Re: workqueue - process context

[Roland Dreier]

    Rene> I have no idea about the module refcounting stuff. Is there
    Rene> a chance that create_workqueue() could increase a reference
    Rene> somewhere so that the module wouldn't be allowed to unload
    Rene> untill after a destroy_workqueue()?

There's no point to doing this, since it's adding complexity to try
and avoid a very obvious and easy to find bug.  Other types of
resource leaks are harder to find, but a module not destroying a
workqueue is going to be trivial to spot and fix.

(There are technical issues as well -- if create_workqueue()
increments the module reference count, then you would never be able to
unload the module if the destroy_workqueue() was in the module_exit
function, because you can never even start to unload a module with a
non-zero ref count).

 - R.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/