[PATCH] timekeeping: Fix missing memory barriers in NMI safe CLOCK_MONOTONIC[_RAW]

[Mathieu Desnoyers]

Commit c7b080e148d9 "timekeeping: Provide fast and NMI safe access to
CLOCK_MONOTONIC[_RAW]" lacks memory barriers.

The following scenario demonstrates a race condition where the reader
can see a corrupted clock value.

Initial conditions:

tkf->seq = 0
tkf->base[0] and tkf->base[1] are initialized.

CPU 0                                      CPU 1
------------                               ----------------
update:
  tkf->seq++
  smb_wmb()
  tkf->seq++ (reordered before update)
                                           reader:
                                             seq = tkf->seq (reads 2)
                                             smp_rmb()
                                             idx = seq & 0x01
                                             now = now(tkf->base[idx]
                                                    # reads base[0]
  update(tkf->base[0], tk)
       # racy concurrent update
                                             smp_rmb()
                                             while (seq != tkf->seq)
                                                    # they are equal
 [ update continues ... ]

In this situation, the reader returns a corrupted value. Adding a
smp_wmb() between update of base[0] and increment of seq, as well as
between update of base[1] and the _following_ increment of seq (next
update call) fixes this.

Introduce raw_write_seqcount_latch() to abstract those barriers rather
than open-coding them in update_fast_timekeeper().

Link: alpine.DEB.2.10.1407122145580.4357@nanos">http://lkml.kernel.org/r/alpine.DEB.2.10.1407122145580.4357@nanos
Fixes: c7b080e148d9 "timekeeping: Provide fast and NMI safe access to CLOCK_MONOTONIC[_RAW]"
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: John Stultz <john.stultz@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
---
 include/linux/seqlock.h   |   11 +++++++++++
 kernel/time/timekeeping.c |    4 ++--
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/include/linux/seqlock.h b/include/linux/seqlock.h
index dcc64b9..c18adee 100644
--- a/include/linux/seqlock.h
+++ b/include/linux/seqlock.h
@@ -236,6 +236,17 @@ static inline void raw_write_seqcount_end(seqcount_t *s)
 }
 
 /*
+ * raw_write_seqcount_latch - redirect readers to even/odd copy
+ * @s: pointer to seqcount_t
+ */
+static inline void raw_write_seqcount_latch(seqcount_t *s)
+{
+	smp_wmb();	/* prior stores before incrementing "sequence" */
+	s->sequence++;
+	smp_wmb();	/* increment "sequence" before following stores */
+}
+
+/*
  * Sequence counter only version assumes that callers are using their
  * own mutexing.
  */
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 47d2caf..2bd73b0 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -317,7 +317,7 @@ static void update_fast_timekeeper(struct clocksource *clk, struct tk_fast *tkf,
 	struct tk_fast_base *base = tkf->base;
 
 	/* Force readers off to base[1] */
-	raw_write_seqcount_begin(&tkf->seq);
+	raw_write_seqcount_latch(&tkf->seq);
 
 	/* Update base[0] */
 	base->clock = clk;
@@ -327,7 +327,7 @@ static void update_fast_timekeeper(struct clocksource *clk, struct tk_fast *tkf,
 	base->mult = mult;
 
 	/* Force readers back to base[0] */
-	raw_write_seqcount_end(&tkf->seq);
+	raw_write_seqcount_latch(&tkf->seq);
 
 	/* Update base[1] */
 	base++;
-- 
1.7.10.4

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/