Dell Venue 8 Pro imaging/eMMC

[Tom Elliott]

Yeah, I did not update 4.1.6. I Updated 4.2.0. This will not work on kernel 4.1.6.

[Wayne Workman]

Well this thread made a whole lot of progress in a short period of time… Job well done. I’m leaving it to you guys to figure out the rest lol.

[AsGF2MX]

@Tom-Elliott I am on the latest git and Kernel 4.2.0 right now and it’s still showing the odd /dev/ram0-15. I ran through dmesg and I see the following points of interest:

multiple occureneces of Not using MMCONFIG

Then:

ACPI PNP0A08:00: _OSC: OSC failed (AE_ERROR); disabling ASPM
ACPI PNP0A08:00 fail to add MMCONFIG information, can’t access extended PCI configuration space under this bridge.

There is a Winbond W83L51xD SD/MMC card interface driver showing a bit further down but that looks like the MicroSD card on the side.

Sidebar - if I want to boot an ISO via iPXE, how would I do so? I’m currently at home and I just realized I haven’t got a single USB drive on me. I do however have ISOs on my ESXi box and also I got FOG. Not trying to go off topic but trying to see if Ubuntu iso or even clonezilla will identify the storage device.

[Sebastian Roth]

@AsGF2MX said:

fail to add MMCONFIG information

MMCONFIG is not MMC! AFAIK this error is unrelated. See here: http://lwn.net/Articles/263288/

Don’t you see anything about mmcblk in dmesg? Like here: https://gist.github.com/thomasdziedzic/01482f33aba5971845b3

Booting ISO from iPXE is possible. Search the forums for “ISO” and you should find a lot of information about it.

[AsGF2MX]

@Uncle-Frank Did not spot the missing C - I thought it was MMC CONFIG. I was searching for mmc in dmesg and that’s all it showed. Searched for blk as well and got nothing. Do you know what kernel generated that particular output? I don’t have any custom settings in my FOG setup except that I am forcing the boot of the 32-bit kernel and not the 64-bit one via the UI so I have to ask if any specific switches or anything is needed anywhere? As usual, I have secure boot disabled as well.

On the ISOs, I did find the some info but I couldn’t make sense of where to put those lines in within the UI. On the other hand I did grab an SD card reader and I have a clean SD card so I will try to make a bootable drive or two in a few hours and hopefully get something back.

[Wayne Workman]

@AsGF2MX http://www.pendrivelinux.com/yumi-multiboot-usb-creator/

[AsGF2MX]

@Wayne-Workman

That did help me to get an active partition but it seems Ubuntu is a non-starter as I couldn’t even boot it. I was then able to just unzip the ISO contents of Clonezilla alternate stable (current off site) and boot.

It is on kernel 3.16.0-4-amd64 and there is an MMC block device shown.

There is indeed and mmcblk0 being shown in clonezilla but I can’t get the 4.2 kernels from yesterday to do the same. I tried to find the header file for the kernel build options but I couldn’t find it from the sourceforge portal.

How can I help troubleshoot this further?

[Wayne Workman]

@AsGF2MX

@Tom-Elliott might have updated the kernel - but sometimes changes need made in the inits too, wonder if he did that?

[AsGF2MX]

@Wayne-Workman

Are you referring to init scripts within the initrd or something else? Since there are no modules here I am not sure how to even get the mmc recognized in dmesg.

[Wayne Workman]

Please post all of the output of this, so we can clearly see it (proper camera positioning and focus).

fdisk -l | more

[AsGF2MX]

@Wayne-Workman

Hopefully this one cuts it. First up, clonezilla:

FOG:

The 8GB device is the SD card I’m using to boot clonezilla.

[Sebastian Roth]

Others seam to have issues with Samsung eMMC disks as well. See here:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1422338
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1471989

But notice the different models, MDGAGC in the first and MCGAFB in the second bug report. Not sure if those were ever solved.

From what I can see in the first screenshots your model is a MBG4GC which I sadly cannot find much about when searching for those error messages. Not sure if this is related: https://dev-nell.com/rpmb-emmc-errors-under-linux.html

I am sorry but I guess this won’t be an easy catch. You/we might need to get in contact with kernel developers. Not 100% sure yet.

Could you please try other live linux systems to see if one of them goes without the errors?

[AsGF2MX]

@Uncle-Frank Perhaps it’s not such a big issue after all:

• My understanding is the RPMB partition is effectively doing what regular desktops do with EEPROM chips - store OEM license keys
• AFAIK Windows does not treat this partition as mountable - it doesn’t show in disk manager (the tablet is out of battery at the moment - will post a pic in a bit)
• Theoretically if I could clone p1-p5 as listed above then I should be all set
• The license or OEM data in the rpmb can remain as is across each unit

So in theory, if we apply Nell’s patch to ignore the rpmb area, we should be able to make this work.

The question is what needs to be done to the initd to get it to work?

@Tom-Elliott

[Wayne Workman]

@AsGF2MX Would it be acceptable for FOG to provide a Windows PE environment via PXE Boot with an installation script for Windows 8.1 for this device to make it automated?

It’s not traditional I know… but what are you going to do? Go pay for a product when FOG can still help you?

I mean… If we created a PXE Option that booted windows PE and just went through a Windows Recovery process, would that be acceptable? It would apply applications and restore settings. I suppose with the FOG Client installed on this device, it’d rename it and join it to the domain later too.

Is this viable - at-least until we can get better support for your device?

[Tom Elliott]

@AsGF2MX from the fog init in debug, what is the output of uname -rm?

[AsGF2MX]

@Tom-Elliott It says 4.2.0 i686

[Sebastian Roth]

Tom pointed me to a discussion here in the forum about a similar issue some time ago. Search for emmc and you’ll find a thread with a user called d4rk3.

Could you please enter a FOG debug session and run the following two commands fdisk -l /dev/mmcblk0 and lsblk /dev/mmcblk0
I guess the output of fdisk will be empty. Would be interesting to see what you get from lsblk.

As far as I can see clonezilla is showing the partitions with fdisk because it is debian based and debian has added the mentioned patch. See here: https://launchpad.net/debian/+source/linux/3.16.7-ckt9-1 (“eMMC: Don’t initialize partitions on RPMB flagged areas”) I downloaded the debian patchset to look through it and it’s got this patch by Nell Hardcastle (https://dev-nell.com/rpmb-emmc-errors-under-linux.html) in it!

@Tom-Elliott Would you mind adding this patch to 4.2? I know you’ve done this for d4rk3 some months ago. Didn’t seam to work for him but I wonder why.

[Tom Elliott]

@Uncle-Frank I’m rebuilding now, should be done in about 5 minutes. I added the patch. Just a note, I don’t have a simple way to patch this to work around the block. I had to manually code the patch, which would only be good on a one-to-one patch of the version as things change quite often between kernel updates.

I, again, won’t know if this works until somebody tests/confirms it as well. Hopefully your suggestion here will be viable and help ALL users out with eMMC SSD Cards.

[Tom Elliott]

Patched mmc drivers and built kernels are now up and running. You can update by choosing the latest kernels in the kernel update list, or by simply rerunning installer if you’re running development versions.

Thank you,

[d4rk3]

I’ll give the kernel a whirl tomorrow at work and report back if anything’s changed.