This is the log as captured by an IRC bot in the channel. The statements
are those of the individual people and might not neccessarily reflect the
policy and legal rules as set forth by the T2 SDE Project.
--- Log opened Fri Apr 11 00:00:57 2008
00:46 -!- hwinkel [email@example.com] has joined #t2
01:38 -!- hwinkel [firstname.lastname@example.org] has quit [Read error: 110 (Connection timed out)]
03:29 -!- jeffdamet1 [email@example.com] has left #t2 
05:24 -!- Enqlave [n=stealth@sourcemage/guru/stealth] has quit ["Dont wait to die to find paradise..."]
07:27 -!- hwinkel [firstname.lastname@example.org] has joined #t2
07:40 < rxr> moin
08:42 < CIA-4> rene * r28671 /trunk/package/gnome2/libgda/libgda.desc: * updated libgda (3.0.3 -> 3.0.4)
08:42 < CIA-4> rene * r28672 /trunk/package/network/dbus/dbus.desc: * updated dbus (1.0.3 -> 1.2.1)
08:46 < CIA-4> rene * r28673 /trunk/package/base/lvm2/lvm2.desc: * updated lvm2 (2.02.33 -> 2.02.34)
08:46 < CIA-4> rene * r28674 /trunk/package/x11/mesa/mesa.desc: * updated mesa (7.0.2 -> 7.0.3)
08:47 < CIA-4> rene * r28675 /trunk/package/x11/mesa-demos/mesa-demos.desc: * updated mesa-demos (7.0.2 -> 7.0.3)
08:47 < CIA-4> rene * r28676 /trunk/package/network/libidn/libidn.desc: * updated libidn (1.6 -> 1.7)
09:19 -!- |beowulf| [i=1014@sourcemage/mage/beowulf] has quit [Remote closed the connection]
09:20 -!- |beowulf| [email@example.com] has joined #t2
09:46 < CIA-4> rene * r28677 /trunk/package/security/strongswan/strongswan.desc: * updated strongswan (4.1.11 -> 4.2.0)
09:59 -!- hwinkel [firstname.lastname@example.org] has quit [Read error: 110 (Connection timed out)]
10:04 -!- hwinkel [email@example.com] has joined #t2
11:14 -!- hwinkel [firstname.lastname@example.org] has quit [Read error: 110 (Connection timed out)]
11:58 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: sparc-kly_, LMJ, __init__, hoss10123, Ragnar|away, @ChanServ, dsoul, valentin, TobiX, mqueiros_, (+2 more, use /NETSPLIT to show all of them)
11:58 -!- Enqlave [email@example.com] has joined #t2
11:58 -!- Enqlave [firstname.lastname@example.org] has quit [Read error: 104 (Connection reset by peer)]
11:58 -!- Enqlave [email@example.com] has joined #t2
11:58 -!- Enqlave [firstname.lastname@example.org] has quit [Client Quit]
11:59 -!- Enqlave [email@example.com] has joined #t2
11:59 -!- Netsplit over, joins: ChanServ
11:59 -!- bfg [firstname.lastname@example.org] has joined #t2
11:59 -!- Netsplit over, joins: mqueiros_, hoss10123, valentin, Ragnar|away, sparc-kly_, Dallur, LMJ, __init__, dogmeat, TobiX (+1 more)
11:59 -!- ServerMode/#t2 [+o ChanServ] by irc.freenode.net
11:59 -!- Enqlave [email@example.com] has quit [Client Quit]
11:59 -!- CIA-53 [n=CIA@22.214.171.124] has joined #t2
12:00 -!- Enqlave [n=stealth@sourcemage/guru/stealth] has joined #t2
12:02 < CIA-4> gweis * r28678 /trunk/package/multimedia/ffmpeg/ffmpeg.conf: * fixed configure parameters for xiv, x264, faac and faad
12:03 -!- CIA-4 [n=CIA@126.96.36.199] has quit 
12:31 < |beowulf|> INIT: version 2.86 booting
12:31 < |beowulf|> INIT: Entering runlevel: 2
12:31 < |beowulf|> ** running rc script **
12:31 < |beowulf|> RC: Previous runlevel: N, new runlevel: 2.
12:31 < |beowulf|> RC: Enter runlevel 2 ...
12:31 < |beowulf|> Mounting /dev/shm and /dev/pts. v
12:32 < |beowulf|> mkdir: cannot create directory `/dev/shm': Read-only file system
12:32 < |beowulf|> mount: mount point /dev/shm does not exist
12:32 < |beowulf|> mkdir: cannot create directory `/dev/pts': Read-only file system :-(
12:32 < |beowulf|> Configuring the /dev filesystem ... v
12:32 < |beowulf|> ln: creating symbolic link `mouse': Read-only file system
12:32 < |beowulf|> rm: cannot remove `fd': Read-only file system
12:32 < |beowulf|> ln: creating symbolic link `core': Read-only file system
12:32 < |beowulf|> ln: cannot remove `fd': Read-only file system
12:32 < |beowulf|> ln: creating symbolic link `stdin': Read-only file system
12:32 < |beowulf|> ln: creating symbolic link `stdout': Read-only file system
12:32 < |beowulf|> ln: creating symbolic link `stderr': Read-only file system
12:32 < |beowulf|> is this supposed to happen?
13:31 < tfl0pz> moin
13:31 -!- Enqlave [n=stealth@sourcemage/guru/stealth] has quit ["Dont wait to die to find paradise..."]
13:35 -!- bfg [firstname.lastname@example.org] has quit ["Kopete 0.11.1 : http://kopete.kde.org"]
13:35 < tfl0pz> is t2 using /var/run/sendsigs.omit?
13:43 -!- bfg [email@example.com] has joined #t2
15:06 < |beowulf|> rxr: are you there?
16:26 < CIA-53> rene * r28679 /trunk/package/gnome2/tomboy/tomboy.desc: * updated tomboy (0.10.0 -> 0.10.1)
16:26 < CIA-53> rene * r28680 /trunk/package/network/tntnet/tntnet.desc: * updated tntnet (188.8.131.52 -> 184.108.40.206)
16:27 < rxr> re
16:27 < rxr> |beowulf|: I guess you use no initrd ?
16:34 < |beowulf|> no initrd, no
16:34 < |beowulf|> the kernel doesn't even have modules
16:38 < rxr> you either need to have some tmpfs entry for /dev in your fstab for the device notes of udev to be creatable ...
16:38 < rxr> I guess 99% of us right now use initrd populating a /dev tmpfs before the main syste mboots ...
16:44 < |beowulf|> dev /dev tmpfs defaults 0 0 <-- will that line in fstab do?
16:45 < rxr> you had it already?
16:45 < |beowulf|> no, just created it...
16:46 < rxr> I just looked at the system init script
16:46 < rxr> I fear it might not yet work anyway
16:47 < rxr> the udevd start woudl need to be moved on top with an explicit /dev mount ...
16:47 < |beowulf|> any reason why it isn't at the top already?
16:48 < rxr> because the deault initrd sets it up for root device mounting with modular kernels
16:48 < rxr> and we just restart udevd from the system then after the filesystem mount and check
16:48 < rxr> guess noone cared so far due fully modular use with initrd these days
17:29 -!- Baldzius [n=Baldzius@220.127.116.11] has joined #t2
17:29 < Baldzius> moin
17:39 < rxr> h ib
17:41 < rxr> hi Baldzius even
17:41 -!- Enqlave [n=stealth@sourcemage/guru/stealth] has joined #t2
17:42 < Baldzius> hi rxr
19:37 -!- hoss10123 [firstname.lastname@example.org] has quit ["peace out"]
19:39 < |beowulf|> rxr: is it sufficient to just call 'mkinitrd 2.6.23-rc3' to build the initrd?
19:48 < rxr> should be
20:11 < |beowulf|> rxr: http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.1.log
<-- any idea what's going wrong here?
20:15 < rxr> probably mostly the not setup udev causing tons of trailing issues
20:15 < rxr> you have not yet got the udev thing work arounded, hm ?
20:17 < |beowulf|> I created an initrd and hoped that would solve it -- but I'm not even sure it's being used
20:19 < rxr> you have not specified it in your boot loader thingy ?
20:19 < rxr> anyway
20:19 < rxr> just leave that alone then
20:19 < rxr> try that:
20:19 < rxr> http://people.exactcode.de/~rene/t2-no-initrd.patch
20:20 * rxr crossing fingers
20:22 < |beowulf|> hm, I don't think I can specify it in the bootloader
20:22 < |beowulf|> at least not so that it is persistent
20:23 < rxr> ok - then try that patch :-)
20:31 < |beowulf|> rxr: http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.2.log
<-- I think /proc isn't mounted yet at that time
20:39 < rxr> is your root on an MD?
20:42 < rxr> ah - sorry the root is mounted at that time
20:42 < |beowulf|> root is an md, yes
20:43 < rxr> the first stray errors are harmless
20:44 < rxr> the proc due to hotplug execution disabling (due using netlink with udev)
20:44 < rxr> and the killall likewise, udevd is not running yet anyway (as oppost to comming form an initrd)
20:44 < rxr> the dev/console stuff is probably just from the logging
20:45 < rxr> can you try to use a device-node identifer for the rootfs in the fstab instead the UUID ?
20:45 < |beowulf|> I'll try...
20:45 < rxr> also a cat /proc/mounts
20:46 < rxr> after the udevd start and the /dev/shm and pts mounts ...
20:46 < |beowulf|> you mean in the system script?
20:46 < rxr> yes
20:53 < |beowulf|> rxr: cat: /proc/mounts: No such file or directory
20:54 < rxr> ah - damn
20:54 < rxr> I guess we uregently should re-improve non initrd support
20:54 < rxr> + mount -n /sys
20:54 < rxr> + mount -n /proc
20:54 < rxr> before mount -n /dev
20:54 < rxr> sorry ...
20:55 < rxr> in the good old days T2 was pure non-initrd (initrd what the heck do we want with that)
20:55 * |beowulf| never used an initrd before today
20:55 < rxr> now that we have fully automatic, and fully modular initrd support we just ignored non-initrd support majorly ...
20:55 < rxr> sorry
20:55 < rxr> I'll fix that in trunk when we got your cube going .-)
20:56 < rxr> btw. /me waiting for my second Apple Cube ...
20:56 < rxr> with replaced CPU (1.2 GHz instead of the original 400-450) and ATi Radeon instead of the original R128
20:57 < rxr> actually I only was seking for a OpenFirmware, Cube compatible Radeon
20:57 < rxr> but as they are that rare, I decided to get a whole cube with it .-)
20:57 < |beowulf|> what do you use it for?
20:57 < rxr> desktop computer
20:57 < rxr> daily work, edit compile it is enough
20:57 < rxr> and for more advanced stuf fI can log into oru central 8-way anyway ...
20:58 < rxr> I just love the cube design ... :-)
20:58 < rxr> (and PPC CPU, too)
20:58 < rxr> but I do not yet ahve the 2nd, "tuned" one, might arrive tomorrow ...
20:59 < |beowulf|> hm, I don't think I have a line that does mount -n /dev
20:59 < rxr> the patch I posted before should had it ,no ?
20:59 < |beowulf|> oh, sorry, wrong system :)
21:00 < |beowulf|> anything else I should add?
21:00 < rxr> maybe another new error reminds me about something missing .-)
21:04 < |beowulf|> http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.3.log
21:06 < rxr> can you add an ls /dev/md* ; ls /dev
21:06 < rxr> to the cat /proc/mounts for debug hints ?
21:10 < rxr> I
21:10 < rxr> m just off for dinner
21:13 < |beowulf|> http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.4.log
22:09 < rxr> can you try (sorry for the try'N error) adding a udevsettle after the udevtrigger ?
22:09 < rxr> btw. you have udev on the box right?
22:09 < rxr> is that some generic build or some custom build ?
22:25 < |beowulf|> it's generic
22:32 < |beowulf|> rxr: http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.5.log
22:33 < rxr> hm - btw. can it be you build your kernel without sysfs support ?
22:33 < rxr> mount: can't find /sys in /etc/fstab or /etc/mtab
22:33 < rxr> oh
22:34 < rxr> fstab:
22:34 < rxr> none /sys sysfs rw 0 0
22:37 < |beowulf|> http://www.hawo.stw.uni-erlangen.de/~siflfran/fstab
<-- anything else missing?
22:39 < rxr> I wonder why my fstab has "rw" for sysfs, strange
22:39 < rxr> I think that could be just "defaults"
22:39 < rxr> I think the fstab should be good now
22:39 < rxr> when I merge that to trunk I'll open code it to explicitly specify the arguemnts to mount anyway, so it will work regardless of missing fstab entires
22:41 < |beowulf|> anything else I should change before I put the harddrive into the qube? :)
22:41 < rxr> maybe I should have locally tested that in Qemu ?
22:52 < |beowulf|> hm, it's in runlevel 3 now but it didn't start a getty on the serial port
22:52 < |beowulf|> and it didn't start networking either
22:53 < rxr> did you enable some serial getty ?
22:53 < |beowulf|> I did
22:54 < rxr> network is configured ?
22:54 < |beowulf|> http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.6.log
22:54 < |beowulf|> I'll take a look at that now
22:55 < rxr> Setting up network devices. v
22:55 < rxr> /sbin/rocknet: line 121: /dev/fd/62: No such file or directory
22:55 < rxr> /sbin/rocknet: line 131: /dev/fd/62: No such file or directory
22:55 < rxr> put this after your custom mount /dev in the init.d/system:
22:55 < rxr> ln -s /proc/self/fd /dev/fd
22:58 < |beowulf|> anything else?
22:58 < rxr> sorry, my last initrd less system was like 5y ago :-(
22:58 < rxr> I wissh I could forsee issues
22:58 < |beowulf|> :)
22:59 < rxr> that should help some scripts liek the netconf ...
22:59 < rxr> but why your serial terminal is not working I did not see ...
22:59 < rxr> you are sure to add it to the /etc/inittab ?
23:01 < |beowulf|> T0:23:respawn:/sbin/agetty -L -f /etc/issue 115200 ttyS0 vt220
23:01 -!- xaw [n=xaw@unaffiliated/xaw] has joined #t2
23:01 [Users #t2]
23:01 [@ChanServ] [ bfg ] [ dsoul ] [ mqueiros_ ] [ sparc-kly_] [ xaw ]
23:01 [ __init__] [ CIA-53 ] [ elias` ] [ mtr ] [ tfl0pz ] [ |beowulf|]
23:01 [ axionix ] [ Dallur ] [ Enqlave] [ Ragnar|away] [ TobiX ]
23:01 [ Baldzius] [ dogmeat] [ LMJ ] [ rxr ] [ valentin ]
23:01 -!- Irssi: #t2: Total of 22 nicks [1 ops, 0 halfops, 0 voices, 21 normal]
23:08 < |beowulf|> rxr: http://www.hawo.stw.uni-erlangen.de/~siflfran/qube.7.log
23:08 < |beowulf|> didn't make much of a difference
23:08 < |beowulf|> except that now there is /dev/fd
23:09 < rxr> mh
23:18 < |beowulf|> why doesn't the prng get seeded?
23:19 < rxr> it should
23:19 < rxr> I would have to review, maybe it is only seeded after the first successful shutdown when some var/seed state is there and you's is lacking it due to a hand install or so
23:24 < |beowulf|> hmmm, the rocknet code looks like a lot of black magic...
23:24 < |beowulf|> no idea what it does with /dev/fd/62
23:26 < rxr> nothing
23:26 < rxr> that is from the shell for job control
23:26 < rxr> job pipe whatever
23:27 < rxr> actually I enver took a closer look
23:27 < rxr> that is what modern linux distributions need this /proc/self -> /dev/fd link for
23:27 < rxr> maybe something is still wrong with the proc mount
23:27 < rxr> or the symlink is not correct ?
23:28 < rxr> Baldzius: I can offer you to help in form of fixing the system init in some VM tomorrow when an article is done I have to finish soonish
23:29 < Baldzius> rxr: huh?
23:31 < rxr> |beowulf|, sorry Baldzius
23:31 < rxr> evil tab completion while looking at another window ...
23:31 < |beowulf|> ok, thank you :)
23:31 < Baldzius> :)
23:49 -!- bfg [email@example.com] has left #t2 
--- Log closed Sat Apr 12 00:00:58 2008