Re: [RFC 2/3] tools/memory-model: Add a litmus test for atomic_set()

[Boqun Feng]

On Tue, Feb 25, 2020 at 02:01:02PM +0100, Luc Maranget wrote:
> Hi,
> 
> As far as I can remember I have implemented atomic_add_unless in herd7.
> 
> As to your test, I have first run a slightly modified version of your test
> as a kernel module (using klitmus7).
> 
> C atomic_add_unless-dependency
> {
>         atomic_t y = ATOMIC_INIT(1);
> }
>   P0(int *x, atomic_t *y, int *z)
> {
>         int r0;
>         r0 = READ_ONCE(*x);
>         if (atomic_add_unless((atomic_t *)y, 2, r0))
>                 WRITE_ONCE(*z, 42);
>         else
>                 WRITE_ONCE(*z, 1);
> }
>   P1(int *x, int *z)
> {
>         int r0;
>         r0 = smp_load_acquire(z);
>         WRITE_ONCE(*x, 1);
> }
> locations [y]
> exists
> (1:r0 = 1 /\ 0:r0 = 1)
> 
> 
> The test is also accepted by herd7, here producing teh same final values
> as actual run on a raspberry PI4B.
> 

Thanks, so I'm planning to make the following change to README file in
memory-model

I will add a separate patch in my v3 patchset of atomic-tests.

Regards,
Boqun

----->8
diff --git a/tools/memory-model/README b/tools/memory-model/README
index fc07b52f2028..d974a96ad273 100644
--- a/tools/memory-model/README
+++ b/tools/memory-model/README
@@ -207,11 +207,15 @@ The Linux-kernel memory model (LKMM) has the following limitations:
 		case as a store release.
 
 	b.	The "unless" RMW operations are not currently modeled:
-		atomic_long_add_unless(), atomic_add_unless(),
-		atomic_inc_unless_negative(), and
-		atomic_dec_unless_positive().  These can be emulated
+		atomic_long_add_unless(), atomic_inc_unless_negative(),
+		and atomic_dec_unless_positive().  These can be emulated
 		in litmus tests, for example, by using atomic_cmpxchg().
 
+		One exception of this limitation is atomic_add_unless(),
+		which is provide directly by herd7 (so no corresponding
+		definition in linux-kernel.def). atomic_add_unless() is
+		modeled by herd7 therefore it can be used in litmus tests.
+
 	c.	The call_rcu() function is not modeled.  It can be
 		emulated in litmus tests by adding another process that
 		invokes synchronize_rcu() and the body of the callback

> --Luc
> 
> > Luc,
> > 
> > Could you have a look at the problem Andrea and I discuss here? It seems
> > that you have done a few things in herd for atomic_add_unless() in
> > particular, and based on the experiments of Andrea and me, seems
> > atomic_add_unless() works correctly. So can you confirm that herd now
> > can handle atomic_add_unless() or there is still something missing?
> > 
> > Thanks!
> > 
> > Regards,
> > Boqun
> > 
> > On Fri, Feb 14, 2020 at 06:40:03PM +0800, Boqun Feng wrote:
> > > On Fri, Feb 14, 2020 at 09:12:13AM +0100, Andrea Parri wrote:
> > > > > @@ -0,0 +1,24 @@
> > > > > +C Atomic-set-observable-to-RMW
> > > > > +
> > > > > +(*
> > > > > + * Result: Never
> > > > > + *
> > > > > + * Test of the result of atomic_set() must be observable to atomic RMWs.
> > > > > + *)
> > > > > +
> > > > > +{
> > > > > +	atomic_t v = ATOMIC_INIT(1);
> > > > > +}
> > > > > +
> > > > > +P0(atomic_t *v)
> > > > > +{
> > > > > +	(void)atomic_add_unless(v,1,0);
> > > > 
> > > > We blacklisted this primitive some time ago, cf. section "LIMITATIONS",
> > > > entry (6b) in tools/memory-model/README; the discussion was here:
> > > > 
> > > >   https://lkml.kernel.org/r/20180829211053.20531-3-paulmck@xxxxxxxxxxxxxxxxxx
> > > > 
> > > 
> > > And in an email replying to that email, you just tried and seemed
> > > atomic_add_unless() works ;-)
> > > 
> > > > but unfortunately I can't remember other details at the moment: maybe
> > > > it is just a matter of or the proper time to update that section.
> > > > 
> > > 
> > > I spend a few time looking into the changes in herd, the dependency
> > > problem seems to be as follow:
> > > 
> > > For atomic_add_unless(ptr, a, u), the return value (true or false)
> > > depends on both *ptr and u, this is different than other atomic RMW,
> > > whose return value only depends on *ptr. Considering the following
> > > litmus test:
> > > 
> > > 	C atomic_add_unless-dependency
> > > 
> > > 	{
> > > 		int y = 1;
> > > 	}
> > > 
> > > 	P0(int *x, int *y, int *z)
> > > 	{
> > > 		int r0;
> > > 		int r1;
> > > 		int r2;
> > > 
> > > 		r0 = READ_ONCE(*x);
> > > 		if (atomic_add_unless(y, 2, r0))
> > > 			WRITE_ONCE(*z, 42);
> > > 		else
> > > 			WRITE_ONCE(*z, 1);
> > > 	}
> > > 
> > > 	P1(int *x, int *y, int *z)
> > > 	{
> > > 		int r0;
> > > 
> > > 		r0 = smp_load_acquire(z);
> > > 
> > > 		WRITE_ONCE(*x, 1);
> > > 	}
> > > 
> > > 	exists
> > > 	(1:r0 = 1 /\ 0:r0 = 1)
> > > 
> > > , the exist-clause will never trigger, however if we replace
> > > "atomic_add_unless(y, 2, r0)" with "atomic_add_unless(y, 2, 1)", the
> > > write on *z and the read from *x on CPU 0 are not ordered, so we could
> > > observe the exist-clause triggered.
> > > 
> > > I just tried with the latest herd, and herd can work out this
> > > dependency. So I think we are good now and can change the limitation
> > > section in the document. But I will wait for Luc's input for this. Luc,
> > > did I get this correct? Is there any other limitation on
> > > atomic_add_unless() now?
> > > 
> > > Regards,
> > > Boqun
> > > 
> > > > Thanks,
> > > >   Andrea