Messages in this thread | | | Date | Tue, 10 Apr 2018 14:34:34 -0700 | From | "Paul E. McKenney" <> | Subject | Re: [PATCH] memory-model: fix cheat sheet typo |
| |
On Tue, Apr 10, 2018 at 11:10:06PM +0200, Paolo Bonzini wrote: > On 10/04/2018 22:32, Paul E. McKenney wrote: > > On Mon, Apr 09, 2018 at 11:42:58AM -0700, Paul E. McKenney wrote: > >> On Mon, Apr 09, 2018 at 06:50:15PM +0200, Paolo Bonzini wrote: > >>> "RWM" should be "RMW", and that's more or less the extent to which I > >>> can claim to change the document. :) In particular, "Self" is not > >>> documented and the difference between "Self" and "SV" is not clear > >>> to me. > >>> > >>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> > >> > >> Applied, though without the questions. ;-) > >> > >> "Self" is for things like smp_load_acquire() and smp_store_release() > >> that order themselves against later and earlier accesses, respectively. > >> This ordering applies to later/earlier access to all variables, not > >> just the one that smp_load_acquire()/smp_store_release() accessed. > >> In contrast, things like smp_mb() order only other accesses, not > >> themselves. Or at least it is impossible to proves whether or not they > >> order themselves because they are not separately visible to other CPUs. > >> > >> "SV" is "same variable", which applies to pretty much anything that > >> accesses a variable, but not to things like smp_mb() which do not. > >> > >> Does that help? > > > > On the perhaps naive assumption that silence means assent, how about > > the following patch? > > Silence meant "I tried thinking of a patch myself, and hadn't come up > yet with a fully satisfactory one"; that's some kind of assent I guess. :)
;-) ;-) ;-)
> Your patch is certainly an improvement!
Glad it helps, and I have queued it for the next merge window. Of course, if a further improvement comes to mind, please do not keep it a secret. ;-)
Thanx, Paul
> Thanks, > > Paolo > > > Thanx, Paul > > > > ------------------------------------------------------------------------ > > > > commit 818e46e8db6cacb099b8640b7f2945a3151c00ab > > Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com> > > Date: Tue Apr 10 13:24:19 2018 -0700 > > > > tools/memory-order: Improve key for SELF and SV > > > > The key for "SELF" was missing completely and the key for "SV" was > > a bit obtuse. This commit therefore adds a key for "SELF" and improves > > the one for "SV". > > > > Reported-by: Paolo Bonzini <pbonzini@redhat.com> > > Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> > > > > diff --git a/tools/memory-model/Documentation/cheatsheet.txt b/tools/memory-model/Documentation/cheatsheet.txt > > index c0eafdaddfa4..d502993ac7d2 100644 > > --- a/tools/memory-model/Documentation/cheatsheet.txt > > +++ b/tools/memory-model/Documentation/cheatsheet.txt > > @@ -26,4 +26,5 @@ Key: C: Ordering is cumulative > > DR: Dependent read (address dependency) > > DW: Dependent write (address, data, or control dependency) > > RMW: Atomic read-modify-write operation > > - SV Same-variable access > > + SELF: Orders self, as opposed to accesses both before and after > > + SV: Orders later accesses to the same variable > > >
| |