Nougat, adb shell+dmesg and future issues?

[kimnach]

--EDIT 30 Jan 2019: the attachment didn't go up yesterday for some reason, now added--

Admittedly, I am not proficient with android/linux, but this is crazy.

I decided that I would dust of the tinker board (not touched in nearly a year) and would use it as a second car pc.  (My goal is that the TB would run android and the following apps: Life360, Waze, Craftsman Garage Door Opener.  The "primary" computer is currently a mini-ITX running Win10 which is primarily used for TunerPro and DriveLine.  The mini-ITX is being replaced with a LattePanda Alpha....yes, I could run bluestacks on the LPA/Windows, but the TB begs to be used.)

Android Hardware: TB, GlobalSat BU-353 GPS antenna, 8" LCD monitor with Egalax resistive touch overlay (USB).

So the issue?  I downloaded the Nougat (14.1.2.50) image and wrote the image to an SD card.  I then installed Google Play Store from which I installed Life360, Waze, the Garage door app, and just a couple of ancillary utilities (e.g., terminal).  All was well until I tried to get the GPS drivers installed.

I planned to follow instructions in the libGPS thread.  Unfortunately, I get as far as running dmesg in the adb shell.  Instead of getting pertinent information (i.e., the device identifier similar to ttyACM0), i get hundreds of lines of errors.  (Image attached.)  Thinking that this was due in some unexplicable way to the ethernet connection (not likely), I ran dmesg on the TB.  Same result--nothing but hundreds of "undelivered transaction xxxxxx, process died" errors.

Other commands (e.g., dmesg | grep | -i usb) are not recognized in android, and lsusb list just the device # and hex ID.

Any thoughts?  Is this somehow a Nougat build issue?  Am I embarking on a path doomed to failure (given my limited proficiency)?

thanks in advance

[kimnach]

Not having gotten past that primary dmesg issue and seeing that i would have to compile a new kernel (I know that will be a trip) to get the egalax driver installed, I decided to start over late last night.

I grabbed the official RPI touchscreen from my 3D printer setup, flashed a new sd card with 14.1.2.50 image, and edited the build.prop file (i.e., sys.hwc.device.primary=DSI and sys.hwc.device.extend=HDMI-A).  I thought that at least I would have a working touch screen.  This was not to be: no touch functionality.  This goes contrary to precision215's video on youtube.  It workd for him, not for me.

I am beginning to think that this dedicated Tinker Board+android+Waze running on a second touch screen in my car is a rabbit hole that even Alice would get lost in.......

back to searching.

[maartinus]

Hi Kimnach, I use TB model 'S' with the standard Android 7 (probably the same build as you) as SatNav device in my car. It is connected to the official 7'' RPI touchscreen and old Navilock GPS USB receiver (chipset u-blox 5). All works fine, Waze and Mapfactor Navigator (OSM) are fully operational and stable.

I do not know what can cause your problems, but here is tip - try to search thread "MIPI-DSI Interface - Known Good Display not working with Tinkerboard - repost". It is from the linux world but Android contains linux kernel.

BTW, I use different GPS USB driver - you can look at my posts in the libGPS thread.

[kimnach]

Hi Kimnach, I use TB model 'S' with the standard Android 7 (probably the same build as you) as SatNav device in my car. It is connected to the official 7'' RPI touchscreen and old Navilock GPS USB receiver (chipset u-blox 5). All works fine, Waze and Mapfactor Navigator (OSM) are fully operational and stable.

I do not know what can cause your problems, but here is tip - try to search thread "MIPI-DSI Interface - Known Good Display not working with Tinkerboard - repost". It is from the linux  world but Android contains linux kernel.

BTW, I use different GPS USB driver - you can look at my posts in the libGPS thread.

thanks for the follow-up.  The RPi touch screen did work with the v13 of TinkerOS-android: i tried that after the lack of success with the v14.

i will look into your posts on libgps.

thanks again,
greg

****** just a note *******
my Raspi 7" touch controller hardware is ver1.1

[maartinus]

You are welcome.

I took a look at your dmesg errors and I remembered something. In another forum, I was talking with somebody about similar problem. Try to google for "Emteria Time sync Rasp3 usb gps". Solution of binder errors was very unexpected:

==============
It is getting strange.
I did film the boot so i can playback what the boot was saying:
It found the Ublox 7 gps receiver. ACM0 cdc_acm 1-1.4:1.0
binder 164:164 transactoin failed 29189.
I changed the ACM to 1. Dont know why, and back to AMC0. Now the GPS status app is working. I can read the exact time from it.
==============

I would just say that original TB Android driver doesn't use build.prop file, but gps.conf (I think).