Anybody working on Android for the Odroid-Go A?

Talking about OS and emulation core development
Post Reply
macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

Cool, good progress. I'm still plugging away at the Wifi issues on AOSP 11, and looking at the suspend issues. Long term I hope to get a 32 bit (aarch32 if Android can do it) version running so as to alleviate some memory constraints.

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

macromorgan wrote:
Thu Dec 17, 2020 1:44 am
Cool, good progress. I'm still plugging away at the Wifi issues on AOSP 11, and looking at the suspend issues. Long term I hope to get a 32 bit (aarch32 if Android can do it) version running so as to alleviate some memory constraints.
The wifi issue, all my usb WiFi sticks work. And looking at serial output wifi connects and gets authenticated but Android says authentication failed. I'm going to look a wpa_supplicant, but may be able to edit the reporting from the driver.

I did have 32bit working on Android 10 but moved to 64bit because I thought I had ram to spare...😓. You don't buy into the 64bit apps are newer, better optimized, and faster? But ram free is usually around 430mb. Gapps launching and keeping 4-6 processes in the background kills it.

One other thing I want to try, I know that we can get rid of the vendor partition and move it to system/vendor. I'm wondering if we can still do that for data too? If we can move everything to one partition we could have two Android builds, a boot partition, and a joint roms partition that both could share? Without having to do extended partitions? Or dual boot it with ubuntu since most 2 or 3 partitions.

Edit: I was able to get audio casting working too. Still trying to fix hardware encoding, it's the only thing stopping screen casting seems like. Trying not to be a quiter and switch to software encoding.
IMG_20201215_234139.jpg
IMG_20201215_234139.jpg (841.49 KiB) Viewed 5316 times
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

And I got screen cast working with hardware encoding!!! Not really excited about screen cast... But more so that this was the first time to really test the hardware encoding and the kernel driver for it. All of the vpu functions should work now.

But screen casting is super laggy on the receiving end casting at 1280x720. Doesn't seem like it's going to be usable for anything interactive. I'm going to try software encoding after work to see if it's better.
IMG_20201216_133521.jpg
IMG_20201216_133521.jpg (769.22 KiB) Viewed 5306 times
IMG_20201216_133717.jpg
IMG_20201216_133717.jpg (704.9 KiB) Viewed 5306 times
IMG_20201216_133459.jpg
IMG_20201216_133459.jpg (596.07 KiB) Viewed 5306 times


Edit: I lied, so i did fix the VPU but i was missing a entry in ueventd giving access to dev/mpp_service. Which caused me to run completely in software mode. But after fixing that, the odroidgo 2 is blazing through some 1080p in a video benchmark test and slow even decoding some 2k videos with graphical errors. I'm shocked by the performance boost. Going to restart to windows and test casting again.
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Okay, here it is, Google Cast with working hardware decoding. It actually isn't that bad. One issue though is that everything looks pixelated compared to when i tested earlier with software decoding. But it may just be my internet as my bandwidth lowers this time of night. Also I didn't get any audio out of my laptop. I don't know if it's from a audio config issue or me not using an actual google cast receiver as audio worked casting to my google home mini. I still need to look back at Miracast support as that would use WIFI-direct and should be faster.






Edit: We need to retest steam link as it was running good with just software decoding.
These users thanked the author notime2d8 for the post (total 2):
odroid (Thu Dec 17, 2020 2:11 pm) • joy (Thu Dec 17, 2020 3:42 pm)
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

User avatar
rooted
Posts: 8999
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 745 times
Been thanked: 422 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rooted »

You keep making milestones and don't give up, awesome progress.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

Freaking sweet, great job!

Still messing with WiFi here. I got things onto a 32 bit branch now (still 64 bit kernel). RAM usage drops for me on a somewhat unoptimized build from 900MB used to 660MB used, so it seems to make quite a difference. Pushed my 32bit branch to my github. https://github.com/macromorgan/device_h ... tree/32bit

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

notime2d8 wrote:
Thu Dec 17, 2020 3:40 am
You don't buy into the 64bit apps are newer, better optimized, and faster?
64bit apps are faster, but they have pointers that are twice as large (affects code size). A pure 64bit Android would work though. What I don’t want is multiarch, where there are 2 versions of most libraries which contributes mostly to the RAM bloat issue on 64bit OSes.

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

I tried Steamlink with the build posted here last and it worked quite well but there were the same controller issues that were also present in the earlier build. Analog control when pushed to maximum right registers as full left, which makes analog control infuriating. Digital works great.

The app doesn’t fit on screen properly, but if you can blag your way past that the performance was good. Just wish Steam would fix the host app on Mac OS big sur now!

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Shoesy wrote:
Fri Dec 18, 2020 8:25 am
I tried Steamlink with the build posted here last and it worked quite well but there were the same controller issues that were also present in the earlier build. Analog control when pushed to maximum right registers as full left, which makes analog control infuriating. Digital works great.

The app doesn’t fit on screen properly, but if you can blag your way past that the performance was good. Just wish Steam would fix the host app on Mac OS big sur now!
Someone else said that about the joystick. I haven't been able to reproduce it, or fix it. We tried different parameters in the boot.ini and android key layout file. Do you know if it does it in PPSSPP in the joystick test?
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

I’ve only noticed in steam. Worked ok in native android stuff / ui

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

macromorgan wrote:
Fri Dec 18, 2020 8:05 am
notime2d8 wrote:
Thu Dec 17, 2020 3:40 am
You don't buy into the 64bit apps are newer, better optimized, and faster?
64bit apps are faster, but they have pointers that are twice as large (affects code size). A pure 64bit Android would work though. What I don’t want is multiarch, where there are 2 versions of most libraries which contributes mostly to the RAM bloat issue on 64bit OSes.
I tried a pure 64bit bit build yesterday and ran into multiple issues. Half of my gapps were 32bit and a lot of libs were linked to 32bit locations like audio for some reason. Seems to be a no go. A 32bit build did work though but I forgot to check ram. 😓

When you get a chance, can you see if you can overclock and boot with the gpu set to 600mhz? It's not stable for me and the GPU crashes and restarts with screen artifacts.
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

valadaa48
Posts: 288
Joined: Mon Feb 17, 2020 1:35 pm
languages_spoken: english
ODROIDs: odroid go advance
Has thanked: 40 times
Been thanked: 96 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by valadaa48 »

So I've tried to revisit mainline and I wish I had committed my package definition when I had it working before with kernel 5.9. It seems now that no matter what I try I can't get the display to initialize. I've tried 5.9 again and I don't get past the boot logo. I tried the 5.10 android repo posted here but the display tries to initialize but I get vertical white lines (I think it's because it's not using simple panel).

To perhaps save me some trial and error iterations, anyone know of a working 5.9 or 5.10 commit for Linux?

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Hardware decoding video test. You can see the different video/audio formats being used. AC-3 and aac(main) audio don't work. AVI, FLV, MKV, MPG, RMVB, and WMV video also don't work. For some reason 2k and 4k videos cause artifacts on the whole screen not just in the video window.



If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Pretty sure most of you knew this but, video out over usb2 is a no go. Looks cool, but the lag is worse than with chromecast over wifi. But the displaylink adapter does do vga, dvi, and hdmi out. I'm cheating a little bit and using the displaylink presenter app though. I'm not sure how it works but it seems to be all software rendering. It may be faster if I can get the drm driver working but failed with that so far. I guess this may be useful for data readouts.

And I tried but I couldn't get this pic to rotate for nothing.
IMG_20201222_234007.jpg
IMG_20201222_234007.jpg (532.7 KiB) Viewed 4975 times
These users thanked the author notime2d8 for the post (total 2):
odroid (Wed Dec 23, 2020 4:17 pm) • Tagruato_Corporation (Wed Dec 23, 2020 9:17 pm)
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Two pics just in case anyone was wondering or hasn't tested it yet. But I tested this with a rtl8821cu combo and also a usb bluetooth 5.0 Realtek 8761B. I was able to do HD audio, file transfer, and connect multiple devices over bluetooth.

IMG_20201223_005306.jpg
IMG_20201223_005306.jpg (529.15 KiB) Viewed 4935 times
IMG_20201223_002647.jpg
IMG_20201223_002647.jpg (447.28 KiB) Viewed 4935 times
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

nex86
Posts: 6
Joined: Tue Dec 29, 2020 8:02 am
languages_spoken: english, german
ODROIDs: Odroid Go Advance Black Editon
Has thanked: 0
Been thanked: 0
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by nex86 »

whats the status on this?
I'm especially interested in running emulators (epsxe, ppsspp, retroarch etc) and some lower end Android games like GTA 3 etc.. stuffs what my old Galaxy S3 was able to handle.

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

Flash the image and give it a go! I don’t think there’s any reason to wait :)

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

nex86 wrote:
Tue Dec 29, 2020 8:04 am
whats the status on this?
I'm especially interested in running emulators (epsxe, ppsspp, retroarch etc) and some lower end Android games like GTA 3 etc.. stuffs what my old Galaxy S3 was able to handle.
Haven’t done any work the last week, still trying to get my new computer set up. Upgraded from an i7-4770 with 32GB to a Ryzen 9 5900x with 128GB. Curious to see how fast this thing can compile a kernel/Android.

My hope is that if I switch to the Rockchip sources from their 4.4 kernel (since removed) for the wifi I will have better luck with Android 11. If I can’t get wifi working I might just give up and try to build for the wifi/bt adapter that works with the super.

User avatar
rooted
Posts: 8999
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 745 times
Been thanked: 422 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rooted »

macromorgan wrote:
nex86 wrote:
Tue Dec 29, 2020 8:04 am
whats the status on this?
I'm especially interested in running emulators (epsxe, ppsspp, retroarch etc) and some lower end Android games like GTA 3 etc.. stuffs what my old Galaxy S3 was able to handle.
Haven’t done any work the last week, still trying to get my new computer set up. Upgraded from an i7-4770 with 32GB to a Ryzen 9 5900x with 128GB. Curious to see how fast this thing can compile a kernel/Android.

My hope is that if I switch to the Rockchip sources from their 4.4 kernel (since removed) for the wifi I will have better luck with Android 11. If I can’t get wifi working I might just give up and try to build for the wifi/bt adapter that works with the super.
35 seconds (defconfig) is my guess to compile the kernel if using a fast SSD, maybe less if using NVME instead of SSD.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

rooted wrote:
Fri Jan 01, 2021 4:30 am
macromorgan wrote:
nex86 wrote:
Tue Dec 29, 2020 8:04 am
whats the status on this?
I'm especially interested in running emulators (epsxe, ppsspp, retroarch etc) and some lower end Android games like GTA 3 etc.. stuffs what my old Galaxy S3 was able to handle.
Haven’t done any work the last week, still trying to get my new computer set up. Upgraded from an i7-4770 with 32GB to a Ryzen 9 5900x with 128GB. Curious to see how fast this thing can compile a kernel/Android.

My hope is that if I switch to the Rockchip sources from their 4.4 kernel (since removed) for the wifi I will have better luck with Android 11. If I can’t get wifi working I might just give up and try to build for the wifi/bt adapter that works with the super.
35 seconds (defconfig) is my guess to compile the kernel if using a fast SSD, maybe less if using NVME instead of SSD.
Using an HDD. 2 minutes to compile and 2.5 minutes to do a full LTO (I’m building with clang). I am almost done with my first virgin Android build so I’ll post times later, but it does about half the build in 15 minutes before it craps out (have to remember how to get vintf to play nice with kernel 5.10).

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

@macromorgan I'm having issues with the OGS screen. It works every time and initializes like it should. But remains blank until android starts and I cycle the power button...works perfect after that. I can't figure out how to cycle or blank the screen during boot. The boot.ini regulator commands don't work, and changing the reset timing doesn't work. Any idea's?
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

User avatar
rooted
Posts: 8999
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 745 times
Been thanked: 422 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rooted »

How much better is Android on the OGS larger screen?

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

rooted wrote:
Sat Jan 02, 2021 5:35 pm
How much better is Android on the OGS larger screen?
So much better. You can watch videos in 480p. You can actually see the menu options and configure Steamlink. Before the UI didn't resize to the 3.5" screen so it was hard to get it to work and you couldn't see any options to configure/tweak it. Remote desktop scales well to the screen and is actually usable. I also haven't noticed any performance drops from being a higher resolution screen.
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

User avatar
rooted
Posts: 8999
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 745 times
Been thanked: 422 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rooted »

Thanks for the update, sounds great.

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

Have you got any ideas what could be happening with my analogue stick issue in Steamlink? Tried it again today using windows host and it still registers full right axis as full left. If you push it slowly it works, but a fast motion right will make it go left.

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

Shoesy wrote:
Sun Jan 03, 2021 3:20 am
Have you got any ideas what could be happening with my analogue stick issue in Steamlink? Tried it again today using windows host and it still registers full right axis as full left. If you push it slowly it works, but a fast motion right will make it go left.
It's only in steamlink? Have you tried another app? Can you see it happen in the PPSSPP analog test? Can you try a newer version of steam link beta from the forums here http://media.steampowered.com/steamlink ... latest.apk

I couldn't get mines to do it. But I'm using a used oem Vita joystick in my oga. We can try adjusting the parameters for it in the boot.ini maybe limiting or increasing the range?
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

I was using vita stick, but have reverted to stock with the same result.

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

Stick works fine in ppsspp

Shoesy
Posts: 93
Joined: Sat Jul 18, 2020 1:26 am
languages_spoken: english
ODROIDs: Odroid Go Advance
Has thanked: 9 times
Been thanked: 8 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by Shoesy »

Tried that later apk - same result.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

Long time no update. I thought I'd share though I figured out why Linux was crashing when using mainline u-boot... basically it's because the binary Op-tee image was reserving some memory that Linux wasn't aware of and if you tried to write to it you'd get a crash. I pushed this commit https://github.com/macromorgan/odroid_g ... eb21b1b86e which is an ugly hack but should fix it.

So your options are either 1) use the fully mainline u-boot with mainline Arm Trusted Firmware which requires no patching; however you either have to nuke your SPI chip (don't worry it's recoverable) to force the device to boot off of SD or you have to hold the reset button on every boot-up or 2) use the Rockchip miniloader with the trust.img and an upstream u-boot image, but if you don't add the reserved memory to the devicetree your kernel will crash during or shortly after boot.

Also, noticed on my Android image that the specific issue with the wifi is actually wpa_supplicant. If the wireless network is not encrypted you can connect to it just fine. I'm still playing with it and trying to get Mesa 21.0-rc working since it brings Open GLES 3.0 to Bifrost chips.
These users thanked the author macromorgan for the post (total 2):
notime2d8 (Thu Jan 21, 2021 1:24 pm) • odroid (Thu Jan 21, 2021 1:59 pm)

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

macromorgan wrote:
Thu Jan 21, 2021 12:59 pm
Long time no update. I thought I'd share though I figured out why Linux was crashing when using mainline u-boot... basically it's because the binary Op-tee image was reserving some memory that Linux wasn't aware of and if you tried to write to it you'd get a crash. I pushed this commit https://github.com/macromorgan/odroid_g ... eb21b1b86e which is an ugly hack but should fix it.

So your options are either 1) use the fully mainline u-boot with mainline Arm Trusted Firmware which requires no patching; however you either have to nuke your SPI chip (don't worry it's recoverable) to force the device to boot off of SD or you have to hold the reset button on every boot-up or 2) use the Rockchip miniloader with the trust.img and an upstream u-boot image, but if you don't add the reserved memory to the devicetree your kernel will crash during or shortly after boot.

Also, noticed on my Android image that the specific issue with the wifi is actually wpa_supplicant. If the wireless network is not encrypted you can connect to it just fine. I'm still playing with it and trying to get Mesa 21.0-rc working since it brings Open GLES 3.0 to Bifrost chips.
There are surprisingly a lot of people who've tried mainline U-boot and ran into that issue. I've also seen it mentioned in the rockchip linux IRC channel unresolved.

I did a diff for wpa supplicant as it worked fine before in Android 10 but I couldn't find what broke authentication. Also it was only for the esp8089, my 3 Realtek usb WiFi sticks work normal.

Can you try this patch to see if it fixes the issue?

https://github.com/followmsi/android_de ... 18d332c95e
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

notime2d8 wrote:
Thu Jan 21, 2021 1:47 pm
macromorgan wrote:
Thu Jan 21, 2021 12:59 pm
Long time no update. I thought I'd share though I figured out why Linux was crashing when using mainline u-boot... basically it's because the binary Op-tee image was reserving some memory that Linux wasn't aware of and if you tried to write to it you'd get a crash. I pushed this commit https://github.com/macromorgan/odroid_g ... eb21b1b86e which is an ugly hack but should fix it.

So your options are either 1) use the fully mainline u-boot with mainline Arm Trusted Firmware which requires no patching; however you either have to nuke your SPI chip (don't worry it's recoverable) to force the device to boot off of SD or you have to hold the reset button on every boot-up or 2) use the Rockchip miniloader with the trust.img and an upstream u-boot image, but if you don't add the reserved memory to the devicetree your kernel will crash during or shortly after boot.

Also, noticed on my Android image that the specific issue with the wifi is actually wpa_supplicant. If the wireless network is not encrypted you can connect to it just fine. I'm still playing with it and trying to get Mesa 21.0-rc working since it brings Open GLES 3.0 to Bifrost chips.
There are surprisingly a lot of people who've tried mainline U-boot and ran into that issue. I've also seen it mentioned in the rockchip linux IRC channel unresolved.

I did a diff for wpa supplicant as it worked fine before in Android 10 but I couldn't find what broke authentication. Also it was only for the esp8089, my 3 Realtek usb WiFi sticks work normal.

Can you try this patch to see if it fixes the issue?

https://github.com/followmsi/android_de ... 18d332c95e
Testing it now.

One more thing I forgot to add, but putting it here for visibility's sake. If you go fully upstream with U-boot (upstream ATF, upstream SPL/TPL, etc) you have to change the value of CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR to 0x4000. Not sure why it's not this way already. If you use the default defconfig (odroid-go2_defconfig) on v2021.01 with just that one change you should be able to flash the u-boot-rockchip.bin that it generates to an SD card with dd (using seek=64 so something like

Code: Select all

dd if=u-boot-rockchip.bin of=/dev/mmcblk0 seek=64
and it will just work... provided you hold the reset button or nuke your SPI flash chip). Obviously you'll need a boot.scr since the boot.ini doesn't work anymore on upstream u-boot...

edit: that patch doesn't fix it. Still not sure what it could be.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

Doing my best to debug the wifi and putting notes in here in case it triggers something in someone else.

It appears to be wpa_supplicant itself that's failing on Android 11 for me with the ESP8089 wifi. Specifically, it's step 2 of the 4-Way handshake that's failing.
Unsuccessful handshake (Android 11):

Code: Select all

01-22 22:49:58.250   581   581 D wpa_supplicant: wlan0:   key_length=16 key_data_length=0
01-22 22:49:58.250   581   581 D wpa_supplicant:   replay_counter - hexdump(len=8): 00 00 00 00 00 00 00 01
01-22 22:49:58.251   581   581 D wpa_supplicant:   key_nonce - hexdump(len=32): 6a c4 54 db d3 5c 23 44 4d e7 73 9d f3 2b 19 be ee 48 d0 e5 40 fb c3 73 c8 bc 84 17 96 a5 8c 41
01-22 22:49:58.251   581   581 D wpa_supplicant:   key_iv - hexdump(len=16): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
01-22 22:49:58.251   581   581 D wpa_supplicant:   key_rsc - hexdump(len=8): 00 00 00 00 00 00 00 00
01-22 22:49:58.251   581   581 D wpa_supplicant:   key_id (reserved) - hexdump(len=8): 00 00 00 00 00 00 00 00
01-22 22:49:58.251   581   581 D wpa_supplicant:   key_mic - hexdump(len=16): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
01-22 22:49:58.251   581   581 D wpa_supplicant: wlan0: State: ASSOCIATED -> 4WAY_HANDSHAKE
01-22 22:49:58.252   581   581 D wpa_supplicant: Notifying state change event to hidl control: 7
01-22 22:49:58.255   581   581 I wpa_supplicant: wlan0: WPA: RX message 1 of 4-Way Handshake from 00:16:fb:b2:b6:c4 (ver=2)
01-22 22:49:58.256   581   581 D wpa_supplicant: RSN: msg 1/4 key data - hexdump(len=0):
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: Renewed SNonce - hexdump(len=32): fd ab 38 70 89 6e 01 05 4e 16 83 8a 4d 2f 04 b2 21 b8 12 f5 29 8c d5 df ef 9a 11 73 25 5f f6 77
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: PTK derivation using PRF(SHA1)
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: PTK derivation - A1=18:fe:34:dc:a2:52 A2=00:16:fb:b2:b6:c4
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: Nonce1 - hexdump(len=32): fd ab 38 70 89 6e 01 05 4e 16 83 8a 4d 2f 04 b2 21 b8 12 f5 29 8c d5 df ef 9a 11 73 25 5f f6 77
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: Nonce2 - hexdump(len=32): 6a c4 54 db d3 5c 23 44 4d e7 73 9d f3 2b 19 be ee 48 d0 e5 40 fb c3 73 c8 bc 84 17 96 a5 8c 41
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: PMK - hexdump(len=32): [REMOVED]
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: PTK - hexdump(len=48): [REMOVED]
01-22 22:49:58.257   581   581 D wpa_supplicant: WPA: KCK - hexdump(len=16): [REMOVED]
01-22 22:49:58.258   581   581 D wpa_supplicant: WPA: KEK - hexdump(len=16): [REMOVED]
01-22 22:49:58.258   581   581 D wpa_supplicant: WPA: TK - hexdump(len=16): [REMOVED]
01-22 22:49:58.258   581   581 D wpa_supplicant: WPA: WPA IE for msg 2/4 - hexdump(len=22): 30 14 01 00 00 0f ac 04 01 00 00 0f ac 04 01 00 00 0f ac 02 00 00
01-22 22:49:58.258   581   581 D wpa_supplicant: WPA: Replay Counter - hexdump(len=8): 00 00 00 00 00 00 00 01
01-22 22:49:58.258   581   581 I wpa_supplicant: wlan0: WPA: Sending EAPOL-Key 2/4
01-22 22:49:58.259   581   581 D wpa_supplicant: WPA: Send EAPOL-Key frame to 00:16:fb:b2:b6:c4 ver=2 mic_len=16 key_mgmt=0x2
01-22 22:49:58.259   581   581 D wpa_supplicant: WPA: EAPOL-Key MIC using HMAC-SHA1
01-22 22:49:58.259   581   581 D wpa_supplicant: WPA: KCK - hexdump(len=16): [REMOVED]
01-22 22:49:58.259   581   581 D wpa_supplicant: WPA: Derived Key MIC - hexdump(len=16): 83 3d 3f 4c be f9 01 a6 c2 6a 05 ce b1 22 19 fe
01-22 22:49:58.259   581   581 D wpa_supplicant: nl80211: Send over control port dest=00:16:fb:b2:b6:c4 proto=0x888e len=121 no_encrypt=1
01-22 22:49:58.261   581   581 D wpa_supplicant: WMM AC: AC mandatory: AC_BE=0 AC_BK=0 AC_VI=0 AC_VO=0
01-22 22:49:58.261   581   581 D wpa_supplicant: WMM AC: U-APSD queues=0x0
01-22 22:49:58.261   581   581 D wpa_supplicant: WMM AC: Valid WMM association, WMM AC is enabled
01-22 22:49:58.261   581   581 I wpa_supplicant: wlan0: CTRL-EVENT-SUBNET-STATUS-UPDATE status=0
01-22 22:49:58.262   581   581 D wpa_supplicant: RTM_NEWLINK: ifi_index=8 ifname=wlan0 wext ifi_family=0 ifi_flags=0x11003 ([UP][LOWER_UP])
01-22 22:49:58.263   581   581 I chatty  : uid=1010(wifi) /vendor/bin/hw/wpa_supplicant identical 1 line
01-22 22:49:58.264   581   581 D wpa_supplicant: RTM_NEWLINK: ifi_index=8 ifname=wlan0 wext ifi_family=0 ifi_flags=0x11003 ([UP][LOWER_UP])
01-22 22:49:58.265   581   581 D wpa_supplicant: nl80211: Drv Event 46 (NL80211_CMD_CONNECT) received for wlan0
01-22 22:49:58.265   581   581 D wpa_supplicant: nl80211: Ignore connect event (cmd=46) when using userspace SME
01-22 22:49:58.266   581   581 D wpa_supplicant: nl80211: Drv Event 36 (NL80211_CMD_REG_CHANGE) received for wlan0
01-22 22:49:58.266   581   581 D wpa_supplicant: nl80211: Regulatory domain change
01-22 22:49:58.267   581   581 D wpa_supplicant:  * initiator=3
01-22 22:49:58.267   581   581 D wpa_supplicant:  * type=0
01-22 22:49:58.267   581   581 D wpa_supplicant:  * alpha2=US
01-22 22:49:58.267   581   581 D wpa_supplicant: wlan0: Event CHANNEL_LIST_CHANGED (27) received
01-22 22:49:58.267   581   581 I wpa_supplicant: wlan0: CTRL-EVENT-REGDOM-CHANGE init=COUNTRY_IE type=COUNTRY alpha2=US
01-22 22:49:58.268   581   581 D wpa_supplicant: wlan0: Updating hw mode
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: Regulatory information - country=US (DFS-FCC)
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 2400-2472 @ 40 MHz 30 mBm
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 5150-5250 @ 80 MHz 23 mBm
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 5250-5350 @ 80 MHz 23 mBm (DFS)
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 5470-5730 @ 160 MHz 23 mBm (DFS)
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 5730-5850 @ 80 MHz 30 mBm
01-22 22:49:58.270   581   581 D wpa_supplicant: nl80211: 57240-71000 @ 2160 MHz 40 mBm
01-22 22:49:58.271   581   581 D wpa_supplicant: nl80211: Added 802.11b mode based on 802.11g information
01-22 22:49:58.271   581   581 D wpa_supplicant: nl80211: Mode IEEE 802.11g: 2412 2417 2422 2427 2432 2437 2442 2447 2452 2457 2462 2467[DISABLED][NO_IR] 2472[DISABLED][NO_IR]
01-22 22:49:58.272   581   581 D wpa_supplicant: nl80211: Mode IEEE 802.11b: 2412 2417 2422 2427 2432 2437 2442 2447 2452 2457 2462 2467[DISABLED][NO_IR] 2472[DISABLED][NO_IR]
01-22 22:49:58.273   581   581 D wpa_supplicant: nl80211: Drv Event 139 (NL80211_CMD_UNKNOWN) received for wlan0
01-22 22:49:58.273   581   581 D wpa_supplicant: wlan0: nl80211: Ignored unknown event (cmd=139)
01-22 22:49:58.306   461   896 I WifiService: updateInterfaceIpState uid=1000
01-22 22:49:58.307   461   541 D WifiService: updateInterfaceIpState: ifaceName=wlan0 mode=-1 previous LOHS mode= -1
01-22 22:49:58.318   461   471 I system_server: Background young concurrent copying GC freed 36130(1439KB) AllocSpace objects, 8(480KB) LOS objects, 16% free, 8542KB/10MB, paused 1.838ms total 146.013ms
01-22 22:49:59.086   581   581 D wpa_supplicant: nl80211: Drv Event 64 (NL80211_CMD_NOTIFY_CQM) received for wlan0
01-22 22:49:59.087   581   581 D wpa_supplicant: nl80211: Beacon loss event
01-22 22:49:59.087   581   581 D wpa_supplicant: wlan0: Event BEACON_LOSS (53) received
01-22 22:49:59.087   581   581 I wpa_supplicant: wlan0: CTRL-EVENT-BEACON-LOSS
01-22 22:50:00.006   604   604 D KeyguardClockSwitch: Updating clock: 1050
01-22 22:50:00.249   581   581 D wpa_supplicant: EAPOL: startWhen --> 0
01-22 22:50:00.250   581   581 D wpa_supplicant: EAPOL: disable timer tick
01-22 22:50:00.250   581   581 D wpa_supplicant: EAPOL: SUPP_PAE entering state CONNECTING
01-22 22:50:00.250   581   581 D wpa_supplicant: EAPOL: enable timer tick
01-22 22:50:00.250   581   581 D wpa_supplicant: EAPOL: txStart
01-22 22:50:00.250   581   581 D wpa_supplicant: WPA: drop TX EAPOL in non-IEEE 802.1X mode (type=1 len=0)
01-22 22:50:01.607   581   581 D wpa_supplicant: RTM_NEWLINK: ifi_index=8 ifname=wlan0 operstate=2 linkmode=1 ifi_family=0 ifi_flags=0x1003 ([UP])
01-22 22:50:01.609   461   478 I EthernetTracker: interfaceLinkStateChanged, iface: wlan0, up: false
01-22 22:50:01.625   581   581 D wpa_supplicant: nl80211: Drv Event 20 (NL80211_CMD_DEL_STATION) received for wlan0
01-22 22:50:01.625   581   581 D wpa_supplicant: nl80211: Delete station 00:16:fb:b2:b6:c4
01-22 22:50:01.636   581   581 D wpa_supplicant: nl80211: Drv Event 39 (NL80211_CMD_DEAUTHENTICATE) received for wlan0
01-22 22:50:01.636   581   581 D wpa_supplicant: nl80211: Deauthenticate event
01-22 22:50:01.636   581   581 D wpa_supplicant: wlan0: Event DEAUTH (11) received
01-22 22:50:01.636   581   581 D wpa_supplicant: wlan0: Deauthentication notification
01-22 22:50:01.637   581   581 D wpa_supplicant: wlan0:  * reason 4 (DISASSOC_DUE_TO_INACTIVITY) locally_generated=1
01-22 22:50:01.637   414   414 I wificond: Regulatory domain changed
01-22 22:50:01.637   581   581 D wpa_supplicant: wlan0:  * address 00:16:fb:b2:b6:c4
01-22 22:50:01.637   581   581 D wpa_supplicant: Deauthentication frame IE(s) - hexdump(len=0): [NULL]
01-22 22:50:01.637   581   581 I wpa_supplicant: wlan0: CTRL-EVENT-DISCONNECTED bssid=00:16:fb:b2:b6:c4 reason=4 locally_generated=1
01-22 22:50:01.639   581   581 I wpa_supplicant: wlan0: WPA: 4-Way Handshake failed - pre-shared key may be incorrect
Compared to a valid connect attempt (Debian Buster):

Code: Select all

Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0:   key_length=16 key_data_length=0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   replay_counter - hexdump(len=8): 00 00 00 00 00 00 00 01
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_nonce - hexdump(len=32): 8a 4c 02 15 78 8c 31 f4 17 04 4b cf e7 37 38 cf 3b e1 ad 4a d2 a3 49 02 f4 ea 7a 86 45 69 5c c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_iv - hexdump(len=16): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_rsc - hexdump(len=8): 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_id (reserved) - hexdump(len=8): 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_mic - hexdump(len=16): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: State: ASSOCIATED -> 4WAY_HANDSHAKE
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: RX message 1 of 4-Way Handshake from 00:16:fb:b2:b6:c4 (ver=2)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: RSN: msg 1/4 key data - hexdump(len=0):
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Renewed SNonce - hexdump(len=32): dc d9 cf 8a b3 8b 80 35 7e 69 12 fd 84 b8 dd ea d3 4f 0d f2 7a 6c 03 15 53 00 29 12 89 1d 80 44
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: PTK derivation using PRF(SHA1)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: PTK derivation - A1=18:fe:34:dc:a2:52 A2=00:16:fb:b2:b6:c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Nonce1 - hexdump(len=32): dc d9 cf 8a b3 8b 80 35 7e 69 12 fd 84 b8 dd ea d3 4f 0d f2 7a 6c 03 15 53 00 29 12 89 1d 80 44
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Nonce2 - hexdump(len=32): 8a 4c 02 15 78 8c 31 f4 17 04 4b cf e7 37 38 cf 3b e1 ad 4a d2 a3 49 02 f4 ea 7a 86 45 69 5c c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: PMK - hexdump(len=32): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: PTK - hexdump(len=48): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: KCK - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: KEK - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: TK - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: WPA IE for msg 2/4 - hexdump(len=22): 30 14 01 00 00 0f ac 04 01 00 00 0f ac 04 01 00 00 0f ac 02 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Replay Counter - hexdump(len=8): 00 00 00 00 00 00 00 01
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: Sending EAPOL-Key 2/4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Send EAPOL-Key frame to 00:16:fb:b2:b6:c4 ver=2 mic_len=16 key_mgmt=0x2
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: EAPOL-Key MIC using HMAC-SHA1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: KCK - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Derived Key MIC - hexdump(len=16): 0c ca c9 6f 55 38 6d 62 e3 ed 25 e7 71 dc 37 aa
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WMM AC: AC mandatory: AC_BE=0 AC_BK=0 AC_VI=0 AC_VO=0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WMM AC: U-APSD queues=0x0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WMM AC: Valid WMM association, WMM AC is enabled
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: CTRL-EVENT-SUBNET-STATUS-UPDATE status=0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Drv Event 46 (NL80211_CMD_CONNECT) received for wlan0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Ignore connect event (cmd=46) when using userspace SME
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Drv Event 36 (NL80211_CMD_REG_CHANGE) received for wlan0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Regulatory domain change
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:  * initiator=3
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:  * type=0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:  * alpha2=US
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: Event CHANNEL_LIST_CHANGED (27) received
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: CTRL-EVENT-REGDOM-CHANGE init=COUNTRY_IE type=COUNTRY alpha2=US
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: Updating hw mode
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Regulatory information - country=US (DFS-FCC)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 2400-2472 @ 40 MHz 30 mBm
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 5150-5250 @ 80 MHz 23 mBm
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 5250-5350 @ 80 MHz 23 mBm (DFS)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 5470-5730 @ 160 MHz 23 mBm (DFS)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 5730-5850 @ 80 MHz 30 mBm
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: 57240-71000 @ 2160 MHz 40 mBm
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Added 802.11b mode based on 802.11g information
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Mode IEEE 802.11g: 2412 2417 2422 2427 2432 2437 2442 2447 2452 2457 2462 2467[DISABLED][NO_IR] 2472[DISABLED][NO_IR]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Mode IEEE 802.11b: 2412 2417 2422 2427 2432 2437 2442 2447 2452 2457 2462 2467[DISABLED][NO_IR] 2472[DISABLED][NO_IR]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: l2_packet_receive: src=00:16:fb:b2:b6:c4 len=163
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: RX EAPOL from 00:16:fb:b2:b6:c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: IEEE 802.1X RX: version=2 type=3 length=159
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0:   EAPOL-Key type=2
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0:   key_info 0x13ca (ver=2 keyidx=0 rsvd=0 Pairwise Install Ack MIC Secure Encr)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0:   key_length=16 key_data_length=64
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   replay_counter - hexdump(len=8): 00 00 00 00 00 00 00 02
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_nonce - hexdump(len=32): 8a 4c 02 15 78 8c 31 f4 17 04 4b cf e7 37 38 cf 3b e1 ad 4a d2 a3 49 02 f4 ea 7a 86 45 69 5c c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_iv - hexdump(len=16): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_rsc - hexdump(len=8): 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_id (reserved) - hexdump(len=8): 00 00 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:   key_mic - hexdump(len=16): 78 a1 f0 9d fa 5c 14 40 ca 27 f8 e6 9e 89 0f a0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: EAPOL-Key MIC using HMAC-SHA1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: RSN: encrypted key data - hexdump(len=64): 2e b4 b0 e9 fe 6b 34 42 c5 28 13 8c e3 38 b5 34 31 bb bf 73 e4 90 e5 3e ba cc 4b 3b 09 cf 27 08 e0 3b ff ae c2 69 93 84 a7 fc 97 6c ee d7 25 02 f4 ad fe a3 a2 a8 02 86 40 5d eb 08 a1 b4 10 f3
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Decrypt Key Data using AES-UNWRAP (KEK length 16)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: decrypted EAPOL-Key key data - hexdump(len=56): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: State: 4WAY_HANDSHAKE -> 4WAY_HANDSHAKE
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: RX message 3 of 4-Way Handshake from 00:16:fb:b2:b6:c4 (ver=2)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: IE KeyData - hexdump(len=56): 30 18 01 00 00 0f ac 04 01 00 00 0f ac 04 02 00 00 0f ac 02 00 0f ac 04 0c 00 36 03 b0 0b 01 dd 16 00 0f ac 01 01 00 7f 77 58 f7 1d 6d 5b 18 68 bd 4f 2e 14 ee 98 eb dd
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: RSN IE in EAPOL-Key - hexdump(len=26): 30 18 01 00 00 0f ac 04 01 00 00 0f ac 04 02 00 00 0f ac 02 00 0f ac 04 0c 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: MDIE in EAPOL-Key - hexdump(len=5): 36 03 b0 0b 01
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: GTK in EAPOL-Key - hexdump(len=24): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: Sending EAPOL-Key 4/4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Send EAPOL-Key frame to 00:16:fb:b2:b6:c4 ver=2 mic_len=16 key_mgmt=0x2
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: EAPOL-Key MIC using HMAC-SHA1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: KCK - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Derived Key MIC - hexdump(len=16): f6 ec 11 9c 88 e4 a2 96 d0 36 04 46 67 a7 3d 87
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: Installing PTK to the driver
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wpa_driver_nl80211_set_key: ifindex=8 (wlan0) alg=3 addr=0x558913c218 key_idx=0 set_tx=1 seq_len=6 key_len=16
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: KEY_DATA - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: KEY_SEQ - hexdump(len=6): 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:    addr=00:16:fb:b2:b6:c4
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: EAPOL: External notification - portValid=1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: State: 4WAY_HANDSHAKE -> GROUP_HANDSHAKE
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: RSN: received GTK in pairwise handshake - hexdump(len=18): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: Group Key - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: Installing GTK to the driver (keyidx=1 tx=0 len=16)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: WPA: RSC - hexdump(len=6): 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wpa_driver_nl80211_set_key: ifindex=8 (wlan0) alg=3 addr=0x556dd0c510 key_idx=1 set_tx=0 seq_len=6 key_len=16
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: KEY_DATA - hexdump(len=16): [REMOVED]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: KEY_SEQ - hexdump(len=6): 00 00 00 00 00 00
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]:    broadcast key
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: WPA: Key negotiation completed with 00:16:fb:b2:b6:c4 [PTK=CCMP GTK=CCMP]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: Cancelling authentication timeout
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: State: GROUP_HANDSHAKE -> COMPLETED
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: Radio work 'sme-connect'@0x5589183480 done in 0.087819 seconds
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: radio_work_free('sme-connect'@0x5589183480): num_active_works --> 0
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: wlan0: CTRL-EVENT-CONNECTED - Connection to 00:16:fb:b2:b6:c4 completed [id=0 id_str=]
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: nl80211: Set wlan0 operstate 0->1 (UP)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: netlink: Operstate: ifindex=8 linkmode=-1 (no change), operstate=6 (IF_OPER_UP)
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: EAPOL: External notification - portValid=1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: EAPOL: External notification - EAP success=1
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: EAPOL: SUPP_PAE entering state AUTHENTICATING
Jan 22 21:18:27 OdroidGoAdvance wpa_supplicant[373]: EAPOL: SUPP_BE entering state SUCCESS
So it appears that at step 2 of the handshake Android gets a response it doesn't understand (Drv Event 139), whereas Debian gets an l2_packet and then completes the handshake.

Also, I have a hunch that the suspend fails to work on the esp8089 because there is nothing implemented for the resume from suspend function in the driver, it's literally just a print debug message and return 0.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

Annoyed the new device doesn’t have a UART 1 since I spent so much time figuring out how to route U-Boot over it. Still, good to see not much else has changed.

Image

I wrote a driver for the panel and committed it to my kernel branch (the driver is HEAVILY based off the st7701s driver already upstream). Note that I still have to implement the backlight controls. I’m also still occasionally playing with the driver for the other panel to try and get it rotated without having to use the RGA. The new panel on the OGS does not even support the MADCTL DSI command, so even if it works it will only be for those devices.

edit: just kidding, backlight works. Really wish I had a schematic...
These users thanked the author macromorgan for the post:
notime2d8 (Wed Feb 03, 2021 6:27 am)

notime2d8
Posts: 318
Joined: Thu Dec 11, 2014 4:40 am
languages_spoken: english
Has thanked: 93 times
Been thanked: 130 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by notime2d8 »

macromorgan wrote:
Wed Feb 03, 2021 5:48 am
Annoyed the new device doesn’t have a UART 1 since I spent so much time figuring out how to route U-Boot over it. Still, good to see not much else has changed.
Same, I was trying to use a trackball mouse or touchpad over I2C and the OGS just got big enough to mount it inside. And maybe Bluetooth or GPS over UART if I got bored.

Edit: creeping on your picture, you're using ethernet? Have you settled on a WiFi module to use?

Edit2: I haven't been able to get internet sharing over USB with an Android phone to work. Seems like it'll be a cool feature.
If you appreciate or used my work, consider buying me a coffee. https://www.paypal.me/notime2d8

User avatar
rooted
Posts: 8999
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 745 times
Been thanked: 422 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rooted »

@macromorgan The display is now glass, you can remove the plastic protective sheet without worrying about scratches. If you already knew continue on.

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

notime2d8 wrote:
Wed Feb 03, 2021 7:11 am
Have you settled on a WiFi module to use?
No, I can’t get wifi working with any adapter due to what I assume is an ABI compatibility issue (Android 11 with an Android 12 kernel).

macromorgan
Posts: 116
Joined: Mon Dec 30, 2019 10:35 am
languages_spoken: english
ODROIDs: Odroid-GO
Has thanked: 6 times
Been thanked: 45 times
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by macromorgan »

macc24 was kind enough to point to this git commit done by valadaa48 which when implemented seems to allow suspend and resume of the USB (by basically disconnecting and reconnecting the USB at suspend/resume). Works for my needs for a network adapter, though I would recommend against using USB storage with this patch applied and suspend/resume enabled.

Thank you valadaa48 for figuring this out and macc24 for pointing me in the right direction. I'm going to talk to the dwc2 team to see if they can figure this out and maybe get a fix upstreamed.

valadaa48's patch: https://github.com/valadaa48/linux/comm ... d38f436807

Applied to my kernel branch: https://github.com/macromorgan/odroid_g ... 9a31c9d348
These users thanked the author macromorgan for the post (total 2):
notime2d8 (Fri Feb 19, 2021 7:04 am) • valadaa48 (Fri Feb 19, 2021 10:33 am)

User avatar
rfrayer
Posts: 44
Joined: Thu Feb 06, 2020 10:13 am
languages_spoken: english
ODROIDs: xu4, hc1
Location: Rutledge, Tn USA
Has thanked: 3 times
Been thanked: 1 time
Contact:

Re: Anybody working on Android for the Odroid-Go A?

Post by rfrayer »

I would love to see ATV ported. lol

Post Reply

Return to “Platform development”

Who is online

Users browsing this forum: No registered users and 1 guest