Re: [PATCH RFC tip/core/rcu 12/15] lib/assoc_array: Remove smp_read_barrier_depends()

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Wed, Oct 11, 2017 at 05:07:05PM +0100, David Howells wrote:
> Paul E. McKenney <paulmck@linux.vnet.ibm.com> wrote:
> 
> > It does not.  In most cases, the barriered version would be
> > smp_store_release().
> 
> Ummm... Is that good enough?  Is:
> 
> 	WRITE_ONCE(x, 1);
> 	WRITE_ONCE(x, 2);
> 
> equivalent to:
> 
> 	smp_store_release(x, 1);
> 	smp_store_release(x, 2);
> 
> if CONFIG_SMP=n?

	smp_store_release(&x, 1);
	smp_store_release(&x, 2);

But yes, give or take that smp_store_release() potentially disables
more compiler optimizations than does WRITE_ONCE().

> (Consider what happens if an interrupt messes with x).

OK, I will bite...  What is your scenario in which an interrupt
gives different results for CONFIG_SMP=n?  The barriers

> If it is good enough, should we be using smp_load_acquire() rather than
> READ_ONCE()?

On x86, that might be OK, give or take that smp_load_acquire() potentially
disables more optimizations than does READ_ONCE().  But on ARM, PowerPC,
MIPS, and so on, smp_load_acquire() emits a memory-barrier instruction
and READ_ONCE() does not.

							Thanx, Paul