Random Boot loop 1.99.49

[an0n981]

After several hours of 'good' usage with 1.99.49, I have started having random boot loops. This reappears randomly anywhere between 5 -30 of booting the phone. Any idea what could be causing this?

What I believe to be the important part of the log:
https://gist.github.com/an0n981/81e5f924ceeb9c78efce

[M66B]

The is probably because if restricting 'connect' and a process that cannot handle the exception that is thrown. An IOException is a normal possible response from 'connect', so this is a combination of a restriction and a bug.

E/AndroidRuntime( 8624): *** FATAL EXCEPTION IN SYSTEM PROCESS: CaptivePortalTracker
E/AndroidRuntime( 8624): java.lang.ClassCastException: java.io.IOException cannot be cast to boolean
E/AndroidRuntime( 8624):        at libcore.io.IoBridge.connect(Native Method)

I have no idea what 'CaptivePortalTracker' is.

[M66B]

Wait, maybe there is a connect variant that expects a boolean.
I will make a test version.

[ckuethe]

I'd guess that it's the process that checks to see is there is a captive
portal on a wireless network, and if so sends the "Wireless Sign-in
Required" notification... but that's just a guess.

On Thu, Feb 20, 2014 at 10:36 AM, Marcel Bokhorst
[email protected]:

The is probably because if restricting 'connect' and a process that cannot
handle the exception that is thrown. An IOException is a normal possible
response from 'connect', so this is a combination of a restriction and a
bug.

E/AndroidRuntime( 8624): *** FATAL EXCEPTION IN SYSTEM PROCESS: CaptivePortalTracker
E/AndroidRuntime( 8624): java.lang.ClassCastException: java.io.IOException cannot be cast to boolean
E/AndroidRuntime( 8624): at libcore.io.IoBridge.connect(Native Method)

I have no idea what 'CaptivePortalTracker' is.

Reply to this email directly or view it on GitHubhttps://github.com//issues/1398#issuecomment-35653798
.

GDB has a 'break' feature; why doesn't it have 'fix' too?

[M66B]

Can you please try this version: http://d-h.st/KzQ

[an0n981]

It may take a little while to let you know if it worked, as I don't know exactly how to reproduce the error

[an0n981]

didn't take that long after all

https://gist.github.com/an0n981/0d58b37e26cdadd259a6

[M66B]

Then you have to find the offending restriction. We know it is 'connect' and probably a system component.

It is not unusual that XPrivacy triggers bugs in other packages, but fortunately it does not occur frequently.

[an0n981]

ok, thanks for looking into it

[an0n981]

Just in case anyone stumbles upon this with the same error. I am pretty confident the issue was caused by restricting Internet: connect for either system ui or com.android.keyguard

[M66B]

It appears to be a bug after all ...

[an0n981]

Ok. But I was unable to reproduce it. I was gonna wait til I got to work to edit my post as it was not system ui or keyguard. After unrestricting 1000/internet/connect it was fine all night. Also after reading up on CaptivePortalTracker it makes sense that the issue wasn't visible at first, since I refuse to connect to my work WiFi. But if I can restrict connect for 1000 again I would be much happier

[an0n981]

Boot loop still active on 1.99.50 with 1000/Internet:connect restricred. Also Internet:connect and ID:/proc still require a reboot

[an0n981]

Edit: boot loop only active when wifi connected

[an0n981]

I was just getting ready to write a novel describing from an IT-Administrator point of view about why I am certain this is a bug that was introduced with .49. But it looks as if you found the problem. As I am however not fluent in java (main reason being I have no patience to find one wrong semi-colon in 100000 lines of code) I will just assume that you should not have missed that.