Cc: Dave Jones <davej@...>, Roland Dreier <rdreier@...>, Linus Torvalds <torvalds@...>, David Miller <davem@...>, <jeremy@...>, <hugh@...>, <mingo@...>, <akpm@...>, <linux-kernel@...>
From: Andrea Arcangeli <andrea@qumranet.com>
Lockdep can't recognize if spinlocks are at a different address. So
trylock avoids lockdep to generate false positives. After lockdep will
be fixed this change can and should be reverted.
Signed-off-by: Andrea Arcangeli <andrea@qumranet.com>
---
On Mon, Aug 04, 2008 at 07:27:28PM +0200, Andrea Arcangeli wrote:
(btw it's not like I forgot to sync, but I sync against the wrong
source tree previously because it was in the bash history, it's a bit
complicate to explain)
In the meantime (as I doubt lockdep will get fixed any time soon) this
will workaround it.
diff -r 3469dce61df1 mm/mmap.c
--- a/mm/mmap.c Tue Jul 29 20:01:28 2008 +0200
+++ b/mm/mmap.c Mon Aug 04 19:41:53 2008 +0200
@@ -2279,8 +2279,13 @@ static void vm_lock_anon_vma(struct anon
/*
* The LSB of head.next can't change from under us
* because we hold the mm_all_locks_mutex.
+ *
+ * spin_lock would confuse lockdep who can't
+ * differentiate between the 'mapping' always changing
+ * address.
*/
- spin_lock(&anon_vma->lock);
+ while (!spin_trylock(&anon_vma->lock))
+ cpu_relax();
/*
* We can safely modify head.next after taking the
* anon_vma->lock. If some other vma in this mm shares
@@ -2310,7 +2315,13 @@ static void vm_lock_mapping(struct addre
*/
if (test_and_set_bit(AS_MM_ALL_LOCKS, &mapping->flags))
BUG();
- spin_lock(&mapping->i_mmap_lock);
+ /*
+ * spin_lock would confuse lockdep who can't
+ * differentiate between the 'mapping' always changing
+ * address.
+ */
+ while (!spin_trylock(&mapping->i_mmap_lock))
+ cpu_relax();
}
}
--
