u-boot-linaro-omap4-panda replaces u-boot-omap4, please remove source and binary packages of the latter from the archive

Bug #613230 reported by Tobin Davis
16
This bug affects 2 people
Affects Status Importance Assigned to Milestone
u-boot-linaro (Ubuntu)
Fix Released
Critical
Unassigned
Maverick
Fix Released
Critical
Unassigned
u-boot-omap4 (Ubuntu)
Fix Released
Undecided
Unassigned
Maverick
Fix Released
Undecided
Unassigned

Bug Description

Binary package hint: u-boot-omap4

After installing a new kernel on omap4, u-boot fails to boot. Flash-kernel updates the boot partition properly, and moves the old kernel to uImage.bak & uInitrd.bak appropriately. The following is the serial output from rebooting after installing a new kernel and running flash-kernel to update the boot partition:

Texas Instruments X-Loader 1.41 (Jun 24 2010 - 13:29:27)
mmc read: Invalid size
Starting OS Bootloader from MMC/SD1 ...

U-Boot 1.1.4 (Jun 24 2010 - 13:11:03)

Load address: 0x80e80000
DRAM: 512 MB
Flash: 0 kB
*** Warning - bad CRC, using default environment

In: serial
Out: serial
Err: serial
Net: KS8851SNL
Hit any key to stop autoboot: 0
mmc read: Invalid size
mmc read: Invalid size

319 bytes read
## Executing script at 80300000

0 bytes read

0 bytes read
## Booting image at 80000000 ...
   Image Name: Ubuntu Kernel
   Image Type: ARM Linux Kernel Image (uncompressed)
   Data Size: 3465280 Bytes = 3.3 MB
   Load Address: 80008000
   Entry Point: 80008000
   Verifying Checksum ... OK
OK
## Loading Ramdisk Image at 81600000 ...
Bad Magic Number
X-Loader hangs

Using an x86 system to backup the fat partition, reformat, and restore the files works.

Another interesting test is to mount the fat partition after the above step, and rename the uImage & uIntrd files to .old, then renaming the .bak files to make them default. u-boot will load the boot.scr script, which should in turn load the renamed uImage and uInitrd, but the .old files get loaded instead.

Sync was executed after each write to SD to ensure the filesystem was updated.

(updated with correct log output)

Tags: armel
Tobin Davis (gruemaster)
description: updated
tags: added: armel
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

For reference, I believe we're using u-boot from http://gitorious.org/pandaboard/u-boot.

There are quite many others u-boot trees around (omapzoom, sakoman) so we should probably try to use one of these trees for the future.

Revision history for this message
Oliver Grawert (ogra) wrote :

thats exactly the plan :) there is an upstreaming branch for the panda that unifies omap3 and 4 in the same (recent) u-boot source, i was planning to test the code between A3 and FF ... feel free to play with it :)

http://omappedia.org/wiki/U-boot_Upstreaming_Project

Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Currently we have two important u-boot trees, one for omap 4 and another for omap 3.

For omap 4 sakoman is improving and pushing the patches upstream already, and I saw that currently we have a very good support upstream (most of sakoman's patches were merged yesterday).

For omap 3 the best tree is still the beagleboard-validation one, at http://gitorious.org/beagleboard-validation/u-boot/. Probably in the future there's going to be some work trying to push the patches upstream, but currently I can't see anyone really focused on that.

So now the question now is which tree should we use for our images? The older u-boot, the one we currently use, is very very old (4 years), and probably every bug we find in the u-boot code base will be also a pain to fix.

As an example, this bug is very important, as it breaks every kernel update, and I can confirm that this doesn't happen with latest upstream u-boot code (hash 69a2a4d9a5884a6f2d04a551308980d452b9b349).

Changed in u-boot-omap4 (Ubuntu):
status: New → Confirmed
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Also, this should just be an issue for this release (10.10), as for 11.04 we will probably be using Linaro's tree (that just got released).

Oliver Grawert (ogra)
Changed in u-boot-omap4 (Ubuntu Maverick):
milestone: none → ubuntu-10.10-beta
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Working well with latest Linaro's u-boot (v2010.06-695-gbd23130-linaro-2).

Texas Instruments X-Loader 1.41 (Jun 24 2010 - 13:29:27)
Starting OS Bootloader from MMC/SD1 ...

U-Boot 2010.06 (Aug 25 2010 - 14:10:39)

CPU : OMAP4430
Board: OMAP4 Panda
I2C: ready
DRAM: 512 MiB
Using default environment

....

Uncompressing Linux... done, booting the kernel.
Linux version 2.6.34 (rsalveti@evatp) (gcc version 4.4.5 20100728 (prerelease) (Ubuntu/Linaro 4.4.4-8ubuntu1hrw1) ) #12 SMP PREEMPT Wed Aug 25 14:05:23 BRT 2010

root@panda-maverick-usb:/# flash-kernel
Using u-boot partition: /dev/mmcblk0p1
Creating backups of uImage and uInitrd... done.
Generating kernel u-boot image... done.
Generating Initramfs u-boot image... done.

Next reboot:

Uncompressing Linux... done, booting the kernel.
Linux version 2.6.34-903-omap4 (root@panda-maverick-usb) (gcc version 4.4.5 20100728 (prerelease) (Ubuntu/Linaro 4.4.4-8ubuntu1) ) #7rsalveti1 SMP PREEMPT Wed Aug 25 00:03:36 BRT 2010

We'll probably just update to linaro's package, but still need to confirm that it works with other boards.

Revision history for this message
Oliver Grawert (ogra) wrote : Re: MIR for u-boot-linaro-omap4-panda because u-boot-omap4 fails to load new kernel after kernel update

turning this bug into a MIR for the linaro package, marking u-boot-omap4 task as wontfix (we will remove the non-linaro package after seed change towards the linaro package)

Changed in u-boot-omap4 (Ubuntu Maverick):
status: Confirmed → Won't Fix
Changed in u-boot-linaro (Ubuntu Maverick):
milestone: none → ubuntu-10.10-beta
summary: - u-boot fails to load new kernel fromm boot partition after kernel update
+ MIR for u-boot-linaro-omap4-panda because u-boot-omap4 fails to load new
+ kernel after kernel update
Changed in u-boot-linaro (Ubuntu Maverick):
importance: Undecided → Critical
Revision history for this message
Oliver Grawert (ogra) wrote :

MIR team, please approve, the package is a bootloader that will never be used directly in an installed environment, onbyl the binary will be installed in our images at build time.
no security or other issues are known, packaging is also fine.

Revision history for this message
Alexander Sack (asac) wrote :

this is fine, but please make it a non-native package.

Changed in u-boot-linaro (Ubuntu Maverick):
status: New → Incomplete
Revision history for this message
Alexander Sack (asac) wrote :

approving MIR to unblock arm team.

opened bug 624533 as RC bug about the non-native packaging.

Changed in u-boot-linaro (Ubuntu Maverick):
status: Incomplete → Fix Committed
Revision history for this message
Martin Pitt (pitti) wrote :

u-boot-linaro promoted

Changed in u-boot-linaro (Ubuntu Maverick):
status: Fix Committed → Fix Released
Revision history for this message
Oliver Grawert (ogra) wrote :

linaro u-boot is now in the new omap4 images and seems to run fine (see below).

please remove u-boot-omap4 from the archive now.

Texas Instruments X-Loader 1.41 (Jun 24 2010 - 13:29:27)
mmc read: Invalid size
Starting OS Bootloader from MMC/SD1 ...

U-Boot 2010.06 (Aug 26 2010 - 13:29:39)

CPU : OMAP4430
Board: OMAP4 Panda
I2C: ready
DRAM: 512 MiB
Using default environment

In: serial
Out: serial
Err: serial
Hit any key to stop autoboot: 0
mmc1 is available
reading boot.scr

319 bytes read
Running bootscript from mmc1 ...
## Executing script at 82000000
reading uImage

3494620 bytes read
reading uInitrd

5952663 bytes read
## Booting kernel from Legacy Image at 80000000 ...
   Image Name: Ubuntu Kernel
   Image Type: ARM Linux Kernel Image (uncompressed)
   Data Size: 3494556 Bytes = 3.3 MiB
   Load Address: 80008000
   Entry Point: 80008000
   Verifying Checksum ... OK
## Loading init Ramdisk from Legacy Image at 81600000 ...
   Image Name: Ubuntu Initrd
   Image Type: ARM Linux RAMDisk Image (gzip compressed)
   Data Size: 5952599 Bytes = 5.7 MiB
   Load Address: 00000000
   Entry Point: 00000000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
OK

Starting kernel ...

Uncompressing Linux... done, booting the kernel.

Changed in u-boot-omap4 (Ubuntu Maverick):
status: Won't Fix → New
summary: - MIR for u-boot-linaro-omap4-panda because u-boot-omap4 fails to load new
- kernel after kernel update
+ u-boot-linaro-omap4-panda replaces u-boot-omap4, please remove source
+ and binary packages of the latter from the archive
Revision history for this message
Steve Langasek (vorlon) wrote :

2010-08-30 17:15:44 INFO Removing candidates:
2010-08-30 17:15:44 INFO u-boot-omap4 L24.7git20100624-0ubuntu1 in maverick
2010-08-30 17:15:44 INFO u-boot-omap4 L24.7git20100624-0ubuntu1 in maverick armel
2010-08-30 17:15:44 INFO Removed-by: Steve Langasek
2010-08-30 17:15:44 INFO Comment: superseded by u-boot-linaro
2010-08-30 17:15:44 INFO 2 packages successfully removed.
2010-08-30 17:15:53 INFO Transaction committed.
2010-08-30 17:15:53 INFO The archive will be updated in the next publishing cycle.

Changed in u-boot-omap4 (Ubuntu Maverick):
status: New → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.