From 993b2ff221999066fcff231590593d0b98f45d32 Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Thu, 23 Oct 2014 20:27:00 -0700 Subject: [PATCH 1/2] futex: Mention key referencing differences between shared and private futexes Update our documentation as of fix 76835b0ebf8 (futex: Ensure get_futex_key_refs() always implies a barrier). Explicitly state that we don't do key referencing for private futexes. Signed-off-by: Davidlohr Bueso Cc: Matteo Franchin Cc: Davidlohr Bueso Cc: Linus Torvalds Cc: Darren Hart Cc: Peter Zijlstra Cc: Paul E. McKenney Acked-by: Catalin Marinas Link: http://lkml.kernel.org/r/1414121220.817.0.camel@linux-t7sj.site Signed-off-by: Thomas Gleixner --- kernel/futex.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/kernel/futex.c b/kernel/futex.c index f3a3a071283c..bbf071f325b8 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -143,9 +143,8 @@ * * Where (A) orders the waiters increment and the futex value read through * atomic operations (see hb_waiters_inc) and where (B) orders the write - * to futex and the waiters read -- this is done by the barriers in - * get_futex_key_refs(), through either ihold or atomic_inc, depending on the - * futex type. + * to futex and the waiters read -- this is done by the barriers for both + * shared and private futexes in get_futex_key_refs(). * * This yields the following case (where X:=waiters, Y:=futex): * @@ -344,13 +343,20 @@ static void get_futex_key_refs(union futex_key *key) futex_get_mm(key); /* implies MB (B) */ break; default: + /* + * Private futexes do not hold reference on an inode or + * mm, therefore the only purpose of calling get_futex_key_refs + * is because we need the barrier for the lockless waiter check. + */ smp_mb(); /* explicit MB (B) */ } } /* * Drop a reference to the resource addressed by a key. - * The hash bucket spinlock must not be held. + * The hash bucket spinlock must not be held. This is + * a no-op for private futexes, see comment in the get + * counterpart. */ static void drop_futex_key_refs(union futex_key *key) { From 30a6b8031fe14031ab27c1fa3483cb9780e7f63c Mon Sep 17 00:00:00 2001 From: Brian Silverman Date: Sat, 25 Oct 2014 20:20:37 -0400 Subject: [PATCH 2/2] futex: Fix a race condition between REQUEUE_PI and task death free_pi_state and exit_pi_state_list both clean up futex_pi_state's. exit_pi_state_list takes the hb lock first, and most callers of free_pi_state do too. requeue_pi doesn't, which means free_pi_state can free the pi_state out from under exit_pi_state_list. For example: task A | task B exit_pi_state_list | pi_state = | curr->pi_state_list->next | | futex_requeue(requeue_pi=1) | // pi_state is the same as | // the one in task A | free_pi_state(pi_state) | list_del_init(&pi_state->list) | kfree(pi_state) list_del_init(&pi_state->list) | Move the free_pi_state calls in requeue_pi to before it drops the hb locks which it's already holding. [ tglx: Removed a pointless free_pi_state() call and the hb->lock held debugging. The latter comes via a seperate patch ] Signed-off-by: Brian Silverman Cc: austin.linux@gmail.com Cc: darren@dvhart.com Cc: peterz@infradead.org Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/1414282837-23092-1-git-send-email-bsilver16384@gmail.com Signed-off-by: Thomas Gleixner --- kernel/futex.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/kernel/futex.c b/kernel/futex.c index bbf071f325b8..63678b573d61 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -647,8 +647,14 @@ static struct futex_pi_state * alloc_pi_state(void) return pi_state; } +/* + * Must be called with the hb lock held. + */ static void free_pi_state(struct futex_pi_state *pi_state) { + if (!pi_state) + return; + if (!atomic_dec_and_test(&pi_state->refcount)) return; @@ -1527,15 +1533,6 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, } retry: - if (pi_state != NULL) { - /* - * We will have to lookup the pi_state again, so free this one - * to keep the accounting correct. - */ - free_pi_state(pi_state); - pi_state = NULL; - } - ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1, VERIFY_READ); if (unlikely(ret != 0)) goto out; @@ -1625,6 +1622,8 @@ retry_private: case 0: break; case -EFAULT: + free_pi_state(pi_state); + pi_state = NULL; double_unlock_hb(hb1, hb2); hb_waiters_dec(hb2); put_futex_key(&key2); @@ -1640,6 +1639,8 @@ retry_private: * exit to complete. * - The user space value changed. */ + free_pi_state(pi_state); + pi_state = NULL; double_unlock_hb(hb1, hb2); hb_waiters_dec(hb2); put_futex_key(&key2); @@ -1716,6 +1717,7 @@ retry_private: } out_unlock: + free_pi_state(pi_state); double_unlock_hb(hb1, hb2); hb_waiters_dec(hb2); @@ -1733,8 +1735,6 @@ out_put_keys: out_put_key1: put_futex_key(&key1); out: - if (pi_state != NULL) - free_pi_state(pi_state); return ret ? ret : task_count; }