[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-bugs-dist
Subject:    Bug#1111: signal bug with session manager
From:       Navindra Umanee <navindra () cs ! mcgill ! ca>
Date:       1999-04-01 2:08:29
[Download RAW message or body]

Montreal Wed Mar 31 21:06:11 1999

Harri Porten <porten@tu-harburg.de> wrote:
> I believe to have tracked down the problem. I don't have a proper fix
> yet but please try the following patch to see if the problem
> dissappears. I hope that you *do* compile from the sources. Otherwise
> we'll have to find someone who has and could do this experiment.
> 
> Harri.
> 
> --- kdelibs/kdecore/kprocctrl.cpp.bak	Thu Apr  1 01:37:30 1999
> +++ kdelibs/kdecore/kprocctrl.cpp	Thu Apr  1 02:54:46 1999
> @@ -76,7 +76,7 @@
>    sigemptyset(&(act.sa_mask));
>    sigaddset(&(act.sa_mask), SIGPIPE);
>    act.sa_flags = 0;
> -  sigaction( SIGPIPE, &act, 0L);
> +  //  sigaction( SIGPIPE, &act, 0L);
>  }
>  
>  void KProcessController::theSigCHLDHandler(int )

Hi, I recompiled libkdecore.* with the above change but it doesn't
seem to have made a difference.  Hmmm.  

Andreas Pour wrote:
[...]
> Even if a parent were to do any signal catching, it doesn't apply to
> children (how could it? the child does not have a copy of the
> signal-handling function).  So, if there is a bug, it seems to be like
> it must not stem from KDE or Gnome bu t from the kernel (shudder,
> shudder!).  I am running kernel 2.0.36 -- which versio n are you
> running?  (I am also suspisicous that this is a kernel bug b/c of the
> value reported by you -- 0x1 is not a legitimate function address!).

I'm using kernel 2.2.2.  I'll let you guys know if 2.0.36 makes a
difference for me.

I'll try to find more details on what exactly is causing the problem
later today/early tomorrow.

Thanks,
Navin.

[prev in list] [next in list] [prev in thread] [next in thread]