On Fri, Apr 07, 2023 at 05:49:02PM -0700, Paul E. McKenney wrote:
> On Fri, Apr 07, 2023 at 03:05:01PM +0200, Jonas Oberhauser wrote:
> >
> >
> > On 4/7/2023 2:12 AM, Joel Fernandes wrote:
> > >
> > >
> > > > On Apr 6, 2023, at 6:34 PM, Paul E. McKenney <paulmck@xxxxxxxxxx> wrote:
> > > >
> > > > On Thu, Apr 06, 2023 at 05:36:13PM -0400, Alan Stern wrote:
> > > > > Paul:
> > > > >
> > > > > I just saw that two of the files in
> > > > > tools/memory-model/litmus-tests have
> > > > > almost identical names:
> > > > >
> > > > > Z6.0+pooncelock+pooncelock+pombonce.litmus
> > > > > Z6.0+pooncelock+poonceLock+pombonce.litmus
> > > > >
> > > > > They differ only by a lower-case 'l' vs. a capital 'L'. It's
> > > > > not at all
> > > > > easy to see, and won't play well in case-insensitive filesystems.
> > > > >
> > > > > Should one of them be renamed?
> > > >
> > > > Quite possibly!
> > > >
> > > > The "L" denotes smp_mb__after_spinlock(). The only code difference
> > > > between these is that Z6.0+pooncelock+poonceLock+pombonce.litmus has
> > > > smp_mb__after_spinlock() and Z6.0+pooncelock+pooncelock+pombonce.litmus
> > > > does not.
> > > >
> > > > Suggestions for a better name? We could capitalize all the letters
> > > > in LOCK, I suppose...
> >
> > I don't think capitalizing LOCK is helpful.
>
> Greek font, then? (Sorry, couldn't resist...)
>
> > To be honest, almost all the names are extremely cryptic to newcomers like
> > me (like, what does Z6.0 mean? Is it some magic incantation?).
> > And that's not something that's easy to fix.
>
> All too true on all counts. Some of the names abbreviate the litmus
> test itself, and there are multiple encodings depending one who/what
> generated the test in question. Others of the names relate to who came
> up with them or the code from which they are derived.
>
> New allegedly universal naming schemes have a rather short half-life.
>
> What would be cool would be a way to structurally compare litmus tests.
> I bet that there are quite a few duplicates, for example.
>
> > The only use case I can think of for spending time improving the names is
> > that sometimes you wanna say something like "oh, this is like
> > Z6.0+pooncelock+pooncelockmb+pombonce". And then people can look up what
> > that is.
> > For that, it's important that the names are easy to disambiguate by humans,
> > and I think Joel's suggestion is an improvement.
> > (and it also fixes the issue brought up by Alan about case-insensitive file
> > systems)
> >
> > >
> > > Z6.0+pooncelock+pooncelockmb+pombonce.litmus ?
>
> I am OK with this one, but then again, I was also OK with the original
> Z6.0+pooncelock+poonceLock+pombonce.litmus. ;-)
FWIW, if I move that smp_mb_after..() a step lower, that also makes the test
work (see below).
If you may look over quickly my analysis of why this smp_mb_after..() is
needed, it is because what I marked as a and d below don't have an hb
relation right?
(*
b ->rf c
d ->co e
e ->hb f
basically the issue is a ->po b ->rf c ->po d does not imply a ->hb d
*)
P0(int *x, int *y, spinlock_t *mylock)
{
spin_lock(mylock);
WRITE_ONCE(*x, 1); // a
WRITE_ONCE(*y, 1); // b
spin_unlock(mylock);
}
P1(int *y, int *z, spinlock_t *mylock)
{
int r0;
spin_lock(mylock);
r0 = READ_ONCE(*y); // c
smp_mb__after_spinlock(); // moving this a bit lower also works fwiw.
WRITE_ONCE(*z, 1); // d
spin_unlock(mylock);
}
P2(int *x, int *z)
{
int r1;
WRITE_ONCE(*z, 2); // e
smp_mb();
r1 = READ_ONCE(*x); // f
}
exists (1:r0=1 /\ z=2 /\ 2:r1=0)
> Would someone like to to a "git mv" send the resulting patch?
Yes I can do that in return as I am thankful in advance for the above
discussion. ;)
thanks,
- Joel
