maemo.org - Talk - [Android] Another Android port: NITdroid

How are you guys getting this to work? I tried installing Calendar/Maps/GTalk/IM/etc. from a ROM dump from a G1 and it always says installation failed while trying to install.

???

Thesandlord

2008-12-30 02:39

Re: Another Android port: NITdroid

Brent, it looks like you need to push it to /system/app/. I have not tried it yet, but...

Also, you had Android in 2007! Wow, you are amazing :)

maxkilla

2008-12-30 04:12

Re: Another Android port: NITdroid

Quote:

Originally Posted by BrentDC (Post 253343)

How are you guys getting this to work? I tried installing Calendar/Maps/GTalk/IM/etc. from a ROM dump from a G1 and it always says installation failed while trying to install.

???

was your dump the dev phone? thats the one where i got my apk files from..

BrentDC

2008-12-30 04:14

Re: Another Android port: NITdroid

Hmm, I wonder if it has something to do with the dev phone apps though? I'll have to try this out (if someone doesn't beat me to it).

Edit: Ah! 2 minutes late.

Could you send me any .apk files you have of Calendar/Maps/GTalk/YouTube/etc. from the dev phone? My email is bchiodo[at]gmail[dot]com. Thank you very much, if you don't mind.

BrentDC

2008-12-30 16:05

Re: Another Android port: NITdroid

Thanks a lot Maxkilla! I'm going to try to see what I can get working; I'll report back if I have any success :)

zuk	2008-12-30 16:05

Re: Another Android port: NITdroid

Is there a github repo or something similar with the nitdroid code? I'm interested in having a look at the source and maybe contributing some patches if possible, but I can't find any info on how best to contribute.

jcl	2008-12-30 18:11

Re: Another Android port: NITdroid

vendor/nokia/nit/BoardConfig.mk reads BOARD_USES_ALSA_AUDIO := false
i guess that should be true, anybody tried to build it like that?
Was it deliberately disabled?

Bundyo

2008-12-30 18:14

Re: Another Android port: NITdroid

Yep, sound is not working yet, so it was disabled.

jcl	2008-12-30 18:18

Re: Another Android port: NITdroid

Must be very close, the driver seems to load correctly (slightly modified kernel) ...
Any clue what is missing?

Bundyo

2008-12-30 18:51

Re: Another Android port: NITdroid

Alsa support? The driver is DSP only (of course I can speak only for 770 and N800).

BrentDC

2008-12-30 23:54

Re: Another Android port: NITdroid

Quote:

Originally Posted by BrentDC (Post 253445)

Thanks a lot Maxkilla! I'm going to try to see what I can get working; I'll report back if I have any success :)

No success.

Maps loads, but doesn't download any map data.
Calendar loads, but just says "Waiting to Sync" (or something like that), and doesn't do anything.
Gmail crashes immediately after loading.
Same for GTalk and YouTube.

Other problems I've noticed with Android:

The touchscreen is way too sensative. Trying to scroll through a list of things is really tough because i always end up selecting something.
The touchscreen also seems to be miscalibrated . A tap of the screen frequently registers to the right or left of where i actually tapped it.
The On-Screen Keyboard is flaky. It doesn't register taps real well and it is impossible to edit text you already entered into, let's say, the browser. Also when trying to scroll through a list, using the Up/Down buttons, it will say the keyboard crashed sometimes. Even though the Keyboard isn't open...
In addition, it would be great if the settings in Dev Tools->Development were persistent through shut downs. Because my N800 is far too unstable if I don't limit the amount of processes allowed to run.

Thanks for all the work going into this. :)

KeeganX

2009-01-01 20:29

Re: Another Android port: NITdroid

Hi, I would like to put android on my N810. The problem is my internal memory card is corrupted. Is there a way I could put in a mini sd card and run android off of that?

sarathmenon

2009-01-02 05:10

Re: Another Android port: NITdroid

Quote:

Originally Posted by Bundyo (Post 253494)

Alsa support? The driver is DSP only (of course I can speak only for 770 and N800).

Maemo stock runs alsa. This is from mine:

Code:

Nokia-N810-WiMAX-43-7:~# cat /proc/asound/cards

 0 [EAC            ]: EAC - OMAP24xx EAC

                      OMAP24xx EAC with codec TLV320AIC33

AFAIK, the driver is fairly open sourced, and a google search points me to the omap-linux mailing list.

Johnx

2009-01-02 05:38

Re: Another Android port: NITdroid

@sarathmenon: That alsa layer sits on top of the dsp though. To use alsa on the Nx80 one needs not only the dsp tasks, but also plenty of alsa libraries.

solca

2009-01-02 06:24

Re: Another Android port: NITdroid

Hi!

New update, hopefully it'll rock!

(Yea theres tons of bugs) :D

sarathmenon

2009-01-02 07:05

Re: Another Android port: NITdroid

Quote:

Originally Posted by Johnx (Post 253982)

@sarathmenon: That alsa layer sits on top of the dsp though. To use alsa on the Nx80 one needs not only the dsp tasks, but also plenty of alsa libraries.

Thanks for clarifying. I haven't looked into it in more detail; I was just poking under /proc out of curiosity. Sucks that we have mostly open hardware, but no good enough free drivers.

jogorman

2009-01-02 07:54

Re: Another Android port: NITdroid

When enabling wireless, and going into WIFI settings, I get a device freeze followed by a shutdown. I grabbed the logcat:

With supplicant

Code:

E/bluetooth_common.cpp(  741): dbus_func_args_timeout_valist: D-Bus error in ListBondings: org.freedesktop.DBus.Error.ServiceUnknown (The name org.bluez was not provided by any .service files)

D/SettingsWifiEnabler( 1112): Received wifi state changed from Enabling to Enabled

I/ActivityManager(  741): Displayed activity com.android.settings/.WirelessSettings: 623 ms

I/ActivityManager(  741): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.wifi.WifiSettings} }

I//system/bin/wpa_supplicant(  795): CTRL-EVENT-STATE-CHANGE id=-1 state=2

V/WifiMonitor(  741): Event [CTRL-EVENT-STATE-CHANGE id=-1 state=2]

V/WifiStateTracker(  741): Changing supplicant state: INACTIVE ==> SCANNING

D/SettingsWifiEnabler( 1112): Received wifi state changed from Enabling to Enabled

Without supplicant

Code:

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-SCAN-RESULTS  Ready

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-STATE-CHANGE id=-1 state=1

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

V/WifiStateTracker(  780): Connection to supplicant established, state=INACTIVE

I/ActivityManager(  780): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.WirelessSettings} }

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

I/ActivityManager(  780): Displayed activity com.android.settings/.WirelessSettings: 540 ms

I/ActivityManager(  780): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.wifi.WifiSettings} }

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-STATE-CHANGE id=-1 state=2

V/WifiMonitor(  780): Event [CTRL-EVENT-STATE-CHANGE id=-1 state=2]

V/WifiStateTracker(  780): Changing supplicant state: INACTIVE ==> SCANNING

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

andreww

2009-01-02 10:46

Re: Another Android port: NITdroid

Quote:

Originally Posted by solca (Post 253986)

Hi!

New update, hopefully it'll rock!

(Yea theres tons of bugs) :D

It's very likely just something stupid I did, but I couldn't get the new release to run out of the box. The rootfs was missing a lot of seemingly important stuff, like /init, /sbin/adbd, and some other misc files (which seem like they might be obsolete).

After copying them from the previous rootfs, it started up. It looks like stlc45xx.ko isn't working for me either. If nobody else is having these problems, maybe I'll try all of the steps again from scratch. As a note, I'm using an n810.

Either way, it feels like huge improvements have been made in the new version. The background is a great touch, and things seem a bit smoother.

Thanks a ton for working on this!

- Andrew

cyrus

2009-01-02 16:21

Re: Another Android port: NITdroid

Quote:

Originally Posted by jogorman (Post 254001)

When enabling wireless, and going into WIFI settings, I get a device freeze followed by a shutdown. I grabbed the logcat:

With supplicant

Code:

E/bluetooth_common.cpp(  741): dbus_func_args_timeout_valist: D-Bus error in ListBondings: org.freedesktop.DBus.Error.ServiceUnknown (The name org.bluez was not provided by any .service files)

D/SettingsWifiEnabler( 1112): Received wifi state changed from Enabling to Enabled

I/ActivityManager(  741): Displayed activity com.android.settings/.WirelessSettings: 623 ms

I/ActivityManager(  741): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.wifi.WifiSettings} }

I//system/bin/wpa_supplicant(  795): CTRL-EVENT-STATE-CHANGE id=-1 state=2

V/WifiMonitor(  741): Event [CTRL-EVENT-STATE-CHANGE id=-1 state=2]

V/WifiStateTracker(  741): Changing supplicant state: INACTIVE ==> SCANNING

D/SettingsWifiEnabler( 1112): Received wifi state changed from Enabling to Enabled

Without supplicant

Code:

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-SCAN-RESULTS  Ready

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-STATE-CHANGE id=-1 state=1

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

V/WifiStateTracker(  780): Connection to supplicant established, state=INACTIVE

I/ActivityManager(  780): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.WirelessSettings} }

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

I/ActivityManager(  780): Displayed activity com.android.settings/.WirelessSettings: 540 ms

I/ActivityManager(  780): Starting activity: Intent { action=android.intent.action.MAIN comp={com.android.settings/com.android.settings.wifi.WifiSettings} }

I//system/bin/wpa_supplicant( 1205): CTRL-EVENT-STATE-CHANGE id=-1 state=2

V/WifiMonitor(  780): Event [CTRL-EVENT-STATE-CHANGE id=-1 state=2]

V/WifiStateTracker(  780): Changing supplicant state: INACTIVE ==> SCANNING

D/SettingsWifiEnabler( 1170): Received wifi state changed from Enabling to Enabled

I get the same thing.

Thesandlord

2009-01-02 19:33

Re: Another Android port: NITdroid

Quote:

Originally Posted by andreww (Post 254018)

I though these were supposed to be inside the kernel now or something.

Anyway, it is not working for me either. so I am guessing either the directions are wrong, or there is a glitch with this release...

Edit: Nope, still won't boot...

Edit 2: I am such an idiot, forgot the external card had to be removed... AND I FLASHED THE WRONG KERNEL. I am a complete fool....

meizirkki

2009-01-02 20:23

Re: Another Android port: NITdroid

Are you sure you remembered to remove the card in external slot???

zehjotkah

2009-01-02 20:28

Re: Another Android port: NITdroid

it worked for me

solca

2009-01-02 20:30

Re: Another Android port: NITdroid

Quote:

Originally Posted by Thesandlord (Post 254092)

Yes, now the new kernels contains an embedded RAM filesystem as other Android devices (G1) which contains init and the other files in /, the userspace tarball that you must untar to mmcblk0p3 contains:

/cache
/data
/nokia
/system

So installation remains the same but you have to flash the new kernel, if you have extra stuff than the above in mmcblk0p3 it simply it is not used.

If you flash the debugging kernel in this release the wifi driver will not load or will hang the device as it is compiled for the normal kernel, I had to figure out how to ship different modules for the debugging kernel yet.

The debugging kernel disables all logos and animations and as usual enables screen console output so if you have problems please try next the debugging kernel and report back.

As there are many new features please report any problem as I probably don't check every single thing. Thx!

Bundyo

2009-01-02 20:43

Re: Another Android port: NITdroid

This embedding the ram fs in the kernel will probably have problems with 770, since you can only have a kernel up to 2MB. How exactly is it done and where can i get the init?

andreww

2009-01-02 21:45

Re: Another Android port: NITdroid

Quote:

Originally Posted by solca (Post 254100)

Got it. For some reason, the kernel I flashed to yesterday was the same as the one I had before. Downloaded it again, made sure the file was different, flashed it and am trying it now.

Thanks again!

cyrus

2009-01-02 21:55

Re: Another Android port: NITdroid

Works now - using the non-debug kernel.

For some reason I find this release to be slower. Is there a reason why Android seems so much more sluggish than Maemo? Is it because it is running off the flash card?

Thanks

chyang

2009-01-02 23:33

Re: Another Android port: NITdroid

I just updated to the newest release 02Jan. And found some issues
1. wireless doesn't work
dmesg shows that it fails to request_firmware, 3826.arm

2. cannot use adb with it
i am using linux in vmware under windows, and previous it works great, but now, it is sluggish when i connect the usb cable, and make my linux responds slowly. dmesg under linux:
[ 2266.959538] usb 2-1: new full speed USB device using uhci_hcd and address 5
[ 2267.366669] usb 2-1: not running at top speed; connect to a high speed hub
[ 2267.366713] usb 2-1: config 1 has 1 interface, different from the descriptor's value: 2
[ 2267.385657] usb 2-1: configuration #1 chosen from 1 choice
[ 2267.399854] scsi9 : SCSI emulation for USB Mass Storage devices
[ 2267.402649] usb-storage: device found at 5
[ 2267.402654] usb-storage: waiting for device to settle before scanning
[ 2272.972216] usb 2-1: reset full speed USB device using uhci_hcd and address 5
[ 2273.459676] usb 2-1: device descriptor read/64, error -84
[ 2274.428339] usb 2-1: device descriptor read/64, error -84
[ 2274.661102] usb 2-1: reset full speed USB device using uhci_hcd and address 5
[ 2277.184100] usb 2-1: device descriptor read/64, error -84
[ 2277.780263] usb 2-1: device descriptor read/64, error -84
[ 2278.013089] usb 2-1: reset full speed USB device using uhci_hcd and address 5
[ 2278.849655] usb 2-1: device descriptor read/8, error -84
[ 2279.169490] usb 2-1: device descriptor read/8, error -71
[ 2279.332287] usb 2-1: USB disconnect, address 5
[ 2279.332287] scsi 9:0:0:0: Device offlined - not ready after error recovery
[ 2279.332287] usb-storage: device scan complete
[ 2285.071326] usb 1-1: new high speed USB device using ehci_hcd and address 6
[ 2285.501987] usb 1-1: config 1 has 1 interface, different from the descriptor's value: 2
[ 2285.513084] usb 1-1: configuration #1 chosen from 1 choice
[ 2285.517470] scsi10 : SCSI emulation for USB Mass Storage devices
[ 2285.523473] usb-storage: device found at 6
[ 2285.523481] usb-storage: waiting for device to settle before scanning
[ 2288.617149] usb 1-1: USB disconnect, address 6
Meanwhile, if connected with usb, the n810 will reboot continously(seems).

3. no suid on /system
thus i cannot get the right permission to run logcat, or some other utility.

Am I the only one meeting those issues? Thanks.

Are

meizirkki

2009-01-03 00:07

Re: Another Android port: NITdroid

I feel it is slower to boot up than earlier version, but there is only few things it is possiblle to nagg about. _Really_ impressive work solca! :)

jogorman

2009-01-03 01:05

Re: Another Android port: NITdroid

Wireless is now working for me now that I have the non-debugging kernel enabled.

adb works just fine for me, not sure what that issue is about. Build seems fast to me, if a bit longer to boot. The power updates are great, having it report battery power is pretty nice.

Problem I am having is, cookies still do not work in the browser. Not sure what is up with that. Anyone having the same issue?

I reloaded the system from scratch (new fs on the card, etc) to make sure that I don't have some carry over incompatibility.

Great work on this. I check the thread daily for updates.

Thesandlord

2009-01-03 01:31

Re: Another Android port: NITdroid

jogorman, the cookies don't work for some reason.

The new release seems slower but smoother... Looks like I am still having that DNS problem. Very good though, almost complete.

And time works! Yeah!!!

Just wondering, how come the EEE PC port took such a little time, when this is taking longer. Are x86 systems easier to work with?

Problems: My computer did not boot when N810 was attached via USB.

adb is not working under windows. It looks like Android is identifing itself as a "USB composite device", which is stupid. How do I enable adb back on the N810?

cyrus

2009-01-03 03:05

Re: Another Android port: NITdroid

solca:

Can you update the links on your page with the latest patch - kernel and user, as well as the latest kernel.config?

Thanks
Awesome work!!

Thesandlord

2009-01-03 04:08

Re: Another Android port: NITdroid

Ok, got adb working in Linux, you just need to enable USB debugging (sigh)... This was in DevTools before, looks like it got moved...

Trying windows now... Yes it works...

solca

2009-01-03 05:51

Re: Another Android port: NITdroid

Quote:

Originally Posted by Bundyo (Post 254102)

This embedding the ram fs in the kernel will probably have problems with 770, since you can only have a kernel up to 2MB. How exactly is it done and where can i get the init?

When you finish compiling Android there is a ramdisk.img file which contains everything in the rootdir.

But I'll include everything in the next tarball as before just that will be unused by N8x0 devices as these will continue using the initramfs aproach.

Quote:

Originally Posted by cyrus (Post 254118)

For some reason I find this release to be slower. Is there a reason why Android seems so much more sluggish than Maemo? Is it because it is running off the flash card?

Quote:

Originally Posted by meizirkki (Post 254138)

I feel it is slower to boot up than earlier version

I found the cause, I didn't use the correct compiler. Will fix for the next release.

Quote:

Originally Posted by jogorman (Post 254147)

Problem I am having is, cookies still do not work in the browser. Not sure what is up with that. Anyone having the same issue?

Cookies don't work since Google update the browser to the november webkit.

Quote:

Originally Posted by Thesandlord (Post 254152)

Just wondering, how come the EEE PC port took such a little time, when this is taking longer. Are x86 systems easier to work with?

The EEE hardware have open source Linux drivers that follow standard Linux APIs that don't require binary blobs to work. The NITs require propietary binary daemons (dsme, bme, kicker) to function, it's open source drivers for basic things like Tahvo and Retu ASICs don't expose standard APIs and just modify directly registers, all the real thing is done in userspace and obfuscated by the propietary daemons.

It's a day and night difference, although it could be much worse, it seems Nokia close the door to other OSes than Maemo.

Quote:

Originally Posted by cyrus (Post 254163)

solca:

Can you update the links on your page with the latest patch - kernel and user, as well as the latest kernel.config?

I update all source patches at the same time as the precompiled downloads or I'm missing something?

Quote:

Originally Posted by Thesandlord (Post 254168)

Ok, got adb working in Linux, you just need to enable USB debugging (sigh)... This was in DevTools before, looks like it got moved...

The normal kernel have adb disabled by default, you have to enable it in settings as any normal Android device, the debugging kernel have it enabled by default.

Bundyo

2009-01-03 08:33

Re: Another Android port: NITdroid

Quote:

Originally Posted by solca (Post 254184)

Cookies don't work since Google update the browser to the november webkit.

Yes, WebKit has its own share of issues... In the WebKit in default browser thread Pronvit mentioned how you can use libcurl to handle cookies, maybe that will help?

chyang

2009-01-03 17:11

Re: Another Android port: NITdroid

After a clean installation, all seem working properly. There must be some directory permission or ownership issues with my previous environment.

But the ro.secure=1 seems to have disabled all the fun of owning the NIT.

solca

2009-01-03 21:26

Re: Another Android port: NITdroid

Quote:

Originally Posted by chyang (Post 254236)

But the ro.secure=1 seems to have disabled all the fun of owning the NIT.

Yes, added to my checklist before releasing... :o

cyrus

2009-01-04 01:00

Re: Another Android port: NITdroid

Quote:

Originally Posted by solca (Post 254184)

I update all source patches at the same time as the precompiled downloads or I'm missing something?

Has the kernel patch not changed then? I have kept each file as you have updated and diffed the old one vs the new one and there are no differences.

stelleg151

2009-01-04 22:09

Re: Another Android port: NITdroid

@Solca & Bundyo: Really great work guys, I've been playing with android and I have wanted to get it working on my n770.

Solca, your n770 debug kernel image fails with that annoying fading nokia screen and restarts.
Bundyo, both of your images get to the verbose linux output bootup screen and successfully mounts my mmc, but fails with:

Code:

Kernel panic - not syncing; No init found. Try passing init= option to kernel.

It has successfully mounted my ext3 partition as readonly root.

There were are few command timeouts for mmci-omap for CMD5 and CMD55, and one for CMD8.

I'm using the current android rootfs image from Solca's site. I tried changing permissions in Ubuntu to read and write, but it's still being read as read-only.

Again great work guys, and kudos to you bundyo for getting this far with the n770, all of my attempted compiles have ended like Solca's, with the fading nokia splash.

All times are GMT. The time now is 18:23.

Page 11 of 14

Show 40 post(s) from this thread on one page

vBulletin® Version 3.8.8