lenovo yoga 2 with Intel Atom Z3745 (baytrail) build android

[Rustam Ivanovich]

Hi all!

Someone already builded android 5.0.x lollipop for Intel Atom x86 64 baytrail? If answer yes, then maybe give to me device tree? Please, please give me :) And, how to switch secure boot off if we can not enter to bios setup? Really need help. 

Thank you.

[Christopher Price]

There is no way to disable UEFI Secure Boot if your BIOS does not offer the option. However, some BIOS panels hide the Secure Boot option by requiring you to set a BIOS password first.

Try setting a BIOS password, then checking the "Boot" and "Startup" menu items to see if a Secure Boot option then appears.

Pretty much all the Android x86 distributions support x86-64 Bay Trail at this point... KitKat is more stable than Lollipop at the moment.

Christopher Price

ConsoleOS.com

[Rustam Ivanovich]

Thank you for answer. Okay i am try make device tree lollipop version for our device and write result here.

пятница, 23 января 2015 г., 18:35:41 UTC+3 пользователь Rustam Ivanovich написал:

[Chih-Wei Huang]

2015-01-26 19:53 GMT+08:00 Rustam Ivanovich <rusti...@gmail.com>:
> Thank you for answer. Okay i am try make device tree lollipop version for
> our device and write result here.

You don't need to create your own device tree.
Just add changes (if any you need) to the android_x86 target.
The baytrail platform should be supported when you
switch the kernel to android-3.18 branch.
At least you should be able to boot to Home
with hardware accelerated OpenGL ES.
Though there are still some bugs to be fixed.

[Rustam Ivanovich]

I am can't change kernel because bios locked, bootloader locked and some troubles. I can read/write system partition.

среда, 28 января 2015 г., 6:30:38 UTC+3 пользователь Chih-Wei Huang написал:

[Chih-Wei Huang]

2015-01-29 3:48 GMT+08:00 Rustam Ivanovich <rusti...@gmail.com>:
> I am can't change kernel because bios locked, bootloader locked and some
> troubles. I can read/write system partition.

I meant you need to rebuild the image with newer kernel.
If you know how to create a device tree, you must know
how to rebuild the kernel. Don't you?
Anyway, search the forum for how to do it.

[Rustam Ivanovich]

Build new kernel? It is no problem, i am known how to make it. But i am can't change in device because secure boot is ON.

четверг, 29 января 2015 г., 4:52:13 UTC+3 пользователь Chih-Wei Huang написал:

[Anton Vosakur]

We use build android-x86-4.4-r3.img - this is the first stable build for our device. Thanks to developer.

We lack drivers for sound, screen rotation, gyroscope, bt, gps.

We would be happy to see the assembly which has drivers for our Lenovo Yoga 2 tablet 10" (windows 8.1 version, only uefi 32).

If we understand correctly, in freely available exist source code. But we are not able to compile.
http://support.lenovo.com/ru/ru/products/tablets/yoga-series/yoga-tablet-2-1050

Thank you.

[Anton Vosakur]

As you now bios on the lenovo yoga 2 is open now (after bios flash 01WT17WW)

We use build android-x86-4.4-r3.img - this is the first stable build for our device. Thanks to developers.

We lack drivers for sound, screen rotation, gyroscope, bt, gps.

We would be happy to see the assembly which has drivers for our Lenovo Yoga 2 tablet 10" (windows 8.1 version, only uefi 32).

If we understand correctly, in freely available exist source code. But we are not able to compile.
http://support.lenovo.com/ru/ru/products/tablets/yoga-series/yoga-tablet-2-1050

Can you compile the build for us?

Thank you.

[pstglia]

Hi Anton,

We use build android-x86-4.4-r3.img - this is the first stable build for our device. Thanks to developers.

We lack drivers for sound, screen rotation, gyroscope, bt, gps. 

Among not working features, audio doesn't require a special driver - snd_hda_intel do the job (according to https://wiki.archlinux.org/index.php/Lenovo_Yoga_2_11)

Could you check if you have this driver loaded? 

busybox lsmod | grep -i snd_hda_intel

If not, you can try to load it:

busybox modprobe snd_hda_intel

 * * *

In any case you can attach logcat / dmesg so we can take a look

Regards,

Pstglia

 

[Anton Vosakur]

Thanks for your reply.

I did what you were told:

Just in case:
In windows our driver is called:
- intel sst audio device (wdm) 
- wolfson microelectronics audio wm5102 (Not very popular, as I understand it.)
Probably our source code (Our model 1051 is similar to 1050)
http://support.lenovo.com/ru/ru/products/tablets/yoga-series/yoga-tablet-2-1050
Thanks again for the response.

[Anton Vosakur]

More 

[Anton Vosakur]

And more

[Povilas Staniulis]

WM5102 itself is supported by the kernel, but I don't think the existing driver will work with Baytrail without patches.

[Anton Vosakur]

Respected Povilas. We are very glad that you are connected to the communication. As far as we know - You and Chih-Wei Huang, are the most authoritatives experts in questions of androd x86.

You are famous in Russia by yours assembles for the T100 (before android-x86-4.4-r3.img yours assemblies was the only who have worked on our lenovo yoga 2 10").
Nowhere on the Internet do not have answers for Lenovo Yoga 2. IMHO my funny video - the first about dualboot on this device.
If you can help us in this case - we would be very happy.

[Anton Vosakur]

https://www.youtube.com/watch?v=cb8Yt89g4Yw

[pstglia]

Hi Anton, Thanks for provided information.

Seems the required driver for your audio works properly is not enabled on config:

$ grep MFD_WM5102 android-x86_defconfig 

# CONFIG_MFD_WM5102 is not set

$ grep MFD_WM5102 android-x86_64_defconfig 

# CONFIG_MFD_WM5102 is not set

We can try to enable this option. So there's a chance you get your audio working.

I'll try to compile a testing ISO with this option enabled. 

If this doesn't work, we can try to look at logcat/dmesg outputs.

Regards,

Pstglia

[Anton Vosakur]

Wow wow wow !!! Super !!! Thanks.

What is required of us? What are the timelines? How can we get this to trial? 

[pstglia]

HI

Em terça-feira, 28 de julho de 2015 03:57:46 UTC-3, Anton Vosakur escreveu:

Wow wow wow !!! Super !!! Thanks.

What is required of us? What are the timelines? How can we get this to trial? 

Here is the testing ISO link:

https://drive.google.com/file/d/0Bx12U5yGNcQCSXEtbEtRQ1FzT3c/view?usp=sharing

Just enabled MFD_WM5102 option. If doesn't work, please provide logcat/dmesg outputs. 

These info will also be useful:

cat /proc/asound/cards

ls -l /system/etc/alsa/*state

Regards,

Pstglia

[Anton Vosakur]

Voila:

[Povilas Staniulis]

As I suspected, just enabling WM5102 support isn't enough.

You need a Intel SST machine driver for WM5102.

Currently Intel SST Linux driver supports Realtek RT5640, RT5645, RT5672 chips and Maxim MAX98090.

see kernel source: /sound/soc/intel

2015 m. liepa 28 d., antradienis 17:24:11 UTC+3, Anton Vosakur rašė:

Voila:

[Anton Vosakur]

see kernel source: /sound/soc/intel

If it is a call to action, what and how we should do? 

[pstglia]

Hi Povilas,

As I suspected, just enabling WM5102 support isn't enough.

You need a Intel SST machine driver for WM5102.

Currently Intel SST Linux driver supports Realtek RT5640, RT5645, RT5672 chips and Maxim MAX98090.

see kernel source: /sound/soc/intel

Thanks for the info. Would you mind answering some questions? 

1) Arch Linux Wiki says their distro has sound working "out of box" on Lenovo Yoga 2 11'' 

https://wiki.archlinux.org/index.php/Lenovo_Yoga_2_11

Could 11'' model be using a different audio device than 10.1''?

2) There's a codec for wm5102 under sound/soc/codecs/. As I could check, this is not compiled unless you select "CONFIG_COMPILE_TEST" and "CONFIG_SND_SOC_ALL_CODECS" (Using menuconfig). This enables CONFIG_SND_SOC_WM5102 as module (creates a module named snd-soc-wm5102.ko)

Could this codec be used for supporting this card? Seems to me Wolfson has included it on kernel mainstream to support this soc (see http://opensource.wolfsonmicro.com/node/6).

Regards,

Pstglia

[Povilas Staniulis]

2015 m. liepa 29 d., trečiadienis 05:25:42 UTC+3, pstglia rašė:

Hi Povilas,

As I suspected, just enabling WM5102 support isn't enough.

You need a Intel SST machine driver for WM5102.

Currently Intel SST Linux driver supports Realtek RT5640, RT5645, RT5672 chips and Maxim MAX98090.

see kernel source: /sound/soc/intel

Thanks for the info. Would you mind answering some questions? 

1) Arch Linux Wiki says their distro has sound working "out of box" on Lenovo Yoga 2 11'' 

https://wiki.archlinux.org/index.php/Lenovo_Yoga_2_11

Could 11'' model be using a different audio device than 10.1''?

It could.

Would need dmesg log or at least lsmod to tell for sure.

 

2) There's a codec for wm5102 under sound/soc/codecs/. As I could check, this is not compiled unless you select "CONFIG_COMPILE_TEST" and "CONFIG_SND_SOC_ALL_CODECS" (Using menuconfig). This enables CONFIG_SND_SOC_WM5102 as module (creates a module named snd-soc-wm5102.ko)

Could this codec be used for supporting this card? Seems to me Wolfson has included it on kernel mainstream to support this soc (see http://opensource.wolfsonmicro.com/node/6).

Yes, the codec could and must be used, but you still need a machine driver.

 

There's a machine driver for WM5102 in the kernel sources for Asus Zenfone 5 it in the kernel sources, but it's for kernel 3.10 and it's for Cloverview platform.

https://github.com/ZenfoneArea/android_kernel_asus_zenfone5/tree/master/linux/kernel/sound/soc/intel/board

 

Regards,

Pstglia

[Anton Vosakur]

I'm here. I check every day communication. When you need something from me, do not forget to tell: Anton - one two three go (and instructions) ))

[pstglia]

Hi Povilas, how are you?

 

2) There's a codec for wm5102 under sound/soc/codecs/. As I could check, this is not compiled unless you select "CONFIG_COMPILE_TEST" and "CONFIG_SND_SOC_ALL_CODECS" (Using menuconfig). This enables CONFIG_SND_SOC_WM5102 as module (creates a module named snd-soc-wm5102.ko)

Could this codec be used for supporting this card? Seems to me Wolfson has included it on kernel mainstream to support this soc (see http://opensource.wolfsonmicro.com/node/6).

Yes, the codec could and must be used, but you still need a machine driver.

 

There's a machine driver for WM5102 in the kernel sources for Asus Zenfone 5 it in the kernel sources, but it's for kernel 3.10 and it's for Cloverview platform.

https://github.com/ZenfoneArea/android_kernel_asus_zenfone5/tree/master/linux/kernel/sound/soc/intel/board

I've ported the sst driver from Lenovo Yoga 2 1050 source code (The one Anton posted on July, 27) to Android-x86 kernel (4.0.8).

Have to make some changes (kernel source was 3.10.20) and copy some requirements.

I compiled it as a module, producing snd-soc-sst-byt-rt5102-mach.ko file

I attached the patches created for this. 

Not sure if this is going to work, anyway.

Hi Anton,

I've created another testing img file with these patches applied: https://drive.google.com/file/d/0Bx12U5yGNcQCNEUxYnNUWUJkS3c/view?usp=sharing 

You can test it if you wish, just keep in mind there's a certain risk of something goes very wrong (losing data or even damaging something)

Probably the required modules won't load automatically (snd-soc-wm5102.ko and snd-soc-sst-byt-rt5102-mach.ko). If that is the case, you'll have to load them manually. 

Suggest you do something like this:

1) Boot img under DEBUG mode (DEBUG option on grub menu)

2) Type exit when message "Type 'exit' to continue booting..." appears

3) When message "Type 'exit' to enter Android..." appears, type "lsmod" and check if these drivers are loaded. If not, load them if these commands:

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

4) Type exit and continue booting. 

Let's hope it works. If not, at least we tried. 

[Anton Vosakur]

In debug mode before typing:

In debug mode after typing:

In terminal in android after all:

Result: no sound, volume +/- no reaction.

PS after "modprobe snd-soc-wm5102" in debug mode was no reaction.

[rbg]

hi pstglia....

was checking the settings for config_mfd_xxxxx yesterday in kernel 4.2-r? and

noticed they were different than the ones in our 4.0.8......so FWIW here are the settings I observed in 4.2-r? (think it was r3 would need

to go back) but you may want to try these..

CONFIG_MFD_ARIZONA=y

CONFIG_MFD_ARIZONA_I2C=y

CONFIG_MFD_ARIZONA_SPI=y

CONFIG_MFD_WM5102=y

CONFIG_MFD_WM5110=y

CONFIG_MFD_WM8997=y

CONFIG_MFD_WM8400=y

CONFIG_MFD_WM831X=y

CONFIG_MFD_WM831X_I2C=y

CONFIG_MFD_WM831X_SPI=y

CONFIG_MFD_WM8350=y

CONFIG_MFD_WM8350_I2C=y

CONFIG_MFD_WM8994=y

[pstglia]

Hi Rbg,

> was checking the settings for config_mfd_xxxxx yesterday in kernel 4.2-r? and
> noticed they were different than the ones in our 4.0.8......so FWIW here are the settings I observed in 4.2-r? (think it was r3 would need
> to go back) but you may want to try these..
>
> CONFIG_MFD_ARIZONA=y
> CONFIG_MFD_ARIZONA_I2C=y
> CONFIG_MFD_ARIZONA_SPI=y
> CONFIG_MFD_WM5102=y
> CONFIG_MFD_WM5110=y
> CONFIG_MFD_WM8997=y
> CONFIG_MFD_WM8400=y
> CONFIG_MFD_WM831X=y
> CONFIG_MFD_WM831X_I2C=y
> CONFIG_MFD_WM831X_SPI=y
> CONFIG_MFD_WM8350=y
> CONFIG_MFD_WM8350_I2C=y
> CONFIG_MFD_WM8994=y

I have some of them enabled on my config file (some as a module):

$ grep ^CONFIG_MFD_[AW] $OUT/obj/kernel/.config
CONFIG_MFD_ARIZONA=y
CONFIG_MFD_ARIZONA_I2C=m
CONFIG_MFD_ARIZONA_SPI=m
CONFIG_MFD_WM5102=y

I also have these 2 (the second from the sst driver I ported): 
 CONFIG_SND_SOC_WM5102=m
CONFIG_SND_SOC_INTEL_BYT_RT5102_MACH=m

As far as I know, the only relevant ones for this hardware are Arizona and 5102 modules/drivers.
But some of these could be pre-requisites, not sure. I just wanted to check logcat and dmesg first

Anton,

When loading it, we had messages the sst driver was registered. I'd like to know what logcat/dmesg shows. Can you send these files (the entire content, not only screenshots)

1) Repeat the steps I pointed on last email (DEBUG mode, modprobe modules, etc)
2) After booting Android, switch to console (ALT+F1) and copy dmesg and logcat and other content:

#logcat
cp /data/log.txt /sdcard/Download/

#dmesg
dmesg > /sdcard/Download/dmesg.txt

# card list
cat /proc/asound/cards > /sdcard/Download/card_list.txt

# alsa config files
ls -l /system/etc/alsa/*state > /sdcard/Download/alsa_cfg_files.txt

3) Switch back to graphical (ALT+F7) and copy these files into a flashdriver or attach them into an e-mail. They will be available on Download directory

Regards
Pstglia

[Anton Vosakur]

#logcat 

#dmesg

# card list

Ok.

# alsa config files

ls -l /system/etc/alsa/*state > /sdcard/Download/alsa_cfg_files.txt

permission denied. I did exactly in alt+f1 (not in terminal with type "su")

Logs: https://drive.google.com/file/d/0B3Q9Tg-pauSsQ3hNWHkxaS1nNms/view?usp=sharing

Regards
Anton

[pstglia]

Still not detecting the card. Wanna try another "combos"? For each you test, after entering graphical ui, check if "cat /proc/asound/cards" returns something different from "--- no soundcards ---"

1) Under debug mode (before that exit msg), load all these modules with modprobe:

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

Type exit, enter graphical out and see what you got. If no sound, try item 2:

2) 

 

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

Another thing: Can you send a "lspci" output?

Regards,

Pstglia

[Anton Vosakur]

 "cat /proc/asound/cards" , 1, 2 - no results. no soundcards.

 

Another thing: Can you send a "lspci" output?

I need more instructions. 

[pstglia]

Another thing: Can you send a "lspci" output?

I need more instructions. 

just type lspci on terminal (ALT+F1). 

[Anton Vosakur]

just type lspci on terminal (ALT+F1). 

 

[pstglia]

Em terça-feira, 4 de agosto de 2015 18:10:26 UTC-3, Anton Vosakur escreveu:

just type lspci on terminal (ALT+F1). 

 

Will try to include some debug messages and build a new iso when I have a spare time to do so

[pstglia]

Hi Anton, how are you?

I've made a change on driver code. I basically commented snd_byt_driver_init (late_initcall)  and snd_byt_driver_exit (module_exit) and replaced these by module_platform_driver. This is what the other drivers are doing ( 5640, 98090).

This is the testing IMG. If you could give a try when possible:

https://drive.google.com/file/d/0Bx12U5yGNcQCMjNpb0NaU2JISm8/view?usp=sharing 

Like before, try loading it on debug mode, before 2nd exit message appears:

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

If this doesn't work, attach here logcat/dmesg.

Thanks!

Pstglia

[Anton Vosakur]

Hello Pstglia. I'm good. Tnks. How are you?
I will try tomorrow - then will report the results. 

[Anton Vosakur]

Result - no sound, But there seems to be some changes in the logs

[Anton Vosakur]

on debug mode, before 2nd exit message appears:

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

result - silence.

[pstglia]

Result - no sound, But there seems to be some changes in the logs

Appears we have to advance from this point:

<3>[  105.003520] sst-acpi 80860F28:00: No matching ASoC machine driver found

Based on info found here (for 5642):

https://bugzilla.kernel.org/show_bug.cgi?id=86581

We should register our sst as a valid machine driver inside sound/soc/intel/sst-acpi.c. The structure with supported drivers is currently populated with this:

        { "10EC5640", "byt-rt5640", "intel/fw_sst_0f28.bin-48kHz_i2s_master" }, --> This for realtek 5640

        { "193C9890", "byt-max98090", "intel/fw_sst_0f28.bin-48kHz_i2s_master" }, --> This for MAXIM

Current plan now is to add your dev info in this structure, build a new img and see what we got. But for this, we have to find which is your VID/PID. Could you execute the command bellow and post it here please?

/sys/bus/acpi/devices

 

Also, if you can, get your WM5102 VID/PID from Windows (Device Manager) .

Thanks again.

[Chih-Wei Huang]

2015-08-11 9:41 GMT+08:00 pstglia <pst...@gmail.com>:

>
> Current plan now is to add your dev info in this structure, build a new img
> and see what we got. But for this, we have to find which is your VID/PID.
> Could you execute the command bellow and post it here please?
>
> /sys/bus/acpi/devices
>

I think you meant

ls -l /sys/bus/acpi/devices


Anton, please DO NOT attach more screenshots.
It's almost useless and wastes bandwidths.
Just attach the text output we asked.
For example, in this case you could
save the output as

ls -l /sys/bus/acpi/devices > /data/acpi.txt

Then copy /data/acpi.txt to usb stick and attach the file.


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

[Chih-Wei Huang]

2015-08-11 9:41 GMT+08:00 pstglia <pst...@gmail.com>:
>
>>

>> Result - no sound, But there seems to be some changes in the logs
>
>
> Appears we have to advance from this point:
>
> <3>[ 105.003520] sst-acpi 80860F28:00: No matching ASoC machine driver
> found
>
> Based on info found here (for 5642):
> https://bugzilla.kernel.org/show_bug.cgi?id=86581
>
> We should register our sst as a valid machine driver inside
> sound/soc/intel/sst-acpi.c. The structure with supported drivers is
> currently populated with this:
>
> { "10EC5640", "byt-rt5640", "intel/fw_sst_0f28.bin-48kHz_i2s_master"
> }, --> This for realtek 5640
> { "193C9890", "byt-max98090",
> "intel/fw_sst_0f28.bin-48kHz_i2s_master" }, --> This for MAXIM
>
> Current plan now is to add your dev info in this structure, build a new img
> and see what we got. But for this, we have to find which is your VID/PID.

From the demsg I saw:

<3>[ 105.051720] i2c i2c-4: Failed to register i2c client MAGN0001:00
at 0x1d (-16)
<3>[ 105.055928] i2c i2c-4: failed to add I2C device MAGN0001:00 from ACPI

After googling I found:

http://www.spinics.net/lists/linux-acpi/msg58643.html
http://www.spinics.net/lists/linux-spi/msg04501.html

Seems Christian Hartmann has some patches we can try.

[Anton Vosakur]

Also, if you can, get your WM5102 VID/PID from Windows (Device Manager) .

 

[Anton Vosakur]

[Anton Vosakur]

Anton, please DO NOT attach more screenshots.
It's almost useless and wastes bandwidths.
Just attach the text output we asked.
For example, in this case you could
save the output as

ls -l /sys/bus/acpi/devices > /data/acpi.txt

Then copy /data/acpi.txt to usb stick and attach the file.

I just saw your previous message. Next time I'll try. 

[Anton Vosakur]

When i attach a photo - you can see what I was doing and if I was mistaken (ex: one letter or character) - you can see it. Without photos - errors can be exponentially.
Other example: 
ls -l /sys/bus/acpi/devices > /data/acpi.txt 
RESULT - "can`t execute: Is a directory"
And i don`t see "/data/acpi.txt"
And I do not even see where "/data"
I need to install rootexplorer, or it has to be seen in the integrated explorer "files" ?

[pstglia]

I suggest you posting your images on Imageshack (or similar) and just put here the url. 

About the cmd I asked. Did you use ls as Chih-Wei pointed? (I forgot to include it). I'm expecting something similar to this:

uid=0(root) gid=0(root)@x86:/ # ls -l /sys/bus/acpi/devices/
lrwxrwxrwx root     root              2015-08-11 21:52 ACPI0003:00 -> ../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A03:00/ACPI0003:00
lrwxrwxrwx root     root              2015-08-11 21:52 APP0001:00 -> ../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A03:00/device:00/APP0001:00
lrwxrwxrwx root     root              2015-08-11 21:52 LNXCPU:00 -> ../../../devices/LNXSYSTM:00/LNXCPU:00
lrwxrwxrwx root     root              2015-08-11 21:52 LNXCPU:01 -> ../../../devices/LNXSYSTM:00/LNXCPU:01
lrwxrwxrwx root     root              2015-08-11 21:52 LNXCPU:02 -> ../../../devices/LNXSYSTM:00/LNXCPU:02
lrwxrwxrwx root     root              2015-08-11 21:52 LNXCPU:03 -> ../../../devices/LNXSYSTM:00/LNXCPU:03
..

Also, Could you get your WM5102 VID/PID from Windows Device Manager?

Regards,

Pstglia

[Anton Vosakur]

>I suggest you posting your images on Imageshack (or similar) and just put here the url. 

No problem. I got it.

>About the cmd I asked. Did you use ls as Chih-Wei pointed? (I forgot to include it). I'm expecting something similar to this:

Yes, I used ls (I've seen remarks of Chih-Wei). https://drive.google.com/file/d/0B3Q9Tg-pauSsaUphZTNMUzNiQms/view?usp=sharing

>Also, Could you get your WM5102 VID/PID from Windows Device Manager?

I have already answered - see in the first image of the previous posts.
repeat:  https://drive.google.com/file/d/0B3Q9Tg-pauSsTHVvSHJJdDh3eDQ/view?usp=sharing

[Alexandru Grigoras]

Hi guys,

I managed to make my yoga2 10' windows8 tablet dual-boot androidx86 and thought to give some help on this thread. I used the last image you uploaded on 9 Aug, these are the results of 'ls -l /sys/bus/acpi/devices' and the Windows Device manager Hardware IDs.

http://pastebin.com/raw.php?i=gjdLGCJr
http://pastebin.com/raw.php?i=NAgR73ih

Sorry, I tried attaching files here but I can't.

[Anton Vosakur]

Welcome to the club Alexandru )))
You also dont`t have any drivers like us?

[pstglia]

Good evening for all

Anton / Alexandru,

  Thanks for the info you provided

I included the VID & PID of WM5102 (VID WM51 PID 0204) inside kernel/sound/soc/intel/sst-acpi.c and created a new testing img:

static struct sst_acpi_mach baytrail_machines[] = {

        { "10EC5640", "byt-rt5640", "intel/fw_sst_0f28.bin-48kHz_i2s_master" },

        { "193C9890", "byt-max98090", "intel/fw_sst_0f28.bin-48kHz_i2s_master" },

        { "WM510204", "byt_wm5102", "intel/fw_sst_0f28.bin-48kHz_i2s_master" },

        {}

};

This attempt is to 5102 sst driver be detected. Probably there's lot more missing, but we can continue investigating.

As before, try loading manually the modules and after entering graphical mode collect logcat/dmesg outputs please .

Check also if "cat /proc/asound/cards" returns something different than "no soundcards"

Here's the img link:

https://drive.google.com/file/d/0Bx12U5yGNcQCRXlfa2g2Mmg2ZTg/view?usp=sharing

Regards,

Pstglia

Ps: I still have to check Chih-Wei link carefully

[Alexandru Grigoras]

Hello,

Just booted an USB stick in Debug mode with the new image provided.

I did modprobe for the following modules:

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

Last module returned the following error message:

<6>[  300.874949] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  300.889635] memio map is not set
<3>[  300.904898] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  300.920049] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  300.935041] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

'cat /proc/asount/cards' returns '--- no soundcards ---'

I attach here dmesg/logcat

[pstglia]

Just booted an USB stick in Debug mode with the new image provided.

I did modprobe for the following modules:

modprobe arizona-spi

modprobe arizona-i2c

modprobe gpio-arizona

modprobe snd-soc-arizona

modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

Last module returned the following error message:

<6>[  300.874949] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  300.889635] memio map is not set
<3>[  300.904898] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  300.920049] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  300.935041] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

'cat /proc/asount/cards' returns '--- no soundcards ---'

I attach here dmesg/logcat

Thank you! Will check logs you provided this weekend 

[pstglia]

Hi,

<3>[  300.889635] memio map is not set

This message is issued by "vlv2_plat_configure_clock" function defined on drivers/misc/vlv2_plat_clock.c (one of the components ported). 

This function is called from snd_byt_mc_probe wm5102 sst driver.

Didn't see any equivalent call on other sst drivers (5640, 98090). Maybe this can be ignored...

 

<3>[  300.904898] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered

Guess this is the reason snd_soc_register_card is returning failure. "Headset-cpu-dai" is defined here:

static struct snd_soc_dai_link byt_dailink[] = {

        [BYT_AUD_AIF1] = {

                .name = "Baytrail Audio",

                .stream_name = "Audio",

                .cpu_dai_name = "Headset-cpu-dai",

                .codec_dai_name = "wm5102-aif1",

                .codec_name = "wm5102-codec",

                .platform_name = "sst-platform",

Both byt-rt5640.c and byt-max98090.c uses "baytrail-pcm-audio" (registered by sst-baytrail-pcm.c)

static struct snd_soc_dai_link byt_rt5640_dais[] = {

        {

                .name = "Baytrail Audio",

                .stream_name = "Audio",

                .cpu_dai_name = "baytrail-pcm-audio",

                .codec_dai_name = "rt5640-aif1",

                .codec_name = "i2c-10EC5640:00",

                .platform_name = "baytrail-pcm-audio",

static struct snd_soc_dai_link byt_max98090_dais[] = {

        {

                .name = "Baytrail Audio",

                .stream_name = "Audio",

                .cpu_dai_name = "baytrail-pcm-audio",

                .codec_dai_name = "HiFi",

                .codec_name = "i2c-193C9890:00",

                .platform_name = "baytrail-pcm-audio",

 

* * * 

Based on the guesses above, I've made a change to 5102 in order to use "baytrail-pcm-audio" both for "cpu_dai_name" and "platform_name". 

I also replaced "snd_soc_register_card" by "devm_snd_soc_register_card" to match the other drivers (in fact, devm_snd_soc_register_card calls snd_soc_register_card, so no big deal)

Created another testing img with proposed changes. If you want to test, here's the link.

https://drive.google.com/file/d/0Bx12U5yGNcQCX0dWdXBtc25iZ2s/view?usp=sharing 

Note: Include this driver just in case (in bold):

modprobe arizona-spi
modprobe arizona-i2c
modprobe gpio-arizona
modprobe snd-soc-arizona
modprobe snd-soc-wm5102

modprobe snd-soc-sst-baytrail-pcm
modprobe snd-soc-sst-byt-rt5102-mach

Ps: I'm trying to do changes based what I google, info you provided by you (Chih-Wei, Povilas, Anton and Alexandru). But it's a work of "try and error". 

More experienced people with real knowledge in coding (audio drivers in special) would be welcome. 

Apparently this driver was wrote by an intel Employee named Omair Mohammed Abdullah. Maybe he could help us (if reachable)

Regards,

Pstglia

[Alexandru Grigoras]

Hi,

Just tried the new build unfortunately still 'no soundcards', attaching the logs.

modprobe snd-soc-sst-baytrail-pcm

<3>[  160.179871] baytrail-pcm-audio baytrail-pcm-audio: error: invalid DMA engine 0
<4>[  160.180253] baytrail-pcm-audio baytrail-pcm-audio: sst_dma_new failed -22
<3>[  160.404641] baytrail-pcm-audio baytrail-pcm-audio: ipc: error DSP boot timeout

modprobe snd-soc-sst-byt-rt5102-mach

<6>[  185.298670] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  185.298964] memio map is not set
<3>[  185.310663] byt_wm5102 byt_wm5102: ASoC: CPU DAI baytrail-pcm-audio not registered
<3>[  185.311074] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  185.311424] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

[pstglia]

Hi,

Just tried the new build unfortunately still 'no soundcards', attaching the logs.

modprobe snd-soc-sst-baytrail-pcm

<3>[  160.179871] baytrail-pcm-audio baytrail-pcm-audio: error: invalid DMA engine 0
<4>[  160.180253] baytrail-pcm-audio baytrail-pcm-audio: sst_dma_new failed -22
<3>[  160.404641] baytrail-pcm-audio baytrail-pcm-audio: ipc: error DSP boot timeout

modprobe snd-soc-sst-byt-rt5102-mach

<6>[  185.298670] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  185.298964] memio map is not set
<3>[  185.310663] byt_wm5102 byt_wm5102: ASoC: CPU DAI baytrail-pcm-audio not registered
<3>[  185.311074] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  185.311424] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

Ok, appears codecs are not being registered. I take a look at Christian Hartmann patches Chih-Wei linked. As I understood, they try to do exactly what we need: Help Registering these codecs (using acpi).

He'd having a NULL pointer exception. But apparently, he has no machine driver. As we have the one ported from Lenovo source, maybe we have better luck.

New testing iso:

https://drive.google.com/file/d/0Bx12U5yGNcQCUm5TelpJVWlrWkU/view?usp=sharing 

As I reverted last change, no need to load snd-soc-sst-baytrail-pcm. Just load these

modprobe arizona-spi
modprobe arizona-i2c
modprobe gpio-arizona
modprobe snd-soc-arizona
modprobe snd-soc-wm5102

modprobe snd-soc-sst-byt-rt5102-mach

Note: Christian Hartmann is still trying to make it work (last message was 6 days ago)

http://www.spinics.net/lists/linux-spi/msg04850.html

He has more chances to achieve this than me (He has some knowledge how driver works). 

[pstglia]

To make changes made so far public to everyone, I've pushed them to github:

https://github.com/pstglia/linux/tree/kernel_4.0.8_Lenovo_Yoga_2

Hello Christian Hartmann, how are you?

We've seem your progress on trying to make WM5102 work on Linux. 

We are trying to do the same aiming Android-x86 (I believe Alsa code is common for both - no specific changes for Android in general)

We are using an Android source code released by Lenovo for 1050/1051 machines, which contains a sst driver for wm5102

(See http://support.lenovo.com/us/en/products/tablets/yoga-series/yoga-tablet-2-1050).

I "ported" this sst driver to Vanilla 4.0.8, currently used by Android-x86 . It compiles and machine driver starts probe function, but due no registered codecs it fails when trying to register sound card.

<6>[  300.874949] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  300.889635] memio map is not set
<3>[  300.904898] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  300.920049] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  300.935041] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

I used some of the patches you posted here and still waiting for some users to test it. 

https://drive.google.com/file/d/0Bx12U5yGNcQCUm5TelpJVWlrWkU/view?usp=sharing 

Maybe there's something we've made that could be useful for you and you could point us to the right direction.

I'm just a curious guy who google around stuff, compare code and try to put it all together. Sometimes it works

If you are interested, here's the link with complete thread (Android-x86 Google group):

https://groups.google.com/forum/#!topic/android-x86/BMSMwG0s-fc

Thanks for your nice work with this!

Regards,

Pstglia

[Paulo Sergio]

thank you for the interest and your mail.

I actually try to add ACPI support to the WM5102 audio codec hub,
which is built into the Lenovo Yoga 851 F.
Currently its a Baytrail system with 32Bit UEFI.

* The i915 video driver faults sometimes,
* BCM2E84 Broadcom Serial Bus Driver over UART Bus Enumerator ACPI id
was missing
* device BCM4752 / LNV4752 Broadcom GPS device fails to init GPIO pins

the SPI controller pxa2xx needs a fix to init correctly and to add the
WM510205 audio codec hub. As there is an error in the firmware, saying
that the SPI controller is no. 2 for the WM5102, but its no.1.
The one patch to select the right chipselect on pxa2xx does solve it
for all broken firmwares (I have mailed lenovo but not seen a new
firmware yet) for x86 where the pxa2xx reads out the wrong SPI
controller value from ACPI (UEFI/ DSDT.dsl)

After this patch I added the ACPI id in drivers/mfd/arizona-core.c
arizona-spi.c etc as it was done for other devices (I used rt5645 as
an example)

but adding only the acpi_id for this audio codec to get it initialized
and bootet correctly, was not all.
It needs more patches.

I have started a discussion on the alsa mailinglist about adding ACPI
support for the WM5102 (in the firmware it has
the non-standardconforme ACPI id of WM510205)

and as of now I am working with the gpiolib aka reading the GPIO pins,
lodena and reset.
These two pins are mandandory and all what I need to - as said by
other developer(s).

The audio codec should work fine now....
...but it does not yet.

I have tried yesterday and today to read out the GPIO pins via
gpiolib, as described it should be possible to map
the two pins in questions with labels etc. Currently the
arizona-spi/arizona-core driver does not find the correct GPIOs and
threrefore fails to init the codec.

IF IT DOES NOT fail anymore, all I have to do, is to add a new machine
driver , which you seems to have already.
I stashed the machine driver, it can also contains not tested bugs
yet. I want to do the machine driver after the audio codec is running
fine

we could or should talk over irc, or just write us.  I am sometimes in
the channel alsa-soc on irc.freeenode.net

cheers
chris

Thanks for your return. We'll stay tuned. 

Meanwhile we can try testing the machine driver even if codecs are not ready (just need to register them). 

Just asking how Lenovo made they work. Apparently there's no significant changes to arizona drivers or sound/soc/codec...

Regards,

Pstglia

[pstglia]

Hi Anton / Alexandru,

Have you tested the last iso?

https://drive.google.com/file/d/0Bx12U5yGNcQCUm5TelpJVWlrWkU/view?usp=sharing 

Some different error message?

Thanks

Pstglia

 

[Anton Vosakur]

Hi Pstgla! I'm on vacation. I'll be back soon. And immediately will look.
What exactly should be typed into the console for testing? (And in debug or in normal mode or both)?

[pstglia]

Hi Anton,

Em terça-feira, 25 de agosto de 2015 04:41:50 UTC-3, Anton Vosakur escreveu:

Hi Pstgla! I'm on vacation. I'll be back soon. And immediately will look.
What exactly should be typed into the console for testing? (And in debug or in normal mode or both)?

 Same as before. Boot under DEBUG mode and load these modules:

modprobe arizona-spi
modprobe arizona-i2c
modprobe gpio-arizona
modprobe snd-soc-arizona
modprobe snd-soc-wm5102
modprobe snd-soc-sst-byt-rt5102-mach

After entering GUI, save a dmesg output:


busybox dmesg > /sdcard/Download/dmesg.txt


And attach this files here (The file itself - pics won't help that much)

You can post directly from bundled Android web browser

Also, check if after loading the modules you still have "no sound cards" as output of "cat /proc/asount/cards"

Enjoy your vacation!

Pstglia

[Mael SONET]

Le mardi 25 août 2015 09:41:50 UTC+2, Anton Vosakur a écrit :
> Hi Pstgla! I'm on vacation. I'll be back soon. And immediately will look.
> What exactly should be typed into the console for testing? (And in debug or in normal mode or both)?

hello !

only for purpose..(because I'm a noob and don't want distrub your work )

I tested the last iso file on my lenovo yoga tablet 2 851-F (WOLFSON device)
but it don't work.

with le official release 4.4.4 R3 of android X86 everything is working like a charm except bluetooth/rotation/sound (WOLFSON).

I think my device is the same but in 8"

would you like a start my slate with android in debug mode ? no risk to brick my device or overwrite my windows partition ?

Regards.

[Tim Wu]

Hi Pstglia,

Just recently discovered this thread so I'm trying to catch up with the current developments. I also have a yoga 2 1051f that I want to get working with android-x86.

So far I've checked out the current kitkat-x86 code and edited my build to pull in the kernel from your github account. After a bit of messing with it, I was able to build an image and attempted to the do the module loading you mentioned above.

Loading snd-soc-sst-byt-rt5102-mach failed with this error:

<6>[ 369.203148] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[ 369.203229] memio map is not set
<3>[ 369.214705] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[ 369.214826] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[ 369.214933] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

The rest of the dmesg output is here: http://pastebin.com/058pGQLX


Is the code on your github the latest so far? I'd actually be pretty curious to help figure out what's going on if you can give me a few pointers as to the current state of what bits of code need to be worked on. Admittedly, I've never actually done any coding in the linux kernel or modules, so I'll probably be a bit slow.

Thanks!
Tim

[pstglia]

Hi Mael,

only for purpose..(because I'm a noob and don't want distrub your work )

Don't disturb at all! Any info you can provide is helpful.

 

I tested the last iso file on my lenovo yoga tablet 2 851-F (WOLFSON device)
but it don't work.

with le official release 4.4.4 R3 of android X86 everything is working like a charm except bluetooth/rotation/sound (WOLFSON).

Can you detail what is going on exactly? Do you have an crash before entering gui? any error message printed?

 

I think my device is the same but in 8"

would you like a start my slate with android in debug mode ? no risk to brick my device or overwrite my windows partition ?

There's always a risk. But if you can boot from a flashdrive (run it without installing), you should have no problems overwriting you partitions.

Not sure if your model uses the same hardware

 

[pstglia]

Hi Tim

Em quarta-feira, 26 de agosto de 2015 13:58:43 UTC-3, Tim Wu escreveu:

Hi Pstglia,

Just recently discovered this thread so I'm trying to catch up with the current developments. I also have a yoga 2 1051f that I want to get working with android-x86.

So far I've checked out the current kitkat-x86 code and edited my build to pull in the kernel from your github account. After a bit of messing with it, I was able to build an image and attempted to the do the module loading you mentioned above.

Loading snd-soc-sst-byt-rt5102-mach failed with this error:

<6>[  369.203148] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  369.203229] memio map is not set
<3>[  369.214705] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  369.214826] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  369.214933] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

The rest of the dmesg output is here: http://pastebin.com/058pGQLX


Is the code on your github the latest so far? I'd actually be pretty curious to help figure out what's going on if you can give me a few pointers as to the current state of what bits of code need to be worked on. Admittedly, I've never actually done any coding in the linux kernel or modules, so I'll probably be a bit slow.

 

This is the latest I've worked on so far. I've included the sst machine driver for Wolfson and some patches from Christian Hartmann (some of them - there are some more) in order to try to make it work.

The error you see is due unregistered codecs (this is what Christian Hartmann is working on). Without these properlly registered, sst driver cannot initialize. 

Consider this as a good opportunity to learn how to debug and apply kernel changes. We are on the same boat: I'm still trying to learn.

I'll check Christian progress and try to include it on source. Here's a thread where they are discussing it:

https://www.mail-archive.com/linu...@vger.kernel.org/msg05613.html

Regards,

Pstglia

[Mael SONET]

ok..if it can help..

when I said than this version it not working, I gave you a bad explanation.
actually everything is working except WOLFSON/bluetooth/rotate.

before the GUI start, i receveive an error message : unable to start RTC device  (I can provide you thevexact message if needed)
however android start.
I saw thant in the devellper I can active a debug mode.
but I don't know where i can find the .log file 

[pstglia]

when I said than this version it not working, I gave you a bad explanation.
actually everything is working except WOLFSON/bluetooth/rotate.

before the GUI start, i receveive an error message : unable to start RTC device  (I can provide you thevexact message if needed)
however android start.
I saw thant in the devellper I can active a debug mode.
but I don't know where i can find the .log file 

Another testing img. Got arizona drivers at Wolfson website (3.14 branch)

If anyone is interested...

https://drive.google.com/file/d/0Bx12U5yGNcQCNkEzMVJoZEJpbnc/view?usp=sharing

Regards

Pstglia

[Alexandru Grigoras]

Hi Pstglia,

Been inactive for a while but today I got the chance to test the latest build. dmesg output attached.

<6>[  317.520411] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  317.520728] memio map is not set
<3>[  317.533215] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  317.533632] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  317.533984] platform byt_wm5102: Driver byt_wm5102 requests probe deferral
<11>[  362.692989] init: could not import file '/init.android_x86.rc' from '/init.rc'
<3>[  362.865563] genirq: Flags mismatch irq 15. 00000083 (volume_up) vs. 00000000 (pata_legacy.1)
<3>[  362.865689] gpio-keys gpio-keys.0.auto: Unable to claim irq 15; error -16
<4>[  362.865793] gpio-keys: probe of gpio-keys.0.auto failed with error -16
<6>[  362.866054] input: gpio-keys as /devices/platform/gpio-keys.1.auto/input/input8
<6>[  362.866289] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  362.866370] memio map is not set
<3>[  362.877945] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  362.878068] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  362.878176] platform byt_wm5102: Driver byt_wm5102 requests probe deferral
<7>[  362.910752] power_supply BATC: uevent
<7>[  362.910764] power_supply BATC: POWER_SUPPLY_NAME=BATC
<3>[  362.922333] baytrail-pcm-audio baytrail-pcm-audio: error: invalid DMA engine 0
<4>[  362.922445] baytrail-pcm-audio baytrail-pcm-audio: sst_dma_new failed -22
<7>[  362.924696] power_supply BATC: prop STATUS=Discharging
<7>[  362.924708] power_supply BATC: prop PRESENT=1
<7>[  362.924715] power_supply BATC: prop TECHNOLOGY=Li-ion
<7>[  362.924720] power_supply BATC: prop CYCLE_COUNT=54
<7>[  362.924726] power_supply BATC: prop VOLTAGE_MIN_DESIGN=3750000
<7>[  362.924731] power_supply BATC: prop VOLTAGE_NOW=4118000
<7>[  362.924737] power_supply BATC: prop CURRENT_NOW=1173000
<7>[  362.924743] power_supply BATC: prop CHARGE_FULL_DESIGN=9300000
<7>[  362.924748] power_supply BATC: prop CHARGE_FULL=8880000
<7>[  362.924754] power_supply BATC: prop CHARGE_NOW=7602000
<7>[  362.924760] power_supply BATC: prop CAPACITY=85
<7>[  362.924766] power_supply BATC: prop CAPACITY_LEVEL=Normal
<7>[  362.924771] power_supply BATC: prop MODEL_NAME=SR Real Battery
<7>[  362.924777] power_supply BATC: prop MANUFACTURER=Intel SR 1
<7>[  362.924782] power_supply BATC: prop SERIAL_NUMBER=123456789
<7>[  362.931946] power_supply ADP1: uevent
<7>[  362.931955] power_supply ADP1: POWER_SUPPLY_NAME=ADP1
<7>[  362.942826] power_supply ADP1: prop ONLINE=0
<11>[  362.946350] init: /dev/hw_random not found
<3>[  363.059822] i2c_hid i2c-SYNA0001:00: failed to retrieve report from device.
<6>[  363.067915] input: SYNA0001:00 06CB:1023 as /devices/platform/80860F41:05/i2c-5/i2c-SYNA0001:00/0018:06CB:1023.0005/input/input9
<6>[  363.073577] hid-multitouch 0018:06CB:1023.0005: input,hidraw4: <UNKNOWN> HID v1.00 Device [SYNA0001:00 06CB:1023] on
<6>[  363.077883] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[  363.082109] memio map is not set
<3>[  363.086929] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[  363.091139] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[  363.095398] platform byt_wm5102: Driver byt_wm5102 requests probe deferral
<3>[  363.162348] baytrail-pcm-audio baytrail-pcm-audio: ipc: error DSP boot timeout
<6>[  363.274199] usbcore: registered new interface driver btusb

--
You received this message because you are subscribed to a topic in the Google Groups "Android-x86" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/android-x86/BMSMwG0s-fc/unsubscribe.
To unsubscribe from this group and all its topics, 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]

Hi Alexandru,

Em segunda-feira, 31 de agosto de 2015 10:00:14 UTC-3, Alexandru Grigoras escreveu:

Hi Pstglia,

Been inactive for a while but today I got the chance to test the latest build. dmesg output attached.

Thaks for testing it. Have you noticed any error message when loading arizona drivers?

If you can, try booting using these extra kernel parameters:

drm.debug=7 log_buf_len=512000

Thanks,

Pstglia

[Alexandru Grigoras]

Hi Pstglia,

when doing modprobe, only the last module returns error messages:

modprobe snd-soc-sst-byt-rt5102-mach
<6>[   75.980207] snd_soc_sst_byt_rt5102_mach: Entry snd_byt_mc_probe
<3>[   75.980502] memio map is not set
<3>[   75.992020] byt_wm5102 byt_wm5102: ASoC: CPU DAI Headset-cpu-dai not registered
<3>[   75.992420] snd_soc_sst_byt_rt5102_mach: snd_soc_register_card failed -517
<6>[   75.992771] platform byt_wm5102: Driver byt_wm5102 requests probe deferral

I have included the dmesg log while booting in drm.debug=7 mode with a log buffer of 16M and the output of lsmod as well

--

[Mael SONET]

Hello;

prepare usb stick with your last iso -> Done
boot in debug mod -> done

load all modprobe :

modprobe arizona-spi
modprobe arizona-i2c
modprobe gpio-arizona
modprobe snd-soc-arizona
modprobe snd-soc-wm5102
modprobe snd-soc-sst-byt-rt5102-mach

-> Done without error
start GUI->Done
switch to consol with alt+F1 -> Done

You will find attached my dmesg.txt

below the result of lspci :

uid=10024(u0_a24) gid=10024(u0_a24) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50024(all_a24)@x86:/ $ lspci
00:00.0 Class 0600: 8086:0f00
00:02.0 Class 0300: 8086:0f31
00:14.0 Class 0c03: 8086:0f35
00:1a.0 Class 1080: 8086:0f18
00:1f.0 Class 0601: 8086:0f1c
uid=10024(u0_a24) gid=10024(u0_a24) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50024(all_a24)@x86:/ $

below the result of : cat /proc/asound/cards

cat /proc/asound/cards                                                                                                              <
--- no soundcards ---
uid=10024(u0_a24) gid=10024(u0_a24) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50024(all_a24)@x86:/ $

Regards.

not easy with an azerty keyboard...

(ps: no access to my external sdcard)
(ps2 : "healthd: battery"  message is a hell !!!)

Regards.

[pstglia]

Thanks Alexandru and Mael for your debug infos.

I couldn't see anything related to any attempt of codec registering ("dai links"?). 

I'll try to understand how to register these dai links.

Will try to contact Chris. Maybe he give us some hint

Regards,

Pstglia

[Jarrod Young]

Any update?

[Jarrod Young]

On Friday, January 23, 2015 at 10:35:41 AM UTC-5, Rustam Ivanovich wrote:
> Hi all!
>
>
> Someone already builded android 5.0.x lollipop for Intel Atom x86 64 baytrail? If answer yes, then maybe give to me device tree? Please, please give me :) And, how to switch secure boot off if we can not enter to bios setup? Really need help. 
>
>
> Thank you.

I would love to test if anyone has more fixes available.

[pstglia]

Hi Jarrod,

I would love to test if anyone has more fixes available.

I'm trying to understand how the drivers work exactly (how to register the codec and start the machine driver)

I'm trying also to compile the kernel provided by Lenovo. If this kernel image can be used with our source, we can check dmesg messages for codecs/sound registering. 

If I can do it, I'll post the img for testing.

Regards,

Pstglia

[Mael SONET]

hi pstglia !

only a little message to thank you for your work and your help.
A lot of guy are being waiting for a way to install android on their yoga tablet...

So, thank you for the time you're spending on this project.

Regards.

[Anton Vosakur]

Join!

From me personally and from all our russian forum threads too.

[pstglia]

Thanks guys. I can't assure I can do this, but I can keep trying. With your help, maybe we can handle it.

I'll try to push the testing iso with Lenovo compiled kernel this weekend.

Regards,

Pstglia

[Anton Vosakur]

Pstglia

One guy from our forum made a kernel with "our sound." He says that it probably could work. Can you try to compile iso for us?
https://cloud.mail.ru/public/Gjzp/tbwgRxg15 

Thank you.

[IamBackX5]

This is compiled linux kernel (debian, ubuntu, e.t.c.), to make something developers need sources, not a kernel. There is: https://cloud.mail.ru/public/CwKu/6Q3TJ3oFJ ;)

суббота, 12 сентября 2015 г., 1:29:29 UTC+5 пользователь Anton Vosakur написал:

[pstglia]

HI all,

Correct. A kernel image won't help in this case, unless it was specific for Android (there are certain config options that must be enabled)

I finished compiling Lenovo kernel and modified a 4.4-3 img file to use it (and compiled modules).

The goal here is:

 - Check with we can boot. I tried here with my hardware (A10 5800K, desktop, not a intel tablet) but got kernel panic on gpio device init. I assumed this was due lack of specific hardware. But something else can be wrong and we can have the same problem on Yoga 2

 - If we can boot, check if you have audio inside Android (as all required drivers for this were compiled built in, there's no need to modprobe anything by now)

 - If we can boot and have sound (booting itself has a small chance of success - getting sound could be a Miracle :D ), I will ask you for some debbuging logs. 

I expect this will help to understand how arizona and 5102 machine driver initialize and register codecs, and start itself.

If you can give a try on it. Here's the ISO link (For anyone else reading this single message, keep in mind this is specific for Lenovo Yoga 2 series)

https://drive.google.com/file/d/0Bx12U5yGNcQCRDBOMTJvbGM4X0E/view?usp=sharing 

Thank you

Pstglia

[Jarrod Young]

Bad news - it isn't booting and doesnt even show the boot screen or anything beyond a "_" that stays on screen forever. Also tried DEBUG mode and same thing.

[pstglia]

Hi Jarrod,

Bad news - it isn't booting and doesnt even show the boot screen or anything beyond a "_" that stays on screen forever. Also tried DEBUG mode and same thing.

Thanks for testing. I'll try to find what's wrong.

Regards,

Pstglia

[Mael SONET]

Hello pstglia.

tied to.(yoga 2 tablet 8")

live mode : does not  work
debug mode: does not work

only a blank screen with a cursor (not bliking), no boot.

Regards.

[pstglia]

Thank you for supporting this.

A quick status. In my latest builds (Lenovo Kernel for Yoga 2) I was able to get rid of initial kernel panic messages.

I'm still receiving kernel panic when loading the i915 driver. Howvever, using "nomodeset" kernel option  I can advance in boot process, but get stucked with some "invalid opcode" messages.

In use a spare time on weekend to play with this. I'll try to boot on my testing hw (an dual core intel processor w/ intel hd graphics). If I managed to boot it, I'll upload another testing ISO.

Here's the config file currently prepared for Lenovo Kernel

[pstglia]

Hi again,

Another testing ISO:

https://drive.google.com/file/d/0Bx12U5yGNcQCZFFBcVZqdWNRRlU/view?usp=sharing

I have 2 different results with this on non-baytrail machines:

1) Dual code with GMA 3000 graphics

Can't load i915 driver. When trying to set modesetting (inteldrmfb) there's some driver problem that causes a kernel panic and boots the machine

2) Core i5 3230M with  Intel HD 4000

Could load i915 and modesetting. Boot doesn't complete because i915 prints some kernel errors.

As this kernel was provided by Lenovo for Yoga 2, I believe i915 was worked specifically for baytrail machines. So, I'm expecting you can enter graphical mode (hopefully). If you got it, try checking if the sound is working.

If so, attach here logcat and dmesg outputs. 

Thanks,

Pstglia

[Jarrod Young]

I am sorry to say that it has the same exact result. :( Thank you for your continued efforts.

[pstglia]

Hi

 

I am sorry to say that it has the same exact result. :( Thank you for your continued efforts.

Can you answer some questions?

1) Have you tried appending "nomodeset" parameter on grub and trying to boot it?

If you can have any activity at this mode, try booting under debug until 2nd "type exit" message appears. And get a dmesg output

2) Connecting a HDMI cable gives you any signal on an external monitor (if this option exists)

Regards,

Pstglia

[Nabil Hassani]

I have a Baytrail Intel proc in an Intel Compute Stick Clone,

4.4 r3 installs well, only sound and BT not working.

BT works with a nano dongle,

I tried creative audio usb which works with most OSes, no work.

The audio chip on stick is Intel of course ;)

Here are the specs:

CPU	Intel® Bay Trail T-CR Atom™ Processor Z3735F (2M Cache, up to 1.83GHz)
CPU Type	Quad Core
Memory	2GB DDR3
Storage	32GB eMMC (24.8GB Free Space)
Bios	Insyde 8MB UEFI
Operating System	Windows 8.1
Display	Via HDMI
Wifi; Bluetooth	802.11bgn; 4.0
Button / Indicator	Power, OS / Power
Input / Output	1 x Micro USB, 1 x USB 2.0, 1 x Micro SD card reader
Dimensions	110.9 * 38 * 9.8 mm
Weight	38 grams
Accessories	HDMI cable, QSG, Charger

[Nabil Hassani]

Actually Creative audio usb worked after I gave it a nudge with the command

dd if=/dev/urandom bs=8000 count=1 | aplay -D plughw:0,0

after I rebooted the usb audio sound works ;)

not ideal, but as it is a stick and I need a monitor for it, it is for home use, so having extras plugged in is not a problem.

on a side note, the BT usb dongle ended up working after reboot too :)

I checked no intel sound driver loaded, so I will next try to load it as suggested above:

busybox lsmod | grep -i snd_hda_intel

[Nabil Hassani]

results:

after loading driver: 

busybox modprobe snd_hda_inte

uid=10023(u0_a23) gid=10023(u0_a23) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50023(all_a23)@x86:/ $ su

uid=0(root) gid=0(root)@x86:/ # busybox lsmod | grep -i snd_hda_intel

snd_hda_intel          24292  0

snd_hda_controller     22904  1 snd_hda_intel

snd_hda_codec         110709  2 snd_hda_intel,snd_hda_controller

snd_pcm                84177  7 snd_hda_intel,snd_hda_controller,snd_hda_codec,snd_usb_audio,snd_soc_rt5640,snd_soc_core,snd_pcm_dmaengine

snd                    64494 11 snd_hda_intel,snd_hda_codec,snd_usb_audio,snd_usbmidi_lib,snd_hwdep,snd_rawmidi,snd_seq_device,snd_soc_core,snd_compress,snd_pcm,snd_timer

uid=0(root) gid=0(root)@x86:/ #

but still no sound card detected:

alsa_aplay -l

aplay: device_list:241: no soundcards found...

uid=0(root) gid=0(root)@x86:/ #

[Nabil Hassani]

And this when both usb sound and intel are loaded:
+++++

creative audio usb

uid=10023(u0_a23) gid=10023(u0_a23) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50023(all_a23)@x86:/ $ su

uid=0(root) gid=0(root)@x86:/ # busybox lsmod | grep -i audio
snd_usb_audio 149617 0
snd_usbmidi_lib 20684 1 snd_usb_audio
snd_hwdep 5521 1 snd_usb_audio
snd_pcm 84177 4 snd_usb_audio,snd_soc_rt5640,snd_soc_core,snd_pcm_dmaengine
snd 64494 9 snd_usb_audio,snd_usbmidi_lib,snd_hwdep,snd_rawmidi,snd_seq_device,snd_soc_core,snd_compress,snd_pcm,snd_timer
uid=0(root) gid=0(root)@x86:/ # alsa_aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: External [SB Live! 24-bit External], device 0: USB Audio [USB Audio]
Subdevices: 0/1
Subdevice #0: subdevice #0
uid=0(root) gid=0(root)@x86:/ #

++++++
now with both creative usb and integrated intel loaded

uid=10023(u0_a23) gid=10023(u0_a23) groups=1015(sdcard_rw),1023(media_rw),1028(sdcard_r),3003(inet),50023(all_a23)@x86:/ $ su

uid=0(root) gid=0(root)@x86:/ # busybox lsmod | grep -i audio
snd_usb_audio 149617 0
snd_usbmidi_lib 20684 1 snd_usb_audio
snd_hwdep 5521 1 snd_usb_audio
snd_pcm 84177 4 snd_usb_audio,snd_soc_rt5640,snd_soc_core,snd_pcm_dmaengine
snd 64494 9 snd_usb_audio,snd_usbmidi_lib,snd_hwdep,snd_rawmidi,snd_seq_device,snd_soc_core,snd_compress,snd_pcm,snd_timer
uid=0(root) gid=0(root)@x86:/ # alsa_aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: External [SB Live! 24-bit External], device 0: USB Audio [USB Audio]
Subdevices: 0/1
Subdevice #0: subdevice #0

uid=0(root) gid=0(root)@x86:/ # busybox lsmod | grep -i snd_hda_intel

1|uid=0(root) gid=0(root)@x86:/ # busybox modprobe snd_hda_intel

uid=0(root) gid=0(root)@x86:/ # busybox lsmod | grep -i snd_hda_intel
snd_hda_intel 24292 0
snd_hda_controller 22904 1 snd_hda_intel
snd_hda_codec 110709 2 snd_hda_intel,snd_hda_controller

snd_pcm 84177 8 snd_hda_intel,snd_hda_controller,snd_hda_codec,snd_usb_audio,snd_soc_rt5640,snd_soc_core,snd_pcm_dmaengine
snd 64494 12 snd_hda_intel,snd_hda_codec,snd_usb_audio,snd_usbmidi_lib,snd_hwdep,snd_rawmidi,snd_seq_device,snd_soc_core,snd_compress,snd_pcm,snd_timer
uid=0(root) gid=0(root)@x86:/ #

So for some reason, Intel audio module gets loaded, but the Intel integrated sound card still does not get to load up :(

[pstglia]

Hi Nabil,

So for some reason, Intel audio module gets loaded, but the Intel integrated sound card still does not get to load up :(

Can you check which is your integrated audio model on Windows or checking it on  /sys/bus/acpi/devices/  devices?

If you find this is not a Wolfson one (WM5102), could you open a new topic please? I think we can manage better this way.

Post also dmesg output and a list of /sys/bus/acpi/devices/.

Thank you!

Pstglia

[Anton Vosakur]

Who can explain why having the source code, it is impossible to do the build for lenovo yoga 2?
http://support.lenovo.com/ru/ru/products/tablets/yoga-series/yoga-tablet-2-1050

We also have images of partitions from android version, obtained by DD.
It may be possible to substitute something in 4.4 P3 to get the result?

[Nabil Hassani]

Hi Pstglia

OK will do.

[pstglia]

Hi Anton,

Em sexta-feira, 25 de setembro de 2015 12:34:39 UTC, Anton Vosakur escreveu:

Who can explain why having the source code, it is impossible to do the build for lenovo yoga 2?
http://support.lenovo.com/ru/ru/products/tablets/yoga-series/yoga-tablet-2-1050

I've been working with these sources. Not all the necessary components are there to compile a full ISO/IMG.

 

We also have images of partitions from android version, obtained by DD.
It may be possible to substitute something in 4.4 P3 to get the result?

What I've been trying to do is compiling the provided kernel with these sources and copy it into a Android-x86 img (The latests imgs I pushed are using it). My goal is getting information in order to understand how audio drivers are working and implement this on the current kernel Android-x86 is using.

As you can see in the latest posts, we can have img booting this img (not on Lenovo Yoga at least).

The latest IMG I pushed was on Sep, 19th. Jarrod tested it but still no img to perform the audio debug. I asked him some questions. Currently waiting for him or someone else to answer it.

That's a hard task (maybe not possible to me to achieve), but I'll keep trying as long as you want to.

Regards,

Pstglia

[Anton Vosakur]

Thank you for the clarification.