Re: 8139too termination

[Robert Kuebel]

> 
> 	tp->diediedie = 1;
> 	wmb();
> 	ret = kill_proc(...);
> 
> and test the flag in rtl8139_thread().
> 

i had something like that in mind.

> The tricky part is teaching the thread to ignore the
> spurious signals - the signal_pending() state needs to be
> cleared.  I think flush_signals() is the way to do this.
> See context_thread() for an example.
> 
>            spin_lock_irq(&curtask->sigmask_lock);
>            flush_signals(curtask);
>            recalc_sigpending(curtask);
>            spin_unlock_irq(&curtask->sigmask_lock);
> 
> The recalc_sigpending() here appears to be unnecessary...
> 

what about changing doing
	spin_lock_irq(&current->sigmask_lock);
	sigfillset(&current->blocked);	/* block all sig's */
	recalc_sigpending(current);
	spin_unlock_irq(&current->sigmask_lock);

instead of 

	spin_lock_irq(&current->sigmask_lock);
	sigemptyset(&current->blocked);  
	recalc_sigpending(current);
	spin_unlock_irq(&current->sigmask_lock);

and replacing the signal_pending() stuff in the loops of
rtl8139_thread() with checks for tp->diediedie?