tdm
Android Expert
I am not sure where diagnostic mode is failing but I thought all vs920 kernels supported it.
-
After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.
Root [KERNEL] Quattrimus Kernel
- Thread starter tdm
- Start date
-
- Tags
- lg spectrum
I am not sure where diagnostic mode is failing but I thought all vs920 kernels supported it.
IMUcarmen
Android Expert
after flashing on my first boot it hung on optimising apps for about ten minutes, after it finished that I started up wifi tether and it rebooted right away. After it booted back up I checked to see if maps and navigation worked and they both did fine. So I checked wifi tether again and it rebooted. So it's still got an issue somewhere.
tdm
Android Expert
Can you pull /proc/last_kmsg after the reboot?
tdm
Android Expert
Oh man, that's ugly. There doesn't seem to be anything usable in the log.
I just tried turning on tethering (both usb and wifi) on my nitro. It's running the same code compiled with the same compiler. No crash.
IMUcarmen
Android Expert
Oh man, that's ugly. There doesn't seem to be anything usable in the log.
I just tried turning on tethering (both usb and wifi) on my nitro. It's running the same code compiled with the same compiler. No crash.
Don't know man. It certainly jack's my phone up big time. I'm about to have to do a clean wipe and flash...again :-\
Maybe it's a matter of me accessing it then. On stock ICS-based ROMs, it was just a matter of using the USB>Internet Connection>Modem" option. Then 2 COM ports would show up in my device manager through which I could edit the phone.
On CM10/aokp, no COM ports show up for me to access. Am I missing something?
tdm
Android Expert
I don't know if the code to turn the modem connection on is even in AOSP (or CM or AOKP). It may be a proprietary vendor thing.
It's possible that this is tied to ECM mode in the USB driver (we switched to RNDIS). That can easily be changed, but it would break tethering. So either I'd need to spin a custom kernel for you or add code to switch on the fly at runtime.
Either way, it's probably a big time sink and I just don't have the motivation. It's easy enough to flash a stock ROM to access NVRAM then flash back when done.
IMUcarmen
Android Expert
Kernel is causing reboots, had 3 this afternoon, didn't have time to pull the file for ya, ended up flashing back to CM10 nightly (in the middle of a run). I'll reinstall the kennel tomorrow and get you the kmsg.
So it's not just you IMU!
Thanks! I was beginning to wonder if me phone was going south. I actually ended up flashing trigger's kernel and it's working great. VERY smooth
Kernel is causing reboots, had 3 this afternoon, didn't have time to pull the file for ya, ended up flashing back to CM10 nightly (in the middle of a run
So it's not just you IMU!
Trying to recreate the reboots, nothing's phasing it! Is there something that could accumulate/cause problems over time?
tdm
Android Expert
That is a possibility. If it's something that accumulates over time, it is likely a memory leak.
As always, a copy of /proc/last_kmsg immediately after reboot is the best way to diagnose.
Hehe, finally broke it
https://www.dropbox.com/s/ce41q9k4yjvhmtb/last_kmsg
Have been running it for several hours non-stop, playing music, browsing, Google play, GPS, maps, tapatalk, texting, etc. finally just rebooted while trying to reply to your last post.
Was also plugged in for awhile and did not get hot like it usually does, so that's a plus!
And another reboot, again while trying to post in tapatalk (weird coincidence?). Had noticed my screen was not turning off, either from inactivity or the power button and rebooted, which fixed it. Was working on posting that here when it rebooted again.
https://www.dropbox.com/s/al3q0dlbzk30v0o/last_kmsg.zip
Let me know if you need more, going back to nightly kernel for now
https://www.dropbox.com/s/al3q0dlbzk30v0o/last_kmsg.zip
Let me know if you need more, going back to nightly kernel for now
tdm
Android Expert
I've been experiencing some instability on my nitro the past couple days after disabling DS Battery Saver. Some of the symptoms are:
* BSOD caused by kgsl running out of file descriptors.
* Reboot caused by failure to set cpu1 frequency from cpu0.
* Excessive battery drain.
I'm instrumenting the kernel to look for the causes of the BSOD and battery drain.
The reboot caused by failure to set cpu1 frequency was fixed once. I took the code out because the latest CAF code removed the offending check. But apparently it's still needed.
* BSOD caused by kgsl running out of file descriptors.
* Reboot caused by failure to set cpu1 frequency from cpu0.
* Excessive battery drain.
I'm instrumenting the kernel to look for the causes of the BSOD and battery drain.
The reboot caused by failure to set cpu1 frequency was fixed once. I took the code out because the latest CAF code removed the offending check. But apparently it's still needed.
tdm
Android Expert
Well, I think it needs more work. Here's what I see testing on my P930:
The fix does seem to increase time in deep sleep (according to CPU spy). But the battery still seems to drain just as fast -- roughly 10%/hour. So perhaps the wifi chip isn't actually going into low power mode.
Additionally, with this code in place, wifi frequently doesn't restart after sleep. I need to go into settings/wireless and the act of scanning for networks seems to restart it.
Is anyone still interested in testing this? Maybe your results will be different, or maybe the VS920 wifi behaves better.