Using ZNC 1.7.1, with znc launched in the default way (so forking).
Enabling the debug mode works, though the stdout file discriptor cannot be found
On disable ZNC crashes with no logs in syslog
steps to reproduce (assuming default prefix for modules):
[12:46:11] -*status- *** An administrator has just turned Debug Mode on. It was enabled by A_D.
[12:46:11] -*status- *** Messages, credentials, and other sensitive data may become exposed to the host during this period.
[12:49:04] -*status- *** An administrator has just turned Debug Mode off. It was enabled by A_D.
[12:49:04] * Disconnected (Remote host closed socket).
Relevant logs from a discussion in #znc with @Zarthus:
<A_D> Zarthus: methinks I found a bug with your admindebug plugin
<Zarthus> if it has to do with the fact you just disconnected i'm not sure if that's a good thing :P
<A_D> yeahhhhhhhhhhh
<A_D> so I turned debug on went looking for the logs, which I couldnt find because it had forked and bad things happen to stdout when that happens
<A_D> and then I turned it off
<A_D> and then znc turned off
<Zarthus> so you ran it with debug on at first, or debug off at first?
<A_D> debug off at first, standard launch
<A_D> sec
<-- spcBNC has quit (Ping timeout: 244 seconds)
<Zarthus> the way i tested it without forking, so that kind of makes sense (I ran debug on and then played around with it)
<A_D> relavent logs: https://paste.snoonet.org/paste/3dfmp
<A_D> and the service file I use
<Zarthus> interesting
<A_D> I should add a pidfile to that >.>
<A_D> no loud noises in syslog
<A_D> what was my disconnect message here?
<A_D> connection closed by remote host?
<Zarthus> (12:49:04) * A_D (A_D@unaffiliated/ad/x-3954327) Quit (Remote host closed the connection)
<A_D> okay then it hard fell over
<A_D> well, not hard. it was nice enough to close sockets
<Zarthus> lol
<A_D> but not nice enough to run quit messages
<Zarthus> I'm not sure what the issue would be beyond "stdout is not there when you fork"
<Zarthus> unlike sane languages, debugging with c++ is a pain in the arse :P
<A_D> stdout should technically still be there?
<Zarthus> yeah, but i don't know what happens to it
<A_D> forked processes get their parent's file descriptors
<Zarthus> can you file an issue on github?
<Zarthus> i'm not sure when i'll get the time and motivation to actually debug it with gdb
<A_D> I will, just wanted to ask about it here first just to be sure that it wasnt already known
<Zarthus> nop, i did enough playtesting that it hasn't fallen over for me, but all of it was without forking a process
<Zarthus> to my memory, at least
<psychon> relevant: https://github.com/znc/znc/blob/master/src/main.cpp#L502-L508
<A_D> ah, so thats what happens to stdout
<Zarthus> A_D: fwiw, the workaround would be to run znc with --foreground
Using ZNC 1.7.1, with znc launched in the default way (so forking).
Enabling the debug mode works, though the stdout file discriptor [cannot be found](https://github.com/znc/znc/blob/master/src/main.cpp#L502-L508)
On disable ZNC crashes with no logs in syslog
steps to reproduce (assuming default prefix for modules):
```
/znc loadmod adminlog
/msg *admindebug enable
/msg *admindebug disable
```
relevant logs from my crash:
```
[12:46:11] -*status- *** An administrator has just turned Debug Mode on. It was enabled by A_D.
[12:46:11] -*status- *** Messages, credentials, and other sensitive data may become exposed to the host during this period.
[12:49:04] -*status- *** An administrator has just turned Debug Mode off. It was enabled by A_D.
[12:49:04] * Disconnected (Remote host closed socket).
```
Relevant logs from a discussion in #znc with @Zarthus:
```
<A_D> Zarthus: methinks I found a bug with your admindebug plugin
<Zarthus> if it has to do with the fact you just disconnected i'm not sure if that's a good thing :P
<A_D> yeahhhhhhhhhhh
<A_D> so I turned debug on went looking for the logs, which I couldnt find because it had forked and bad things happen to stdout when that happens
<A_D> and then I turned it off
<A_D> and then znc turned off
<Zarthus> so you ran it with debug on at first, or debug off at first?
<A_D> debug off at first, standard launch
<A_D> sec
<-- spcBNC has quit (Ping timeout: 244 seconds)
<Zarthus> the way i tested it without forking, so that kind of makes sense (I ran debug on and then played around with it)
<A_D> relavent logs: https://paste.snoonet.org/paste/3dfmp
<A_D> and the service file I use
<Zarthus> interesting
<A_D> I should add a pidfile to that >.>
<A_D> no loud noises in syslog
<A_D> what was my disconnect message here?
<A_D> connection closed by remote host?
<Zarthus> (12:49:04) * A_D (A_D@unaffiliated/ad/x-3954327) Quit (Remote host closed the connection)
<A_D> okay then it hard fell over
<A_D> well, not hard. it was nice enough to close sockets
<Zarthus> lol
<A_D> but not nice enough to run quit messages
<Zarthus> I'm not sure what the issue would be beyond "stdout is not there when you fork"
<Zarthus> unlike sane languages, debugging with c++ is a pain in the arse :P
<A_D> stdout should technically still be there?
<Zarthus> yeah, but i don't know what happens to it
<A_D> forked processes get their parent's file descriptors
<Zarthus> can you file an issue on github?
<Zarthus> i'm not sure when i'll get the time and motivation to actually debug it with gdb
<A_D> I will, just wanted to ask about it here first just to be sure that it wasnt already known
<Zarthus> nop, i did enough playtesting that it hasn't fallen over for me, but all of it was without forking a process
<Zarthus> to my memory, at least
<psychon> relevant: https://github.com/znc/znc/blob/master/src/main.cpp#L502-L508
<A_D> ah, so thats what happens to stdout
<Zarthus> A_D: fwiw, the workaround would be to run znc with --foreground
```
<!-- github-sync-id:343319386 -->
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Using ZNC 1.7.1, with znc launched in the default way (so forking).
Enabling the debug mode works, though the stdout file discriptor cannot be found
On disable ZNC crashes with no logs in syslog
steps to reproduce (assuming default prefix for modules):
relevant logs from my crash:
Relevant logs from a discussion in #znc with @Zarthus: