Desktop Username password

[jigar-f]

No description provided.

[jigar-f]

@atavism You start reviewing this PR.

[jigar-f]

@atavism thanks for approving, Did you able to find out about the crash?

[atavism]

@jigar-f I haven't been able to figure out what's causing the app to crash. I actually haven't experienced the same issue in awhile testing the latest changes. So I think we can just wait to see if it resurfaces

[atavism]

I discovered an issue where it looks like the timer in the VPNChangeNotifier isn't being stopped when it reaches the max number of ticks..

Jul 16 16:34:43.809 - 0m20s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized
Jul 16 16:34:44.810 - 0m21s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized
Jul 16 16:34:45.809 - 0m22s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized
Jul 16 16:34:46.810 - 0m23s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized
Jul 16 16:34:47.810 - 0m24s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized
Jul 16 16:34:48.810 - 0m25s DEBUG app.settings: settings.go:670 Could not get value for userPro
flutter: startup status proxy true config true success false
flutter: flashlight fail initialized

[jigar-f]

t

This is not an issue, We run the timer till the flashlight is initialized.

[jigar-f]

signal 16 received but handler not on signal stack
mp.gsignal stack [0x14000084000 0x1400008c000], mp.g0 stack [0x16ba88000 0x16bc8b000], sp=0x14000857678
fatal error: non-Go code set up signal handler without SA_ONSTACK flag

@atavism I am still getting crash, Any idea whats going on?

[atavism]

@atavism I am still getting crash, Any idea whats going on?

No, I am still having trouble getting the app to crash again myself. What are you doing just before the crash happens -- switching tabs, turning the VPN on, etc? It might be helpful to add additional logging to try to pinpoint where the crash is happening

[jigar-f]

It crashes for me at random places, I have already added some logging to track it down.

[jigar-f]

@atavism This might be the issue

[atavism]

I don't think this code to handle signals is the problem area since it hasn't caused the app to crash previously. It's likely some code that changed recently..

For instance, I wonder if we maybe introduced a bug somewhere fetching and storing user data.. the fact we see this error consistently in the logs points to an issue where a thread may be attempting to re-lock a mutex without first unlocking it or something:

goroutine 283 gp=0x14000ba3a40 m=nil [sync.Mutex.Lock]:
runtime.gopark(0x14000622e48?, 0x2?, 0xa?, 0x0?, 0x14000622e1c?)
	runtime/proc.go:402 +0xc8 fp=0x14000622cd0 sp=0x14000622cb0 pc=0x15fa0f6b8
runtime.selectgo(0x14000622e48, 0x14000622e18, 0x14000622e08?, 0x0, 0x15fa19c70?, 0x1)
	runtime/select.go:327 +0x614 fp=0x14000622de0 sp=0x14000622cd0 pc=0x15fa22c34

have you tried commenting out the new code or calls that are being made to see if the app still crashes with those disabled?

[atavism]

Have we checked if the app crashes on main? Let me do that now..

[atavism]

Ok, this is interesting: the app is crashing for me on main, but it no longer does if I go back to commit 4c7a68a. Looks like this started happening with 042186e?

So I'm pretty sure the crash is unrelated to the changes here. Given that, I guess we could re-enable the signal handling and continue using proxied.ParallelForIdempotent to get this merged?

[jigar-f]

Sure, I have re-enbale signal and added proxied.ParallelForIdempotent

[jigar-f]

I am fine with merging.

[jigar-f]

To me it looks like this might be crash, Since this actually throws painc.

runtime stack:
runtime.throw({0x15c9fc20e?, 0x0?})
	runtime/panic.go:1023 +0x40 fp=0x14000051490 sp=0x14000051460 pc=0x15bce13a0
runtime.sigNotOnStack(0x10, 0x14000051538, 0x14000080008)
	runtime/signal_unix.go:1065 +0x118 fp=0x140000514c0 sp=0x14000051490 pc=0x15bcfbe08
runtime.adjustSignalStack(0x10, 0x14000080008, 0x14000051568)
	runtime/signal_unix.go:592 +0x25c fp=0x14000051530 sp=0x140000514c0 pc=0x15bcfac6c
runtime.sigtrampgo(0x10, 0x140000516d0, 0x14000051738)
	runtime/signal_unix.go:480 +0x8c fp=0x140000515b0 sp=0x14000051530 pc=0x15bcfa7ac
runtime.sigtrampgo(0x10, 0x140000516d0, 0x14000051738)
	<autogenerated>:1 +0x1c fp=0x140000515e0 sp=0x140000515b0 pc=0x15bd20adc
runtime.sigtramp()
	runtime/sys_darwin_arm64.s:227 +0x4c fp=0x140000516a0 sp=0x140000515e0 pc=0x15bd1f5ac

[jigar-f]

Here is ticket-:https://github.com/getlantern/engineering/issues/1504

[atavism]

Thanks, @jigar-f! I will take a closer look at this over the weekend

[jigar-f]

@atavism All changes are added, SHould I go ahead and merge this?

[atavism]

@atavism All changes are added, SHould I go ahead and merge this?

Thanks for making those changes! Sounds great! Merging..