[GTALUG] Android / fastboot / ROM problem

[Russell Reiter]

On Mon, Jun 24, 2019, 10:01 AM Evan Leibovitch, <evan at telly.org> wrote:

>
>
> On Mon, 24 Jun 2019 at 09:29, Russell Reiter <rreiter91 at gmail.com> wrote:
>
>> On Sun, Jun 23, 2019, 8:35 PM Evan Leibovitch via talk <talk at gtalug.org>
>> wrote:
>>
>>>
>
>>  Since its OS stopped being upgraded in 2015, I'd installed LineageOS on
>>> it some time ago. A few months ago when I went to upgrade, the install from
>>> TWRP didn't take and it just bootlooped on the Lineage animation about 15
>>> seconds in.
>>>
>>
>> Was it the latest version of TWRP?
>>
>
> Yes. There appears to be no problem flashing recovery.
>
> [...]
>>
>> fastboot -w update image-nakasi-ktu84p.zip
>>> target didn't report max-download-size
>>> sending 'bootloader' (2100 KB)...
>>> OKAY [  0.262s]
>>> writing 'bootloader'...
>>> OKAY [  0.601s]
>>> finished. total time: 0.863s
>>> rebooting into bootloader...
>>> OKAY [  0.015s]
>>> finished. total time: 0.065s
>>> extracting android-info.txt (0 MB)...
>>> extracting boot.img  (4 MB)...
>>> target didn't report max-download-size
>>>
>>
>> The archive doesn't sign correctly.
>>
>> archive does not contain 'boot.sig'
>>> archive does not contain 'dtbo.img'
>>>
>>
Its not the install script, that just passed parameters to fastboot. I
jumped the gun because I did have F28 problems with adb persistance in
allowing fresh connections to the same device.  That was a key sig issue.

.



>> The link below mentions a problem with dtbo.img.
>>
>> archive does not contain 'dt.img'
>>> extracting recovery.img (5 MB)...
>>> archive does not contain 'recovery.sig'
>>> extracting system.img (628 MB)...
>>> archive does not contain 'system.sig'
>>> archive does not contain 'vbmeta.img'
>>> archive does not contain 'vendor.img'
>>> wiping userdata...
>>> Couldn't parse erase-block-size '0x'.
>>> Couldn't parse logical-block-size '0x'.
>>>
>> mke2fs 1.44.6 (5-Mar-2019)
>>> /tmp/TemporaryFile-fnXNLB:
>>>
>>
>> Looks like the install script was out of date and wasn't linking things
>> correctly.
>>
>
>> Unimplemented ext2 library function while setting up superblock
>>> /usr/lib/android-sdk/platform-tools/mke2fs failed with status 1
>>> mke2fs failed: 1
>>> error: Cannot generate image for userdata
>>>
>>
>>
>> At this point, if it's not bricked, it looks a bit more like software to
>> me. A little research found similar problems.
>>
>>
>> https://forum.xda-developers.com/pixel/help/mke2fs-error-flash-factory-images-pixel-t3734828
>>
>
> Thanks! Lots there to check out. Guess these problems weren't limited to
> Nexus devices.
>

I looked at an example flash script from your site link and saw that all
the script did was initiate fastboot. Adb fastboot now supports dynamic
partioning as part of project treble for AOSP.

Android from 8.0 onwards has internal hw support for dynamic partitioning.
Others, like myself on my Motorola Nexus 6 got a software nudge into
realizing those AOSP HAL targets.

You might be caught in the changeup depending on your base release version.
You could count the number of partitions on the device and compare it with
the number of img files in the zip file.

adb pull /proc/partitions

If the counts match you might get away with the last post on the link I
sent which will flash all partion images again.

Otherwise it looks like you'll need the full SDK and will have to provide
the example corrected symlink to mke2fs  in order to write the correct
partition info to the superblock, so the reformatted partitions match the
flash images.

Personally I got a pretty good performance boost from upgrading my Nexus 6
to Pie using the PixelExperience rom for this phone. The second aftermarket
battery which is feeding Pie is outperforming the original battery shipped
by Motorola. I'd be surprised if the AM product was better than Motorola
stock, but it could happen.

I'd be curious to see what, if anything works for you?


> - Evan
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gtalug.org/pipermail/talk/attachments/20190625/a33fbbb7/attachment.html>