[PATCH] kprobes: changed from using spinlock to mutex

Since Kprobes runtime exception handlers is now lock free as this code path is
now using RCU to walk through the list, there is no need for the
register/unregister{_kprobe} to use spin_{lock/unlock}_isr{save/restore}.  The
serialization during registration/unregistration is now possible using just a
mutex.

In the above process, this patch also fixes a minor memory leak for x86_64 and
powerpc.

Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 43 insertions, 48 deletions

diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 3897630..f14ccd3 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -48,7 +48,7 @@
 static struct hlist_head kprobe_table[KPROBE_TABLE_SIZE];
 static struct hlist_head kretprobe_inst_table[KPROBE_TABLE_SIZE];
 
-static DEFINE_SPINLOCK(kprobe_lock);	/* Protects kprobe_table */
+static DECLARE_MUTEX(kprobe_mutex);	/* Protects kprobe_table */
 DEFINE_SPINLOCK(kretprobe_lock);	/* Protects kretprobe_inst_table */
 static DEFINE_PER_CPU(struct kprobe *, kprobe_instance) = NULL;
 
@@ -167,7 +167,7 @@ static inline void reset_kprobe_instance(void)
 
 /*
  * This routine is called either:
- * 	- under the kprobe_lock spinlock - during kprobe_[un]register()
+ * 	- under the kprobe_mutex - during kprobe_[un]register()
  * 				OR
  * 	- with preemption disabled - from arch/xxx/kernel/kprobes.c
  */
@@ -420,7 +420,6 @@ static inline void add_aggr_kprobe(struct kprobe *ap, struct kprobe *p)
 /*
  * This is the second or subsequent kprobe at the address - handle
  * the intricacies
- * TODO: Move kcalloc outside the spin_lock
  */
 static int __kprobes register_aggr_kprobe(struct kprobe *old_p,
 					  struct kprobe *p)
@@ -442,25 +441,6 @@ static int __kprobes register_aggr_kprobe(struct kprobe *old_p,
 	return ret;
 }
 
-/* kprobe removal house-keeping routines */
-static inline void cleanup_kprobe(struct kprobe *p, unsigned long flags)
-{
-	arch_disarm_kprobe(p);
-	hlist_del_rcu(&p->hlist);
-	spin_unlock_irqrestore(&kprobe_lock, flags);
-	arch_remove_kprobe(p);
-}
-
-static inline void cleanup_aggr_kprobe(struct kprobe *old_p,
-		struct kprobe *p, unsigned long flags)
-{
-	list_del_rcu(&p->list);
-	if (list_empty(&old_p->list))
-		cleanup_kprobe(old_p, flags);
-	else
-		spin_unlock_irqrestore(&kprobe_lock, flags);
-}
-
 static int __kprobes in_kprobes_functions(unsigned long addr)
 {
 	if (addr >= (unsigned long)__kprobes_text_start
@@ -472,7 +452,6 @@ static int __kprobes in_kprobes_functions(unsigned long addr)
 int __kprobes register_kprobe(struct kprobe *p)
 {
 	int ret = 0;
-	unsigned long flags = 0;
 	struct kprobe *old_p;
 	struct module *mod;
 
@@ -484,18 +463,17 @@ int __kprobes register_kprobe(struct kprobe *p)
 			(unlikely(!try_module_get(mod))))
 		return -EINVAL;
 
-	if ((ret = arch_prepare_kprobe(p)) != 0)
-		goto rm_kprobe;
-
 	p->nmissed = 0;
-	spin_lock_irqsave(&kprobe_lock, flags);
+	down(&kprobe_mutex);
 	old_p = get_kprobe(p->addr);
 	if (old_p) {
 		ret = register_aggr_kprobe(old_p, p);
 		goto out;
 	}
 
-	arch_copy_kprobe(p);
+	if ((ret = arch_prepare_kprobe(p)) != 0)
+		goto out;
+
 	INIT_HLIST_NODE(&p->hlist);
 	hlist_add_head_rcu(&p->hlist,
 		       &kprobe_table[hash_ptr(p->addr, KPROBE_HASH_BITS)]);
@@ -503,10 +481,8 @@ int __kprobes register_kprobe(struct kprobe *p)
   	arch_arm_kprobe(p);
 
 out:
-	spin_unlock_irqrestore(&kprobe_lock, flags);
-rm_kprobe:
-	if (ret == -EEXIST)
-		arch_remove_kprobe(p);
+	up(&kprobe_mutex);
+
 	if (ret && mod)
 		module_put(mod);
 	return ret;
@@ -514,29 +490,48 @@ rm_kprobe:
 
 void __kprobes unregister_kprobe(struct kprobe *p)
 {
-	unsigned long flags;
-	struct kprobe *old_p;
 	struct module *mod;
+	struct kprobe *old_p, *cleanup_p;
 
-	spin_lock_irqsave(&kprobe_lock, flags);
+	down(&kprobe_mutex);
 	old_p = get_kprobe(p->addr);
-	if (old_p) {
-		/* cleanup_*_kprobe() does the spin_unlock_irqrestore */
-		if (old_p->pre_handler == aggr_pre_handler)
-			cleanup_aggr_kprobe(old_p, p, flags);
-		else
-			cleanup_kprobe(p, flags);
+	if (unlikely(!old_p)) {
+		up(&kprobe_mutex);
+		return;
+	}
 
-		synchronize_sched();
+	if ((old_p->pre_handler == aggr_pre_handler) &&
+		(p->list.next == &old_p->list) &&
+		(p->list.prev == &old_p->list)) {
+		/* Only one element in the aggregate list */
+		arch_disarm_kprobe(p);
+		hlist_del_rcu(&old_p->hlist);
+		cleanup_p = old_p;
+	} else if (old_p == p) {
+		/* Only one kprobe element in the hash list */
+		arch_disarm_kprobe(p);
+		hlist_del_rcu(&p->hlist);
+		cleanup_p = p;
+	} else {
+		list_del_rcu(&p->list);
+		cleanup_p = NULL;
+	}
 
-		if ((mod = module_text_address((unsigned long)p->addr)))
-			module_put(mod);
+	up(&kprobe_mutex);
 
-		if (old_p->pre_handler == aggr_pre_handler &&
-				list_empty(&old_p->list))
+	synchronize_sched();
+	if ((mod = module_text_address((unsigned long)p->addr)))
+		module_put(mod);
+
+	if (cleanup_p) {
+		if (cleanup_p->pre_handler == aggr_pre_handler) {
+			list_del_rcu(&p->list);
 			kfree(old_p);
-	} else
-		spin_unlock_irqrestore(&kprobe_lock, flags);
+		}
+		down(&kprobe_mutex);
+		arch_remove_kprobe(p);
+		up(&kprobe_mutex);
+	}
 }
 
 static struct notifier_block kprobe_exceptions_nb = {