Ubuntu eMMC flash issue on v11 board

Bug #977766 reported by Sunil Kamath
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
IglooCommunity
New
Low
Jayeeta Bandyopadhyay

Bug Description

Getting following error while flashing Ubuntu 12.03-1 release image on v11 board.

LCD ERROR: Process raw image failed 152.

We are able to create sd card image and also able to flash on v7 boards.

But on v11 boards (tried on 2 of them), flashing process stops while using both Windows/Ubuntu flash tools.
The size of the image is 3GB. And it consistently stops at 58% while flashing on eMMC with LCD error.

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

riff verbose trace is as below:

sunil@sunil-HP-Compaq-6910p:~/SnowBall_Sunil/ubuntu$ sudo riff -v -f snowball-ubuntu-2012.03-1.img

 ----------------------- riff - Raw Image File Flasher -------------------------
 Version: 0.4.2
 Flash a device. Try `riff --help' for more information.
 -------------------------------------------------------------------------------
19:35:08 - Using config file /usr/share/riff/config
19:35:08 - Listening on USB for device connection...
19:35:12 - Device detected on USB@0
19:35:12 - Connected Device@0
19:35:12 - Initializing device...
19:35:12 - UNKNOWN token
19:35:12 - PWR_MGT token
19:35:12 - PWR_MGT sent
19:35:12 - X-LOADER token
19:35:12 - X-LOADER sent
19:35:12 - MEM_INIT token
19:35:12 - MEM_INIT sent
19:35:12 - UNKNOWN token
19:35:12 - NORMAL token
19:35:12 - NORMAL sent
19:35:12 - Initializing finished
19:35:12 - Starting initialization of LCD...
19:35:12 - LCD initialization finished successfully
19:35:12 - Progress listener added successfully
19:35:12 - Loader successfully started
19:35:12 - Version: $Id: Dec 9 2011 10:03:12 FlashLoader 8500 HW:V2 SECURITY:false LDR:u8500-android-2.3_v2.12-19-ge5c
19:35:12 - Protocol: 10
19:35:12 - Flashing raw image...
19:35:12 - Filesize: 3221225472
51.0 % 19:39:37 - LCD ERROR: Process raw image failed 152
19:39:37 - Shutting down device...
19:39:37 - Device shutdown finished successfully
19:39:37 - Disconnect detected on USB@0
19:39:37 - Disconnected Device@0
19:39:37 - Destroying LCD context...
19:39:37 - LCD context destroyed successfully

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

I am in touch with toolset team.

Latest update from them is:

Since V7 works it seems like something is wrong with the V11 board.
Perhaps we need to update the boot binaries (config pack)

I will see if I can borrow a V11 board and test tomorrow.

Revision history for this message
Kalle Vahlman (kvahlman) wrote :

For what it is worth, V10 works as well with the release image.

One problem I've noticed is that through an USB hub the flashing does _not_ succeed with V10 but does with a V5 board. It fails to even start though, so it's probably not the same issue but worth verifying I think.

Revision history for this message
Patrik Klinger (patrik-klinger) wrote :

Sunil, what is you recommended importance level of this bug and should it be mapped to the 2012.04 release?

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

I am currently in touch with toolset team for the same.
So far they feel issue may not be from riff.
Lets target it for 2012.05.
We anyways have sdcard image created for our work. But its important that we get to work by flashing directly on emmc.

Changed in igloocommunity:
milestone: none → 2012.05
importance: Undecided → Medium
Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

The riff tool version used is as below:

sunil@sunil-HP-Compaq-6910p:~/SnowBall_Sunil/ubuntu$ dpkg -l | grep riff
ii riff 0.4.2 A flashing tool initially developed for igloocommunity.org and snowball devices.
ii riff-configpack 0.4.4-1 Configpack for riff
ii riff-loadercomm 0.4.3 Loader communication driver library, Loader communication module library and loader binary

On the same host PC and tool I was able to flash v7 but not v11.

As mentioned, issue comes even while using windows tool.

Working with toolset team for checking real root cause of the issue.

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

Comment from toolset team:

I borrowed a V11 board and have successfully flashed it with snowball-ubuntu-2012.03-1.img (Host PC = Ubuntu 10.04 64bit).

Windows version of riff also works fine.

C:\Documents and Settings\eolabor\My Documents\Riff-Windows-0.5.0>riff -f snowball-ubuntu-2012.03-1.img

----------------------- riff - Raw Image File Flasher -------------------------
Version: 0.5.0
Flash a device. Try `riff --help' for more information.
-------------------------------------------------------------------------------
15:24:48 - Listening on USB for device connection...
15:26:06 - Connected Device@0
15:26:06 - Flashing raw image...
15:26:06 - Filesize: 3221225472
100.0 %
15:34:38 - Flashing finished successfully
15:34:39 - Disconnected Device@0

C:\Documents and Settings\eolabor\My Documents\Riff-Windows-0.5.0>

Changed in igloocommunity:
assignee: nobody → Sunil Kamath (sunil-kamath)
Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

We are suspecting more on emmc itself.
On emmc bad blocks will he not be handled from flashloader/riff as emmc internally shall handle them.
hence need to check emmc is working properly or not. (on the board on which issue observed).
Also need to test with various v11 boards.
lets continue investigating after 12.04 release. (use that image for test).
targetted for 12.05.
Issue priority need to be reduced if its reproducible on only handful number of v11 boards.

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

I can also see new bug raised:
https://bugs.launchpad.net/igloocommunity/+bug/984647

But seems like its on v4 board. But similar observation in that bugreport as well.

Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

This can be related to hardware. As its reproducible only in some h/w. But this can be confirmed only with the a detailed error message. Implementation of "https://bugs.launchpad.net/igloocommunity/+bug/985474" will give better picture on this.

Changed in igloocommunity:
milestone: 2012.05 → 2012.06
Revision history for this message
Sunil Kamath (sunil-kamath) wrote :

Further tests done. Tried on several other boards. This issue is coming in only one specific v11 board. Seems to be h/w issue.

Revision history for this message
Jayeeta Bandyopadhyay (jayeeta) wrote :

When flashed on one V11 board got following error message, while flashing works fine with other H/Ws

 ----------------------- riff - Raw Image File Flasher ------------------------- Version: 0.5.1 Flash a device. Try `riff --help' for more information.
 -------------------------------------------------------------------------------
20:15:54 - Listening on USB for device connection...
20:15:58 - Connected Device@0
20:15:58 - Flashing raw image...
20:15:58 - Filesize: 3221225472
0.0 % 20:16:00 - LCD ERROR: Process raw image failed 152 20:16:00 - LCD ERROR 152 : Internal malicious state detected in IO layer. Report this problem.
20:16:00 - Disconnected Device@0

Changed in igloocommunity:
milestone: 2012.06 → 2012.07
Changed in igloocommunity:
assignee: Sunil Kamath (sunil-kamath) → Jayeeta Bandyopadhyay (jayeeta)
Revision history for this message
Lee Jones (lag) wrote :

I have this problem too.

How is it progressing?

Revision history for this message
Lee Jones (lag) wrote :

I can confirm that this only happens with one of my v11s.

... and always at 34%:

34.0 % 10:22:48 - LCD ERROR: Process raw image failed 152

So I'm guessing there are some bad blocks there and Riff doesn't know how to deal with them?

Can I suggest that we raise the priority of this bug please, as it's bricking Snowballs.

Changed in igloocommunity:
milestone: 2012.07 → 2012.09
Changed in igloocommunity:
importance: Medium → Low
Changed in igloocommunity:
milestone: 2012.09 → none
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.