[DaGuy]

When I try to launch an instance of NSTask from a detached thread, my test app quits and sends the message below to the console.


TomcatWrapper has exited due to signal 5 (SIGTRAP).


I think is because NSTask initiates and runs on its own thread and since I'm launching NSTask from within a detached thread then things get kinda loopy somewhere.

How about it guys, any ideas?

Thanks!

[Chuckit]

Are you certain that this will work even in the main thread? Because really, I can't see any obvious way that being in a separate thread would hurt NSTask like you seem to be saying. As the name implies, the task executes not just in another thread, but as an entirely separate task. It doesn't share memory with the calling program, so it can't really conflict with it.

[DaGuy]

I hear you. I guess this error is due to some other reason. Let me poke around some more.

So launching a thread from within another thread is fine. Not so different (at least in principle) from the main thread detaching threads. What about recursively detaching threads? i.e., one thread, launches a thread and that thread launches another thread and son on... Say N times. N finite :-)

I know the complexity would increase greatly and I would bet that their is rarely, if any, need for something like this but I just want to understand this topic a little better.

Is it equivalent from a resource allocation perspective (memory, CPU) to having the main thread detach N threads?


Thanks!