[ROM][GEN2]CyanogenMod 7 (Android 2.3.7)

[Guest Urpo]

Bah.... had to wipe everything and format the whole sdcard after before I was able to reinstall the rom.

is that the market app or the zip i attached above?

The market blade wifi fix app. As the market blade wifi fix app seems to work for others, it must be then either Toggle 2G/3G app, or the combination of Toggle 2G/3G and market blade wifi fix app. I suspect the combination, don't have the time to test this now.

certainly with the wpa_supplicant zip above, i have no wake issues, however it doesn't actually work - it never turns off wifi!

Yes, confirmed.Going to try still some wifi mgmt app.

[Guest asm19]

Guys, still one of the problems is the fact that MMS still does not work without resorting to the Internet, for those who do not have Internet tariff spend money ... I've tried installing APNs manually but still the same. Still no solution for this?

I recently spoke with the creator of the ROM in Portuguese Tejo based Cyanogen, this ROM have him MMS to work properly and asked him what he did to solve it, he replied:

"Well, the MMS run through a data connection that is configured through the APN. As a general rule, the Android 2.3 will only send if the Data (I mean, all APN) are active. What was done was in a Tejo "cheating" to this system, allowing the app would activate the MMS data and linking their logo to the APN, turning off immediately. "

sej7278 if it were possible to solve this problem was spectacular in the same way

Thanks

Help please...

[Guest sm4tik]

did you pull in the full patch from gerrit - not just the kernel blob but the libcopybit etc?

cd device/zte/blade/

git fetch http://review.cyanogenmod.com/p/CyanogenMod/android_device_zte_blade refs/changes/52/7852/3 && git format-patch -1 --stdout FETCH_HEAD > /tmp/2635p3.patch

git apply /tmp/2635p3.patch  

otherwise it does look like some of the files from your error message are base android headers etc. so might be out-of-date/knackered due to sync issues.

I cherry-picked the commit and all files are there, so it should be fine. I think it might be a mix of building with testing vendor libs without the kernel commit and using ccache. I'll do a clean build and see if the problem still exists. Now to figure out how to do a clean build of the android part only without having to rebuild the whole thing.. duh.

[Guest sej7278]

I cherry-picked the commit and all files are there, so it should be fine. I think it might be a mix of building with testing vendor libs without the kernel commit and using ccache. I'll do a clean build and see if the problem still exists. Now to figure out how to do a clean build of the android part only without having to rebuild the whole thing.. duh.

i'd give up mate, without the android git repo's, if you didn't already have a working setup its going to be difficult to figure out if its a problem with whatever you're building or a problem with your development environment, you need to delete your ~/.ccache and out directories for a start.

at the moment i'm just syncing what few commits are in the cyanogenmod repo's and pulling the odd patch from gerrit and github and making sure i don't break anything in the android repo i've got downloaded.

i can't believe google haven't put up a mirror (hell it wasn't even the server that was hacked!) i'm not entirely convinced kernel.org should be hosting android really.

anyway there's nothing much to do until tom_g makes a patch4 with fixed wifi reconnects :unsure:

Edited September 13, 2011 by sej7278

[Guest sm4tik]

i'd give up mate, without the android git repo's, if you didn't already have a working setup its going to be difficult to figure out if its a problem with whatever you're building or a problem with your development environment, you need to delete your ~/.ccache and out directories for a start.

at the moment i'm just syncing what few commits are in the cyanogenmod repo's and pulling the odd patch from gerrit and github and making sure i don't break anything in the android repo i've got downloaded.

i can't believe google haven't put up a mirror (hell it wasn't even the server that was hacked!) i'm not entirely convinced kernel.org should be hosting android really.

anyway there's nothing much to do until tom_g makes a patch4 with fixed wifi reconnects :unsure:

I never give up :D

Anyway, I'm doing this just for myself and learning as I go, so no actual harm done. I hope I'll get it fixed at some point as I've put quite few good hours into it. Nothing special, but things I like to have/discard during the build process, not editing the update.zip each time. It's been a good learning process btw :) I think I'll 'make clean' and let it rebuild the whole thing over night.

I guess I could just replace the kernel from my previous succesful build, but that wouldn't be any fun, would it ;)

[Guest sej7278]

patch3 stays on when the screen is off, until 15mins is up, then goes off and never comes back on again (until you reboot, or apparently turn it off and on again using the setting).

patch3 with the wpa_supplicant fix reconnects after sleep mainly as it never sleeps lol.

the wpa_supplicant fix worked fine for 2.6.32 n179+, it disconnected with the screen, turned off after 15mins and came back instantly. i guess it doesn't work with patch3 as the new kernel has a different wifi subsystem.

can anyone confirm patch3's behaviour, as i just removed the wpa_supplicant fix and rebooted and after a few hours i've noticed that wifi was on the whole time, although it won't reconnect (so confirmed the wpa has gone!) it hasn't used much battery though.

so i'm confused now. i thought patch3 went off and never came back on, but it looks like it just disconnects from the ap,

[Guest Geree]

can anyone confirm patch3's behaviour, as i just removed the wpa_supplicant fix and rebooted and after a few hours i've noticed that wifi was on the whole time, although it won't reconnect (so confirmed the wpa has gone!) it hasn't used much battery though.

so i'm confused now. i thought patch3 went off and never came back on, but it looks like it just disconnects from the ap,

Yes, it just disconnects from the AP, but you have to reconnect manually. I didn't have to reboot anytime. (expect after the automatic shutdown...)

[Guest sej7278]

Yes, it just disconnects from the AP, but you have to reconnect manually. I didn't have to reboot anytime. (expect after the automatic shutdown...)

thanks, thought i was going mad for a minute there!

when it disconnects i guess it goes into powersaving as 5 hours of wifi should have used a lot more than 7% battery.

[Guest Geree]

I'm having a wakelock problem again with patchset 3. I killed every apps, turned off wifi, but I still can wake the phone up with the back, home, menu buttons...

I don't know what caused this, in the morning it was fine. :S

[Guest sm4tik]

Just an observation I made earlier. The camera zoom does work, but not all the way up to 2x. I thought it didn't crop the picture at all, but now looking at it closer I think it gets "stuck" at somewhere around 1.5x. Nothing to do with the new kernel so I'll freely let this post get buried :)

[Guest Geree]

This is what "dumpsys power" says (when phone is not connected via usb):

Power Manager State:

  mIsPowered=false mPowerState=3 mScreenOffTime=234395 ms

  mPartialCount=1

  mWakeLockState=

  mUserState=SCREEN_BRIGHT_BIT SCREEN_ON_BIT 

  mPowerState=SCREEN_BRIGHT_BIT SCREEN_ON_BIT 

  mLocks.gather=

  mNextTimeout=56610994 now=56605052 5s from now

  mDimScreen=true mStayOnConditions=0

  mScreenOffReason=2 mUserState=3

  mBroadcastQueue={-1,-1,-1}

  mBroadcastWhy={0,0,0}

  mPokey=1 mPokeAwakeonSet=false

  mKeyboardVisible=false mUserActivityAllowed=true

  mKeylightDelay=6000 mDimDelay=47000 mScreenOffDelay=7000

  mPreventScreenOn=false  mScreenBrightnessOverride=-1  mButtonBrightnessOverride=-1

  mScreenOffTimeoutSetting=60000 mMaximumScreenOffTimeout=2147483647

  mLastScreenOnTime=96180196

  mBroadcastWakeLock=UnsynchronizedWakeLock(mFlags=0x1 mCount=0 mHeld=false)

  mStayOnWhilePluggedInScreenDimLock=UnsynchronizedWakeLock(mFlags=0x6 mCount=0 mHeld=false)

  mStayOnWhilePluggedInPartialLock=UnsynchronizedWakeLock(mFlags=0x1 mCount=0 mHeld=false)

  mPreventScreenOnPartialLock=UnsynchronizedWakeLock(mFlags=0x1 mCount=0 mHeld=false)

  mProximityPartialLock=UnsynchronizedWakeLock(mFlags=0x1 mCount=0 mHeld=false)

  mProximityWakeLockCount=0

  mProximitySensorEnabled=false

  mProximitySensorActive=false

  mProximityPendingValue=-1

  mLastProximityEventTime=87653246

  mLightSensorEnabled=true

  mLightSensorValue=14.0 mLightSensorPendingValue=-1.0

  mLightSensorScreenBrightness=35 mLightSensorButtonBrightness=255 mLightSensorKeyboardBrightness=0

  mUseSoftwareAutoBrightness=true

  mAutoBrightessEnabled=true

  mScreenBrightness: animating=false targetValue=35 curValue=35.0 delta=2.3333333


mLocks.size=1:

  PARTIAL_WAKE_LOCK              'GpsLocationProvider' activated (minState=0, uid=1000, pid=232)


mPokeLocks.size=1:

    poke lock 'PhoneApp': POKE_LOCK_IGNORE_CHEEK_EVENTS 

from that I assume the partial gps wakelock causes the issue.. can be a Maps bug?

edit: pid 232 belongs to system_server process (sorry for the long post)

Edited September 13, 2011 by Geree

[Guest krauli]

Guys, still one of the problems is the fact that MMS still does not work without resorting to the Internet, for those who do not have Internet tariff spend money ... I've tried installing APNs manually but still the same. Still no solution for this?

I recently spoke with the creator of the ROM in Portuguese Tejo based Cyanogen, this ROM have him MMS to work properly and asked him what he did to solve it, he replied:

"Well, the MMS run through a data connection that is configured through the APN. As a general rule, the Android 2.3 will only send if the Data (I mean, all APN) are active. What was done was in a Tejo "cheating" to this system, allowing the app would activate the MMS data and linking their logo to the APN, turning off immediately. "

sej7278 if it were possible to solve this problem was spectacular in the same way

Thanks

is there any solution for this, yet?

[Guest paulocostavct]

Hi.

I have the same problem with mms, i'm from Portugal too, and would be awesome if this could be solved.

[Guest asm19]

Hi.

I have the same problem with mms, i'm from Portugal too, and would be awesome if this could be solved.

yeah, i'm also from Portugal. Please correct this situation.

[Guest Kamikaze.]

So, what do you guys think, 7.1 stable enough? I'm currently on Stir Fry because CM7 was so buggy, and I'm wondering, is it time to go back. Honestly, I don't give a s*** of gps or MMS, so...

EDIT: Nah, whatever, already upgraded ;)

Edited September 14, 2011 by Kamikaze.

[Guest hlubepet]

So, what do you guys think, 7.1 stable enough? I'm currently on Stir Fry because CM7 was so buggy, and I'm wondering, is it time to go back. Honestly, I don't give a s*** of gps or MMS, so...

mee too... more than a day with patch3 version and totally happy...

[Guest Pelemanne]

mee too... more than a day with patch3 version and totally happy...

update-cm-7.1.0-RC1-Blade-KANG-n179-2635p3-signed is ready and steady.

[Guest hlubepet]

update-cm-7.1.0-RC1-Blade-KANG-n179-2635p3-signed is ready and steady.

and the difference is (are)?

difference are nothing because Im using this version allready from begining :-P

"me too" was for the mms and gps thing :)

Edited September 14, 2011 by hlubepet

[Guest Pelemanne]

and the difference is (are)?

difference are nothing because Im using this version allready from begining :-P

"me too" was for the mms and gps thing :)

Compared to previous versions.

Edited September 14, 2011 by Pelemanne

[Guest hlubepet]

Compared to previous versions.

i know, tried more than enought count of rom... finally cm7 is usuable for me... gr8!!

[Guest sej7278]

....

mLocks.size=1:

  PARTIAL_WAKE_LOCK              'GpsLocationProvider' activated (minState=0, uid=1000, pid=232)


mPokeLocks.size=1:

    poke lock 'PhoneApp': POKE_LOCK_IGNORE_CHEEK_EVENTS 

from that I assume the partial gps wakelock causes the issue.. can be a Maps bug?

edit: pid 232 belongs to system_server process (sorry for the long post)

looks like it, maps is so bad i refuse to use it anymore, seems to be the root of all evil when it comes to gps reboots (that don't happen with other gps apps) and wakelocks. so not a kernel issue it seems.

as far as patch3 readiness, i don't think i'll go back to the regular nightlies now, the only bug is wifi reconnect it seems.

Edited September 14, 2011 by sej7278

[Guest hankhandsome]

Patch 3 actually works for me now - no more force close on WiFi selection. And it's great - the battery life is shaping up to be pretty good as well. Only lost about 3 percent last night with some texting, gaming and WiFi web surfing.

[Guest Tom G]

I can confirm this :D, great work to tom and the rest of the guys working on the .35 kernel

I however have one bug and that is either cifs in the kernel is missing or isnt working because when I try to mound a share it says device not found, just to make sure it wasnt my server I tried loading the cifs module and it mounted the share fine

cifs is built as a module. You need to load the module first. The old kernel was the same.

It seems battery took a hit with the new kernel!

I am using patch3 (thanks sej!) and in 15min I am already at 97% (8s of wifi,22 of screen, the rest is Android OS and inactivity).

Anyone having the same problem?

at 88% after 23 hours on battery here. Looks ok to me.

After an upgrade to 2.6.35 and dalvik/cache wipe, debug.composition. type=gpu (what gsf use to speed up scrolling in menus) seems to be working :-)

debug.composition.type=gpu does nothing. Even in gsf it does nothing. It is a debug setting from CAF which has no meaning in AOSP/CM7, but gpu is not a valid setting for it. Removing the line entirely will give the same results in gsf. Link

Trying sej7278's Patch3 build, openvpn dies after authentication and successfully getting the routing information.

Logcat said:-

D/OpenVPN-DaemonMonitor[/sdcard/openvpn/ngauchai.ovpn]-daemon-stdout(  753): Tue Sep 13 17:46:50 2011 Note: Cannot open TUN/TAP dev /dev/tun: No such file or directory (errno=2)

D/OpenVPN-DaemonMonitor[/sdcard/openvpn/ngauchai.ovpn]-daemon-stdout(  753): Tue Sep 13 17:46:51 2011 Note: Attempting fallback to kernel 2.2 TUN/TAP interface

D/OpenVPNDaemonEnabler(  753): Received OpenVPN daemon state changed from Unknown to DisabledD/OpenVPN-DaemonMonitor[/sdcard/openvpn/ngauchai.ovpn]-mgmt(  753): terminated

Any idea? Thanks.

TUN support will be missing in the kernel. I'll fix it.

Wifi reconnects good for me. I'll get the awext fix submitted soon.

[Guest targetbsp]

Played Kairosofts latest for 4 and a half hours last night so gaming battery life is at least as good as previous kernel (which varied between 4 and 5 hours of gaming). And the standby life seems perfectly good too.

[Guest Revenge]

Hello people, im using the last Official Nightly 179.

I don't have issues in gps or wifi like some people talks.

But i have one issue, that is my phone doesn't go to deep sleep. You can test that, if the phone wakes up by pressing the home key for example. It should only wake up if pressing the power button.

Can someone check if you have the same issue? Many Thanks

Here is my logcat: http://pastebin.com/YZTwAwhR