[11/99] futex: Move drop_futex_key_refs out of spinlocked region

From: Greg KH
Date: Fri Nov 06 2009 - 17:41:33 EST

2.6.31-stable review patch. If anyone has any objections, please let us know.

From: Darren Hart <dvhltc@xxxxxxxxxx>

commit 89061d3d58e1f0742139605dc6a7950aa1ecc019 upstream.

When requeuing tasks from one futex to another, the reference held
by the requeued task to the original futex location needs to be
dropped eventually.

Dropping the reference may ultimately lead to a call to
"iput_final" and subsequently call into filesystem- specific code -
which may be non-atomic.

It is therefore safer to defer this drop operation until after the
futex_hash_bucket spinlock has been dropped.

Originally-From: Helge Bahmann <hcb@xxxxxxxxxxxxxxx>
Signed-off-by: Darren Hart <dvhltc@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Eric Dumazet <eric.dumazet@xxxxxxxxx>
Cc: Dinakar Guniguntala <dino@xxxxxxxxxx>
Cc: John Stultz <johnstul@xxxxxxxxxxxxxxxxxx>
Cc: Sven-Thorsten Dietrich <sdietrich@xxxxxxxxxx>
Cc: John Kacur <jkacur@xxxxxxxxxx>
LKML-Reference: <4AD7A298.5040802@xxxxxxxxxx>
Signed-off-by: Ingo Molnar <mingo@xxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxx>

kernel/futex.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -1028,7 +1028,6 @@ static inline
void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key,
struct futex_hash_bucket *hb)
- drop_futex_key_refs(&q->key);
q->key = *key;

@@ -1226,6 +1225,7 @@ retry_private:
if (ret == 1) {
+ drop_count++;
ret = get_futex_value_locked(&curval2, uaddr2);
if (!ret)
@@ -1304,6 +1304,7 @@ retry_private:
if (ret == 1) {
/* We got the lock. */
requeue_pi_wake_futex(this, &key2, hb2);
+ drop_count++;
} else if (ret) {
/* -EDEADLK */

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/