Hi! Because of lack of time I was unable to continue my development of LineageOS/Android Wear 2.0 for Moto 360. I created this thread to find someone who can help fix SGX drivers (probably lack of some code in Android 8.0(, because Android does not start because of them. I'm publishing all repos and sources I've already created.
Firstly we need to focus on booting LineageOS, because It is the easiest way to boot Android 7 on Moto 360. Google didn't publish sources for Android Wear 2.0, so we will need to port it using image from other watch and our Android 7.1 drivers.
https://github.com/artas182x/android_device_motorola_minnow (branch cm-14.1)
https://github.com/artas182x/kernel-omap/tree/android-7.1-wear (branch android-7.1-wear)
https://github.com/artas182x/vendor_motorola_minnow (branch master)
Logcat with SGX error: https://pastebin.com/WKDhjEzX
If you have a problem with SGX Driver you can take inspire from this project: Unlegacy-Android. They took an Omap4 Device (Samsung Galaxy Nexus) and try to import all the necessary source from omapzoom repository to build a fully functional android 7.1.
retsef said:
If you have a problem with SGX Driver you can take inspire from this project: Unlegacy-Android. They took an Omap4 Device (Samsung Galaxy Nexus) and try to import all the necessary source from omapzoom repository to build a fully functional android 7.1.
Click to expand...
Click to collapse
Moto 360 user newer SGX driver and newer kernel than Galaxy Nexus.
Wysłane z iPhone za pomocą Tapatalk
I Hope this project Will continue, because i want wear 2.0 in my moto 360!!
I am very helpful with troubleshooting builds that boot.
I've worked on GS6 for sprint.
I've done a TON of testing for moto e 2015.
I currently have GS7 and moto 360 1st gen.
Please let me know how I can help get lineage/wear 2.0 going.
I will help as much as I can and willing to donate monetarily.
Good to see people working on this. Just got my moto 360 and was bummed to see the lack of development for it.
Let's make this happen! Going to Google your issue and see of it has dropped up before.
Thanks for all the time you've spent already dev!
I also just made my interface cable. Not tested yet. Will tonight.
I also sent a pm to a Dev who has ported wear 2.0 to the LG g watch. Hopefully I hear back and he can help us out!
Edit:
I got a response from another Dev. He had ported AW 2.0 for the LG G Watch.
Mentioned it might be easier to use another os and swap out libs and other files. I will copy text and post his full reply next.
His response:
"Unfortunately, I have no experience of working with OMAP devices.
Anyway, why are you building lineage? I think, using system image of another device as a base and patching it with device-specific configs and libs would be easier"
Thoughts on this @artas182x ?
crazynapkinman said:
His response:
"Unfortunately, I have no experience of working with OMAP devices.
Anyway, why are you building lineage? I think, using system image of another device as a base and patching it with device-specific configs and libs would be easier"
Thoughts on this @artas182x ?
Click to expand...
Click to collapse
Hmm Moto 360 is only one smartwatch based on OMAP3. What's more it's has newer kernel and SGX drivers than Galaxy Nexus and other OMAP Android devices. I don't know how to fix SGX problem. Probably it's the last problem we need to solve. Everything else should work.
Unfortunately, we don't have open-source SGX drivers for Moto 360. There are some on the internet but in earlier version than we need. So we need to find how to run prebuilt drivers.
All my sources are on my GitHub, you should be able to build Lineage OS from that files.
What do you mean by "prebuilt drivers"? The drivers already on device?
And we would need to make everything work with those existing drivers rather than getting new drivers that work with everything else?
@artas182x could you point me in the direction for the sgx drivers online? Are they Intel drivers? Or....?
Sorry for all questions. Trying to research and see what I can do or find someone to help us.
I did just find this one:
https://github.com/nemomobile/ti-omap3-sgx
What about this one?
https://01.org/intel-softwareguard-extensions/downloads/intel-sgx-linux-2.0-release
Hey to both of you. I have had a moto 360 for a while and am more than willing to help. Do we wanna maybe setup a telegram chat or something similar. For easy communication.
dylanneve1 said:
Hey to both of you. I have had a moto 360 for a while and am more than willing to help. Do we wanna maybe setup a telegram chat or something similar. For easy communication.
Click to expand...
Click to collapse
Sure. I have never used it. But ok
Im rly glade to see that some ppl are trying to bring Wear 2.0 on the Moto 360 Gen1, i would like to help but i can´t program ^^
So i just can say thanks
Great to see some news
I am willing to help if you need any testers etc.
Keep it up guyz!
Spoke with Motorola today.
I took a picture of the convo. No luck...
Does anybody know if there is some android based device compatible with our hardware? If it's popular in community we could help devs with development and maybe be able to get working driver...
Deleted
Here's the wiki.
I have a nook color! But is that too old?
What about droid?
https://t.me/joinchat/FPu8TkfPJE8URYWg9x7HxA
Here's a group link
crazynapkinman said:
Here's the wiki.
I have a nook color! But is that too old?
What about droid?
Click to expand...
Click to collapse
Seems like Moto Droid 4 has OMAP as well and working 7.1. But idk if that OMAP is compatible.
https://forum.xda-developers.com/droid-4/development/rom-cm14-1-nougat-t3499338
viktorkovarik said:
Seems like Moto Droid 4 has OMAP as well and working 7.1. But idk if that OMAP is compatible.
https://forum.xda-developers.com/droid-4/development/rom-cm14-1-nougat-t3499338
Click to expand...
Click to collapse
It's omap4 not sure
Related
[This is not the place to say "this is awesome" or "thanks!"]
[DEVS ONLY]
I want to start this thread to keep up the progress on the port.
the bad news is that the SDK is incomplete for now, so (like other devices)
we will have to write our own code for the OS, the nook community
has done a wonderful job writing their own libraries and stuffs, so we will have to do the same.
Instead of pursuing different goals, let's focus on one thing at a time.
since GSM and CDMA versions are already out, it seems we suffer the same bugs, so for now let's unite strength and knowledge to overcome these.
I propose that the first goal to fix is the SurfaceFlinger, so we could at least see the apps, the buttons and the notifications.
it is currently throwing this:
04-06 22:01:35.495: ERROR/Surface(2960): dequeueBuffer failed (Out of memory)
this could lead us that it might coudln't start because of some malloc malfunction or something.
also that pvrsrvinit bugs me a lot.
update:
8/APR/2011
since DiP7 could fix SurfaceFlinger and other things using a different build from the GSM kernel, we have to dig deeper
======================================
CURRENT GOAL
Rebuild Kernel
======================================
Current approaches:
*none
Post any finding, guessing or anything, and please, please don't be afraid to ask anything you have a doubt, as a Dev you should not know everything, so we can help us each other
------------------------------------------
Google Easter egg:
while searching some info about the android.mk file , I put it on the chrome bar to search for that term, but instead I went to
http://android.mk
an easter egg web page from google lol
******TOOLS*********
How to send text and Keystrokes via ADB
http://bradchow.blogspot.com/2011/02/send-intent-and-key-event-by-adb.html
use DroidExplorer to easily access your device from your computer and makes changes from it
http://de.codeplex.com/
The Android Boot Process
http://www.androidenea.com/2009/06/android-boot-process-from-power-on.html
---Kernel Tools----
CPU Datasheet
http://forum.xda-developers.com/showthread.php?t=745877
Samsung GIT
http://android.git.kernel.org/?p=kernel/samsung.git;a=summary
PowerVR SDK
http://www.imgtec.com/powervr/insider/powervr-sdk.asp
Source code of samsung firmwares(keep and eye on this)
http://opensource.samsung.com/
Asus pad honeycomb Kernel Source
http://forum.xda-developers.com/showthread.php?t=1026528
ellokomen said:
======================================
CURRENT GOAL
Find why SurfaceFlinger is not working
======================================
Click to expand...
Click to collapse
Mmmk.. Let me tell you a story.
A long time ago in a galaxy far away... No, that will take too long. In a nutshell, one third of the answer is here, another third is here and the rest is here. I'm not trying to be intentionally vague, I just haven't figured out how these three fit together yet.
Of course, there could also be some bit of code that I missed... some telling line in a debug log that I overlooked or some driver or library that I could have decompiled to sift through its juicy secrets. There could easily be a fix that would take seconds to add and make the whole thing fall in line...
Or we might have to work it from the ground up.
Either way it will happen. It's just a matter of whether it will happen next week, or next month.
(Watch it be a misplaced semi-colon, or a bad symlink... that's how these things go.)
updated approaches and new tools have been added
spacemoose1 said:
Mmmk.. Let me tell you a story.
A long time ago in a galaxy far away... No, that will take too long. In a nutshell, one third of the answer is here, another third is here and the rest is here. I'm not trying to be intentionally vague, I just haven't figured out how these three fit together yet.
Of course, there could also be some bit of code that I missed... some telling line in a debug log that I overlooked or some driver or library that I could have decompiled to sift through its juicy secrets. There could easily be a fix that would take seconds to add and make the whole thing fall in line...
Or we might have to work it from the ground up.
Either way it will happen. It's just a matter of whether it will happen next week, or next month.
(Watch it be a misplaced semi-colon, or a bad symlink... that's how these things go.)
Click to expand...
Click to collapse
The kernel source that you have linked to is 2.6.35.7 for the Nexus S gingerbread and is not fully maintained by samsung but rather by google.
However I do not think you are wrong that there is a problem somewhere in the kernel. The kernel that has been released for the galaxy tab is a mess of horrid code, I have had to re-write parts of kernel drivers just to get them to work under linux, I would not be surprised if similar patches are needed for honeycomb
lilstevie said:
The kernel source that you have linked to is 2.6.35.7 for the Nexus S gingerbread and is not fully maintained by samsung but rather by google.
However I do not think you are wrong that there is a problem somewhere in the kernel. The kernel that has been released for the galaxy tab is a mess of horrid code, I have had to re-write parts of kernel drivers just to get them to work under linux, I would not be surprised if similar patches are needed for honeycomb
Click to expand...
Click to collapse
you mean that you made a port of a Linux Distro into the tab?
ellokomen said:
you mean that you made a port of a Linux Distro into the tab?
Click to expand...
Click to collapse
yes click here for the thread on the port of ubuntu
Current kernel source
Do you guys have another link to the current spacemoose kernel source? The download link seems to be corrupted and won't untar. I want to get in on the fun
noobporter said:
Do you guys have another link to the current spacemoose kernel source? The download link seems to be corrupted and won't untar. I want to get in on the fun
Click to expand...
Click to collapse
here it is, bear in mind that this is for CDMA devices
Unfortunatey, we have 4 other honeycomb threads.
Not trying to be rude but spacemoose updates us in the cdma forums AND we have russian rom updates in the gsm forums.
I really dislike the idea of this thread, there is enough clutter amongst the other threads. Do we really need one more place to browse..
The first posts in the roms thread are kept updated by devs.. Is this not enough??
daml said:
Unfortunatey, we have 4 other honeycomb threads.
Not trying to be rude but spacemoose updates us in the cdma forums AND we have russian rom updates in the gsm forums.
I really dislike the idea of this thread, there is enough clutter amongst the other threads. Do we really need one more place to browse..
The first posts in the roms thread are kept updated by devs.. Is this not enough??
Click to expand...
Click to collapse
yeah but we need a place for the other devs to share their milestones, here is a place for technical discussion amongst us, to share the knowledge etc...
the other threads are flooded from non devs messages, so it´s kind of difficult to read 14 pages of information when the 80% is people complaining not making it boot
lilstevie said:
The kernel source that you have linked to is 2.6.35.7 for the Nexus S gingerbread and is not fully maintained by samsung but rather by google.
Click to expand...
Click to collapse
Yes, and it contains some support for our device (s5pc110), and some more that can be added (pvr) and the architecture necessary to fully support HC without patching the build itself. If we work only towards patching the system build to communicate with the hardware, we won't be able to run AOSP hc versions when the source drops without going through the same painstaking process of hacking the system to function (while creating numerous faults causing FCs in the process). If we build a new kernel, we can get the hardware to communicate in the way future android versions want it to and we can then do what we want with ease.
noobporter said:
Do you guys have another link to the current spacemoose kernel source? The download link seems to be corrupted and won't untar. I want to get in on the fun
Click to expand...
Click to collapse
D'oh! Nobody told me, LOL... I'll get another copy up.
spacemoose1 said:
D'oh! Nobody told me, LOL... I'll get another copy up.
Click to expand...
Click to collapse
http://forum.xda-developers.com/showthread.php?t=1026528
hey spacemoose! the first portion of honeycomb source... The kernel source of the Asus EEE Pad Transformer... maybe it helps you with a few kernel issues, even if it's for another device... It's honeycomb!
Flokey said:
http://forum.xda-developers.com/showthread.php?t=1026528
hey spacemoose! the first portion of honeycomb source... The kernel source of the Asus EEE Pad Transformer... maybe it helps you with a few kernel issues, even if it's for another device... It's honeycomb!
Click to expand...
Click to collapse
Digging through it now.
spacemoose1 said:
Yes, and it contains some support for our device (s5pc110), and some more that can be added (pvr) and the architecture necessary to fully support HC without patching the build itself. If we work only towards patching the system build to communicate with the hardware, we won't be able to run AOSP hc versions when the source drops without going through the same painstaking process of hacking the system to function (while creating numerous faults causing FCs in the process). If we build a new kernel, we can get the hardware to communicate in the way future android versions want it to and we can then do what we want with ease.
Click to expand...
Click to collapse
Not enough really, PVR kernel module sources have been released from samsung for our device, and is available in update1 zip.
The kernel panics and we have no framebuffer from the nexus s, believe me that is the kernel I want to be running for my project, it is cleaner nicer and things are implemented overall better. unless you know of a solution for kernel debuging over usb
lilstevie said:
Not enough really, PVR kernel module sources have been released from samsung for our device, and is available in update1 zip.
The kernel panics and we have no framebuffer from the nexus s, believe me that is the kernel I want to be running for my project, it is cleaner nicer and things are implemented overall better. unless you know of a solution for kernel debuging over usb
Click to expand...
Click to collapse
Kernel debugging over USB = adb shell cat /proc/kmsg
You can make any kernel work for any device as long as you add the **** it needs. Just takes time. Working on it now.
Goal and tools updated*
spacemoose1 said:
Kernel debugging over USB = adb shell cat /proc/kmsg
You can make any kernel work for any device as long as you add the **** it needs. Just takes time. Working on it now.
Click to expand...
Click to collapse
You don't know what a kernel panic is do you?
lilstevie said:
You don't know what a kernel panic is do you?
Click to expand...
Click to collapse
I thought that linux throws a dump log when it makes a kernel panic specifying the memory address and the cause of crash
Is there a way that someone can make a Lollipop ROM for SM-T210R? Preferably CM 12.1 or a stock ROM. I have seen multiple people on these forums ask when this will happen. I've seen the Galaxy Tab 2 AND the ORIGINAL Galaxy Tab get Lollipop before Tab 3. The Tab 4 already got it...
Thanks, and I hope that we will see a day where everyone will have the latest version of Android on their devices...
TrendingTech said:
Is there a way that someone can make a Lollipop ROM for SM-T210R? Preferably CM 12.1 or a stock ROM. I have seen multiple people on these forums ask when this will happen. I've seen the Galaxy Tab 2 AND the ORIGINAL Galaxy Tab get Lollipop before Tab 3. The Tab 4 already got it...
Thanks, and I hope that we will see a day where everyone will have the latest version of Android on their devices...
Click to expand...
Click to collapse
Tab 3 has closed source I wouldnt expect any CM in the future
Sent from my white GT-I9305
why?
Minto107 said:
Tab 3 has closed source I wouldnt expect any CM in the future
Sent from my white GT-I9305
Click to expand...
Click to collapse
How come, I don't understand, there are serveral other ROM's for KITKAT, why would it be impossible to build a Lollipop ROM? I have NO idea how to build a rom, otherwise I would give it a try.
What makes it so difficult?
We don't have drivers to build any custom ROM rather than Samsung based one. We probably can have MIUI based on Sammy but if you want custom ROMs you've chosen a wrong tablet. Tab 2 7 would do a job correctly
Sent from my white GT-I9305
Minto107 said:
We don't have drivers to build any custom ROM rather than Samsung based one. We probably can have MIUI based on Sammy but if you want custom ROMs you've chosen a wrong tablet. Tab 2 7 would do a job correctly
Sent from my white GT-I9305
Click to expand...
Click to collapse
isn't there a Galaxy Tab 3 Lite lollipop ROM that we can port over to SM-T210R?
TrendingTech said:
isn't there a Galaxy Tab 3 Lite lollipop ROM that we can port over to SM-T210R?
Click to expand...
Click to collapse
None of TAB 3 family devices got 5.x so there's nothing to port. Tab 3 7,8,10 won't be updated to 5.x
Wysłane z mojego Lenovo TAB S8-50L przy użyciu Tapatalka
Tab 3 8.0 has unofficial cm-12.1
Tab 3 8.0 has unofficial cm-12.1 and tab 3 10.1 also got it,
Grynot said:
Tab 3 8.0 has unofficial cm-12.1 and tab 3 10.1 also got it,
Click to expand...
Click to collapse
Omg but 8.0 uses Exynos 10.1 uses Intel and 7 uses ****ty Marvel CPU! You don't have source so no CM for you
Wysłane z mojego Lenovo TAB S8-50L przy użyciu Tapatalka
So, im just stuck with effin 4.2? Since i originally rooted over a year ago, i have then tried to go back to stock to update, but ii can never get an ota update.....im just frustrated with all this. Trying to teach myself and getting stuck, SUCKS!!! I KNOW I DONT USE it to its fullest potential.....ANY help would rock....oh and would help my brain hurt LESS!! Thanks! !
rheecollinskelley said:
So, im just stuck with effin 4.2? Since i originally rooted over a year ago, i have then tried to go back to stock to update, but ii can never get an ota update.....im just frustrated with all this. Trying to teach myself and getting stuck, SUCKS!!! I KNOW I DONT USE it to its fullest potential.....ANY help would rock....oh and would help my brain hurt LESS!! Thanks! !
Click to expand...
Click to collapse
You know that Tab 3 8.0 is only 4.2.2 supported, not even close to KitKat, although you can find yourselves SM-T21x Kit Kat roms. Lollipop will not be supported by any chance, even for custom roms man!
It really depends on the amount of work you are willing to do.
Some parts are binary blob only *at this time*, but may change in the future.
For example, the Vivante GPU driver has some patches under review for the mainline Linux kernel and a open source driver in progress (look up etnaviv).
If you *really* want the Tab 3 7.0" to have Lollipop, you have to be willing to do some work. Maybe try to get some of the drivers that are used in the device into the mainline kernel. Maybe get basic functionality working.
In any case, it really is up to the community at this point, and we are part of the community.
vorpalblade77 said:
It really depends on the amount of work you are willing to do.
Some parts are binary blob only *at this time*, but may change in the future.
For example, the Vivante GPU driver has some patches under review for the mainline Linux kernel and a open source driver in progress (look up etnaviv).
If you *really* want the Tab 3 7.0" to have Lollipop, you have to be willing to do some work. Maybe try to get some of the drivers that are used in the device into the mainline kernel. Maybe get basic functionality working.
In any case, it really is up to the community at this point, and we are part of the community.
Click to expand...
Click to collapse
Yeah but XDA is full of noobs that only asks when, why it takes so long time... Bringing Lollipop to TAB 3 7inch is very hard work... This tablet wasn't a beast, since release date it was low end tablet with such tragic performance. For such noobs who just want and doesn't want to help I've got an advice. Get Nexus 7 2013... 6.0 on board, noob friendly
Wysłane z mojego Lenovo TAB S8-50L przy użyciu Tapatalka
As I know so far, Marvell didn't release source codes for Galaxy Tab 3, so it's impossible to build CM12 for this device.
DroidTwe4kz said:
As I know so far, Marvell didn't release source codes for Galaxy Tab 3, so it's impossible to build CM12 for this device.
Click to expand...
Click to collapse
I *thought* everything that was necessary to build a working ROM was in Samsung's source dumps. Maybe not all the information required, but all the source code.
For that matter, I was under the impression that the unofficial Cyanogenmod port was working (Cyanogenmod 11) with an unofficial kernel.
Am I wrong?
If I am, then that might be why I have yet to successfully build a booting recovery, but other people have.
vorpalblade77 said:
I *thought* everything that was necessary to build a working ROM was in Samsung's source dumps. Maybe not all the information required, but all the source code.
For that matter, I was under the impression that the unofficial Cyanogenmod port was working (Cyanogenmod 11) with an unofficial kernel.
Am I wrong?
If I am, then that might be why I have yet to successfully build a booting recovery, but other people have.
Click to expand...
Click to collapse
@gr8nole built unofficial CM11 from his own source. He wrote device specific tree I think.
vorpalblade77 said:
I *thought* everything that was necessary to build a working ROM was in Samsung's source dumps. Maybe not all the information required, but all the source code.
For that matter, I was under the impression that the unofficial Cyanogenmod port was working (Cyanogenmod 11) with an unofficial kernel.
Am I wrong?
If I am, then that might be why I have yet to successfully build a booting recovery, but other people have.
Click to expand...
Click to collapse
The only sources that Samsung provides is kernel sources. That is not nearly enough to build a completely working rom.
DroidTwe4kz said:
@gr8nole built unofficial CM11 from his own source. He wrote device specific tree I think.
Click to expand...
Click to collapse
Yes, I did build from my own device tree, but there were other devs that contributed a lot of the early work on that tree. I'm not going to try to mention all the names because I don't want to leave anyone out. But, suffice it to say, it wasn't just me. :good:
To wrap up this post, the reason I gave up on CM for that device is because the Marvel hardware was too difficult (for me) to get working properly. Things like Wifi and hardware acceleration never worked well (or at all) and I don't have the ability to get them working any better. With essentially no other Marvel based devices to use as a guide, it was beyond my capabilities.
gr8nole said:
The only sources that Samsung provides is kernel sources. That is not nearly enough to build a completely working rom.
Yes, I did build from my own device tree, but there were other devs that contributed a lot of the early work on that tree. I'm not going to try to mention all the names because I don't want to leave anyone out. But, suffice it to say, it wasn't just me. :good:
To wrap up this post, the reason I gave up on CM for that device is because the Marvel hardware was too difficult (for me) to get working properly. Things like Wifi and hardware acceleration never worked well (or at all) and I don't have the ability to get them working any better. With essentially no other Marvel based devices to use as a guide, it was beyond my capabilities.
Click to expand...
Click to collapse
You already did more than enough for our device and I'm thankful for that. Someone will continue to build CM based on your source.
gr8nole said:
To wrap up this post, the reason I gave up on CM for that device is because the Marvel hardware was too difficult (for me) to get working properly. Things like Wifi and hardware acceleration never worked well (or at all) and I don't have the ability to get them working any better. With essentially no other Marvel based devices to use as a guide, it was beyond my capabilities.
Click to expand...
Click to collapse
For hardware acceleration, did you ever look at the etnaviv project? I know that they are working on getting patches upstreamed for the GPU on the SM-T210 (Vivante GC 1000)?
For that matter, the Marvell SD8787 wifi chip has firmware mainlined (as of 2011, updated in 2015) and has driver support in "drivers/net/wireless/mwifiex/sdio.h" as of commit 4a7f5db1 (take a look at CONFIG_MWIFIEX). I don't know if its worth attempting to backport or not (some issues may have been fixed). Maybe look at the linux-firmware tree?
I'm pretty certain that at one point Marvell was looking at getting a device tree for the PXA986/PXA988 into the mainline kernel, but I don't think they succeeded.
Parts on the SM-T210R that I haven't found in the mainline kernel is the touchscreen. I haven't looked at everything else yet, and I'm not *absolutely* certain that the touchscreen isn't in the mainline kernel either (I might be looking at a config option and assuming the wrong thing from Samsung's sources).
EDIT1: I've been trying to build cyanogenmod on systems that aren't officially supported by AOSP, which is why I may have never been able to build it myself and *actually* contribute and check to see if etnaviv works, and if the mainline driver works better. Anyway, thanks for your work gr8nole (and everyone else's).
EDIT2: For SD8787 bluetooth support, look at CONFIG_BT_MRVL/CONFIG_BT_MRVL_SDIO
vorpalblade77 said:
For hardware acceleration, did you ever look at the etnaviv project? I know that they are working on getting patches upstreamed for the GPU on the SM-T210 (Vivante GC 1000)?
For that matter, the Marvell SD8787 wifi chip has firmware mainlined (as of 2011, updated in 2015) and has driver support in "drivers/net/wireless/mwifiex/sdio.h" as of commit 4a7f5db1 (take a look at CONFIG_MWIFIEX). I don't know if its worth attempting to backport or not (some issues may have been fixed). Maybe look at the linux-firmware tree?
I'm pretty certain that at one point Marvell was looking at getting a device tree for the PXA986/PXA988 into the mainline kernel, but I don't think they succeeded.
Parts on the SM-T210R that I haven't found in the mainline kernel is the touchscreen. I haven't looked at everything else yet, and I'm not *absolutely* certain that the touchscreen isn't in the mainline kernel either (I might be looking at a config option and assuming the wrong thing from Samsung's sources).
EDIT1: I've been trying to build cyanogenmod on systems that aren't officially supported by AOSP, which is why I may have never been able to build it myself and *actually* contribute and check to see if etnaviv works, and if the mainline driver works better. Anyway, thanks for your work gr8nole (and everyone else's).
EDIT2: For SD8787 bluetooth support, look at CONFIG_BT_MRVL/CONFIG_BT_MRVL_SDIO
Click to expand...
Click to collapse
I haven't looked at any of this in a long time, and have no real desire to at this point.
gr8nole said:
I haven't looked at any of this in a long time, and have no real desire to at this point.
Click to expand...
Click to collapse
Fair enough. I'll probably look into once I know I have my build environment set up correctly. E.g., I can build a recovery and have it work correctly.
This is a test build I take no responsibility for what this may or may not do to your device.
Click to expand...
Click to collapse
NEED TESTERS
currently i don't have the device to test i will be getting my device in the next week .
Click to expand...
Click to collapse
1. Install CWM Recovery
2. Do Full Wipe If coming from another ROM
3. Flash ROM
4. Flash Google Apps (my recommendation is to install minipal gapps due to /system partition size limits)
5. Reboot
Click to expand...
Click to collapse
]Not Booting
Click to expand...
Click to collapse
Code:
[B]Source[/B]
Here are the device tree :
Device Tree (from cm13 Nexus 9) : https://github.com/vickdu31/android_device_xiaomi_mocha
Kernel Tree (from Xiaomi KK) : https://github.com/vickdu31/android_kernel_xiaomi_mocha
Vendor Tree (old Xiaomi KK) : https://github.com/vickdu31/android_vendor_xiaomi_mocha
Thank you
@faust93 for cwm
XDA:DevDB Information
CM-13, ROM for the Xiaomi Mi Pad
Contributors
Rohit99, tank0412, vickdu31
ROM OS Version: 6.0.x Marshmallow
Based On: CyanogenMod
Version Information
Status: Testing
Created 2016-12-10
Last Updated 2016-12-13
Reserved
Edit 1:initial build didn't boot for me, no adb can't get log
Ok im gonna try it.
If you want, i can leave logcat for you
OoSTARTERoO said:
Ok im gonna try it.
If you want, i can leave logcat for you
Click to expand...
Click to collapse
Can get logcat? Try please.
Ofc, it doesn't boot. And don't create threads without testing.
Sudeep Duhoon said:
Can get logcat? Try please.
Click to expand...
Click to collapse
unfortunately....I couldn't get logcat too...
Sudeep Duhoon said:
Can get logcat? Try please.
Click to expand...
Click to collapse
we can't get logcat. It doesn't boot, i mean there are problems with kernel or init scripts or both. Kernel needs patches.
Nihhaar said:
Ofc, it doesn't boot.
Click to expand...
Click to collapse
As I thought
Prebuilt kernel from cm11
Sudeep Duhoon said:
As I thought
Click to expand...
Click to collapse
hey can i use prebuilt kernel from cm11 source @Nihhaar
Rohit99 said:
hey can i use prebuilt kernel from cm11 source @Nihhaar
Click to expand...
Click to collapse
Always try to build kernel with source. I think you can't, but you can try.
Please edit the link in OP, its me who miswrote it... : /android_device_xioami_mocha --> /android_device_xiaomi_mocha
You need to fix up SELinux and the init scripts. Taking a shield tablet device tree, renaming things to mocha and changing the fstab isn't enough. As far as I can tell the shieldtablet device tree also has dependency on a shield common tree. There's also no value in creating a new thread each time you push new commits
Our kernel from Xiaomi is also pretty useless without commit history. The device tree is also from shield tablet, not Nexus 9.
We also have a tree for building TWRP with OmniROM that you can compile TWRP 3.0.2.0 with.
What kernel repo is upstream one for tegra android? Like CAF for quallcom.
I've tried to merge xiaomi changes on l4t kernel https://github.com/HighwayStar/android_kernel_xiaomi_mocha
Merged dts files and arch/boot/march-tegra with some required drivers, but still cant boot it. Found somewhere here that l4t kernels is not for android, but for GNU/Linux, but what kernel is for android?
Cyanogenmod's shield kernel?
highwaystar_ru said:
What kernel repo is upstream one for tegra android? Like CAF for quallcom.
I've tried to merge xiaomi changes on l4t kernel https://github.com/HighwayStar/android_kernel_xiaomi_mocha
Merged dts files and arch/boot/march-tegra with some required drivers, but still cant boot it. Found somewhere here that l4t kernels is not for android, but for GNU/Linux, but what kernel is for android?
Cyanogenmod's shield kernel?
Click to expand...
Click to collapse
git://nv-tegra.nvidia.com/linux-3.10.git
I used rel-tn8-l-r7-shieldtablet8 to make a start on doing this. You could also use android_kernel_shield from CyanogenMod repo as it comes from the same place. But it will have a lot of extra commits for things not useful for mocha.
Here's what I have so far:
https://github.com/harrynowl/nvidia_tegra-3.10
Harrynowl said:
git://nv-tegra.nvidia.com/linux-3.10.git
I used rel-tn8-l-r7-shieldtablet8 to make a start on doing this. You could also use android_kernel_shield from CyanogenMod repo as it comes from the same place. But it will have a lot of extra commits for things not useful for mocha.
Here's what I have so far:
https://github.com/harrynowl/nvidia_tegra-3.10
Click to expand...
Click to collapse
Oh, intresting, cloned same url, but started other branch l4t/l4t-r21.5.
What status of your tree? Cant boot? We need minimal bootable kernel to be able to get dmesg and make further progress.
highwaystar_ru said:
Oh, intresting, cloned same url, but started other branch l4t/l4t-r21.5.
What status of your tree? Cant boot? We need minimal bootable kernel to be able to get dmesg and make further progress.
Click to expand...
Click to collapse
Untested, been doing device bring up. That kernel won't work yet anyway it's missing atmel, synaptic and battery firmware. (Plus the code that goes with it)
Some devs have one that gets to recovery but only 1 CPU core can come online and the CPU gets a bit hot. Unfortunately their commit history is hard to follow so I've not had a proper look at it yet
My my...it's been a while since I've seen a cringe-worthy dev thread on XDA. So OP just used the device tree from Nexus 9 (a 64-bit device with a different processor and nothing in common with the Mi Pad other than the fact that they use Nvidia SoCs and are both tablets), unmodified KITKAT kernel sources from Xiaomi (which takes work to make it boot, even on KK) and KITKAT userspace blobs to compile the ROM, and for some reason decided to share it without testing. The 10-post rule is meant to keep such people out of the development sub-forums, but obviously that's not working.
I do not understand why the moderators are allowing this thread to remain open. It's just a thread by a wannabe dev offering a non-bootable ROM that has the potential to brick the devices of many newbies who may decide to try this.
This thread is even more useless than the other "DEVS-ONLY" thread someone else made. Honestly, the entire Mi Pad forum is filled with people who either expect Android 7.1.1 to magically appear on their tabs just because Xiaomi released outdated kernel sources or just wanna create threads like this for the "thanks or whatever". The never ask for ETA rule seems to just float over their heads.
Now to some useful information. I believe, even if we patch the kernel enough for it to theoretically boot cm13, it probably won't, considering the outdated userspace blobs for the Mi Pad. Unless someone has the knowledge to take their tab apart and figure out how to get UART data, it will be virtually impossible for the mi pad kernel to boot any new version of Android. I heard another interesting approach taken by some russian devs at the Xiaomi forums. They've managed to port the kernel from the Nvidia shield (and apparently it's booting). Now this would mean we can use the userspace blobs from the Shield to at least get the ROM booting (most peripherals won't work, but that's step 2). I do not know whether they use GitHub or if they've even open-sourced their work, but now we know that it's possible, so that could be a possible method of approach.
EDIT: LOL. I just read the OP again. The "dev" expects donations to help him work harder. This from a person who posted his way into somehow compiling his first Android build without even having the damn device.
drakonizer said:
My my...it's been a while since I've seen a cringe-worthy dev thread on XDA. So OP just used the device tree from Nexus 9 (a 64-bit device with a different processor and nothing in common with the Mi Pad other than the fact that they use Nvidia SoCs and are both tablets), unmodified KITKAT kernel sources from Xiaomi (which takes work to make it boot, even on KK) and KITKAT userspace blobs to compile the ROM, and for some reason decided to share it without testing. The 10-post rule is meant to keep such people out of the development sub-forums, but obviously that's not working.
I do not understand why the moderators are allowing this thread to remain open. It's just a thread by a wannabe dev offering a non-bootable ROM that has the potential to brick the devices of many newbies who may decide to try this.
This thread is even more useless than the other "DEVS-ONLY" thread someone else made. Honestly, the entire Mi Pad forum is filled with people who either expect Android 7.1.1 to magically appear on their tabs just because Xiaomi released outdated kernel sources or just wanna create threads like this for the "thanks or whatever". The never ask for ETA rule seems to just float over their heads.
Now to some useful information. I believe, even if we patch the kernel enough for it to theoretically boot cm13, it probably won't, considering the outdated userspace blobs for the Mi Pad. Unless someone has the knowledge to take their tab apart and figure out how to get UART data, it will be virtually impossible for the mi pad kernel to boot any new version of Android. I heard another interesting approach taken by some russian devs at the Xiaomi forums. They've managed to port the kernel from the Nvidia shield (and apparently it's booting). Now this would mean we can use the userspace blobs from the Shield to at least get the ROM booting (most peripherals won't work, but that's step 2). I do not know whether they use GitHub or if they've even open-sourced their work, but now we know that it's possible, so that could be a possible method of approach.
EDIT: LOL. I just read the OP again. The "dev" expects donations to help him work harder. This from a person who posted his way into somehow compiling his first Android build without even having the damn device.
Click to expand...
Click to collapse
I agree, this device has no development threads. Only 2 give me thanks and here's my donation link threads.
They have indeed used GitHub, but I don't believe they want to share the repo just yet so I won't post it publicly. It has many problems to sort and also sadly, they didn't keep the commit history from Nvidia as it was initialised via zip DL. It does boot to recovery though.
Harrynowl said:
I agree, this device has no development threads. Only 2 give me thanks and here's my donation link threads.
They have indeed used GitHub, but I don't believe they want to share the repo just yet so I won't post it publicly. It has many problems to sort and also sadly, they didn't keep the commit history from Nvidia as it was initialised via zip DL. It does boot to recovery though.
Click to expand...
Click to collapse
Oh! I didn't realize we were talking about the same person/team. I'd love to start developing for this device. In fact, I'm working on KEXEC-hardboot and Multirom right now, but I'd like to follow what you guys are doing and contribute as much as I can. Is there an IRC channel/IM app you guys use for communication? If so, I'd appreciate it if you could PM the info.
Hello,
As some people heard, i really try to get Android Wear 2.0 for our beloved SmartWatch 3.
And it's possible, sources, device trees, all exist to build a working own compiled Android Wear 2.0.
So this thread tells about the current process of the port
PROGRESS
Code:
Repo synced
First build finished (not buildable yet due no blobs included
GAPPS Porting in progress
Fixed to 99% the device tree
Updating Blobs to M1D64T in progress
Using own Device tree
SOURCES
Device Tree for tetra:
https://github.com/Sony-SmartWatch-3-OSS/android_device_tree_tetra_eng_prerelease
Device Kernel for tetra (unmodified):
https://android.googlesource.com/kernel/bcm/+/android-wear-6.0.1_r0.99
(we can use MM kernel on Nougat too, maybe some patches needed but that's okay)
Prebuilts/Blobs for tetra:
https://github.com/Sony-SmartWatch-3-OSS/vendor_prebuilts_tetra
For Discussions go to this thread please: https://forum.xda-developers.com/smartwatch-3/general/companion-thread-android-wear-2-0-t3565693
If you want to help me on this project, write to me via telegram: https://t.me/PDesire
I need that so much any ETA ?
By the way keep up the good work !
Nice to see someone is working on this :victory:
Sent from my Sony Xperia Z5 using XDA Labs
Thanks for your job.
Enviado desde mi Redmi Note 3 mediante Tapatalk
great!!! thanks
Dear PDesire,,
My support can be only moral..... :crying:
And I hoping good news soon from you
Thx again for your great hard work :good:
thankx you great hard work and good luck my friend
I checked out your device tree. It looks like it's based off an Open Devices tree. I don't think we need the following lines
Code:
include device/sony/common-headers/KernelHeaders.mk
-include device/sony/common-kernel/KernelConfig.mk
in BoardConfig.mk.
Also, what about the HALs? Asteroid OS used the system.img from a stock build, but we will need to build from source. I do not think it is open source, as it is not available on google git servers.
EDIT: I do not think building from source will work. This is what I found the Google Android building group:
Certain components of Android (and therefore Wear) come from third parties and have licenses which require the source code to be made available (e.g. the Linux kernel). The components made available via the AOSP repositories for releases like Wear are, afaik, the components which have that type of license.
Wear is a release which has taken the same path as Honeycomb. I can understand that for open source fans and people looking to tinker with a working system it’s somewhat disappointing to see, but those are the choices Google has made.
Click to expand...
Click to collapse
Looks like the only way to go forward is to port it from a pre-compiled version for another watch, like they did for the Galaxy Gear devices.
EDIT 2: Refer:
https://groups.google.com/forum/#!s...nce/android-building/R0r9BBn4M3g/-KexIWTCDwAJ
https://groups.google.com/forum/#!s...nce/android-building/vx0Hcy88Ofs/USxAUaKVAwAJ
These sources are incomplete and only include GPL licensed projects. It is not the complete source, just like google releases preview sources for AOSP preview builds (M, N etc). I suggest you not waste your quota in downloading the source.
drakonizer said:
I checked out your device tree. It looks like it's based off an Open Devices tree. I don't think we need the following lines
Code:
include device/sony/common-headers/KernelHeaders.mk
-include device/sony/common-kernel/KernelConfig.mk
in BoardConfig.mk.
Also, what about the HALs? Asteroid OS used the system.img from a stock build, but we will need to build from source. I do not think it is open source, as it is not available on google git servers.
Click to expand...
Click to collapse
Yeah I've seen these, and already fixed. Patch comes soon.
Things for WLAN and BT for example are on google repo too and if a library or something is missing we can use it from M1D64T
EDIT: I do not think building from source will work. This is what I found the Google Android building group:
Looks like the only way to go forward is to port it from a pre-compiled version for another watch, like they did for the Galaxy Gear devices.
EDIT 2: Refer:
https://groups.google.com/forum/#!s...nce/android-building/R0r9BBn4M3g/-KexIWTCDwAJ
https://groups.google.com/forum/#!s...nce/android-building/vx0Hcy88Ofs/USxAUaKVAwAJ
These sources are incomplete and only include GPL licensed projects. It is not the complete source, just like google releases preview sources for AOSP preview builds (M, N etc). I suggest you not waste your quota in downloading the source.
Click to expand...
Click to collapse
Lel then I can't understand why the build is working xD at least I am building in time and already 14% finished
I know some libs we have to grab from the Marshmallow FW but that shouldn't be a big deal
At least we also need some things from Preview 2 so we can get our AW 2.0 Preview 2 working
So I try my best to grab much from source and if required also from FW, but all in all its possible for our SmartWatch 3 to get Android Wear 2.0
Your PDesire
but there's one thing i don't understand ... why r u using preview 2 instead of a feature complete newer version or the final version?
xpatze85x said:
but there's one thing i don't understand ... why r u using preview 2 instead of a feature complete newer version or the final version?
Click to expand...
Click to collapse
Because there aren't any sources of the final version.
@PDesire :
You're the man...keep up a good work ???
Hey @PDesire , any news? I'm sooooo hyped!!!
Thank bro I hope
Thanks for this!
So current status is:
Code:
Repo synced
First Test build stopped to fix things on device tree
Expected in future:
Code:
Blobs dump
Fully fixed Device tree
Creating vendor for tetra
Adding overlays for framework to device tree
Some mix with armv7-a-neon generic
Gapps Dump
Expected in far future:
Code:
Creating an OSS project for tetra (like sonyxperiadev but for tetra)
So guys I will do more tomorrow as I am awake already from Friday 7AM - 3AM Saturday (right now GMT +1) and working on it and I need my sleep xD
So Goodnight guys and I can confirm our dream of Android Wear 2.0 for SmartWatch 3 comes nearer
Your PDesire
PDesire said:
So Goodnight guys and I can confirm our dream of Android Wear 2.0 for SmartWatch 3 comes nearer
Click to expand...
Click to collapse
And I can confirm my donation is coming when a working build is ready, and I suppose others are coming, isn't it guys? ???
Is it possible that this could give us Android pay or will that only work with official AW2?
Sent from my SM-G920F using XDA-Developers Legacy app
Neo Cortex said:
Is it possible that this could give us Android pay or will that only work with official AW2?
Sent from my SM-G920F using XDA-Developers Legacy app
Click to expand...
Click to collapse
No. That'll only work on an official firmware, unfortunately...
Thank you very much, hope you will finish it without problems~!
Hi, I'm working on porting LineageOS to the J327P (j3popltespr), I have the kernel ready but the problem that I've encountered is the vendor tree. There seems to be no vendor tree available for the j3popltespr. I found a vendor tree for the 2016 J3 (github.com/djeman/android_device_samsung_j3xnlte - link to GitHub for the j3xnlte) which is a similar model in regards to being the predecessor of this device but otherwise have different CPUs (j3xnlte/J320 having an Exynos 3475 or Spreadtrum SC9830 CPU and j3poplte/J327P having a Qualcomm MSM8937 CPU) I don't know if I could use that vendor tree or if I'd have to manually get components from my own phone. There seems to be a phone named the BQ Aquaris U which has the same CPU and is officially supported by Lineage which might be a good place to start. Any help would be appreciated, thank you!
HUAWEI HONOR 7C also uses MSM8937 CPU.
AgentICS said:
Hi, I'm working on porting LineageOS to the J327P (j3popltespr), I have the kernel ready but the problem that I've encountered is the vendor tree. There seems to be no vendor tree available for the j3popltespr. I found a vendor tree for the 2016 J3 (github.com/djeman/android_device_samsung_j3xnlte - link to GitHub for the j3xnlte) which is a similar model in regards to being the predecessor of this device but otherwise have different CPUs (j3xnlte/J320 having an Exynos 3475 or Spreadtrum SC9830 CPU and j3poplte/J327P having a Qualcomm MSM8937 CPU) I don't know if I could use that vendor tree or if I'd have to manually get components from my own phone. There seems to be a phone named the BQ Aquaris U which has the same CPU and is officially supported by Lineage which might be a good place to start. Any help would be appreciated, thank you!
Click to expand...
Click to collapse
Hey did you ever have any luck? Or need me to mess around with mine or her anything for you?
EirikrHinnRauthi said:
Hey did you ever have any luck? Or need me to mess around with mine or her anything for you?
Click to expand...
Click to collapse
User sands207 was leading dev work on this device until a big fiasco caused him to cease development. I contacted him a few years ago after I messed up the baseband on my j327p and couldn't figure out how to make wifi work anymore. He graciously helped me out. Perhaps he still has some files that may be of use to restart new dev work.
funkybluemonkey3 said:
User sands207 was leading dev work on this device until a big fiasco caused him to cease development. I contacted him a few years ago after I messed up the baseband on my j327p and couldn't figure out how to make wifi work anymore. He graciously helped me out. Perhaps he still has some files that may be of use to restart new dev work.
Click to expand...
Click to collapse
Oh no! Do you know what kind of fiasco? Is it OK you think for me to @ him in here to check in?
EirikrHinnRauthi said:
Oh no! Do you know what kind of fiasco? Is it OK you think for me to @ him in here to check in?
Click to expand...
Click to collapse
It's a long read but this was my starting point for the j327p. https://forum.xda-developers.com/t/...-mobile-sprint-sprint-xas-and-virgin.3637741/ I believe it was him and another dev SonderTech that were the gurus for this device. I forget whom I actually messaged. If you read through all the links and pages it looks like they got it as far as 6.1.0 with quite a performance boost. I think what was preventing it further was something do with the kernel or device tree for the next version never being released for j327p.