From: Alan Stern <stern@rowland.harvard.edu>
To: LKMM Maintainers -- Akira Yokosawa <akiyks@gmail.com>,
Andrea Parri <andrea.parri@amarulasolutions.com>,
Boqun Feng <boqun.feng@gmail.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.vnet.ibm.com>,
Peter Zijlstra <peterz@infradead.org>,
Will Deacon <will.deacon@arm.com>
Cc: Kernel development list <linux-kernel@vger.kernel.org>
Subject: [PATCH 2/2] tools/memory-model: Add write ordering by release-acquire and by locks
Date: Thu, 21 Jun 2018 13:27:12 -0400 (EDT) [thread overview]
Message-ID: <Pine.LNX.4.44L0.1806211322160.2381-100000@iolanthe.rowland.org> (raw)
More than one kernel developer has expressed the opinion that the LKMM
should enforce ordering of writes by release-acquire chains and by
locking. In other words, given the following code:
WRITE_ONCE(x, 1);
spin_unlock(&s):
spin_lock(&s);
WRITE_ONCE(y, 1);
or the following:
smp_store_release(&x, 1);
r1 = smp_load_acquire(&x); // r1 = 1
WRITE_ONCE(y, 1);
the stores to x and y should be propagated in order to all other CPUs,
even though those other CPUs might not access the lock s or be part of
the release-acquire chain. In terms of the memory model, this means
that rel-rf-acq-po should be part of the cumul-fence relation.
All the architectures supported by the Linux kernel (including RISC-V)
do behave this way, albeit for varying reasons. Therefore this patch
changes the model in accordance with the developers' wishes.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
---
[as1871]
tools/memory-model/Documentation/explanation.txt | 81 +++++++++++++++++++++++
tools/memory-model/linux-kernel.cat | 2
2 files changed, 82 insertions(+), 1 deletion(-)
Index: usb-4.x/tools/memory-model/linux-kernel.cat
===================================================================
--- usb-4.x.orig/tools/memory-model/linux-kernel.cat
+++ usb-4.x/tools/memory-model/linux-kernel.cat
@@ -66,7 +66,7 @@ let ppo = to-r | to-w | fence
(* Propagation: Ordering from release operations and strong fences. *)
let A-cumul(r) = rfe? ; r
-let cumul-fence = A-cumul(strong-fence | po-rel) | wmb
+let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po
let prop = (overwrite & ext)? ; cumul-fence* ; rfe?
(*
Index: usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================
--- usb-4.x.orig/tools/memory-model/Documentation/explanation.txt
+++ usb-4.x/tools/memory-model/Documentation/explanation.txt
@@ -1897,3 +1897,84 @@ non-deadlocking executions. For example
Is it possible to end up with r0 = 36 at the end? The LKMM will tell
you it is not, but the model won't mention that this is because P1
will self-deadlock in the executions where it stores 36 in y.
+
+In the LKMM, locks and release-acquire chains cause stores to
+propagate in order. For example:
+
+ int x, y, z;
+
+ P0()
+ {
+ WRITE_ONCE(x, 1);
+ smp_store_release(&y, 1);
+ }
+
+ P1()
+ {
+ int r1;
+
+ r1 = smp_load_acquire(&y);
+ WRITE_ONCE(z, 1);
+ }
+
+ P2()
+ {
+ int r2, r3, r4;
+
+ r2 = READ_ONCE(z);
+ smp_rmb();
+ r3 = READ_ONCE(x);
+ r4 = READ_ONCE(y);
+ }
+
+If r1 = 1 and r2 = 1 at the end, then both r3 and r4 must also be 1.
+In other words, the smp_store_release() read by the smp_load_acquire()
+together act as a sort of inter-processor fence, forcing the stores to
+x and y to propagate to P2 before the store to z does, regardless of
+the fact that P2 doesn't execute any release or acquire instructions.
+This conclusion would hold even if P0 and P1 were on the same CPU, so
+long as r1 = 1.
+
+We have mentioned that the LKMM treats locks as acquires and unlocks
+as releases. Therefore it should not be surprising that something
+analogous to this ordering also holds for locks:
+
+ int x, y;
+ spinlock_t s;
+
+ P0()
+ {
+ spin_lock(&s);
+ WRITE_ONCE(x, 1);
+ spin_unlock(&s);
+ }
+
+ P1()
+ {
+ int r1;
+
+ spin_lock(&s);
+ r1 = READ_ONCE(x):
+ WRITE_ONCE(y, 1);
+ spin_unlock(&s);
+ }
+
+ P2()
+ {
+ int r2, r3;
+
+ r2 = READ_ONCE(y);
+ smp_rmb();
+ r3 = READ_ONCE(x);
+ }
+
+If r1 = 1 at the end (implying that P1's critical section executes
+after P0's) and r2 = 1, then r3 must be 1; the ordering of the
+critical sections forces the store to x to propagate to P2 before the
+store to y does.
+
+In both versions of this scenario, the store-propagation ordering is
+not required by the operational model. However, it does happen on all
+the architectures supporting the Linux kernel, and kernel developers
+seem to expect it; they have requested that this behavior be included
+in the LKMM.
next reply other threads:[~2018-06-21 17:27 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-21 17:27 Alan Stern [this message]
2018-06-21 18:04 ` [PATCH 2/2] tools/memory-model: Add write ordering by release-acquire and by locks Peter Zijlstra
2018-06-22 3:34 ` Paul E. McKenney
2018-06-22 8:08 ` Peter Zijlstra
2018-06-22 8:09 ` Will Deacon
2018-06-22 9:55 ` Will Deacon
2018-06-22 10:31 ` Peter Zijlstra
2018-06-22 10:38 ` Will Deacon
2018-06-22 11:25 ` Andrea Parri
2018-06-22 16:40 ` Paul E. McKenney
2018-06-22 18:09 ` Alan Stern
2018-06-22 18:30 ` Will Deacon
2018-06-22 19:11 ` Alan Stern
2018-06-22 20:53 ` Paul E. McKenney
2018-07-04 11:53 ` Will Deacon
2018-06-25 8:19 ` Andrea Parri
2018-07-03 17:28 ` Alan Stern
2018-07-04 11:28 ` Paul E. McKenney
2018-07-04 12:13 ` Will Deacon
2018-07-05 14:23 ` Alan Stern
2018-07-05 15:31 ` Paul E. McKenney
2018-07-04 12:11 ` Will Deacon
2018-07-05 14:00 ` Andrea Parri
2018-07-05 14:44 ` Will Deacon
2018-07-05 15:16 ` Daniel Lustig
2018-07-05 15:35 ` Daniel Lustig
2018-07-05 14:21 ` Alan Stern
2018-07-05 14:46 ` Will Deacon
2018-07-05 14:57 ` Alan Stern
2018-07-05 15:15 ` Will Deacon
2018-07-05 15:09 ` Andrea Parri
2018-07-06 20:37 ` Alan Stern
2018-07-06 21:10 ` Paul E. McKenney
2018-07-09 16:52 ` Will Deacon
2018-07-09 17:29 ` Daniel Lustig
2018-07-09 19:18 ` Alan Stern
2018-07-05 15:31 ` Paul E. McKenney
2018-07-05 15:39 ` Andrea Parri
2018-07-05 16:58 ` Paul E. McKenney
2018-07-05 17:06 ` Andrea Parri
2018-07-05 15:44 ` Daniel Lustig
2018-07-05 16:22 ` Will Deacon
2018-07-05 16:56 ` Paul E. McKenney
2018-07-05 18:12 ` Daniel Lustig
2018-07-05 18:38 ` Andrea Parri
2018-07-05 18:44 ` Andrea Parri
2018-07-05 23:32 ` Paul E. McKenney
2018-07-05 23:31 ` Paul E. McKenney
2018-07-06 9:25 ` Will Deacon
2018-07-06 14:14 ` Paul E. McKenney
2018-06-25 7:32 ` Peter Zijlstra
2018-06-25 8:29 ` Andrea Parri
2018-06-25 9:06 ` Peter Zijlstra
2018-06-22 9:06 ` Andrea Parri
2018-06-22 19:23 ` Alan Stern
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Pine.LNX.4.44L0.1806211322160.2381-100000@iolanthe.rowland.org \
--to=stern@rowland.harvard.edu \
--cc=akiyks@gmail.com \
--cc=andrea.parri@amarulasolutions.com \
--cc=boqun.feng@gmail.com \
--cc=dhowells@redhat.com \
--cc=j.alglave@ucl.ac.uk \
--cc=linux-kernel@vger.kernel.org \
--cc=luc.maranget@inria.fr \
--cc=npiggin@gmail.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=will.deacon@arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).