tools/memory-model: Heuristics using data_race() must handle all values

Data loaded for use by some sorts of heuristics can tolerate the
occasional erroneous value.  In this case the loads may use data_race()
to give the compiler full freedom to optimize while also informing KCSAN
of the intent.  However, for this to work, the heuristic needs to be
able to tolerate any erroneous value that could possibly arise.  This
commit therefore adds a paragraph spelling this out.

Signed-off-by: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>

diff --git a/tools/memory-model/Documentation/access-marking.txt b/tools/memory-model/Documentation/access-marking.txt
index d96fe20ed582ae9c2c5f53763263600595acc42c..82a489988726987061006e06544a111bcd1a8fda 100644 (file)
--- a/tools/memory-model/Documentation/access-marking.txt
+++ b/tools/memory-model/Documentation/access-marking.txt
@@ -126,6 +126,11 @@ consistent errors, which in turn are quite capable of breaking heuristics.
 Therefore use of data_race() should be limited to cases where some other
 code (such as a barrier() call) will force the occasional reload.
 
+Note that this use case requires that the heuristic be able to handle
+any possible error.  In contrast, if the heuristics might be fatally
+confused by one or more of the possible erroneous values, use READ_ONCE()
+instead of data_race().
+
 In theory, plain C-language loads can also be used for this use case.
 However, in practice this will have the disadvantage of causing KCSAN
 to generate false positives because KCSAN will have no way of knowing