Post Reply

Stuck at "Data.img" thru odin. Possible EMMC brick fix! Thanks to Jaymoon pg.#44

OP caspersfi

29th February 2012, 10:53 PM   |  #271  
Member
Thanks Meter: 4
 
52 posts
Join Date:Joined: Oct 2010
More
I have the same issue with my phone and it was definately CWM touch that messed it up . Samsung has a one year warranty on these phones.....and even though mine was bought as soon as they were available which was last june I called samsung and was told my phone is indeed under warranty until december of this year by it`s lot number...so as you can imagine I am sending mine directly to sammy free shipping out and back She expected to have my phone back to me in five to seven days !!! Samsung support number 1-800-726-7864 Hope this helps !!!
The Following User Says Thank You to D_A_M_A_G_E For This Useful Post: [ View ]
29th February 2012, 10:59 PM   |  #272  
Dchibro's Avatar
Senior Member
San Diego, CA
Thanks Meter: 281
 
1,032 posts
Join Date:Joined: Jul 2011
More
Quote:
Originally Posted by sfhub

Clearlyif jtag is not writing successfully then something low-level is wrong. I can believe that fake flashing cwm touch onto a 3.0 kernel and restoring or wiping did something out of bounds rendering portions of the emmc inaccessible, but i would find it hard to believe this is coincindental emmc failure due to quality issues with the emmc. It is simply too specific, both in reproducability, symptoms, and prerequisites.

Well, it's certainly possible that he recently worked on a phone that had a bad EMMC, coincidentally around the time all this mess started happening. It sounds like we should start getting reports from people who sent their phones in from JTAG service soon. My money says JTAG gets them back up and running, where that blue light means a good thing not bad.
29th February 2012, 11:05 PM   |  #273  
connexion2005's Avatar
Senior Member
Flag Texas
Thanks Meter: 714
 
994 posts
Join Date:Joined: Feb 2009
Donate to Me
More
You are correct in your statement sfthub. JTAG is about as low a level initialization of reading and writing from the EMMC chip as you can possibly go. It's pure binary serial communication between the EMMC/oneNAND and CPU so if I am experiencing the inability to write at that level, then yes the EMMC chip is truly problematic and it is a hardware problem. With that said I'm not going to state the ROM flashing experience is to blame here as it would be highly unlikely but at the same time I have seen very odd things happen such as ROM's causing documented digitizer problems on the HD2 when Android or Windows Phone 7 are installed. A while back on the I9000M platform a leaked firmware was causing similar issues to the oneNAND and it was documented as true hardware failure as I could never pull one out of it's partition problems either.

EDIT: Also the blue light only represents that the boot.bin zone is intact which in nearly all cases it is as the primary bootloader is NOT on the EMMC chip but yet on a smaller segmented individual chip. In all cases where I can't write to the EMMC chip, I'm still able to write to the primary bootloader chip without issue.
29th February 2012, 11:31 PM   |  #274  
nanos1982's Avatar
Junior Member
Thanks Meter: 0
 
4 posts
Join Date:Joined: Oct 2010
Quote:
Originally Posted by D_A_M_A_G_E

I have the same issue with my phone and it was definately CWM touch that messed it up . Samsung has a one year warranty on these phones.....and even though mine was bought as soon as they were available which was last june I called samsung and was told my phone is indeed under warranty until december of this year by it`s lot number...so as you can imagine I am sending mine directly to sammy free shipping out and back She expected to have my phone back to me in five to seven days !!! Samsung support number 1-800-726-7864 Hope this helps !!!

Come to think of it, I did use touch CWM to attempt a nand restore. Insurance and warranties are wonderful things. If you're going to be a mad scientist, of course you're gonna kill a few subjects. Let me know what happens with the sammy warranty process. I'm curious to see the outcome.
29th February 2012, 11:34 PM   |  #275  
Dchibro's Avatar
Senior Member
San Diego, CA
Thanks Meter: 281
 
1,032 posts
Join Date:Joined: Jul 2011
More
Man, to hear that while highly unlikely, there "could" be a kernel/recovery scenario that could permanently damage the internal memory chip to the point where even JTAG can't write to it is a bit unnerving.
29th February 2012, 11:48 PM   |  #276  
Recognized Contributor
Thanks Meter: 6,844
 
4,757 posts
Join Date:Joined: Oct 2008
Quote:
Originally Posted by connexion2005

With that said I'm not going to state the ROM flashing experience is to blame here as it would be highly unlikely but at the same time I have seen very odd things happen such as ROM's causing documented digitizer problems on the HD2 when Android or Windows Phone 7 are installed. A while back on the I9000M platform a leaked firmware was causing similar issues to the oneNAND and it was documented as true hardware failure as I could never pull one out of it's partition problems either

To clarify, we are not talking about the act of flashing a rom causing emmc issues, we are talking about partition operations issued from recovery which are interfacing with a new linux 3.0 kernel. This brings into the the picture the possibility these are device driver level issues being exposed.

Imo it is equally hard to believe all these people who combined cwm touch with the ics linux 3.0 kernel are the only ones suddenly having an issue with just the /data partition, and only if they performed a wipe or restore using previously described setup. In addition nobody who has used the previous gb linux 2.6 kernels packed with cwm has ever reported such an issue.

There are too many specifics, imo, for this to just be arbitrary emmc failure. I'm not saying emmc isn't hoarked in some way, clearly it is, as your jtag bit blasting is failing, just imo the previously mentioned actors are somehow involved.
Last edited by sfhub; 29th February 2012 at 11:58 PM.
1st March 2012, 03:46 AM   |  #277  
connexion2005's Avatar
Senior Member
Flag Texas
Thanks Meter: 714
 
994 posts
Join Date:Joined: Feb 2009
Donate to Me
More
I agree 100% although hard to believe, the specific way these phones died are all the same as everyone was working with a specific build of ICS at the time this damage occurred. I've seen strange like this happen and be associated with flashing a file. In nearly all cases it's also the data zone as well that has the issue and not the entire chip. However in some cases I've seen the entire chip be un-writable...
1st March 2012, 03:58 AM   |  #278  
Recognized Contributor
Thanks Meter: 6,844
 
4,757 posts
Join Date:Joined: Oct 2008
So perhaps you could add your expertise, what is different about the data partition making more prone to these issues. Is it on a separate emmc chip? Or is everything on one chip that gets segmented? Is the issue with the section of emmc memory containing /data or is it with the meta data the keeps track of the sections?

I noticed the major/minor dev numbers changed for /data between the 2.6 and 3.0 kernel while the other partitions remained the same. That might just be a red herring.
2nd March 2012, 01:36 AM   |  #279  
connexion2005's Avatar
Senior Member
Flag Texas
Thanks Meter: 714
 
994 posts
Join Date:Joined: Feb 2009
Donate to Me
More
There is nothing exclusive about the /data partition. It resides on the EMMC chip with all other partitions (except for the primary bootloader). The fact that it would be the zone failing the most is beyond me other than the fact it most likely sees more I/O than any other segment...
2nd March 2012, 01:43 AM   |  #280  
Senior Member
Milwaukee, WI
Thanks Meter: 89
 
226 posts
Join Date:Joined: Nov 2011
More
well mine will get to cwm, and download mode. most people have just the blue light i hear. Im kind of out of options regarding this. this post might be my last resort.


http://forum.xda-developers.com/show...4&postcount=10

Post Reply Subscribe to Thread

Tags
brick, odin
Previous Thread Next Thread
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes