On Thursday 20 March 2003 09:01, Thiago Macieira wrote:
> Ralf Nolden wrote:
> >Anyone who has a clue what's going on here ?
>
> Actually, I think I do now. I had seen someone complain about this error
> before, but my inspection to the code turned out that the situation was
> close to impossible.
>
> >kdecore (KSocket): ERROR: KExtendedSocket::connectionEvent() called but no
> >data available!
>
> This provided an insight:
> >#7  0x415aae87 in KMCupsManager::completePrinterShort ()
> >   from /usr/lib/kde3/kdeprint_cups.so
>
> That function creates a socket with asynchronous connection and kills it
> off after 2 seconds. I don't know where the fault is, but I guess some
> signals are being emitted and then received by the deleted object. That
> could explain the error above.

   From what I can tell, KExtendedSocket async calls continue to run after the 
destructor is called.  If this is indeed the case, please apply a patch to 
KExtendedSocket to cancel asynchronous actions in the destructor.  I put a 
hack in the cups code for now.

-- 

George Staikos