Re: [PATCH tip/core/rcu 5/6] Documentation/memory-barriers.txt: Need barriers() for some control dependencies

[Paul E. McKenney]

On Mon, Feb 17, 2014 at 07:29:55PM -0800, Josh Triplett wrote:
> On Mon, Feb 17, 2014 at 05:21:37PM -0800, Paul E. McKenney wrote:
> > On Mon, Feb 17, 2014 at 04:45:11PM -0800, Josh Triplett wrote:
> > > On Mon, Feb 17, 2014 at 04:17:40PM -0800, Paul E. McKenney wrote:
> > > > On Mon, Feb 17, 2014 at 04:02:47PM -0800, Josh Triplett wrote:
> > > > > On Mon, Feb 17, 2014 at 02:58:16PM -0800, Paul E. McKenney wrote:
> > > > > > On Mon, Feb 17, 2014 at 01:46:06PM -0800, Josh Triplett wrote:
> > > > > > > On Mon, Feb 17, 2014 at 01:26:52PM -0800, Paul E. McKenney wrote:
> > > > > > > > From: "Paul E. McKenney" <paulmck@xxxxxxxxxxxxxxxxxx>
> > > > > > > > 
> > > > > > > > Current compilers can "speculate" stores in the case where both legs
> > > > > > > > of the "if" statement start with identical stores.  Because the stores
> > > > > > > > are identical, the compiler knows that the store will unconditionally
> > > > > > > > execute regardless of the "if" condition, and so the compiler is within
> > > > > > > > its rights to hoist the store to precede the condition.  Such hoisting
> > > > > > > > destroys the control-dependency ordering.  This ordering can be restored
> > > > > > > > by placing a barrier() at the beginning of each leg of the "if" statement.
> > > > > > > > This commit adds this requirement to the control-dependencies section.
> > > > > > > > 
> > > > > > > > Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
> > > > > > > 
> > > > > > > This is starting to become a rather unreasonable level of fighting the
> > > > > > > compiler.  ("Threads cannot be implemented as a library" indeed.)  This
> > > > > > > doesn't seem like a reasonable thing to require users to do.  Is there
> > > > > > > really no other way to cope with this particular bit of "help" from the
> > > > > > > compiler?
> > > > > > 
> > > > > > Well, we could use smp_mb() instead of barrier(), but that was the
> > > > > > sort of thing that Peter Zijlstra was trying to avoid.
> > > > > 
> > > > > Yeah, that's not an improvement.  The goal would be to make the code no
> > > > > more complex than it already needs to be with ACCESS_ONCE; changing
> > > > > "barrier()" to something else doesn't help (quite apart from smp_mb()
> > > > > being suboptimal).
> > > > > 
> > > > > > That said, I do sympathize completely with your position here -- it is
> > > > > > just that it is better to have our compiler-fights documented that
> > > > > > not, right?
> > > > > 
> > > > > Sure, better to document them, but better still to not have them.  Is
> > > > > there some other way we could avoid this one entirely?
> > > > 
> > > > We could try change the standard so as to outlaw pulling common code from
> > > > both legs of an "if" statement, but that will be a serious uphill battle.
> > > 
> > > And insufficient given widespread use of existing compilers.
> > 
> > Fair point...
> > 
> > > > Or perhaps do something to warn the developer about the possibility of
> > > > this happening.
> > > > 
> > > > Other thoughts?
> > > 
> > > Might be worth bringing this up with the GCC folks to find out if
> > > there's something obvious we're missing.  (For non-obvious values of
> > > "obvious".)
> > 
> > Non-obvious values of "obvious" -- I have no idea what that means, but
> > it does have a nice counter-intuitive sound to it, doesn't it?  ;-)
> 
> I'm hoping for something that we'll consider obvious in hindsight.
> 
> > This conversation has started, albeit with much more noise and smoke
> > than signal or light.
> 
> Whereabouts is that conversation taking place?

Here you go: https://lkml.org/lkml/2014/2/6/193

Not too bad, only about 80 messages thus far.  Spirited at times, though.  ;-)

							Thanx, Paul

--
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/