[BUG] Process terminate race condition?

Michael Rueger michael at squeakland.org
Fri Jun 18 21:32:24 UTC 2004

Hi all,

just puzzling over some process termination issue on Windows.
I'm running a couple of processes in the background receiving and 
sending data through sockets.
Under certain circumstances those processes get terminated from a 
control process. I first terminate the process and then close the 
associated sockets.

I then occasionally get a DNU bottomContext in Proces>>terminate, 
because the suspended context is nil.
Could this be a racce condition of the process becoming active due to 
Semaphoe signals on the sockstes between the isActive test and the 
actual unwinding and termination logic?


More information about the Squeak-dev mailing list