• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Root [KERNEL] Quattrimus Kernel

Haven't checked it either but I'm down 5% in like 3 hours. Used to be like 15 -20 in that time.

Sent from my VS920 4G using Tapatalk 2
 
Almost 2 hours off charger, only 1% drop! Mostly standby but still a huge improvement, and that was with WiFi on!

e3a5y3y2.jpg
 
Completely random reboots with the google gcc and linaro on CM10.1 M2. The GCC one rebooted while the phone was screen off and no power or activities connected. My phone has always been finicky, but this one was just weird. It seems to have calmed down since I've left it off the charger. Log attached for when that one happened when the screen was off.

http://androidfiles.massivefilehost.com/misc/last_kmsg.txt
 
Neph81 said:
Completely random reboots with the google gcc and linaro on CM10.1 M2. The GCC one rebooted while the phone was screen off and no power or activities connected. My phone has always been finicky, but this one was just weird. It seems to have calmed down since I've left it off the charger. Log attached for when that one happened when the screen was off.

http://androidfiles.massivefilehost.com/misc/last_kmsg.txt
Click to expand...

Looks like a null ptr bug in the scheduler. I'll try to track it down later.
 
tdm said:
That's ugly. Don't know if I can track it down. :(
Click to expand...

I know. As far as I could tell, it was some kind of kernel panic but that's all I got out of it.

Edit: I found a couple of lines that you may or may not have seen.

Code: 
[ 4390.904846] Kernel panic - not syncing: Fatal exception...
a bunch of stuff here (ie. exception stacks, etc.)
[ 4391.205017] subsystem-fatal-8x60: Q6 NMI was sent.
...
[ 4397.245239] Unable to handle kernel NULL pointer dereference at virtual address 00000024

It goes through the same thing a few more times before the end.
 
I also had a couple reboots on AOKP. Screen was off, phone unplugged, just sitting on the desk. Have been running it for 2 days without a problem, just started tonight.

WiFi was on, linaro build, will try the other tomorrow if it keeps rebooting.

http://db.tt/ucxNXjgk
 
Another log. This one had a very specific error message though.
Code: 
[  378.979553] ------------[ cut here ]------------
[B][  378.983245] kernel BUG at kernel/workqueue.c:523![/B]
[  378.988037] Internal error: Oops - BUG: 0 [#1] PREEMPT SMP
[  378.993408] Modules linked in:
[  378.996459] CPU: 0    Tainted: G        W    (3.0.31-quattrimus+ #40)
[  379.002990] PC is at __cancel_work_timer+0x14c/0x17c
[  379.007812] LR is at wait_on_work+0xb4/0x1c0
[  379.012207] pc : c01d9338    lr : c01d8284    psr: 40000013
[  379.012207] sp : da645cf8  ip : b319b31a  fp : da645db0
[  379.022949] r10: 00000000  r9 : 00000000  r8 : 00000001
[  379.028045] r7 : d9d499b0  r6 : d9d499ec  r5 : da644000  r4 : d9d499a0
[  379.034667] r3 : 00000300  r2 : b31ab31a  r1 : 00000000  r0 : 00000000
[  379.041076] Flags: nZcv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
[  379.048461] Control: 10c5787d  Table: 507bc06a  DAC: 00000015
[  379.054077]
http://androidfiles.massivefilehost.com/misc/last_kmsg3.txt
 
Sorry I don't see anything in any of those logs that looks like a reproducible problem.

I'm guessing that either there's a very rare and subtle memory corruption issue or the hardware is unstable.
 
tdm said:
Sorry I don't see anything in any of those logs that looks like a reproducible problem.

I'm guessing that either there's a very rare and subtle memory corruption issue or the hardware is unstable.
Click to expand...
I'm going with unstable hardware. Like I said, mine has always been finicky. It's ok. CM10.1 M2 runs great with the exception of bad battery life. I tried some of the newer ones and ran into more and more issues. They are all most likely tied to my phone and not the software.
 
Hey Ortrigger, quick question. Are you using HD Widgets with TDMs kernel? I'm trying to track things down and this seems to be the common pattern between what reboots and what doesn't. It's still only a hunch at the moment.
 
Neph81 said:
Hey Ortrigger, quick question. Are you using HD Widgets with TDMs kernel? I'm trying to track things down and this seems to be the common pattern between what reboots and what doesn't. It's still only a hunch at the moment.
Click to expand...

I have never used HD widgets. (Not that I don't want to, I'm just too cheap) Honestly, I'm not sure what could have caused my reboot. :(
 
A user on XDA sent me a "gimpy" p930 -- one that seems to reboot at the slightest provocation on ICS/JB but is reportedly rock solid on GB. One of the difficulties I've been having is that most of the time, the phone reboots into a weird state and needs to reboot again to do a full reset. This wipes out last_kmsg.

I've made a change to the ram console code that starts appending where it left off last time instead of starting clean. This means last_kmsg has the ability to capture several successive reboots (as long as recovery isn't invoked -- the recovery kernel doesn't have that code).

Let's see what happens over the next couple of days .. hopefully I can track down more of these issues.
 
tdm said:
A user on XDA sent me a "gimpy" p930 -- one that seems to reboot at the slightest provocation on ICS/JB but is reportedly rock solid on GB. One of the difficulties I've been having is that most of the time, the phone reboots into a weird state and needs to reboot again to do a full reset. .
Click to expand...

That sounds similar to the soft rebooting than hounded us on the gb roms. When it would finally boot back up it wouldn't replay the notification bar, so a full power off and back on was required.
 
I just installed a kernel with the CPU voltage table and 8901_s0/8901_s1 regulator voltages from GB.

Stability test is running now. Let's see how long it lasts... ;)
 
Neph81 said:
I'm going with unstable hardware. Like I said, mine has always been finicky. It's ok. CM10.1 M2 runs great with the exception of bad battery life. I tried some of the newer ones and ran into more and more issues. They are all most likely tied to my phone and not the software.
Click to expand...
FWIW, the latest CM10.1 nightlies are running real smooth for me and much improved battery life over M2 thanks to TDMs WiFi commits
 
  • Like
Reactions: tdm
You must log in or register to reply here.
Back
Top Bottom