Hello Waiman, Marek

Thanks for the review.

I've never used lockdep for debug but it seems preferable to
keep that feature working. It could be look like this:

diff --git a/include/linux/mutex.h b/include/linux/mutex.h
index f7611c092db7..574f6de6084d 100644
--- a/include/linux/mutex.h
+++ b/include/linux/mutex.h

@@ -22,6 +22,8 @@
  #include <linux/cleanup.h>
  #include <linux/mutex_types.h>

+struct device;
+
  #ifdef CONFIG_DEBUG_LOCK_ALLOC
  # define __DEP_MAP_MUTEX_INITIALIZER(lockname)			\
  		, .dep_map = {					\

@@ -115,10 +117,31 @@ do {							\

  #ifdef CONFIG_DEBUG_MUTEXES

+int debug_devm_mutex_init(struct device *dev, struct mutex *lock);
+
+#define devm_mutex_init(dev, mutex)			\
+({							\
+	int ret;					\
+	mutex_init(mutex);				\
+	ret = debug_devm_mutex_init(dev, mutex);	\
+	ret;						\
+})
+
  void mutex_destroy(struct mutex *lock);

  #else

+/*
+* When CONFIG_DEBUG_MUTEXES is off mutex_destroy is just a nop so
+* there's no really need to register it in devm subsystem.
+*/
+#define devm_mutex_init(dev, mutex)			\
+({							\
+	typecheck(struct device *, dev);		\
+	mutex_init(mutex);				\
+	0;						\
+})
+
  static inline void mutex_destroy(struct mutex *lock) {}

  #endif

diff --git a/kernel/locking/mutex-debug.c b/kernel/locking/mutex-debug.c
index bc8abb8549d2..967a5367c79a 100644
--- a/kernel/locking/mutex-debug.c
+++ b/kernel/locking/mutex-debug.c

@@ -19,6 +19,7 @@
  #include <linux/kallsyms.h>
  #include <linux/interrupt.h>
  #include <linux/debug_locks.h>
+#include <linux/device.h>

  #include "mutex.h"

@@ -89,6 +90,16 @@ void debug_mutex_init(struct mutex *lock, const char 

*name,
  	lock->magic = lock;
  }

+static void devm_mutex_release(void *res)
+{
+	mutex_destroy(res);
+}
+
+int debug_devm_mutex_init(struct device *dev, struct mutex *lock)
+{
+	return devm_add_action_or_reset(dev, devm_mutex_release, lock);
+}
+
  /***
   * mutex_destroy - mark a mutex unusable
   * @lock: the mutex to be destroyed
-- 
2.25.1



And now I would drop the the refactoring patch with moving down 
mutex_destroy. devm block is big enough to be declared standalone.


On 3/7/24 19:44, Marek Behún wrote:

On Thu, 7 Mar 2024 08:39:46 -0500
Waiman Long [off-list ref] wrote:

quoted

On 3/7/24 04:56, Marek Behún wrote:

quoted

On Thu, Mar 07, 2024 at 05:40:26AM +0300, George Stark wrote:

quoted

Using of devm API leads to a certain order of releasing resources.
So all dependent resources which are not devm-wrapped should be deleted
with respect to devm-release order. Mutex is one of such objects that
often is bound to other resources and has no own devm wrapping.
Since mutex_destroy() actually does nothing in non-debug builds
frequently calling mutex_destroy() is just ignored which is safe for now
but wrong formally and can lead to a problem if mutex_destroy() will be
extended so introduce devm_mutex_init()

Signed-off-by: George Stark <redacted>
Signed-off-by: Christophe Leroy <redacted>
---
   Hello Christophe. Hope you don't mind I put you SoB tag because you helped alot
   to make this patch happen.

   include/linux/mutex.h        | 13 +++++++++++++
   kernel/locking/mutex-debug.c | 22 ++++++++++++++++++++++
   2 files changed, 35 insertions(+)

diff --git a/include/linux/mutex.h b/include/linux/mutex.h
index f7611c092db7..9bcf72cb941a 100644
--- a/include/linux/mutex.h
+++ b/include/linux/mutex.h

@@ -22,6 +22,8 @@
   #include <linux/cleanup.h>
   #include <linux/mutex_types.h>

+struct device;
+
   #ifdef CONFIG_DEBUG_LOCK_ALLOC
   # define __DEP_MAP_MUTEX_INITIALIZER(lockname)			\
   		, .dep_map = {					\

@@ -115,10 +117,21 @@ do {							\

   #ifdef CONFIG_DEBUG_MUTEXES

+int devm_mutex_init(struct device *dev, struct mutex *lock);
   void mutex_destroy(struct mutex *lock);

   #else

+static inline int devm_mutex_init(struct device *dev, struct mutex *lock)
+{
+	/*
+	 * since mutex_destroy is nop actually there's no need to register it
+	 * in devm subsystem.
+	 */
+	mutex_init(lock);
+	return 0;
+}
+
   static inline void mutex_destroy(struct mutex *lock) {}

   #endif

diff --git a/kernel/locking/mutex-debug.c b/kernel/locking/mutex-debug.c
index bc8abb8549d2..c9efab1a8026 100644
--- a/kernel/locking/mutex-debug.c
+++ b/kernel/locking/mutex-debug.c

@@ -19,6 +19,7 @@
   #include <linux/kallsyms.h>
   #include <linux/interrupt.h>
   #include <linux/debug_locks.h>
+#include <linux/device.h>

   #include "mutex.h"

@@ -104,3 +105,24 @@ void mutex_destroy(struct mutex *lock)
   }

   EXPORT_SYMBOL_GPL(mutex_destroy);
+
+static void devm_mutex_release(void *res)
+{
+	mutex_destroy(res);
+}
+
+/**
+ * devm_mutex_init - Resource-managed mutex initialization
+ * @dev:	Device which lifetime mutex is bound to
+ * @lock:	Pointer to a mutex
+ *
+ * Initialize mutex which is automatically destroyed when the driver is detached.
+ *
+ * Returns: 0 on success or a negative error code on failure.
+ */
+int devm_mutex_init(struct device *dev, struct mutex *lock)
+{
+	mutex_init(lock);
+	return devm_add_action_or_reset(dev, devm_mutex_release, lock);
+}
+EXPORT_SYMBOL_GPL(devm_mutex_init);

Hi George,

look at
https://lore.kernel.org/lkml/7013bf9e-2663-4613-ae61-61872e81355b@redhat.com/ (local)
where Matthew and Hans explain that devm_mutex_init needs to be a macro
because of the static lockdep key.

so this should be something like:

static inline int __devm_mutex_init(struct device *dev, struct mutex *mutex,
				    const char *name,
				    struct lock_class_key *key)
{
	__mutex_init(mutex, name, key);
	return devm_add_action_or_reset(dev, devm_mutex_release, mutex);
}

#define devm_mutex_init(dev, mutex)				\
do {								\
	static struct lock_class_key __key;			\
								\
	__devm_mutex_init(dev, (mutex), #mutex, &__key);	\
} while (0);


Marek

Making devm_mutex_init() a function will make all the devm_mutex share
the same lockdep key. Making it a macro will make each caller of
devm_mutex_init() have a distinct lockdep key. It all depends on whether
all the devm_mutexes have the same lock usage pattern or not and whether
it is possible for one devm_mutex to be nested inside another. So either
way can be fine depending on the mutex usage pattern. My suggestion is
to use a function, if possible, unless it will cause a false positive
lockdep splat as there is a limit on the maximum # of lockdep keys that
can be used.

devm_mutex_init() should behave like other similar function
initializing stuff with resource management. I.e. it should behave like
mutex_init(), but with resource management.

mutex_init() is a macro generating static lockdep key for each instance,
so devm_mutex_init() should also generate static lockdep key for each
instance.

Marek

-- 
Best regards
George