[Kindle touch]help, debricking , can't enter diags mode - Page 2

geekmaster · 08-22-2012, 05:35 PM

Quote:

Originally Posted by Zephyrus

Thank you. Tried but failed.

Actually I suspected another thing.

After I set the bootmode to diags by using "fastboot setvar bootmode diags", I try to check the var I set by "fastboot getvar bootmode" and it remains as main.... but when I try to reboot by "fastboot reboot", the machine response and restarted (the tree screen flashed). Don't know if this is related or is just the same as other machines here.

Thanks.

That is NOT normal behavior. It could happen if the fastboot portion of your u-boot bist partition on mmc is damaged, or if your fastboot utility is damaged. Perhaps fastboot does not return an error status when if fails while writing to the bootmode var (similar to the false success report when flashing mmcblk0p1 with fastboot).

I do not know what else might cause that abnormal fastboot behavior.

Zephyrus · 08-22-2012, 05:42 PM

Quote:

Originally Posted by geekmaster

That is NOT normal behavior. It could happen if the fastboot portion of your u-boot bist partition on mmc is damaged, or if your fastboot utility is damaged. Perhaps fastboot does not return an error status when if fails while writing to the bootmode var (similar to the false success report when flashing mmcblk0p1 with fastboot).

I do not know what else might cause that abnormal fastboot behavior.

Hmm....seems I've encountered a difficult situation.

The mfgtool is working perfect. I can get into fastboot mode and using the tool.
Sometimes the mass storage device can be detected, but showing no media in device manage (drive letter assigned though), so no chance to put file directly into the kindle.
The fastboot utility return no error for setvar bootmode, I tried to get into diags via mfgtool, but nothing positive.

It seems the better solution for me is sent it back to amazon...... T_T
I'll do it if no luck land on it after two days......thanks everyone...

geekmaster · 08-22-2012, 05:45 PM

Quote:

Originally Posted by Zephyrus

Is it a sign that some other partitions crashed? eg. bootloader or mbr or what....

or would it be a sign of hardware failure?

As you can see in the additional text in my post above, it COULD be a damaged bootloader (or a failed fastboot setvar command, or other things).

Zephyrus · 08-22-2012, 05:53 PM

Quote:

Originally Posted by geekmaster

As you can see in the additional text in my post above, it COULD be a damaged bootloader (or a failed fastboot setvar command, or other things).

I used the "check" command to check both the bootloader and mbr.....both return okay....

Is it possible to flash a new bootloader img to fix it?

Or are there any other tool, such as atk can be use?

geekmaster · 08-22-2012, 06:35 PM

Quote:

Originally Posted by Zephyrus

I used the "check" command to check both the bootloader and mbr.....both return okay....

Is it possible to flash a new bootloader img to fix it?

Or are there any other tool, such as atk can be use?

ATK is for K3 and earlier, and MfgTool is for K4/K5. You would need to create a new MfgTool profile to flash the bootloader. The current MfgTool profiles are only for booting temporarily to main/diags/fastboot.

knc1 · 08-22-2012, 06:47 PM

Actually, the ATK is for any i.MX SoC that Freescale has produced.

The limits are that we have not developed directions here for using it with other than the K3.
No real reason to, the other established procedures here work just fine for the other machines.

Now K3Flasher - that is limited to the K3 because it only has support for the K3 SoC hard coded. It does not contain the details required to setup the other i.MX SoCs.

geekmaster · 08-22-2012, 10:14 PM

Quote:

Originally Posted by knc1

Actually, the ATK is for any i.MX SoC that Freescale has produced.

ATK uses the VID/PID for the K3 and earlier. MfgTool uses the DIFFERENT VID/PID for the K4/K5. There may be protocol differences as well.

Does ATK (as configured in the download links in this forum) also recognize and support the K4/K5 VID/PID?

knc1 · 08-22-2012, 10:45 PM

I don't know about what might be posted here, I got mine from Freescale as a source bundle.

The iROM protocol is the same, with the same command set, with differences in the initial time out and the watchdog settings.

The "RAM kernel" is specific to the i.MX model, the ATK source has a long list of them.
But the result is the same command set and protocol across all models.

That is two (2) command sets and two (2) protocols total. The iROM command set and protocol and the "RAM kernel" command set and protocol.

Register set-up and dynamic ram setups will also vary.

Freescale has some nice, bit-level, application notes on both the stock source and about making derivative command processors (which is what the "RAM kernel" is, a command processor).

Most of the above is exposed in the *.h files.
Except for the various downloaded "RAM kernels" - those are more than an *.h file.

Which is why I had to get the license terms clear with Freescale last week.

These two pieces (*.h files and command processors) are all that I want to re-use from their sources for a project of my own.
The *.h files do not have distribution restrictions.
The "RAM kernels" (and/or derivates) can be re-distributed in object form as a part of an application for a licensed i.MX machine (which Kindles are).
I.E: You can't ship the "RAM kernels" loose - they have to be embedded in the application.

dmhs2004 · 09-06-2012, 09:31 PM

@zklcy I think your contribution was very important, helped me after 4 days of searching without results. In my humble opinion, this additional step should be included in "A small debricking guide for" noobs "_ V4" a cordial greeting from Colombia.

adi23 · 10-24-2012, 03:28 AM

Quote:

Originally Posted by Zephyrus

I used the "check" command to check both the bootloader and mbr.....both return okay....

Is it possible to flash a new bootloader img to fix it?

Or are there any other tool, such as atk can be use?

As geekmaster said, probably you have a corrupted u-boot bist or idme vars.

Just load the new custom u-boot (with fastboot included) and check all idme vars, and write them somewhere to keep them safe, if you still have them, or rewrite the corrupt idme with the old one (check Mfg logs).

If still doesn't working, reflash the u-boot with fastboot (i have used main.bin from MfgConfig) and check/set again all idme vars and try again.

eureka · 10-24-2012, 07:55 AM

Quote:

Originally Posted by knc1

Actually, the ATK is for any i.MX SoC that Freescale has produced.

...at the time of ATK release.

Quote:

Originally Posted by knc1

The limits are that we have not developed directions here for using it with other than the K3.
No real reason to, the other established procedures here work just fine for the other machines.

Now K3Flasher - that is limited to the K3 because it only has support for the K3 SoC hard coded. It does not contain the details required to setup the other i.MX SoCs.

ATK as well supports only part of available Freescale SOCs. Amount of work for implementing support of protocols for new Freescale SOCs in ATK is roughly the same as implementing it in K3Flasher (for example).

Quote:

Originally Posted by knc1

The iROM protocol is the same, with the same command set, with differences in the initial time out and the watchdog settings.

Wrong. There are more differences.

It's true that someone can improve ATK to support new SOCs. So you are right in general. But with presence of MfgTool and cross-platform lightweight imx_usb_loader it's just pointless.

knc1 · 10-24-2012, 09:47 AM

Quote:

Originally Posted by eureka

It's true that someone can improve ATK to support new SOCs. So you are right in general. But with presence of MfgTool and cross-platform lightweight imx_usb_loader it's just pointless.

You have taken something completely out of context if you are trying to put those words into my mouth.

I NEVER proposed improving ATK support.

In case you haven't noticed, that is not Open Source code.
The source is owned by Freescale and any derivatives become the property of Freescale.

I do not work on, not even "for hire", non-OpenSource code.
Last time I did that ("for hire") was 1991.

08-22-2012, 10:45 PM	#23
knc1 Going Viral Posts: 17,212 Karma: 18210809 Join Date: Feb 2012 Location: Central Texas Device: No K1, PW2, KV, KOA	I don't know about what might be posted here, I got mine from Freescale as a source bundle. The iROM protocol is the same, with the same command set, with differences in the initial time out and the watchdog settings. The "RAM kernel" is specific to the i.MX model, the ATK source has a long list of them. But the result is the same command set and protocol across all models. That is two (2) command sets and two (2) protocols total. The iROM command set and protocol and the "RAM kernel" command set and protocol. Register set-up and dynamic ram setups will also vary. Freescale has some nice, bit-level, application notes on both the stock source and about making derivative command processors (which is what the "RAM kernel" is, a command processor). Most of the above is exposed in the .h files. Except for the various downloaded "RAM kernels" - those are more than an .h file. Which is why I had to get the license terms clear with Freescale last week. These two pieces (.h files and command processors) are all that I want to re-use from their sources for a project of my own. The .h files do not have distribution restrictions. The "RAM kernels" (and/or derivates) can be re-distributed in object form as a part of an application for a licensed i.MX machine (which Kindles are). I.E: You can't ship the "RAM kernels" loose - they have to be embedded in the application. Last edited by knc1; 08-22-2012 at 11:09 PM.

09-06-2012, 09:31 PM	#24
dmhs2004 Junior Member Posts: 3 Karma: 10 Join Date: Feb 2011 Device: dmhs2004@hotmail.com	Thanks zklcy @zklcy I think your contribution was very important, helped me after 4 days of searching without results. In my humble opinion, this additional step should be included in "A small debricking guide for" noobs "_ V4" a cordial greeting from Colombia.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Kindle 4 Crashed in Diags Mode	drajov	Amazon Kindle	6	07-25-2012 03:34 PM
Kindle Touch semi brick(diags only)	MooseBear	Kindle Developer's Corner	10	07-19-2012 05:58 PM
Kindle Touch Stuck at diags	sole	Kindle Developer's Corner	13	05-01-2012 08:47 AM
kindle touch simple debricking for help	non_fox	Amazon Kindle	4	04-18-2012 08:32 PM
Troubleshooting simple kindle touch (and k4nt) debricking method	geekmaster	Amazon Kindle	0	03-02-2012 10:31 PM

08-22-2012, 06:47 PM	#21
knc1 Going Viral Posts: 17,212 Karma: 18210809 Join Date: Feb 2012 Location: Central Texas Device: No K1, PW2, KV, KOA	Actually, the ATK is for any i.MX SoC that Freescale has produced. The limits are that we have not developed directions here for using it with other than the K3. No real reason to, the other established procedures here work just fine for the other machines. Now K3Flasher - that is limited to the K3 because it only has support for the K3 SoC hard coded. It does not contain the details required to setup the other i.MX SoCs.