Re: [patch 2/3] scheduler: add full memory barriers upon taskswitch at runqueue lock/unlock

[Steven Rostedt]

On Mon, 2010-02-01 at 17:39 -0500, Steven Rostedt wrote:

> 
> 	   CPU 0		   CPU 1
> 	-----------		-----------
> 				< same thread >
> 				clear_bit(mm_cpumask)
> 
> 	sys_membarrier();
> 	<miss sending CPU 1 mb()>
> 	return back to user-space
> 
> 	modify data
> 
> 				< CPU 0 data is read > <<- corruption
> 
> 				schedule()
> 
> 

> (posted in a previous email)
> 
> sys_membarrier(void) {
> 
> 	again:
> 		tmp_mask = mm_cpumask(current->mm);
> 		smp_mb();
> 		rcu_read_lock(); /* ensures validity of cpu_curr(cpu) tasks */
> 		for_each_cpu(cpu, tmp_mask) {
> 			spin_lock_irq(&cpu_rq(cpu)->lock);
> 			ret = current->mm == cpu_curr(cpu)->mm;
> 			spin_unlock_irq(&cpu_rq(cpu)->lock);
> 			if (ret)
> 			smp_call_function_single(cpu, membarrier_ipi, NULL, 1);
> 		}
> 		rcu_read_unlock();
> 		smp_mb();
> 		if (tmp_mask != mm_cpumask(current->mm)) {
> 			/* do check for signals here */
> 			goto again;
> 		}

Doh, this misses the above case if a mb() is not before the caller.
Nevermind, it looks like the only other case is to iterate over all
CPUs.

-- Steve


--
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/