[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20201105220017.15410-1-paulmck@kernel.org>
Date: Thu, 5 Nov 2020 14:00:10 -0800
From: paulmck@...nel.org
To: linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org,
kernel-team@...com, mingo@...nel.org
Cc: stern@...land.harvard.edu, parri.andrea@...il.com, will@...nel.org,
peterz@...radead.org, boqun.feng@...il.com, npiggin@...il.com,
dhowells@...hat.com, j.alglave@....ac.uk, luc.maranget@...ia.fr,
akiyks@...il.com, "Paul E . McKenney" <paulmck@...nel.org>
Subject: [PATCH memory-model 1/8] tools: memory-model: Document that the LKMM can easily miss control dependencies
From: Alan Stern <stern@...land.harvard.edu>
Add a small section to the litmus-tests.txt documentation file for
the Linux Kernel Memory Model explaining that the memory model often
fails to recognize certain control dependencies.
Suggested-by: Akira Yokosawa <akiyks@...il.com>
Signed-off-by: Alan Stern <stern@...land.harvard.edu>
Reviewed-by: Joel Fernandes (Google) <joel@...lfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@...nel.org>
---
tools/memory-model/Documentation/litmus-tests.txt | 17 +++++++++++++++++
1 file changed, 17 insertions(+)
diff --git a/tools/memory-model/Documentation/litmus-tests.txt b/tools/memory-model/Documentation/litmus-tests.txt
index 2f840dc..8a9d5d2 100644
--- a/tools/memory-model/Documentation/litmus-tests.txt
+++ b/tools/memory-model/Documentation/litmus-tests.txt
@@ -946,6 +946,23 @@ Limitations of the Linux-kernel memory model (LKMM) include:
carrying a dependency, then the compiler can break that dependency
by substituting a constant of that value.
+ Conversely, LKMM sometimes doesn't recognize that a particular
+ optimization is not allowed, and as a result, thinks that a
+ dependency is not present (because the optimization would break it).
+ The memory model misses some pretty obvious control dependencies
+ because of this limitation. A simple example is:
+
+ r1 = READ_ONCE(x);
+ if (r1 == 0)
+ smp_mb();
+ WRITE_ONCE(y, 1);
+
+ There is a control dependency from the READ_ONCE to the WRITE_ONCE,
+ even when r1 is nonzero, but LKMM doesn't realize this and thinks
+ that the write may execute before the read if r1 != 0. (Yes, that
+ doesn't make sense if you think about it, but the memory model's
+ intelligence is limited.)
+
2. Multiple access sizes for a single variable are not supported,
and neither are misaligned or partially overlapping accesses.
--
2.9.5
Powered by blists - more mailing lists