Re: Plain accesses and data races in the Linux Kernel Memory Model

From: Andrea Parri <andrea.parri@amarulasolutions.com>
To: Alan Stern <stern@rowland.harvard.edu>
Cc: LKMM Maintainers -- Akira Yokosawa <akiyks@gmail.com>,
	Boqun Feng <boqun.feng@gmail.com>,
	Daniel Lustig <dlustig@nvidia.com>,
	David Howells <dhowells@redhat.com>,
	Jade Alglave <j.alglave@ucl.ac.uk>,
	Luc Maranget <luc.maranget@inria.fr>,
	Nicholas Piggin <npiggin@gmail.com>,
	"Paul E. McKenney" <paulmck@linux.ibm.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Will Deacon <will.deacon@arm.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	Nick Desaulniers <ndesaulniers@google.com>,
	linux-kernel@vger.kernel.org
Subject: Re: Plain accesses and data races in the Linux Kernel Memory Model
Date: Fri, 18 Jan 2019 16:56:38 +0100	[thread overview]
Message-ID: <20190118155638.GA24442@andrea> (raw)
In-Reply-To: <Pine.LNX.4.44L0.1901181005110.1425-100000@iolanthe.rowland.org>

On Fri, Jan 18, 2019 at 10:10:22AM -0500, Alan Stern wrote:
> On Thu, 17 Jan 2019, Andrea Parri wrote:
> 
> > > Can the compiler (maybe, it does?) transform, at the C or at the "asm"
> > > level, LB1's P0 in LB2's P0 (LB1 and LB2 are reported below)?
> > > 
> > > C LB1
> > > 
> > > {
> > > 	int *x = &a;
> > > }
> > > 
> > > P0(int **x, int *y)
> > > {
> > > 	int *r0;
> > > 
> > > 	r0 = rcu_dereference(*x);
> > > 	*r0 = 0;
> > > 	smp_wmb();
> > > 	WRITE_ONCE(*y, 1);
> > > }
> > > 
> > > P1(int **x, int *y, int *b)
> > > {
> > > 	int r0;
> > > 
> > > 	r0 = READ_ONCE(*y);
> > > 	rcu_assign_pointer(*x, b);
> > > }
> > > 
> > > exists (0:r0=b /\ 1:r0=1)
> > > 
> > > 
> > > C LB2
> > > 
> > > {
> > > 	int *x = &a;
> > > }
> > > 
> > > P0(int **x, int *y)
> > > {
> > > 	int *r0;
> > > 
> > > 	r0 = rcu_dereference(*x);
> > > 	if (*r0)
> > > 		*r0 = 0;
> > > 	smp_wmb();
> > > 	WRITE_ONCE(*y, 1);
> > > }
> > > 
> > > P1(int **x, int *y, int *b)
> > > {
> > > 	int r0;
> > > 
> > > 	r0 = READ_ONCE(*y);
> > > 	rcu_assign_pointer(*x, b);
> > > }
> > > 
> > > exists (0:r0=b /\ 1:r0=1)
> > > 
> > > LB1 and LB2 are data-race free, according to the patch; LB1's "exists"
> > > clause is not satisfiable, while LB2's "exists" clause is satisfiable.
> 
> A relatively simple solution to this problem would be to say that 
> smp_wmb doesn't order plain writes.

It seems so; I don't have other solutions to suggest ATM.  (But, TBH,
I'm still in the process of reviewing/testing these changes... )

And yes, this is a pain! : I don't have the exact statistics, but I'm
willing to believe that removing this order will take us back ~99% of
the current (~500!) uses of smp_wmb() ;-/

Oh, well, maybe we'll find a better solution one day: after all, that
one doesn't seem worse than what the current LKMM has to say! ;-)

> 
> I think the rest of the memory model would then be okay.  However, I am
> open to arguments that this approach is too complex and we should
> insist on the same kind of strict ordering for race avoidance that the
> C11 standard uses (i.e., conflicting accesses separated by full memory
> barriers or release & acquire barriers or locking).

Indeed;  maybe, we've just found another reason to obsolete smp_wmb()! ;-)

  Andrea

> 
> Alan
> 