini () from /usr/pkg/lib/libgthread-2.0.so.0
#4  0x in ?? ()

I am deleting glib2 and dependents and trying again.


Same thing.  Rebuilding everything with symbols to further debug.


Well, I half expected this.  I built with symbols and it didn't dump core.
Now what?


Can you check if libgthread exports any symbols from libgcc? I wonder
if that is where the problem comes from...


You're a bit above my pay grade.  I understand what you are suggesting 
but I am not sure how to check.


By the way, I added this to my mk,conf:

MKSTRIPSYM=no
MKDEBUG=YES
INSTALL_UNSTRIPPED=YES

But I still get the exact same backtrace from gdb.

--
D'Arcy J.M. Cain <da...@netbsd.org>
http://www.NetBSD.org/ IM:da...@vex.net

Re: Seg Faults building devel/gobject-inspection with latest 8.0_BETA

[D'Arcy Cain]

On 06/21/17 17:02, D'Arcy Cain wrote:

On 06/21/17 11:04, D'Arcy Cain wrote:

#0  0x70e05a605663 in ?? ()
#1  0x70e05a200585 in ?? () from /usr/pkg/lib/libgthread-2.0.so.0
#2  0x70e0665e9b60 in ?? ()
#3  0x70e05a200669 in _fini () from /usr/pkg/lib/libgthread-2.0.so.0
#4  0x in ?? ()

I am deleting glib2 and dependents and trying again.


Same thing.  Rebuilding everything with symbols to further debug.


Well, I half expected this.  I built with symbols and it didn't dump 
core.  Now what?


--
D'Arcy J.M. Cain <da...@netbsd.org>
http://www.NetBSD.org/ IM:da...@vex.net

Re: Seg Faults building devel/gobject-inspection with latest 8.0_BETA

[D'Arcy Cain]

On 06/21/17 11:04, D'Arcy Cain wrote:

#0  0x70e05a605663 in ?? ()
#1  0x70e05a200585 in ?? () from /usr/pkg/lib/libgthread-2.0.so.0
#2  0x70e0665e9b60 in ?? ()
#3  0x70e05a200669 in _fini () from /usr/pkg/lib/libgthread-2.0.so.0
#4  0x in ?? ()

I am deleting glib2 and dependents and trying again.


Same thing.  Rebuilding everything with symbols to further debug.

--
D'Arcy J.M. Cain <da...@netbsd.org>
http://www.NetBSD.org/ IM:da...@vex.net

Re: Seg Faults building devel/gobject-inspection with latest 8.0_BETA

[D'Arcy Cain]

On 06/20/17 12:44, Thomas Klausner wrote:

I'm using python-3.6 by default (instead of python-2.7) and don't see
the gio segfaults.


Well, I don't understand then.  I am seeing it on two systems, current 
and 7.1.0_PATCH using Python 2.7.13, 3.6.1 and 3.5.3.  It's always a 
Python core file that gets created.  Here's a sample.


#0  0x70e05a605663 in ?? ()
#1  0x70e05a200585 in ?? () from /usr/pkg/lib/libgthread-2.0.so.0
#2  0x70e0665e9b60 in ?? ()
#3  0x70e05a200669 in _fini () from /usr/pkg/lib/libgthread-2.0.so.0
#4  0x in ?? ()

I am deleting glib2 and dependents and trying again.

--
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net

Re: Seg Faults building devel/gobject-inspection with latest 8.0_BETA

[D'Arcy Cain]

On 06/20/17 12:44, Thomas Klausner wrote:

I'm using python-3.6 by default (instead of python-2.7) and don't see
the gio segfaults.


3.6.?

--
D'Arcy J.M. Cain 
http://www.NetBSD.org/ IM:da...@vex.net