Rootstock fails during Stage 2

GoogleGroups
1

Im trying to build a lucid image for my beagleboard. After a lot of
failing right after rootstock starts stage 2 i jumped on irc and rcn-
ee told me that i had to get the newest rootstock using bzr (Im
running ubuntu 9.10 on my dev machine). So i did and it worked.
Started stage 2 and began grabbing packages. I'll post what was output
from when the last package was downloaded until it just stops:

Get:392 http://ports.ubuntu.com lucid/main libgl1-mesa-dri
7.7-3ubuntu1 [1819kB]
Fetched 133MB in 3min 14s (683kB/
s)
Extracting templates from packages: 100%
Preconfiguring packages ...
modprobe: FATAL: Could not load /lib/modules/2.6.32-14-versatile/
modules.dep: No such file or directory

modprobe: FATAL: Could not load /lib/modules/2.6.32-14-versatile/
modules.dep: No such file or directory

Selecting previously deselected package libfuse2.
(Reading database ... 11089 files and directories currently
installed.)
Unpacking libfuse2 (from .../libfuse2_2.8.1-1.1ubuntu2_armel.deb) ...
Selecting previously deselected package fuse-utils.
Unpacking fuse-utils (from .../fuse-
utils_2.8.1-1.1ubuntu2_armel.deb) ...
Selecting previously deselected package libntfs-3g54.
Unpacking libntfs-3g54 (from .../
libntfs-3g54_1%3a2009.4.4-1ubuntu5_armel.deb) ...
Setting up libfuse2 (2.8.1-1.1ubuntu2) ...

Setting up fuse-utils (2.8.1-1.1ubuntu2) ...
creating fuse group...
udev active, skipping device node creation.
reload: Unable to connect to Upstart: Failed to connect to socket /com/
ubuntu/upstart: Connection refused
update-initramfs: deferring update (trigger activated)

Processing triggers for libc-bin ...
ldconfig deferred processing now taking place
Processing triggers for initramfs-tools ...
Selecting previously deselected package ntfs-3g.
(Reading database ... 11127 files and directories currently
installed.)
Unpacking ntfs-3g (from .../
ntfs-3g_1%3a2009.4.4-1ubuntu5_armel.deb) ...
Selecting previously deselected package x11-common.
Unpacking x11-common (from .../x11-
common_1%3a7.5+1ubuntu8_all.deb) ...
Selecting previously deselected package libexpat1.
Unpacking libexpat1 (from .../libexpat1_2.0.1-7ubuntu1_armel.deb) ...
Selecting previously deselected package libfreetype6.
Unpacking libfreetype6 (from .../
libfreetype6_2.3.11-1ubuntu2_armel.deb) ...
Selecting previously deselected package ttf-dejavu-core.
Unpacking ttf-dejavu-core (from .../ttf-dejavu-
core_2.30-2_all.deb) ...

I do see now that there is a modprobe problem and some other stuff in
there.
I've tried running rootstock 5-6 times now i think and it usually
stops during the dejavu-core unpacking.
It has also stopped 2 times at usbutils. And it stops without any
error messages. I have to ctrl+C to exit and then it cleans up.

This is what im using to run rootstock:

sudo /home/infected/tempbzr/project-rootstock/rootstock --fqdn
beagleboard --login ubuntu --password temppwd --imagesize 2G --seed
wget,nano,linux-firmware,wireless-tools,usbutils,lxde,gdm,build-
essential --dist lucid --serial ttyS2 --script fixup.sh --kernel-image
http://rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.7-l7.0/linux-image-2.6.32.7-l7.0_1.0lucid_armel.deb

Anyone know what im doing wrong?

2

Tor Stokka <infected.sith@gmail.com> [2010-02-25 00:40:31]:

Started stage 2 and began grabbing packages. I'll post what was output
from when the last package was downloaded until it just stops:

You can modify rootstock script, for example you can add "set +x" somewhere
to see what's doing and why it fails.

modprobe: FATAL: Could not load /lib/modules/2.6.32-14-versatile/
modules.dep: No such file or directory

Maybe you can try to use this kernel:

http://ports.ubuntu.com/ubuntu-ports/pool/main/l/linux/linux-image-2.6.32-14-versatile_2.6.32-14.20_armel.deb

I've tried running rootstock 5-6 times now i think and it usually
stops during the dejavu-core unpacking.
It has also stopped 2 times at usbutils. And it stops without any
error messages. I have to ctrl+C to exit and then it cleans up.

If you try to build it with minimal seed, does it works? Can you try with
karmic instead of lucid, does it works?

-- ynezz

3

Im trying to build a lucid image for my beagleboard. After a lot of
failing right after rootstock starts stage 2 i jumped on irc and rcn-
ee told me that i had to get the newest rootstock using bzr (Im
running ubuntu 9.10 on my dev machine). So i did and it worked.
Started stage 2 and began grabbing packages. I'll post what was output
from when the last package was downloaded until it just stops:

Get:392 http://ports.ubuntu.com lucid/main libgl1-mesa-dri
7.7-3ubuntu1 [1819kB]
Fetched 133MB in 3min 14s (683kB/
s)
Extracting templates from packages: 100%

Preconfiguring packages ...
modprobe: FATAL: Could not load /lib/modules/2.6.32-14-versatile/
modules.dep: No such file or directory

These are safe, as qemu is using a zImage with no modules...

I do see now that there is a modprobe problem and some other stuff in
there.
I've tried running rootstock 5-6 times now i think and it usually
stops during the dejavu-core unpacking.
It has also stopped 2 times at usbutils. And it stops without any
error messages. I have to ctrl+C to exit and then it cleans up.

This is what im using to run rootstock:

sudo /home/infected/tempbzr/project-rootstock/rootstock --fqdn
beagleboard --login ubuntu --password temppwd --imagesize 2G --seed
wget,nano,linux-firmware,wireless-tools,usbutils,lxde,gdm,build-
essential --dist lucid --serial ttyS2 --script fixup.sh --kernel-image
http://rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.7-l7.0/linux-image-2.6.32.7-l7.0_1.0lucid_armel.deb

drop the '--login ubuntu --password temppwd' part... It's kinda buggy
in bzr trunk right now, and it's going to be replaced... On first boot
it'll create a new user/password thru ubuntu's oem-config gui...

lxde brings in a lot of packages, bump the --imagesize from 2G to
4G... lxde actually brings in ALOT of gnome, xfce4 is basically
lighter with lucid...

Regards,

4

Tried what you suggested, still the same :confused:
BUT now i checked the system monitor to see that qemu is running one
of my CPU cores at 100% while just sitting there apparently locked up.
So i think i need another version of qemu. im currently using whatever
version is in the repo

5

Something's fishy... I just successfully completed a bzr rootstock
(rev 82) run in preparation for an alpha-3 rootfs dump to rcn-ee.net
tomorrow.

Using Karmic as the base os on the x86 machine.. (old celeron, so it's 32 bit..)

sudo ./rootstock --fqdn beagleboard --imagesize 2G --seed
wget,nano,linux-firmware,wireless-tools,usbutils \
--dist lucid --serial ttyS2 --script fixup.sh \
--kernel-image http://rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.9-l9.1/linux-image-2.6.32.9-l9.1_1.0lucid_armel.deb

I would drop your 'lxde' option form the seed, anything that gives
qemu a reason to run longer to install more packages is a bad idea in
my mind.. In my experience, running apt-get on the beagle later works
better for me...

Regards,

6

i removed lxde and gdm from the seed and it worked. Now to get this
thing up and running
Thanks!

7

Robert Nelson escreveu:

I would drop your 'lxde' option form the seed, anything that gives
qemu a reason to run longer to install more packages is a bad idea in
my mind.. In my experience, running apt-get on the beagle later works
better for me...

I have to agree. I usually install just the basic and wireless tools (or
not even that as I have a usb-ethernet adapter that works fine) and once
I have network, it's apt-get all the way.
Next step will be a way to automate the instalation of the proprietary
bits (sgx, codecs, etc).

[]s

Adilson.

8

My current plan of attack is to get the ubuntu guys to pull in an omap
defconfig (one config for all omap3/4 boards,
beagle/overo/igep2/etc..) for 10.10, then we can use a dkms build
script to build the sgx modules from the TI SDK...

For 10.04 (2.6.32), there was just too many upstream patches needed
and the ubuntu arm developers where too busy with freescale/marvel
hardware..

Regards,

9

sigh...
Now i get this while booting:
init: ureadahead main process (40) terminated with status 5
init: plymouth main process (41) terminated with status 71
mountall: Could not connect to Plymouth
init: procps main process (62) terminated with status 255
fsck from util-linux-ng 2.17
rootfs: clean, 23264/79200 files, 127223/313267 blocks
init: plymouth-splash pre-start process (186) terminated with status
111
[ 26.496032] phy0 -> rt2500usb_init_eeprom: Error - Invalid RT
chipset detected.
[ 26.509521] phy0 -> rt2x00lib_probe_dev: Error - Failed to allocate
device.
init: plymouth-log main process (229) terminated with status 111

Tried with 2 different kernels. The first one this not have the errors
with rt2500 other than that the error messages are the same.

Kernels:
http://www.rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.7-l7.0/linux-image-2.6.32.7-l7.0_1.0lucid_armel.deb
http://www.rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.9-l9.1/linux-image-2.6.32.9-l9.1_1.0lucid_armel.deb

10

Robert Nelson escreveu:

My current plan of attack is to get the ubuntu guys to pull in an omap
defconfig (one config for all omap3/4 boards,
beagle/overo/igep2/etc..) for 10.10, then we can use a dkms build
script to build the sgx modules from the TI SDK...

For 10.04 (2.6.32), there was just too many upstream patches needed
and the ubuntu arm developers where too busy with freescale/marvel
hardware..

That's my idea as well. As a matter of fact, I do work for Canonical and
was the first person to be hired to work on the mobile team. I don't
work on this team any longer but do know all the people there. Just to
be clear, my interest on the beagleboard is personal, it has nothing to
do with Canonical official work.

[]s

Adilson.

11

Remember, this is only 'alpha-3' so we have time to fix the bugs...
(10.04 = april...)

sigh...
Now i get this while booting:
init: ureadahead main process (40) terminated with status 5

not sure...

init: plymouth main process (41) terminated with status 71
mountall: Could not connect to Plymouth

plymouth needs kms/framebuffer and initramfs i believe...

early patch, i'm messing with:
http://bazaar.launchpad.net/~robertcnelson/project-rootstock/initramfs/revision/57
and boot.scr needs to be updated (simple) to also load the initramfs
into ram..

init: procps main process (62) terminated with status 255

not sure what it's for, same warning was in karmic...

fsck from util-linux-ng 2.17
rootfs: clean, 23264/79200 files, 127223/313267 blocks
init: plymouth-splash pre-start process (186) terminated with status
111
[ 26.496032] phy0 -> rt2500usb_init_eeprom: Error - Invalid RT
chipset detected.
[ 26.509521] phy0 -> rt2x00lib_probe_dev: Error - Failed to allocate
device.

what's your device's vip/pid/name.. do you have firmware installed...

note: i still haven't completely sync'd my defconfig with ubuntu's, so
wifi devices don't work exactly the same as in x86 lucid..

init: plymouth-log main process (229) terminated with status 111

Tried with 2 different kernels. The first one this not have the errors
with rt2500 other than that the error messages are the same.

Kernels:
http://www.rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.7-l7.0/linux-image-2.6.32.7-l7.0_1.0lucid_armel.deb
http://www.rcn-ee.net/deb/kernel/beagle/lucid/v2.6.32.9-l9.1/linux-image-2.6.32.9-l9.1_1.0lucid_armel.deb

Regards,

12

I tried the alpha 3 you uploaded, still the same errors while booting.
What do i need to add in the boot.scr to load initramfs into ram?