Re: [TESTING] - NAND OMGB (1.2.3)
 

...Can we get back to testing OMGB?

Thanks!!


Edit - on that topic, I had a loss of audio again. Incoming message didn't make any sound just vibrated. It turned the speaker on and made that odd noise - like low amounts of feedback, as if the speaker was turned on at full blast, but no input going to it.

Hopefully this log captures the moment of audio failure... I grabbed the log immediately after the failed notification.

Edit2 - crap, even tho I grabbed logs and this time put them to the SD card because pastebin failed last time, it seems like this time it failed because the logs I posted appear to be from HaRET. Oy... GetLogs hasn't failed like this before, or I'm just going nuts...:drunken:

Edit3 - Well, it certainly does appear to be from HaRET, but it definitely has the classic audio problem at the end of the log. Hum... I swear the date/timestamp matches when I had the problem, and I definitely was running on NAND.

Re: [TESTING] - NAND OMGB (1.2.3)
 

You have some interesting issues. Well i notice that your device is having a touch time sleeping on the last few lines but then it gets cut off. Now, when it becomes unresponsive, is it a complete lock? where you have to physically reset the device? If so, then there was a kernel panic. A panic will lock up the phone and force you to reset. However, the next step is crucial. When the phone comes back up from the reset, you have to "adb shell" into your phone. then type the following.

cat /proc/last_kmsg > /sdcard/panic.txt

This will create a panic.txt on your sdcard which you can upload to us here. Panics are usually a result of either SMD bugs or possibly something new.

Also there is a CDMA error on the logs.

"01-14 08:35:34.360 E/CDMA ( 221): Wrong network. Can not return a string." I dont think ive ever seen that. Maybe muzling has seen it since he has explored a lot of other cdma issue. Either way that will require some research.

Re: [TESTING] - NAND OMGB (1.2.3)
 

This is odd. This is the same error i'm having now with the new build. However i have never faced it with the 12/24. Also these are haret logs but maybe those new fixes after 12/24 are breaking more stuff.

Bottomline is that the media manger went ape shit. and start sending these out to the logs.

01-13 18:17:11.793 W/AudioTrack( 83): obtainBuffer timed out (is the CPU pegged?) 0x539b0 user=000012c0, server=00000000

Also, i'm having varied result with different csvs. :cussing:

Re: [TESTING] - NAND OMGB (1.2.3)
 

Ping me if you want some help testing/triaging. You know where to find me ;).

Re: [TESTING] - NAND OMGB (1.2.3)
 

+1 on this.

If you need help working through CSVs I'm more than happy to play the part of guinea pig via PM. :toothy7:

Re: [TESTING] - NAND OMGB (1.2.3)
 

+1, i would be interested too :-)

Re: [TESTING] - NAND OMGB (1.2.3)
 

The phone doesn't lock up. I can still navigate all menus using the keyboard. But the screen is unresponsive to touch, or it behaves oddly, for example, if I touch an icon on the homescreen to open an app it instead swipes to another homescreen or it ignores my touch.

I rebooted my phone and ran the adb command you suggested, results are attached.

I'd like to inform that I applied a task29 to my phone and started from scratch with the 12/24 build, the screen problem persists.

Thanks for your help!

Re: [TESTING] - NAND OMGB (1.2.3)
 

Try the task29. If there was a dirty install there is a chance that some stuff will be broken. Task29 will guarantee a clean delicious install.

since there is no kernel panic and we really cant tell from the logs, then best way to approach this is to see if there is a sequence that reproduces the issue. The only weird thing about your logs is that the phone cant sleep. It could be either wifi being on, data sync or anything that has a wake lock open. Anyways, keep trying dood. :dontknow:

Re: [TESTING] - NAND OMGB (1.2.3)
 

I had to revert back some code to begin from scratch. It's not the same issue as haret so this is going to take some research. I can't release this build without some sort of hack around this. It would set us back so for now i'm just going to continue butchering away. :confused2:

Edit: I saw that sometimes the missing ringinger has nothing to do with acoustic, but the error below.

"E/RingtoneManager( 202): Failed to open ringtone null"

This unfortunately is deep in android and i'm not sure if it was on the 12/24 builds. But it only happens randomly or mostly early on.

Re: [TESTING] - NAND OMGB (1.2.3)
 

Already did my good friend, the panic.txt I sent you is from my install of 12/24 after performing task29.

I'll keep fiddling and reporting back.

thanks for your help!