Surface 3 (Atom x7) does not boot

[Bernard]

I have a Surface 3 4/128 in my possession at the moment. It's a slick piece of kit, but all the Android-x86 builds do not boot.

I see "Detecting Android-x86..." and then light on USB flash drive looks like it loads the system, but no GUI presents.

I've tried Sandman01's Surface builds as well as the new 4.4-r3. This would definitely be the tablet I'd settle with if I could get Android working.

How can I help?

[Fajar A. Nugraha]

On Wed, Jul 22, 2015 at 6:51 AM, Bernard <ma...@bernarddime.com> wrote:
> I have a Surface 3 4/128 in my possession at the moment. It's a slick piece of kit, but all the Android-x86 builds do not boot.
>
> I see "Detecting Android-x86..." and then light on USB flash drive looks like it loads the system, but no GUI presents.

Did it say something like "found on /dev/sda1" ?

>
> I've tried Sandman01's Surface builds as well as the new 4.4-r3. This would definitely be the tablet I'd settle with if I could get Android working.
>
> How can I help?

If it says "found" but the GUI is not loaded, I'd try testing vesa mode.

If it keeps on stuck at "detecting", whatever device you run it from
(e.g. sd card) does not have the drivers loaded. Try using another
device (e.g. usb drive).

--
Fajar

[Bernard]

Yes. Tried vesa mode as well (I think I did it correctly)

[Bernard]

Anyone get this working on S3?

[Jim]

Bernard於 2015年7月22日星期三 UTC+8上午7時51分49秒寫道：

No, I try on my Surface 3 and get black screen too. Waiting for someone can do this working.

[rbg]

FWIW: search these forums....

FWIW:

Here's 4.0 build: 

https://mega.co.nz/#!RUV3gAiJ!nKnKTPS94UtMNAGcrvbeAjoh8FntJ-WqHZF2mw2J36s

Here's 4.1rc8 build (very broken, broken wifi): 

https://mega.co.nz/#!UIcSVarR!Mb6I1oCvE5IfwzfCpNWSHVbheDqn1DbxoE9xUqbmBKo

https://drive.google.com/file/d/0B43N_UcsJhpjOEVRRFdMZGNKMWc/view

[Bernard]

Tried all 3 images, none of them work.

Detects Android-x86, finds it, then goes to black screen with no GUI and stays there.

[pstglia]

Hi,

 

Tried all 3 images, none of them work.

Detects Android-x86, finds it, then goes to black screen with no GUI and stays there.

Can you get logcat / dmesg outputs? This will help trying to identify what's wrong/missing

Regards,

Pstglia

[Bernard]

I have done considerable searching for the past hour, and there is no straightforward guide to save logcat/dmesg. Honestly, there isn't. There's posts here and there of people doing it in 50 different ways.

If you can please tell me what to do, I can do this immediately. I'll keep searching, but I'm not finding anything.

[pstglia]

Em sábado, 1 de agosto de 2015 22:08:08 UTC-3, Bernard escreveu:

I have done considerable searching for the past hour, and there is no straightforward guide to save logcat/dmesg. Honestly, there isn't. There's posts here and there of people doing it in 50 different ways.

If you can please tell me what to do, I can do this immediately. I'll keep searching, but I'm not finding anything.

This is one of the threads with a procedure to do this (see 2 last posts from Mauro Rossi)

https://groups.google.com/forum/#!searchin/android-x86/cp$20$2Fdata$2Flog.txt/android-x86/8IZLSYGRjic/JaTr38e0VEAJ 

Regards,

Pstglia

[Bernard]

Tried that, no go.

When I boot in debud mode, it gets to the point where it says hit exit to continue booting. If I do that, I cannot pull up anything with Alt-F2, etc, it just gets stuck on the black screen.

Running any of those commands returns errors.

[pstglia]

 

Tried that, no go.

When I boot in debud mode, it gets to the point where it says hit exit to continue booting. If I do that, I cannot pull up anything with Alt-F2, etc, it just gets stuck on the black screen.

Running any of those commands returns errors.

Ok. We can try to get at least logcat by now. This will require you to install Android-x86 with "writable /system" (this can be selected during install)

1) After installing, run installed Android-x86 under debug mode and type the 2 exits until you got stucked with black screen.

2) Reboot your machine with boot it in another Operating Linux SO 

3) mount Android-x86 installed partition (Ubuntu does this automatically), and locate the data directory. There you'll find a file named log.txt, which is saved logcat.

You can also use Windows to access this file with a program called "ext2explore" (if your partition was formated to ext2, ext3 or ext4). If you formated as FAT, just select Android-x86 installed partition/drive and get the file

Regards,

Pstglia

 

[Bernard]

Yea, I'll do that later. I was just hoping this could be done using 2 USB drives, without a full install of Android + bootloader.

[pstglia]

Em sábado, 1 de agosto de 2015 23:29:54 UTC-3, Bernard escreveu:

Yea, I'll do that later. I was just hoping this could be done using 2 USB drives, without a full install of Android + bootloader.

You can try to install it on your 2nd USB drive. Did this once.  

[Bernard]

I'm not even able to get a log.

Installed Android-x86 to partition as FAT32 so I can access it within Windows. When I try booting to it in debug mode, it only gets to the FIRST "type exit to continue booting." I type exit and it ends at the black screen.

When I go back into Windows and browse into the FAT32 partition and into the android-4.4-r3 folder, 5 files exist exactly like they do on the live USB (data.img, initrd.img, kernel, ramdisk.img, system.img). It seems like it didn't even get far enough for data and system folders to be populated?

There's no log anywhere.

On Saturday, August 1, 2015 at 6:54:38 PM UTC-7, pstglia wrote:

[Bernard]

[Saipudin Haris]

How to mount data.img on Windows:

1. install Ext2FSD ----> http://sourceforge.net/projects/ext2fsd/files/latest/download?source=recommended
2. download and run OSFMount ------> http://www.osforensics.com/downloads/osfmount.exe (32bit)
3. Mount DATA.IMG using OSFMount.
4. double click on mounted data.img to open in windows explorer.

[Saipudin Haris]

I use these methode to get some data, and tombstones/crash log file from data.img.

[Bernard]

Thank you Saipuidin,

Prior to 4.4-r3, when I would install Android on an ext4 partition and then use ext2fsd in windows, it always showed up as folders mimicking the file system as if you were using the android device. 

Did things change with 4.4-r3, keeping data packaged within an image file now? Or is it just acting this way because I chose FAT32 this time?

[Bernard]

I have data.img mounted in Windows, but I cannot find a log file anywhere, in any of the nested folders.

[Mike R]

You're not going to..... Why would it be placed in a Static File?

On 8/2/2015 5:27 PM, Bernard wrote:

I have data.img mounted in Windows, but I cannot find a log file anywhere, in any of the nested folders.

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-x86...@googlegroups.com.
To post to this group, send email to andro...@googlegroups.com.
Visit this group at http://groups.google.com/group/android-x86.
For more options, visit https://groups.google.com/d/optout.

[pstglia]

Em domingo, 2 de agosto de 2015 19:27:42 UTC-3, Bernard escreveu:

I have data.img mounted in Windows, but I cannot find a log file anywhere, in any of the nested folders.

Should be inside data directory. File name is log.txt. 

Strange that every time I install here,  /system  is a directory of the partition I selected to install Android (I always select it as writable)

bash-4.3$ pwd

/media/caerrrrrr-aaaaaa-zzzzzz-yyyyyy-xxxxxxxxxxxxx/android-2015-06-28/data

bash-4.3$ ls -l log.txt

-rw------- 1 root root 672800 Aug  2 16:04 log.txt

bash-4.3$ mount | grep cae

/dev/sda8 on /media/caerrrrrr-aaaaaa-zzzzzz-yyyyyy-xxxxxxxxxxxxx type ext4 (rw,nosuid,nodev,uhelper=udisks)

However, I use iso files. If you are using img/efi files, this can be different. 

[Bernard]

Well, there's nothing.

Like I said, it will say "Detecting Android-x86," find it, go through what seems like initial boot stuff when in debug mode. Than I type exit to continue booting (first exit option) and it freezes with black screen there. I never get to a second option to type exit again. 

I have to power it off by holding the power button down at that point. From there, I mount the disk within Windows and don't find anything.

What can I do at this point?

[Mike R]

Just Wondering, and guess not sure?!?!?! 

Does it Actually Store Information inside of the Data.IMG, and System.IMG??  Without the special setup the C.W. Mentioned Earlier??

If So things have changed alot, and for ALL my READING, I missed it...

8-)

Really just wondering...

Mike

[Mike R]

Oh, N/M, Guess that's the BASH command?

Mike

On 8/2/2015 6:03 PM, pstglia wrote:

[Saipudin Haris]

Hi Bernard,

If the data.img contains no logs, you could try to install android x86 into a harddrive partition or a flash drive partition.
it could be possible the logs located on your android x86 system partitions. refer to pstglia instruction on previous post how to install
Note: for the logs to be created, your android x86 system partitions must have "writeable" access.

Could you upload your data.img to google drive and provide the link so every body could "see" what's inside?
Please compress your data.img using 7zip before upload.

Also If you can, could you dump your android x86 system partition and upload it too?
you could use isobuster (or other raw partition save to image software) to dump your android x86 system partition. 
If yes please compress your dumped android x86 system partition using 7zip
then upload it to google drive so every body can "see" what's inside.

[Bernard]

I did install Android-x86 to the Surface 3's drive; that's what I'm talking about in the last few posts. It is an eMMC drive, could that be the issue?

I just tried mounting the system partition as well using OSFMount and couldn't find anything in there that's worthwhile either.

Also, what's up with the new 2GB limit for /data in the 4.4-r3 installer? I've always formatted a partition ext4 with whatever size I wanted and then installed Android-x86 to it without formatting. 

[Mike R]

Hey, Sorry To interject here, I'm not a DEV, and been "trying" to learn this for along time, The 2Gig Limit is your installing without R/W permissions, My Install Right Now has an 8Gig Partition, Which is actually getting to SMALL!! 

Not Sure, I haven't used the new install I'm still on RC2.... Cause Fuzzy Made my Synaptics Touchpad into a Touchscreen Emulator... I Get Everyone else has a "touchscreen" but I don't, And I've Asked a Million times on what I need to edit to rebuild the kernel (The Latest) the way I need it and Really get no answers...

So the Best I can Say From awhile ago, Is if you install W/O R/W permissions, your stuck on 2Gigs....

Now Believe I could be wrong, just saying what happened to me on my Old Intel 1750 Inspiron,.,,

Mike

--

[Bernard]

You are wrong in this case. During install I made sure (wiped the partition and installed a second time to make sure) that I chose YES to R/W permissions.

It only seems to happen with the 4.4-r3 release img. You're allowed to choose between a range of 512MB-200MB. Let me see what rbg's 7/21 iso build does. 

[Saipudin Haris]

Could you dump your eMMC drive partition which contains the installed Android x86?
If yes, then dump it, compress it using 7zip, upload to google drive and provide
the link so every body can "see" what happend with your android x86 installation.

[Mike R]

Ok, Wow, That's Kinda Strange Because The R2 will do Whatever Partition You make For it, But you said IMG, So I'm gonna Assume it's an UEFI Format, Which I Know Nothing, But will keep reading to see how it works out..

[Bernard]

Ok, so I reinstalled using rbg's 7-21-15 4.4-r3 ISO.

There's something different going on between the EFI and ISO installers. The ISO installed like I've known Android-x96 to install: didn't ask to specify size for data img, and when viewing the partition within Windows using ext2fsb, the data and system parts are in folders that are browsable, not packaged in .img files. Not sure what the change was for.

Saipudin, can I just package the data and system folders into a .7z file? Here's the thing though, the data folder is completely empty. I don't think the Android-x86 boot process even gets far enough to populate data?

[Mike R]

Ya and On the new RBG Build, Could Be Better?, I haven't heard of a Limit On the GIG's but maybe it's a UEFI thing, Or The Builder of the new installer?, Not Sure really...I think I would have heard about it tho....  Like I said I'll be Listening..

On 8/2/2015 7:31 PM, Bernard wrote:

[Mike R]

Ya, I'm gonna guess, because it was installed R/W???

As I said In My Last Post...Maybe Missed or Something on the 1st Install, Or Not EXT4, Dunno....

Mike

[Mike R]

Wasn't

[Mike R]

In Any case sounds like you got it, have fun!  8-)

On 8/2/2015 7:47 PM, Bernard wrote:

[Mike R]

And On This Exact issue, I'm gonna simply say, If it ASKS for the the SIZE of the install, You ARE going READ ONLY......

You said yourself, Why It's happening Dunno, But that's the Slandered for an Install, It's Either R/O With A Limit, OR R/W with no limit, It Won't or shouldn't ask on a true root install....

Hope It helps...

On 8/2/2015 7:47 PM, Bernard wrote:

[Bernard]

Then there's something wrong with the 4.4-r3 UEFI installer. I tried twice, making sure to choose Yes for R/W permissions, and it still asked to choose size. I then tried rbg's test build ISO and it installed fine.

Still doesn't change the fact that I can't get any logs and there's no data within the partition.

[Mike R]

Ok, But now I'm Confused, did you have the install R/W?  And it could be the new installer build, It works on some but not others, I've got a r8 build over on my site, but, like i said, I don't use them just provide, Maybe Paulo, Or someone can help better, But I do know the DL's I have provided by RBG are very Experimental...

You'll have to ask C.W. On the R3 build I Believe, I'm thinking your gonna have to go of the Official Builds, To see what works.... And if they don't you gotta google and go straight with the test builds....

As I Said, I'm just trying to help....

Mike

[Mike R]

If your being asked to check size , regardless Your installing on R/O.... Talk to the DEVS to get them to fix it, and Until Then step back To an Older Version a Search the Group to get an install answer....

On 8/2/2015 8:10 PM, Bernard wrote:

[Bernard]

I have android installed now using rbg's test build. Still doesn't boot on this Surface 3 and I'm trying to get logs for the devs, but I can't. I've mounted partitions in Windows and there's absolutely no files in the data folder. It seems like Android-x86 doesn't get far enough along in the boot process or something to populate data. 

Maybe it's the eMMC drive or the Atom-x7 Z8700 that's causing these issues. I'd like to help but I don't know what I can do. I can upload the system and data folders in .7z archive, but the system is probably exactly what it's like as if you were building an image on a Linux machine, and the data folder is blank.

[Bernard]

Hey Devs,

So I installed Android-x86 to the eMMC drive of the Surface 3 in every combination possible: NTFS, FAT32, ext3, ext4 and I get the same result everytime. It doesn't seem to get far enough along in the boot process to populate any data. When it freezes on black screen and I go back and mount the partition in Windows, the data folder is completely empty.

I was able to set up a tripod and catch the entire screen after trial and error in debug mode. Attached is all I see before the screen goes black and the tablet freezes.

[Saipudin Haris]

Chih-Wei Huang (12/06/13) response on Android x86 crashing:

> 1. What could possibly be the reason of crashing ? 

Anything is possible.

> 2. What are the approaches of identifying causes of crash ?

Check the logcat.

If you boot (Android x86 in) the DEBUG mode, it's recorded in
/data/log.txt.

=========
https://groups.google.com/forum/#!topic/android-x86/600nkl6mfwA

[Bernard]

> 2. What are the approaches of identifying causes of crash ?

Check the logcat.

If you boot (Android x86 in) the DEBUG mode, it's recorded in
/data/log.txt.

The data folder is EMPTY, even after booting in DEBUG mode. 

[Saipudin Haris]

How to Obtain Crash Information for Android* App on the Android platform

On the Android platform, you can use a debuggerd daemon process for collecting error information when applications crash. This process can detect application crashes and then store error information in files or print it to a log so that you can analyze it and locate problems. These files are named tombstones_xx and stored in sequence within the folder /data/tombstone/. There can only be ten files in the system, so the earliest file generated will be replaced once the maximum number of files has been reached.

=====
https://software.intel.com/en-us/android/articles/how-to-obtain-crash-information-for-android-app-on-x86-platform


Android Tombstones & It's Locations

Working with Android, every time a process that runs on native side crashes, we have some small pieces of your stack in files called tombstones.

The tombstones are located at /data/tombstones as isolated files (one files represents one crash) or you can see them in your logcat:

root@android:/ # find . |grep tombs

./data/tombstones
./data/tombstones/tombstone_00
./data/tombstones/tombstone_01
root@android:/ #

The tombstone inform you about:

1. Build fingerprint
2. Crashed process and PIDs
3. Terminated signal and fault address
4. CPU registers
5. Call stack
6. Stack content of each call

=====
http://bytesthink.com/blog/?p=133

On Friday, August 7, 2015 at 2:59:45 AM UTC+7, Bernard wrote:

[Bernard]

You sure do insist on fighting me about this.

[Saipudin Haris]

Could you confirm again that you have an android x86 "writable /system" like pstglia instructions?

Could you dump the drive partition and zip it with 7zip then upload it to google drive?
If yes, I'll wait for your the google drive link for the dumped partition.

could you post your grub/syslinux configuration?

[Saipudin Haris]

Could you provide the folder content screenshot for number 1 and number 2 (see in attached image)

On Friday, August 7, 2015 at 3:43:06 AM UTC+7, Bernard wrote:

[Bernard]

1) Screenshot of my folders

2) Screenshot of my grub.cfg. It's the same one I use for most of my Android-x86 installs.

[Saipudin Haris]

I'll try to replicate your configuration on my mbp, i'll update once i finish doing it.

[Bernard]

I have this exact same setup running perfectly on a Surface Pro 3 i3-4xxx. It was running perfectly on a Toshiba Portege Z10t i7-4610Y. I believe it will work fine on your MBP.

It's something with the Atom x7, if I was to guess.

[Saipudin Haris]

I wan't to trace the log creation failure by replicating you configuration.
i'll notify you once i finish testing it, if it worth enough to report.

in the mean while you can talk to others in this group.

[rbg]

check his Secure Boot settings....??

[Bernard]

Secure boot is off.

[rbg]

ok thanks....and live usb runs / boots fine correct?

[Bernard]

No, the Surface 3 does not boot at all, Live USB or installed to eMMC. I installed it to eMMC so I could try and get a log, but it's not happening.

[rbg]

IMO

so if live doesn't boot at all...

you created a fat32 partition, created an ext4 partition...

copied .img files / kernel / etc. to a folder on the fat32 partition and copied

the boot info to the ext4 partition, made changes in the windows bootmge to boot from ext4 partition

that has the android-x86 boot info on it pointing to the folder on the fat32 partition

it doesn't boot at all and/or gives errors / etc.

so I feel pstglia is correct in that you need to provide Logcat / dmesg / log.txt form the

boot from disk, without it there is no way to actually see what's going on....and the photos

you have supplied really don't show anything as far as what is really happening.....

check these forums as others have solved your issue of getting the logcat / dmesg / etc. when using

debug mode, dual flash devices / etc.

does ubuntu live flash boot correctly / install and boot correctly....if not why not.....

others have successfully booted, installed, provided logs from android-x86 to surfacePro 3...you need to do some

research / work and learn what you may be doing wrong or provide the logs.....

[Bernard]

Hi rbg,

Please let me clarify a few things you said.

I did not “copy .img files to a folder. What I did, and what I have always done to install Android-x86 is this:

1. Create a small ~100MB FAT32 partition that houses grub with manual configuration for Android-x86 (let’s call it dev/sda5), so that I don’t need to add grub to the Windows ESP and can make changes easily. The Android-x86 boot parameter points to dev/sda6).
2. Create whatever size ext4 partition that I want following the grub partition, where I’m installing Android-x86 to (let’s call it sda6).
3. Boot live USB and install Android-x86 to sda6.
4. From there, I use rEFInd bootloader so I can use a custom theme. It sees Windows Boot Manager and Grub (so I have Windows and Android icons). I choose whichever one I want at boot, and it boots the corresponding OS.

This has worked flawlessly on various hardware to boot and run Android-x86. At this point, every machine I’ve tried is a Core M/i3/i5/i7 (Ivy Bridge, Haswell, Broadwell).

What I am talking about here is the SURFACE 3 with an ATOM Z8700 SoC, not the Surface Pro 3. I have a Surface Pro 3 that is successfully running Android-x86 using this exact procedure.

The Surface 3 however, does not boot Android-x86 whether I run Live USB or installed on the eMMC drive. It just stops booting and freezes on a black screen. I tried installing Android-x86 to various partition formats to see if it makes a difference. It doesn’t.

I am telling you, when I install Android-x86 and select Yes to R/W, even after it says it’s successful, it does not boot. If I browse the partition using ext2fsd, the data folder is empty.

I would love to get you all logcats, but every method you’ve provided to me has come up fruitless. Since I’ve successfully installed this to lots of devices now, I doubt it’s my install process. It’s either (I think) the Atom x7 or the eMMC.

[Mike R]

I'm not sure if it will help but I believe pstglia and/or Mauro Rossi have a debug Dameon that they have used to Write a report AS the system Boots, Even With the Lock-up your describing, I've had the blank data myself way back, Might either want to ask them or search the forum on it...

Hope it might help.

Mike

--

[Chih-Wei Huang]

I think it's clear the booting can't pass to the
android init -- that's why /data/ is empty.
It's probably stuck at some kernel drivers,
possibly the display driver.

It's a big challenge to debug such a problem.
Usually you need a UART port to get the
kernel dump message to know what happened.

The possible fact is the platform is too new
to be supported by the kernel (4.0) we use.
If I were you, I'll check if the latest kernel (4.2-rcX)
has display driver that supports this platform.

> --
> You received this message because you are subscribed to the Google Groups
> "Android-x86" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to android-x86...@googlegroups.com.
> To post to this group, send email to andro...@googlegroups.com.
> Visit this group at http://groups.google.com/group/android-x86.
> For more options, visit https://groups.google.com/d/optout.



--

Chih-Wei
Android-x86 project
http://www.android-x86.org

[rbg]

think we are missing something here........??????????

"No, the Surface 3 does not boot at all, Live USB or installed to eMMC. I installed it to eMMC so I could try and get a log, but it's not happening."

then later you say...

"I am telling you, when I install Android-x86 and select Yes to R/W, even after it says it’s successful, it does not boot. If I browse the partition using ext2fsd, the data folder is empty."

so I'll move on and just repeat...

pstglia is correct in that you need to provide Logcat / dmesg / log.txt form the

boot from disk, without it there is no way to actually see what's going on....and the photos

you have supplied really don't show anything as far as what is really happening.....

[Chih-Wei Huang]

2015-08-07 10:53 GMT+08:00 rbg <rbg.hp...@gmail.com>:
> think we are missing something here........??????????
>
> "No, the Surface 3 does not boot at all, Live USB or installed to eMMC. I
> installed it to eMMC so I could try and get a log, but it's not happening."
>
> then later you say...
>
> "I am telling you, when I install Android-x86 and select Yes to R/W, even
> after it says it’s successful, it does not boot. If I browse the partition
> using ext2fsd, the data folder is empty."
>
> so I'll move on and just repeat...

Well, I see no contradiction in Bernard's statement.

The booting process has a little different
between normal boot (live or non-live) and installer.
In normal boot the android-x86 init script
will run the auto-detect function to load
all necessary drivers it detects.
I think that's the time it get stuck.

For the installer the auto-detect step
is skipped so he can still install android-x86
successfully. But he can't boot the installed one
to the android ui due to the above reason.

[rbg]

well guess my wife is right again...she always says I take things too literally......

but hey if someone tells me that "No, the Surface 3 does not boot at all, Live USB or installed to eMMC"

I assume they mean that they cannot boot the Live USB and after going through some procedure to install it by

copying stuff to a partition, cannot boot it installed..........

But I see your point.........my bad!!!!!!!!

[rbg]

could you check this build....there should be a folder on the build called boot/S3-android-x86-boot

if not there...just stop and report results, if there read the file S3-note.txt and follow the directions, see if it boots ok

for you and post results....thanks

https://drive.google.com/file/d/0B2evBKvIpdYkZ0pRc1MwV2ZUd1U/view?usp=sharing

On Thursday, August 6, 2015 at 8:59:22 PM UTC-4, Bernard wrote:

[Bernard]

Unfortunately, no change. I tried Live USB and then installed it to eMMC as well. After typing "exit" in debug mode to continue booting, it seemed to hang longer before going to the black screen and locking up.

Data folder was empty when browsing the partition.

Thanks for building an image so fast. I'm going to go to bed, have a good night rbg.

[rbg]

FYI -

maybe the issue on surface 3 display....??

https://bugzilla.kernel.org/show_bug.cgi?id=97941

note that was with kernel 4.1 so maybe not til kernel 4.2 for a fix...???

[Saipudin Haris]

I think the issue is related to processor bugs/errata -------> http://www.intel.com/content/www/us/en/processors/atom/atom-z8000-spec-update.html

Quote from Bernard:
==============

"What I am talking about here is the SURFACE 3 with an ATOM Z8700 SoC, not the Surface Pro 3. I have a Surface Pro 3 that is successfully running Android-x86 using this exact procedure."

[rbg]

certainly it's all tied together...as

workaround i915.edp_vswing=2 on boot command line

was shown to work on some Ubuntu systems to temporarily solve the black screen issue

the Z8000 specification includes both the Atom Z5 & Z7, that are used on the surface 3

[Bernard]

Tried putting that in for boot parameter, didn't do anything.

[Dejay Nickerous]

This could be a widespread problem for the Intel Atom SOC processors.

My Lenovo MIIX 2 8" uses a Z3740 SOC, and does exactly the same thing.

[Bernard]

rbg, are you able to make a test build with 4.2 kernel?

[rbg]

not yet, but having built mrossi's 4.1 kernel and adding all the needed sp3 patches was / is on android-x86.net

somewhere.....though it's outdated you may want to try it.......

real key(to me) is following where the drm developers are going with the solutions.....

the article I posted was 2 months old....you need to really follow the intel-gfx-bugs
and drm i915 commits by vetter to see how the issue is progressing.....

IMO

[Mikerx86]

Go to the new download forum and follow links I'd imagine be the easiest way.

Sent with AquaMail for Android
http://www.aqua-mail.com

--

[Елліот Луцяк]

On Thursday, August 6, 2015 at 2:11:06 PM UTC-7, Bernard wrote:
> 1) Screenshot of my folders
>
>
> 2) Screenshot of my grub.cfg. It's the same one I use for most of my Android-x86 installs.

Edit grub.cfg so that "androidboot.hardware=android_x86" reads "androidboot.hardware=x86"

[Bernard]

That didn't change anything.

[Chih-Wei Huang]

First of all, many thanks to Bernard who
donated a Surface 3 tablet to me so
I can work on it.

I just made some primary tests.
The bad news is the i915 display driver
is broken that results in black screen on loading
as I expected. It seems to be a known issue
and not fixed yet:
https://bugzilla.kernel.org/show_bug.cgi?id=97941

Besides, there is no driver for audio and touch, either.

The Surface 3 is a new device.
It seems the open source community
still need more time to catch it up.
Stay tuned.

[Saipudin Haris]

I'm still suspecting the blanking display issue is related to a known Intel Processor Bugs/Errata.

http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/atom-z8000-spec-update.pdf

[James Davison]

Any progress on this, I am having the same problem.

[DZ]

Click on the link in Chih-We's post. The thread explains the progress being made.

[Jim]

Hello, thanks for your working.

But how can I download this kernel?

or it's on beta kernel only for log file?

Very hope android-x86 on Surface 3.

Chih-Wei Huang於 2015年8月20日星期四 UTC+8上午10時42分33秒寫道：

[Milan Parikh]

It seems there is a fix for the the black screen issue at least: https://bugzilla.kernel.org/show_bug.cgi?id=97941