rtmutex-Fix-CONFIG_DEBUG_RT_MUTEX-lock-underflow-war.patch

From 2cafb0a670cbd6571dec957bde90e908819d01d9 Mon Sep 17 00:00:00 2001
From: john stultz <johnstul@us.ibm.com>
Date: Fri, 3 Jul 2009 08:44:25 -0500
Subject: [PATCH] rtmutex: Fix CONFIG_DEBUG_RT_MUTEX lock underflow warnings

commit 2c237cf3487394a46a92c002d8c15f19a8e4faf7 in tip.

So if I enable CONFIG_DEBUG_RT_MUTEXES with 2.6.24.7-rt14, I tend to
quickly see a number of BUG warnings when running Java tests:

BUG: jxeinajar/3383: lock count underflow!
Pid: 3383, comm: jxeinajar Not tainted 2.6.24-ibmrt2.5john #3

Call Trace:
 [<ffffffff8107208d>] rt_mutex_deadlock_account_unlock+0x5d/0x70
 [<ffffffff817d6aa5>] rt_read_slowunlock+0x35/0x550
 [<ffffffff8107173d>] rt_mutex_up_read+0x3d/0xc0
 [<ffffffff81072a99>] rt_up_read+0x29/0x30
 [<ffffffff8106e34e>] do_futex+0x32e/0xd40
 [<ffffffff8107173d>] ? rt_mutex_up_read+0x3d/0xc0
 [<ffffffff81072a99>] ? rt_up_read+0x29/0x30
 [<ffffffff8106f370>] compat_sys_futex+0xa0/0x110
 [<ffffffff81010a36>] ? syscall_trace_enter+0x86/0xb0
 [<ffffffff8102ff04>] cstar_do_call+0x1b/0x65

INFO: lockdep is turned off.
---------------------------
| preempt count: 00000001 ]
| 1-level deep critical section nesting:
----------------------------------------
... [<ffffffff817d8e42>] .... __spin_lock_irqsave+0x22/0x60
......[<ffffffff817d6a93>] ..   ( <= rt_read_slowunlock+0x23/0x550)

After some debugging and with Steven's help, we realized that with
rwlocks, rt_mutex_deadlock_account_lock can be called multiple times in
parallel (where as in most cases the mutex must be held by the caller to
to call the function). This can cause integer lock_count value being
used to be non-atomically incremented.

The following patch converts lock_count to a atomic_t and resolves the
warnings.

Signed-off-by: John Stultz <johnstul@us.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Clark Williams <williams@redhat.com>
Cc: dvhltc <dvhltc@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

diff --git a/include/linux/sched.h b/include/linux/sched.h
index a958544..7e50287 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1479,7 +1479,7 @@ struct task_struct {
 #define MAX_PREEMPT_TRACE 25
 #define MAX_LOCK_STACK	MAX_PREEMPT_TRACE
 #ifdef CONFIG_DEBUG_PREEMPT
-	int lock_count;
+	atomic_t lock_count;
 # ifdef CONFIG_PREEMPT_RT
 	struct rt_mutex *owned_lock[MAX_LOCK_STACK];
 # endif
diff --git a/kernel/fork.c b/kernel/fork.c
index 8453327..81fcb2a 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1171,7 +1171,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 	if (retval)
 		goto bad_fork_cleanup_io;
 #ifdef CONFIG_DEBUG_PREEMPT
-	p->lock_count = 0;
+	atomic_set(&p->lock_count, 0);
 #endif
 
 	if (pid != &init_struct_pid) {
diff --git a/kernel/rtmutex-debug.c b/kernel/rtmutex-debug.c
index 2381c7a..e7e6314 100644
--- a/kernel/rtmutex-debug.c
+++ b/kernel/rtmutex-debug.c
@@ -178,7 +178,7 @@ void
 rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task)
 {
 #ifdef CONFIG_DEBUG_PREEMPT
-	if (task->lock_count >= MAX_LOCK_STACK) {
+	if (atomic_read(&task->lock_count) >= MAX_LOCK_STACK) {
 		if (!debug_locks_off())
 			return;
 		printk("BUG: %s/%d: lock count overflow!\n",
@@ -187,16 +187,16 @@ rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task)
 		return;
 	}
 #ifdef CONFIG_PREEMPT_RT
-	task->owned_lock[task->lock_count] = lock;
+	task->owned_lock[atomic_read(&task->lock_count)] = lock;
 #endif
-	task->lock_count++;
+	atomic_inc(&task->lock_count);
 #endif
 }
 
 void rt_mutex_deadlock_account_unlock(struct task_struct *task)
 {
 #ifdef CONFIG_DEBUG_PREEMPT
-	if (!task->lock_count) {
+	if (!atomic_read(&task->lock_count)) {
 		if (!debug_locks_off())
 			return;
 		printk("BUG: %s/%d: lock count underflow!\n",
@@ -204,9 +204,9 @@ void rt_mutex_deadlock_account_unlock(struct task_struct *task)
 		dump_stack();
 		return;
 	}
-	task->lock_count--;
+	atomic_dec(&task->lock_count);
 #ifdef CONFIG_PREEMPT_RT
-	task->owned_lock[task->lock_count] = NULL;
+	task->owned_lock[atomic_read(&task->lock_count)] = NULL;
 #endif
 #endif
 }
-- 
1.7.1.1