[PATCH] lockdep: better lock debugging
Ingo Molnar [Mon, 3 Jul 2006 07:24:33 +0000 (00:24 -0700)]
Generic lock debugging:

 - generalized lock debugging framework. For example, a bug in one lock
   subsystem turns off debugging in all lock subsystems.

 - got rid of the caller address passing (__IP__/__IP_DECL__/etc.) from
   the mutex/rtmutex debugging code: it caused way too much prototype
   hackery, and lockdep will give the same information anyway.

 - ability to do silent tests

 - check lock freeing in vfree too.

 - more finegrained debugging options, to allow distributions to
   turn off more expensive debugging features.

There's no separate 'held mutexes' list anymore - but there's a 'held locks'
stack within lockdep, which unifies deadlock detection across all lock
classes.  (this is independent of the lockdep validation stuff - lockdep first
checks whether we are holding a lock already)

Here are the current debugging options:

CONFIG_DEBUG_MUTEXES=y
CONFIG_DEBUG_LOCK_ALLOC=y

which do:

 config DEBUG_MUTEXES
          bool "Mutex debugging, basic checks"

 config DEBUG_LOCK_ALLOC
         bool "Detect incorrect freeing of live mutexes"

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

25 files changed:
drivers/char/sysrq.c
include/asm-generic/mutex-null.h
include/linux/debug_locks.h [new file with mode: 0644]
include/linux/init_task.h
include/linux/mm.h
include/linux/mutex-debug.h
include/linux/mutex.h
include/linux/rtmutex.h
include/linux/sched.h
init/main.c
kernel/exit.c
kernel/fork.c
kernel/mutex-debug.c
kernel/mutex-debug.h
kernel/mutex.c
kernel/mutex.h
kernel/rtmutex-debug.c
kernel/rtmutex-debug.h
kernel/rtmutex.c
kernel/rtmutex.h
kernel/sched.c
lib/Makefile
lib/debug_locks.c [new file with mode: 0644]
lib/spinlock_debug.c
mm/vmalloc.c

index a064ee9..e31f079 100644 (file)
@@ -151,7 +151,7 @@ static struct sysrq_key_op sysrq_mountro_op = {
 static void sysrq_handle_showlocks(int key, struct pt_regs *pt_regs,
                                struct tty_struct *tty)
 {
-       mutex_debug_show_all_locks();
+       debug_show_all_locks();
 }
 static struct sysrq_key_op sysrq_showlocks_op = {
        .handler        = sysrq_handle_showlocks,
index 5cf8b7c..254a126 100644 (file)
 #ifndef _ASM_GENERIC_MUTEX_NULL_H
 #define _ASM_GENERIC_MUTEX_NULL_H
 
-/* extra parameter only needed for mutex debugging: */
-#ifndef __IP__
-# define __IP__
-#endif
-
-#define __mutex_fastpath_lock(count, fail_fn)        fail_fn(count __RET_IP__)
-#define __mutex_fastpath_lock_retval(count, fail_fn)  fail_fn(count __RET_IP__)
-#define __mutex_fastpath_unlock(count, fail_fn)       fail_fn(count __RET_IP__)
-#define __mutex_fastpath_trylock(count, fail_fn)      fail_fn(count)
-#define __mutex_slowpath_needs_to_unlock()           1
+#define __mutex_fastpath_lock(count, fail_fn)          fail_fn(count)
+#define __mutex_fastpath_lock_retval(count, fail_fn)   fail_fn(count)
+#define __mutex_fastpath_unlock(count, fail_fn)                fail_fn(count)
+#define __mutex_fastpath_trylock(count, fail_fn)       fail_fn(count)
+#define __mutex_slowpath_needs_to_unlock()             1
 
 #endif
diff --git a/include/linux/debug_locks.h b/include/linux/debug_locks.h
new file mode 100644 (file)
index 0000000..6a70478
--- /dev/null
@@ -0,0 +1,69 @@
+#ifndef __LINUX_DEBUG_LOCKING_H
+#define __LINUX_DEBUG_LOCKING_H
+
+extern int debug_locks;
+extern int debug_locks_silent;
+
+/*
+ * Generic 'turn off all lock debugging' function:
+ */
+extern int debug_locks_off(void);
+
+/*
+ * In the debug case we carry the caller's instruction pointer into
+ * other functions, but we dont want the function argument overhead
+ * in the nondebug case - hence these macros:
+ */
+#define _RET_IP_               (unsigned long)__builtin_return_address(0)
+#define _THIS_IP_  ({ __label__ __here; __here: (unsigned long)&&__here; })
+
+#define DEBUG_LOCKS_WARN_ON(c)                                         \
+({                                                                     \
+       int __ret = 0;                                                  \
+                                                                       \
+       if (unlikely(c)) {                                              \
+               if (debug_locks_off())                                  \
+                       WARN_ON(1);                                     \
+               __ret = 1;                                              \
+       }                                                               \
+       __ret;                                                          \
+})
+
+#ifdef CONFIG_SMP
+# define SMP_DEBUG_LOCKS_WARN_ON(c)                    DEBUG_LOCKS_WARN_ON(c)
+#else
+# define SMP_DEBUG_LOCKS_WARN_ON(c)                    do { } while (0)
+#endif
+
+#ifdef CONFIG_DEBUG_LOCKING_API_SELFTESTS
+  extern void locking_selftest(void);
+#else
+# define locking_selftest()    do { } while (0)
+#endif
+
+#ifdef CONFIG_LOCKDEP
+extern void debug_show_all_locks(void);
+extern void debug_show_held_locks(struct task_struct *task);
+extern void debug_check_no_locks_freed(const void *from, unsigned long len);
+extern void debug_check_no_locks_held(struct task_struct *task);
+#else
+static inline void debug_show_all_locks(void)
+{
+}
+
+static inline void debug_show_held_locks(struct task_struct *task)
+{
+}
+
+static inline void
+debug_check_no_locks_freed(const void *from, unsigned long len)
+{
+}
+
+static inline void
+debug_check_no_locks_held(struct task_struct *task)
+{
+}
+#endif
+
+#endif
index 3a25695..678c1a9 100644 (file)
@@ -124,7 +124,6 @@ extern struct group_info init_groups;
        .cpu_timers     = INIT_CPU_TIMERS(tsk.cpu_timers),              \
        .fs_excl        = ATOMIC_INIT(0),                               \
        .pi_lock        = SPIN_LOCK_UNLOCKED,                           \
-       INIT_RT_MUTEXES(tsk)                                            \
 }
 
 
index 7517952..990957e 100644 (file)
@@ -14,6 +14,7 @@
 #include <linux/prio_tree.h>
 #include <linux/fs.h>
 #include <linux/mutex.h>
+#include <linux/debug_locks.h>
 
 struct mempolicy;
 struct anon_vma;
@@ -1034,13 +1035,6 @@ static inline void vm_stat_account(struct mm_struct *mm,
 }
 #endif /* CONFIG_PROC_FS */
 
-static inline void
-debug_check_no_locks_freed(const void *from, unsigned long len)
-{
-       mutex_debug_check_no_locks_freed(from, len);
-       rt_mutex_debug_check_no_locks_freed(from, len);
-}
-
 #ifndef CONFIG_DEBUG_PAGEALLOC
 static inline void
 kernel_map_pages(struct page *page, int numpages, int enable)
index 8b5769f..70a2609 100644 (file)
@@ -7,17 +7,11 @@
  * Mutexes - debugging helpers:
  */
 
-#define __DEBUG_MUTEX_INITIALIZER(lockname) \
-       , .held_list = LIST_HEAD_INIT(lockname.held_list), \
-         .name = #lockname , .magic = &lockname
+#define __DEBUG_MUTEX_INITIALIZER(lockname)                            \
+       , .magic = &lockname
 
-#define mutex_init(sem)                __mutex_init(sem, __FUNCTION__)
+#define mutex_init(sem)                __mutex_init(sem, __FILE__":"#sem)
 
 extern void FASTCALL(mutex_destroy(struct mutex *lock));
 
-extern void mutex_debug_show_all_locks(void);
-extern void mutex_debug_show_held_locks(struct task_struct *filter);
-extern void mutex_debug_check_no_locks_held(struct task_struct *task);
-extern void mutex_debug_check_no_locks_freed(const void *from, unsigned long len);
-
 #endif
index f1ac507..caafecd 100644 (file)
@@ -50,8 +50,6 @@ struct mutex {
        struct list_head        wait_list;
 #ifdef CONFIG_DEBUG_MUTEXES
        struct thread_info      *owner;
-       struct list_head        held_list;
-       unsigned long           acquire_ip;
        const char              *name;
        void                    *magic;
 #endif
@@ -76,10 +74,6 @@ struct mutex_waiter {
 # define __DEBUG_MUTEX_INITIALIZER(lockname)
 # define mutex_init(mutex)                     __mutex_init(mutex, NULL)
 # define mutex_destroy(mutex)                          do { } while (0)
-# define mutex_debug_show_all_locks()                  do { } while (0)
-# define mutex_debug_show_held_locks(p)                        do { } while (0)
-# define mutex_debug_check_no_locks_held(task)         do { } while (0)
-# define mutex_debug_check_no_locks_freed(from, len)   do { } while (0)
 #endif
 
 #define __MUTEX_INITIALIZER(lockname) \
index fa4a3b8..5d41dee 100644 (file)
@@ -29,8 +29,6 @@ struct rt_mutex {
        struct task_struct      *owner;
 #ifdef CONFIG_DEBUG_RT_MUTEXES
        int                     save_state;
-       struct list_head        held_list_entry;
-       unsigned long           acquire_ip;
        const char              *name, *file;
        int                     line;
        void                    *magic;
@@ -98,14 +96,6 @@ extern int rt_mutex_trylock(struct rt_mutex *lock);
 
 extern void rt_mutex_unlock(struct rt_mutex *lock);
 
-#ifdef CONFIG_DEBUG_RT_MUTEXES
-# define INIT_RT_MUTEX_DEBUG(tsk)                                      \
-       .held_list_head = LIST_HEAD_INIT(tsk.held_list_head),           \
-       .held_list_lock = SPIN_LOCK_UNLOCKED
-#else
-# define INIT_RT_MUTEX_DEBUG(tsk)
-#endif
-
 #ifdef CONFIG_RT_MUTEXES
 # define INIT_RT_MUTEXES(tsk)                                          \
        .pi_waiters     = PLIST_HEAD_INIT(tsk.pi_waiters, tsk.pi_lock), \
index aaf7233..bdabeee 100644 (file)
@@ -865,10 +865,6 @@ struct task_struct {
        struct plist_head pi_waiters;
        /* Deadlock detection and priority inheritance handling */
        struct rt_mutex_waiter *pi_blocked_on;
-# ifdef CONFIG_DEBUG_RT_MUTEXES
-       spinlock_t held_list_lock;
-       struct list_head held_list_head;
-# endif
 #endif
 
 #ifdef CONFIG_DEBUG_MUTEXES
index d604dfe..fc73e1c 100644 (file)
@@ -47,6 +47,7 @@
 #include <linux/key.h>
 #include <linux/unwind.h>
 #include <linux/buffer_head.h>
+#include <linux/debug_locks.h>
 
 #include <asm/io.h>
 #include <asm/bugs.h>
@@ -511,6 +512,13 @@ asmlinkage void __init start_kernel(void)
        console_init();
        if (panic_later)
                panic(panic_later, panic_param);
+       /*
+        * Need to run this when irqs are enabled, because it wants
+        * to self-test [hard/soft]-irqs on/off lock inversion bugs
+        * too:
+        */
+       locking_selftest();
+
 #ifdef CONFIG_BLK_DEV_INITRD
        if (initrd_start && !initrd_below_start_ok &&
                        initrd_start < min_low_pfn << PAGE_SHIFT) {
index 7f7ef22..c595db1 100644 (file)
@@ -933,10 +933,9 @@ fastcall NORET_TYPE void do_exit(long code)
        if (unlikely(current->pi_state_cache))
                kfree(current->pi_state_cache);
        /*
-        * If DEBUG_MUTEXES is on, make sure we are holding no locks:
+        * Make sure we are holding no locks:
         */
-       mutex_debug_check_no_locks_held(tsk);
-       rt_mutex_debug_check_no_locks_held(tsk);
+       debug_check_no_locks_held(tsk);
 
        if (tsk->io_context)
                exit_io_context();
index 9064bf9..1cd46a4 100644 (file)
@@ -919,10 +919,6 @@ static inline void rt_mutex_init_task(struct task_struct *p)
        spin_lock_init(&p->pi_lock);
        plist_head_init(&p->pi_waiters, &p->pi_lock);
        p->pi_blocked_on = NULL;
-# ifdef CONFIG_DEBUG_RT_MUTEXES
-       spin_lock_init(&p->held_list_lock);
-       INIT_LIST_HEAD(&p->held_list_head);
-# endif
 #endif
 }
 
index a92de14..5569766 100644 (file)
 #include <linux/spinlock.h>
 #include <linux/kallsyms.h>
 #include <linux/interrupt.h>
+#include <linux/debug_locks.h>
 
 #include "mutex-debug.h"
 
 /*
- * We need a global lock when we walk through the multi-process
- * lock tree. Only used in the deadlock-debugging case.
- */
-DEFINE_SPINLOCK(debug_mutex_lock);
-
-/*
- * All locks held by all tasks, in a single global list:
- */
-LIST_HEAD(debug_mutex_held_locks);
-
-/*
- * In the debug case we carry the caller's instruction pointer into
- * other functions, but we dont want the function argument overhead
- * in the nondebug case - hence these macros:
- */
-#define __IP_DECL__            , unsigned long ip
-#define __IP__                 , ip
-#define __RET_IP__             , (unsigned long)__builtin_return_address(0)
-
-/*
- * "mutex debugging enabled" flag. We turn it off when we detect
- * the first problem because we dont want to recurse back
- * into the tracing code when doing error printk or
- * executing a BUG():
- */
-int debug_mutex_on = 1;
-
-/*
  * Must be called with lock->wait_lock held.
  */
-void debug_mutex_set_owner(struct mutex *lock,
-                          struct thread_info *new_owner __IP_DECL__)
+void debug_mutex_set_owner(struct mutex *lock, struct thread_info *new_owner)
 {
        lock->owner = new_owner;
-       DEBUG_LOCKS_WARN_ON(!list_empty(&lock->held_list));
-       if (debug_mutex_on) {
-               list_add_tail(&lock->held_list, &debug_mutex_held_locks);
-               lock->acquire_ip = ip;
-       }
 }
 
-void debug_mutex_init_waiter(struct mutex_waiter *waiter)
+void debug_mutex_lock_common(struct mutex *lock, struct mutex_waiter *waiter)
 {
        memset(waiter, MUTEX_DEBUG_INIT, sizeof(*waiter));
        waiter->magic = waiter;
@@ -87,9 +54,10 @@ void debug_mutex_free_waiter(struct mutex_waiter *waiter)
 }
 
 void debug_mutex_add_waiter(struct mutex *lock, struct mutex_waiter *waiter,
-                           struct thread_info *ti __IP_DECL__)
+                           struct thread_info *ti)
 {
        SMP_DEBUG_LOCKS_WARN_ON(!spin_is_locked(&lock->wait_lock));
+
        /* Mark the current thread as blocked on the lock: */
        ti->task->blocked_on = waiter;
        waiter->lock = lock;
@@ -109,13 +77,10 @@ void mutex_remove_waiter(struct mutex *lock, struct mutex_waiter *waiter,
 
 void debug_mutex_unlock(struct mutex *lock)
 {
+       DEBUG_LOCKS_WARN_ON(lock->owner != current_thread_info());
        DEBUG_LOCKS_WARN_ON(lock->magic != lock);
        DEBUG_LOCKS_WARN_ON(!lock->wait_list.prev && !lock->wait_list.next);
        DEBUG_LOCKS_WARN_ON(lock->owner != current_thread_info());
-       if (debug_mutex_on) {
-               DEBUG_LOCKS_WARN_ON(list_empty(&lock->held_list));
-               list_del_init(&lock->held_list);
-       }
 }
 
 void debug_mutex_init(struct mutex *lock, const char *name)
@@ -123,10 +88,8 @@ void debug_mutex_init(struct mutex *lock, const char *name)
        /*
         * Make sure we are not reinitializing a held lock:
         */
-       mutex_debug_check_no_locks_freed((void *)lock, sizeof(*lock));
+       debug_check_no_locks_freed((void *)lock, sizeof(*lock));
        lock->owner = NULL;
-       INIT_LIST_HEAD(&lock->held_list);
-       lock->name = name;
        lock->magic = lock;
 }
 
index bdab13a..babfbdf 100644 (file)
  * More details are in kernel/mutex-debug.c.
  */
 
-extern spinlock_t debug_mutex_lock;
-extern struct list_head debug_mutex_held_locks;
-extern int debug_mutex_on;
-
-/*
- * In the debug case we carry the caller's instruction pointer into
- * other functions, but we dont want the function argument overhead
- * in the nondebug case - hence these macros:
- */
-#define __IP_DECL__            , unsigned long ip
-#define __IP__                 , ip
-#define __RET_IP__             , (unsigned long)__builtin_return_address(0)
-
 /*
  * This must be called with lock->wait_lock held.
  */
-extern void debug_mutex_set_owner(struct mutex *lock,
-                                 struct thread_info *new_owner __IP_DECL__);
+extern void
+debug_mutex_set_owner(struct mutex *lock, struct thread_info *new_owner);
 
 static inline void debug_mutex_clear_owner(struct mutex *lock)
 {
        lock->owner = NULL;
 }
 
-extern void debug_mutex_init_waiter(struct mutex_waiter *waiter);
+extern void debug_mutex_lock_common(struct mutex *lock,
+                                   struct mutex_waiter *waiter);
 extern void debug_mutex_wake_waiter(struct mutex *lock,
                                    struct mutex_waiter *waiter);
 extern void debug_mutex_free_waiter(struct mutex_waiter *waiter);
 extern void debug_mutex_add_waiter(struct mutex *lock,
                                   struct mutex_waiter *waiter,
-                                  struct thread_info *ti __IP_DECL__);
+                                  struct thread_info *ti);
 extern void mutex_remove_waiter(struct mutex *lock, struct mutex_waiter *waiter,
                                struct thread_info *ti);
 extern void debug_mutex_unlock(struct mutex *lock);
-extern void debug_mutex_init(struct mutex *lock, const char *name);
-
-#define debug_spin_lock_save(lock, flags)              \
-       do {                                            \
-               local_irq_save(flags);                  \
-               if (debug_mutex_on)                     \
-                       spin_lock(lock);                \
-       } while (0)
-
-#define debug_spin_unlock_restore(lock, flags)         \
-       do {                                            \
-               if (debug_mutex_on)                     \
-                       spin_unlock(lock);              \
-               local_irq_restore(flags);               \
-               preempt_check_resched();                \
-       } while (0)
+extern void debug_mutex_init(struct mutex *lock, const char *name,
+                            struct lock_class_key *key);
 
 #define spin_lock_mutex(lock, flags)                   \
        do {                                            \
                struct mutex *l = container_of(lock, struct mutex, wait_lock); \
                                                        \
                DEBUG_LOCKS_WARN_ON(in_interrupt());    \
-               debug_spin_lock_save(&debug_mutex_lock, flags); \
-               spin_lock(lock);                        \
+               local_irq_save(flags);                  \
+               __raw_spin_lock(&(lock)->raw_lock);     \
                DEBUG_LOCKS_WARN_ON(l->magic != l);     \
        } while (0)
 
 #define spin_unlock_mutex(lock, flags)                 \
        do {                                            \
-               spin_unlock(lock);                      \
-               debug_spin_unlock_restore(&debug_mutex_lock, flags);    \
+               __raw_spin_unlock(&(lock)->raw_lock);   \
+               local_irq_restore(flags);               \
+               preempt_check_resched();                \
        } while (0)
-
-#define DEBUG_OFF()                                    \
-do {                                                   \
-       if (debug_mutex_on) {                           \
-               debug_mutex_on = 0;                     \
-               console_verbose();                      \
-               if (spin_is_locked(&debug_mutex_lock))  \
-                       spin_unlock(&debug_mutex_lock); \
-       }                                               \
-} while (0)
-
-#define DEBUG_BUG()                                    \
-do {                                                   \
-       if (debug_mutex_on) {                           \
-               DEBUG_OFF();                            \
-               BUG();                                  \
-       }                                               \
-} while (0)
-
-#define DEBUG_LOCKS_WARN_ON(c)                         \
-do {                                                   \
-       if (unlikely(c && debug_mutex_on)) {            \
-               DEBUG_OFF();                            \
-               WARN_ON(1);                             \
-       }                                               \
-} while (0)
-
-#ifdef CONFIG_SMP
-# define SMP_DEBUG_LOCKS_WARN_ON(c)                    DEBUG_LOCKS_WARN_ON(c)
-#else
-# define SMP_DEBUG_LOCKS_WARN_ON(c)                    do { } while (0)
-#endif
-
index 101ceeb..3aad0b7 100644 (file)
@@ -17,6 +17,7 @@
 #include <linux/module.h>
 #include <linux/spinlock.h>
 #include <linux/interrupt.h>
+#include <linux/debug_locks.h>
 
 /*
  * In the DEBUG case we are using the "NULL fastpath" for mutexes,
@@ -38,7 +39,7 @@
  *
  * It is not allowed to initialize an already locked mutex.
  */
-void fastcall __mutex_init(struct mutex *lock, const char *name)
+__always_inline void fastcall __mutex_init(struct mutex *lock, const char *name)
 {
        atomic_set(&lock->count, 1);
        spin_lock_init(&lock->wait_lock);
@@ -56,7 +57,7 @@ EXPORT_SYMBOL(__mutex_init);
  * branch is predicted by the CPU as default-untaken.
  */
 static void fastcall noinline __sched
-__mutex_lock_slowpath(atomic_t *lock_count __IP_DECL__);
+__mutex_lock_slowpath(atomic_t *lock_count);
 
 /***
  * mutex_lock - acquire the mutex
@@ -79,7 +80,7 @@ __mutex_lock_slowpath(atomic_t *lock_count __IP_DECL__);
  *
  * This function is similar to (but not equivalent to) down().
  */
-void fastcall __sched mutex_lock(struct mutex *lock)
+void inline fastcall __sched mutex_lock(struct mutex *lock)
 {
        might_sleep();
        /*
@@ -92,7 +93,7 @@ void fastcall __sched mutex_lock(struct mutex *lock)
 EXPORT_SYMBOL(mutex_lock);
 
 static void fastcall noinline __sched
-__mutex_unlock_slowpath(atomic_t *lock_count __IP_DECL__);
+__mutex_unlock_slowpath(atomic_t *lock_count);
 
 /***
  * mutex_unlock - release the mutex
@@ -120,18 +121,17 @@ EXPORT_SYMBOL(mutex_unlock);
  * Lock a mutex (possibly interruptible), slowpath:
  */
 static inline int __sched
-__mutex_lock_common(struct mutex *lock, long state __IP_DECL__)
+__mutex_lock_common(struct mutex *lock, long state, unsigned int subclass)
 {
        struct task_struct *task = current;
        struct mutex_waiter waiter;
        unsigned int old_val;
        unsigned long flags;
 
-       debug_mutex_init_waiter(&waiter);
-
        spin_lock_mutex(&lock->wait_lock, flags);
 
-       debug_mutex_add_waiter(lock, &waiter, task->thread_info, ip);
+       debug_mutex_lock_common(lock, &waiter);
+       debug_mutex_add_waiter(lock, &waiter, task->thread_info);
 
        /* add waiting tasks to the end of the waitqueue (FIFO): */
        list_add_tail(&waiter.list, &lock->wait_list);
@@ -173,7 +173,7 @@ __mutex_lock_common(struct mutex *lock, long state __IP_DECL__)
 
        /* got the lock - rejoice! */
        mutex_remove_waiter(lock, &waiter, task->thread_info);
-       debug_mutex_set_owner(lock, task->thread_info __IP__);
+       debug_mutex_set_owner(lock, task->thread_info);
 
        /* set it to 0 if there are no waiters left: */
        if (likely(list_empty(&lock->wait_list)))
@@ -183,32 +183,28 @@ __mutex_lock_common(struct mutex *lock, long state __IP_DECL__)
 
        debug_mutex_free_waiter(&waiter);
 
-       DEBUG_LOCKS_WARN_ON(list_empty(&lock->held_list));
-       DEBUG_LOCKS_WARN_ON(lock->owner != task->thread_info);
-
        return 0;
 }
 
 static void fastcall noinline __sched
-__mutex_lock_slowpath(atomic_t *lock_count __IP_DECL__)
+__mutex_lock_slowpath(atomic_t *lock_count)
 {
        struct mutex *lock = container_of(lock_count, struct mutex, count);
 
-       __mutex_lock_common(lock, TASK_UNINTERRUPTIBLE __IP__);
+       __mutex_lock_common(lock, TASK_UNINTERRUPTIBLE, 0);
 }
 
 /*
  * Release the lock, slowpath:
  */
-static fastcall noinline void
-__mutex_unlock_slowpath(atomic_t *lock_count __IP_DECL__)
+static fastcall inline void
+__mutex_unlock_common_slowpath(atomic_t *lock_count)
 {
        struct mutex *lock = container_of(lock_count, struct mutex, count);
        unsigned long flags;
 
-       DEBUG_LOCKS_WARN_ON(lock->owner != current_thread_info());
-
        spin_lock_mutex(&lock->wait_lock, flags);
+       debug_mutex_unlock(lock);
 
        /*
         * some architectures leave the lock unlocked in the fastpath failure
@@ -218,8 +214,6 @@ __mutex_unlock_slowpath(atomic_t *lock_count __IP_DECL__)
        if (__mutex_slowpath_needs_to_unlock())
                atomic_set(&lock->count, 1);
 
-       debug_mutex_unlock(lock);
-
        if (!list_empty(&lock->wait_list)) {
                /* get the first entry from the wait-list: */
                struct mutex_waiter *waiter =
@@ -237,11 +231,20 @@ __mutex_unlock_slowpath(atomic_t *lock_count __IP_DECL__)
 }
 
 /*
+ * Release the lock, slowpath:
+ */
+static fastcall noinline void
+__mutex_unlock_slowpath(atomic_t *lock_count)
+{
+       __mutex_unlock_common_slowpath(lock_count);
+}
+
+/*
  * Here come the less common (and hence less performance-critical) APIs:
  * mutex_lock_interruptible() and mutex_trylock().
  */
 static int fastcall noinline __sched
-__mutex_lock_interruptible_slowpath(atomic_t *lock_count __IP_DECL__);
+__mutex_lock_interruptible_slowpath(atomic_t *lock_count);
 
 /***
  * mutex_lock_interruptible - acquire the mutex, interruptable
@@ -264,11 +267,11 @@ int fastcall __sched mutex_lock_interruptible(struct mutex *lock)
 EXPORT_SYMBOL(mutex_lock_interruptible);
 
 static int fastcall noinline __sched
-__mutex_lock_interruptible_slowpath(atomic_t *lock_count __IP_DECL__)
+__mutex_lock_interruptible_slowpath(atomic_t *lock_count)
 {
        struct mutex *lock = container_of(lock_count, struct mutex, count);
 
-       return __mutex_lock_common(lock, TASK_INTERRUPTIBLE __IP__);
+       return __mutex_lock_common(lock, TASK_INTERRUPTIBLE, 0);
 }
 
 /*
@@ -285,7 +288,8 @@ static inline int __mutex_trylock_slowpath(atomic_t *lock_count)
 
        prev = atomic_xchg(&lock->count, -1);
        if (likely(prev == 1))
-               debug_mutex_set_owner(lock, current_thread_info() __RET_IP__);
+               debug_mutex_set_owner(lock, current_thread_info());
+
        /* Set it back to 0 if there are no waiters: */
        if (likely(list_empty(&lock->wait_list)))
                atomic_set(&lock->count, 0);
index 7e1ed48..aeb2d91 100644 (file)
 #define mutex_remove_waiter(lock, waiter, ti) \
                __list_del((waiter)->list.prev, (waiter)->list.next)
 
-#define DEBUG_LOCKS_WARN_ON(c)                         do { } while (0)
 #define debug_mutex_set_owner(lock, new_owner)         do { } while (0)
 #define debug_mutex_clear_owner(lock)                  do { } while (0)
-#define debug_mutex_init_waiter(waiter)                        do { } while (0)
 #define debug_mutex_wake_waiter(lock, waiter)          do { } while (0)
 #define debug_mutex_free_waiter(waiter)                        do { } while (0)
-#define debug_mutex_add_waiter(lock, waiter, ti, ip)   do { } while (0)
+#define debug_mutex_add_waiter(lock, waiter, ti)       do { } while (0)
 #define debug_mutex_unlock(lock)                       do { } while (0)
 #define debug_mutex_init(lock, name)                   do { } while (0)
 
-/*
- * Return-address parameters/declarations. They are very useful for
- * debugging, but add overhead in the !DEBUG case - so we go the
- * trouble of using this not too elegant but zero-cost solution:
- */
-#define __IP_DECL__
-#define __IP__
-#define __RET_IP__
-
+static inline void
+debug_mutex_lock_common(struct mutex *lock, struct mutex_waiter *waiter)
+{
+}
index 4aa8a2c..353a853 100644 (file)
@@ -26,6 +26,7 @@
 #include <linux/interrupt.h>
 #include <linux/plist.h>
 #include <linux/fs.h>
+#include <linux/debug_locks.h>
 
 #include "rtmutex_common.h"
 
@@ -45,8 +46,6 @@ do {                                                          \
                console_verbose();                              \
                if (spin_is_locked(&current->pi_lock))          \
                        spin_unlock(&current->pi_lock);         \
-               if (spin_is_locked(&current->held_list_lock))   \
-                       spin_unlock(&current->held_list_lock);  \
        }                                                       \
 } while (0)
 
@@ -105,14 +104,6 @@ static void printk_task(task_t *p)
                printk("<none>");
 }
 
-static void printk_task_short(task_t *p)
-{
-       if (p)
-               printk("%s/%d [%p, %3d]", p->comm, p->pid, p, p->prio);
-       else
-               printk("<none>");
-}
-
 static void printk_lock(struct rt_mutex *lock, int print_owner)
 {
        if (lock->name)
@@ -128,222 +119,6 @@ static void printk_lock(struct rt_mutex *lock, int print_owner)
                printk_task(rt_mutex_owner(lock));
                printk("\n");
        }
-       if (rt_mutex_owner(lock)) {
-               printk("... acquired at:               ");
-               print_symbol("%s\n", lock->acquire_ip);
-       }
-}
-
-static void printk_waiter(struct rt_mutex_waiter *w)
-{
-       printk("-------------------------\n");
-       printk("| waiter struct %p:\n", w);
-       printk("| w->list_entry: [DP:%p/%p|SP:%p/%p|PRI:%d]\n",
-              w->list_entry.plist.prio_list.prev, w->list_entry.plist.prio_list.next,
-              w->list_entry.plist.node_list.prev, w->list_entry.plist.node_list.next,
-              w->list_entry.prio);
-       printk("| w->pi_list_entry: [DP:%p/%p|SP:%p/%p|PRI:%d]\n",
-              w->pi_list_entry.plist.prio_list.prev, w->pi_list_entry.plist.prio_list.next,
-              w->pi_list_entry.plist.node_list.prev, w->pi_list_entry.plist.node_list.next,
-              w->pi_list_entry.prio);
-       printk("\n| lock:\n");
-       printk_lock(w->lock, 1);
-       printk("| w->ti->task:\n");
-       printk_task(w->task);
-       printk("| blocked at:  ");
-       print_symbol("%s\n", w->ip);
-       printk("-------------------------\n");
-}
-
-static void show_task_locks(task_t *p)
-{
-       switch (p->state) {
-       case TASK_RUNNING:              printk("R"); break;
-       case TASK_INTERRUPTIBLE:        printk("S"); break;
-       case TASK_UNINTERRUPTIBLE:      printk("D"); break;
-       case TASK_STOPPED:              printk("T"); break;
-       case EXIT_ZOMBIE:               printk("Z"); break;
-       case EXIT_DEAD:                 printk("X"); break;
-       default:                        printk("?"); break;
-       }
-       printk_task(p);
-       if (p->pi_blocked_on) {
-               struct rt_mutex *lock = p->pi_blocked_on->lock;
-
-               printk(" blocked on:");
-               printk_lock(lock, 1);
-       } else
-               printk(" (not blocked)\n");
-}
-
-void rt_mutex_show_held_locks(task_t *task, int verbose)
-{
-       struct list_head *curr, *cursor = NULL;
-       struct rt_mutex *lock;
-       task_t *t;
-       unsigned long flags;
-       int count = 0;
-
-       if (!rt_trace_on)
-               return;
-
-       if (verbose) {
-               printk("------------------------------\n");
-               printk("| showing all locks held by: |  (");
-               printk_task_short(task);
-               printk("):\n");
-               printk("------------------------------\n");
-       }
-
-next:
-       spin_lock_irqsave(&task->held_list_lock, flags);
-       list_for_each(curr, &task->held_list_head) {
-               if (cursor && curr != cursor)
-                       continue;
-               lock = list_entry(curr, struct rt_mutex, held_list_entry);
-               t = rt_mutex_owner(lock);
-               WARN_ON(t != task);
-               count++;
-               cursor = curr->next;
-               spin_unlock_irqrestore(&task->held_list_lock, flags);
-
-               printk("\n#%03d:            ", count);
-               printk_lock(lock, 0);
-               goto next;
-       }
-       spin_unlock_irqrestore(&task->held_list_lock, flags);
-
-       printk("\n");
-}
-
-void rt_mutex_show_all_locks(void)
-{
-       task_t *g, *p;
-       int count = 10;
-       int unlock = 1;
-
-       printk("\n");
-       printk("----------------------\n");
-       printk("| showing all tasks: |\n");
-       printk("----------------------\n");
-
-       /*
-        * Here we try to get the tasklist_lock as hard as possible,
-        * if not successful after 2 seconds we ignore it (but keep
-        * trying). This is to enable a debug printout even if a
-        * tasklist_lock-holding task deadlocks or crashes.
-        */
-retry:
-       if (!read_trylock(&tasklist_lock)) {
-               if (count == 10)
-                       printk("hm, tasklist_lock locked, retrying... ");
-               if (count) {
-                       count--;
-                       printk(" #%d", 10-count);
-                       mdelay(200);
-                       goto retry;
-               }
-               printk(" ignoring it.\n");
-               unlock = 0;
-       }
-       if (count != 10)
-               printk(" locked it.\n");
-
-       do_each_thread(g, p) {
-               show_task_locks(p);
-               if (!unlock)
-                       if (read_trylock(&tasklist_lock))
-                               unlock = 1;
-       } while_each_thread(g, p);
-
-       printk("\n");
-
-       printk("-----------------------------------------\n");
-       printk("| showing all locks held in the system: |\n");
-       printk("-----------------------------------------\n");
-
-       do_each_thread(g, p) {
-               rt_mutex_show_held_locks(p, 0);
-               if (!unlock)
-                       if (read_trylock(&tasklist_lock))
-                               unlock = 1;
-       } while_each_thread(g, p);
-
-
-       printk("=============================================\n\n");
-
-       if (unlock)
-               read_unlock(&tasklist_lock);
-}
-
-void rt_mutex_debug_check_no_locks_held(task_t *task)
-{
-       struct rt_mutex_waiter *w;
-       struct list_head *curr;
-       struct rt_mutex *lock;
-
-       if (!rt_trace_on)
-               return;
-       if (!rt_prio(task->normal_prio) && rt_prio(task->prio)) {
-               printk("BUG: PI priority boost leaked!\n");
-               printk_task(task);
-               printk("\n");
-       }
-       if (list_empty(&task->held_list_head))
-               return;
-
-       spin_lock(&task->pi_lock);
-       plist_for_each_entry(w, &task->pi_waiters, pi_list_entry) {
-               TRACE_OFF();
-
-               printk("hm, PI interest held at exit time? Task:\n");
-               printk_task(task);
-               printk_waiter(w);
-               return;
-       }
-       spin_unlock(&task->pi_lock);
-
-       list_for_each(curr, &task->held_list_head) {
-               lock = list_entry(curr, struct rt_mutex, held_list_entry);
-
-               printk("BUG: %s/%d, lock held at task exit time!\n",
-                      task->comm, task->pid);
-               printk_lock(lock, 1);
-               if (rt_mutex_owner(lock) != task)
-                       printk("exiting task is not even the owner??\n");
-       }
-}
-
-int rt_mutex_debug_check_no_locks_freed(const void *from, unsigned long len)
-{
-       const void *to = from + len;
-       struct list_head *curr;
-       struct rt_mutex *lock;
-       unsigned long flags;
-       void *lock_addr;
-
-       if (!rt_trace_on)
-               return 0;
-
-       spin_lock_irqsave(&current->held_list_lock, flags);
-       list_for_each(curr, &current->held_list_head) {
-               lock = list_entry(curr, struct rt_mutex, held_list_entry);
-               lock_addr = lock;
-               if (lock_addr < from || lock_addr >= to)
-                       continue;
-               TRACE_OFF();
-
-               printk("BUG: %s/%d, active lock [%p(%p-%p)] freed!\n",
-                       current->comm, current->pid, lock, from, to);
-               dump_stack();
-               printk_lock(lock, 1);
-               if (rt_mutex_owner(lock) != current)
-                       printk("freeing task is not even the owner??\n");
-               return 1;
-       }
-       spin_unlock_irqrestore(&current->held_list_lock, flags);
-
-       return 0;
 }
 
 void rt_mutex_debug_task_free(struct task_struct *task)
@@ -395,85 +170,41 @@ void debug_rt_mutex_print_deadlock(struct rt_mutex_waiter *waiter)
               current->comm, current->pid);
        printk_lock(waiter->lock, 1);
 
-       printk("... trying at:                 ");
-       print_symbol("%s\n", waiter->ip);
-
        printk("\n2) %s/%d is blocked on this lock:\n", task->comm, task->pid);
        printk_lock(waiter->deadlock_lock, 1);
 
-       rt_mutex_show_held_locks(current, 1);
-       rt_mutex_show_held_locks(task, 1);
+       debug_show_held_locks(current);
+       debug_show_held_locks(task);
 
        printk("\n%s/%d's [blocked] stackdump:\n\n", task->comm, task->pid);
        show_stack(task, NULL);
        printk("\n%s/%d's [current] stackdump:\n\n",
               current->comm, current->pid);
        dump_stack();
-       rt_mutex_show_all_locks();
+       debug_show_all_locks();
+
        printk("[ turning off deadlock detection."
               "Please report this trace. ]\n\n");
        local_irq_disable();
 }
 
-void debug_rt_mutex_lock(struct rt_mutex *lock __IP_DECL__)
+void debug_rt_mutex_lock(struct rt_mutex *lock)
 {
-       unsigned long flags;
-
-       if (rt_trace_on) {
-               TRACE_WARN_ON_LOCKED(!list_empty(&lock->held_list_entry));
-
-               spin_lock_irqsave(&current->held_list_lock, flags);
-               list_add_tail(&lock->held_list_entry, &current->held_list_head);
-               spin_unlock_irqrestore(&current->held_list_lock, flags);
-
-               lock->acquire_ip = ip;
-       }
 }
 
 void debug_rt_mutex_unlock(struct rt_mutex *lock)
 {
-       unsigned long flags;
-
-       if (rt_trace_on) {
-               TRACE_WARN_ON_LOCKED(rt_mutex_owner(lock) != current);
-               TRACE_WARN_ON_LOCKED(list_empty(&lock->held_list_entry));
-
-               spin_lock_irqsave(&current->held_list_lock, flags);
-               list_del_init(&lock->held_list_entry);
-               spin_unlock_irqrestore(&current->held_list_lock, flags);
-       }
+       TRACE_WARN_ON_LOCKED(rt_mutex_owner(lock) != current);
 }
 
-void debug_rt_mutex_proxy_lock(struct rt_mutex *lock,
-                              struct task_struct *powner __IP_DECL__)
+void
+debug_rt_mutex_proxy_lock(struct rt_mutex *lock, struct task_struct *powner)
 {
-       unsigned long flags;
-
-       if (rt_trace_on) {
-               TRACE_WARN_ON_LOCKED(!list_empty(&lock->held_list_entry));
-
-               spin_lock_irqsave(&powner->held_list_lock, flags);
-               list_add_tail(&lock->held_list_entry, &powner->held_list_head);
-               spin_unlock_irqrestore(&powner->held_list_lock, flags);
-
-               lock->acquire_ip = ip;
-       }
 }
 
 void debug_rt_mutex_proxy_unlock(struct rt_mutex *lock)
 {
-       unsigned long flags;
-
-       if (rt_trace_on) {
-               struct task_struct *owner = rt_mutex_owner(lock);
-
-               TRACE_WARN_ON_LOCKED(!owner);
-               TRACE_WARN_ON_LOCKED(list_empty(&lock->held_list_entry));
-
-               spin_lock_irqsave(&owner->held_list_lock, flags);
-               list_del_init(&lock->held_list_entry);
-               spin_unlock_irqrestore(&owner->held_list_lock, flags);
-       }
+       TRACE_WARN_ON_LOCKED(!rt_mutex_owner(lock));
 }
 
 void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter)
@@ -493,14 +224,11 @@ void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter)
 
 void debug_rt_mutex_init(struct rt_mutex *lock, const char *name)
 {
-       void *addr = lock;
-
-       if (rt_trace_on) {
-               rt_mutex_debug_check_no_locks_freed(addr,
-                                                   sizeof(struct rt_mutex));
-               INIT_LIST_HEAD(&lock->held_list_entry);
-               lock->name = name;
-       }
+       /*
+        * Make sure we are not reinitializing a held lock:
+        */
+       debug_check_no_locks_freed((void *)lock, sizeof(*lock));
+       lock->name = name;
 }
 
 void rt_mutex_deadlock_account_lock(struct rt_mutex *lock, task_t *task)
index 7612fbc..14193d5 100644 (file)
@@ -9,20 +9,16 @@
  * This file contains macros used solely by rtmutex.c. Debug version.
  */
 
-#define __IP_DECL__            , unsigned long ip
-#define __IP__                 , ip
-#define __RET_IP__             , (unsigned long)__builtin_return_address(0)
-
 extern void
 rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task);
 extern void rt_mutex_deadlock_account_unlock(struct task_struct *task);
 extern void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter);
 extern void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter);
 extern void debug_rt_mutex_init(struct rt_mutex *lock, const char *name);
-extern void debug_rt_mutex_lock(struct rt_mutex *lock __IP_DECL__);
+extern void debug_rt_mutex_lock(struct rt_mutex *lock);
 extern void debug_rt_mutex_unlock(struct rt_mutex *lock);
 extern void debug_rt_mutex_proxy_lock(struct rt_mutex *lock,
-                                     struct task_struct *powner __IP_DECL__);
+                                     struct task_struct *powner);
 extern void debug_rt_mutex_proxy_unlock(struct rt_mutex *lock);
 extern void debug_rt_mutex_deadlock(int detect, struct rt_mutex_waiter *waiter,
                                    struct rt_mutex *lock);
index 45d6101..91b699a 100644 (file)
@@ -161,8 +161,7 @@ static int rt_mutex_adjust_prio_chain(task_t *task,
                                      int deadlock_detect,
                                      struct rt_mutex *orig_lock,
                                      struct rt_mutex_waiter *orig_waiter,
-                                     struct task_struct *top_task
-                                     __IP_DECL__)
+                                     struct task_struct *top_task)
 {
        struct rt_mutex *lock;
        struct rt_mutex_waiter *waiter, *top_waiter = orig_waiter;
@@ -357,7 +356,7 @@ static inline int try_to_steal_lock(struct rt_mutex *lock)
  *
  * Must be called with lock->wait_lock held.
  */
-static int try_to_take_rt_mutex(struct rt_mutex *lock __IP_DECL__)
+static int try_to_take_rt_mutex(struct rt_mutex *lock)
 {
        /*
         * We have to be careful here if the atomic speedups are
@@ -384,7 +383,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock __IP_DECL__)
                return 0;
 
        /* We got the lock. */
-       debug_rt_mutex_lock(lock __IP__);
+       debug_rt_mutex_lock(lock);
 
        rt_mutex_set_owner(lock, current, 0);
 
@@ -402,8 +401,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock __IP_DECL__)
  */
 static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
                                   struct rt_mutex_waiter *waiter,
-                                  int detect_deadlock
-                                  __IP_DECL__)
+                                  int detect_deadlock)
 {
        struct rt_mutex_waiter *top_waiter = waiter;
        task_t *owner = rt_mutex_owner(lock);
@@ -454,7 +452,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
        spin_unlock(&lock->wait_lock);
 
        res = rt_mutex_adjust_prio_chain(owner, detect_deadlock, lock, waiter,
-                                        current __IP__);
+                                        current);
 
        spin_lock(&lock->wait_lock);
 
@@ -526,7 +524,7 @@ static void wakeup_next_waiter(struct rt_mutex *lock)
  * Must be called with lock->wait_lock held
  */
 static void remove_waiter(struct rt_mutex *lock,
-                         struct rt_mutex_waiter *waiter  __IP_DECL__)
+                         struct rt_mutex_waiter *waiter)
 {
        int first = (waiter == rt_mutex_top_waiter(lock));
        int boost = 0;
@@ -568,7 +566,7 @@ static void remove_waiter(struct rt_mutex *lock,
 
        spin_unlock(&lock->wait_lock);
 
-       rt_mutex_adjust_prio_chain(owner, 0, lock, NULL, current __IP__);
+       rt_mutex_adjust_prio_chain(owner, 0, lock, NULL, current);
 
        spin_lock(&lock->wait_lock);
 }
@@ -595,7 +593,7 @@ void rt_mutex_adjust_pi(struct task_struct *task)
        get_task_struct(task);
        spin_unlock_irqrestore(&task->pi_lock, flags);
 
-       rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task __RET_IP__);
+       rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task);
 }
 
 /*
@@ -604,7 +602,7 @@ void rt_mutex_adjust_pi(struct task_struct *task)
 static int __sched
 rt_mutex_slowlock(struct rt_mutex *lock, int state,
                  struct hrtimer_sleeper *timeout,
-                 int detect_deadlock __IP_DECL__)
+                 int detect_deadlock)
 {
        struct rt_mutex_waiter waiter;
        int ret = 0;
@@ -615,7 +613,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
        spin_lock(&lock->wait_lock);
 
        /* Try to acquire the lock again: */
-       if (try_to_take_rt_mutex(lock __IP__)) {
+       if (try_to_take_rt_mutex(lock)) {
                spin_unlock(&lock->wait_lock);
                return 0;
        }
@@ -629,7 +627,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
 
        for (;;) {
                /* Try to acquire the lock: */
-               if (try_to_take_rt_mutex(lock __IP__))
+               if (try_to_take_rt_mutex(lock))
                        break;
 
                /*
@@ -653,7 +651,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
                 */
                if (!waiter.task) {
                        ret = task_blocks_on_rt_mutex(lock, &waiter,
-                                                     detect_deadlock __IP__);
+                                                     detect_deadlock);
                        /*
                         * If we got woken up by the owner then start loop
                         * all over without going into schedule to try
@@ -680,7 +678,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
        set_current_state(TASK_RUNNING);
 
        if (unlikely(waiter.task))
-               remove_waiter(lock, &waiter __IP__);
+               remove_waiter(lock, &waiter);
 
        /*
         * try_to_take_rt_mutex() sets the waiter bit
@@ -711,7 +709,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
  * Slow path try-lock function:
  */
 static inline int
-rt_mutex_slowtrylock(struct rt_mutex *lock __IP_DECL__)
+rt_mutex_slowtrylock(struct rt_mutex *lock)
 {
        int ret = 0;
 
@@ -719,7 +717,7 @@ rt_mutex_slowtrylock(struct rt_mutex *lock __IP_DECL__)
 
        if (likely(rt_mutex_owner(lock) != current)) {
 
-               ret = try_to_take_rt_mutex(lock __IP__);
+               ret = try_to_take_rt_mutex(lock);
                /*
                 * try_to_take_rt_mutex() sets the lock waiters
                 * bit unconditionally. Clean this up.
@@ -769,13 +767,13 @@ rt_mutex_fastlock(struct rt_mutex *lock, int state,
                  int detect_deadlock,
                  int (*slowfn)(struct rt_mutex *lock, int state,
                                struct hrtimer_sleeper *timeout,
-                               int detect_deadlock __IP_DECL__))
+                               int detect_deadlock))
 {
        if (!detect_deadlock && likely(rt_mutex_cmpxchg(lock, NULL, current))) {
                rt_mutex_deadlock_account_lock(lock, current);
                return 0;
        } else
-               return slowfn(lock, state, NULL, detect_deadlock __RET_IP__);
+               return slowfn(lock, state, NULL, detect_deadlock);
 }
 
 static inline int
@@ -783,24 +781,24 @@ rt_mutex_timed_fastlock(struct rt_mutex *lock, int state,
                        struct hrtimer_sleeper *timeout, int detect_deadlock,
                        int (*slowfn)(struct rt_mutex *lock, int state,
                                      struct hrtimer_sleeper *timeout,
-                                     int detect_deadlock __IP_DECL__))
+                                     int detect_deadlock))
 {
        if (!detect_deadlock && likely(rt_mutex_cmpxchg(lock, NULL, current))) {
                rt_mutex_deadlock_account_lock(lock, current);
                return 0;
        } else
-               return slowfn(lock, state, timeout, detect_deadlock __RET_IP__);
+               return slowfn(lock, state, timeout, detect_deadlock);
 }
 
 static inline int
 rt_mutex_fasttrylock(struct rt_mutex *lock,
-                    int (*slowfn)(struct rt_mutex *lock __IP_DECL__))
+                    int (*slowfn)(struct rt_mutex *lock))
 {
        if (likely(rt_mutex_cmpxchg(lock, NULL, current))) {
                rt_mutex_deadlock_account_lock(lock, current);
                return 1;
        }
-       return slowfn(lock __RET_IP__);
+       return slowfn(lock);
 }
 
 static inline void
@@ -948,7 +946,7 @@ void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
                                struct task_struct *proxy_owner)
 {
        __rt_mutex_init(lock, NULL);
-       debug_rt_mutex_proxy_lock(lock, proxy_owner __RET_IP__);
+       debug_rt_mutex_proxy_lock(lock, proxy_owner);
        rt_mutex_set_owner(lock, proxy_owner, 0);
        rt_mutex_deadlock_account_lock(lock, proxy_owner);
 }
index 1e0fca1..a1a1dd0 100644 (file)
@@ -10,9 +10,6 @@
  * Non-debug version.
  */
 
-#define __IP_DECL__
-#define __IP__
-#define __RET_IP__
 #define rt_mutex_deadlock_check(l)                     (0)
 #define rt_mutex_deadlock_account_lock(m, t)           do { } while (0)
 #define rt_mutex_deadlock_account_unlock(l)            do { } while (0)
index d5e3707..48c1faa 100644 (file)
@@ -30,6 +30,7 @@
 #include <linux/capability.h>
 #include <linux/completion.h>
 #include <linux/kernel_stat.h>
+#include <linux/debug_locks.h>
 #include <linux/security.h>
 #include <linux/notifier.h>
 #include <linux/profile.h>
@@ -3142,12 +3143,13 @@ void fastcall add_preempt_count(int val)
        /*
         * Underflow?
         */
-       BUG_ON((preempt_count() < 0));
+       if (DEBUG_LOCKS_WARN_ON((preempt_count() < 0)))
+               return;
        preempt_count() += val;
        /*
         * Spinlock count overflowing soon?
         */
-       BUG_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK-10);
+       DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK-10);
 }
 EXPORT_SYMBOL(add_preempt_count);
 
@@ -3156,11 +3158,15 @@ void fastcall sub_preempt_count(int val)
        /*
         * Underflow?
         */
-       BUG_ON(val > preempt_count());
+       if (DEBUG_LOCKS_WARN_ON(val > preempt_count()))
+               return;
        /*
         * Is the spinlock portion underflowing?
         */
-       BUG_ON((val < PREEMPT_MASK) && !(preempt_count() & PREEMPT_MASK));
+       if (DEBUG_LOCKS_WARN_ON((val < PREEMPT_MASK) &&
+                       !(preempt_count() & PREEMPT_MASK)))
+               return;
+
        preempt_count() -= val;
 }
 EXPORT_SYMBOL(sub_preempt_count);
@@ -4690,7 +4696,7 @@ void show_state(void)
        } while_each_thread(g, p);
 
        read_unlock(&tasklist_lock);
-       mutex_debug_show_all_locks();
+       debug_show_all_locks();
 }
 
 /**
index 10c13c9..4f5d019 100644 (file)
@@ -11,7 +11,7 @@ lib-$(CONFIG_SMP) += cpumask.o
 
 lib-y  += kobject.o kref.o kobject_uevent.o klist.o
 
-obj-y += sort.o parser.o halfmd4.o iomap_copy.o
+obj-y += sort.o parser.o halfmd4.o iomap_copy.o debug_locks.o
 
 ifeq ($(CONFIG_DEBUG_KOBJECT),y)
 CFLAGS_kobject.o += -DDEBUG
diff --git a/lib/debug_locks.c b/lib/debug_locks.c
new file mode 100644 (file)
index 0000000..0ef01d1
--- /dev/null
@@ -0,0 +1,45 @@
+/*
+ * lib/debug_locks.c
+ *
+ * Generic place for common debugging facilities for various locks:
+ * spinlocks, rwlocks, mutexes and rwsems.
+ *
+ * Started by Ingo Molnar:
+ *
+ *  Copyright (C) 2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
+ */
+#include <linux/rwsem.h>
+#include <linux/mutex.h>
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/debug_locks.h>
+
+/*
+ * We want to turn all lock-debugging facilities on/off at once,
+ * via a global flag. The reason is that once a single bug has been
+ * detected and reported, there might be cascade of followup bugs
+ * that would just muddy the log. So we report the first one and
+ * shut up after that.
+ */
+int debug_locks = 1;
+
+/*
+ * The locking-testsuite uses <debug_locks_silent> to get a
+ * 'silent failure': nothing is printed to the console when
+ * a locking bug is detected.
+ */
+int debug_locks_silent;
+
+/*
+ * Generic 'turn off all lock debugging' function:
+ */
+int debug_locks_off(void)
+{
+       if (xchg(&debug_locks, 0)) {
+               if (!debug_locks_silent) {
+                       console_verbose();
+                       return 1;
+               }
+       }
+       return 0;
+}
index 93c15ee..3de2ccf 100644 (file)
@@ -8,38 +8,35 @@
 
 #include <linux/spinlock.h>
 #include <linux/interrupt.h>
+#include <linux/debug_locks.h>
 #include <linux/delay.h>
+#include <linux/module.h>
 
 static void spin_bug(spinlock_t *lock, const char *msg)
 {
-       static long print_once = 1;
        struct task_struct *owner = NULL;
 
-       if (xchg(&print_once, 0)) {
-               if (lock->owner && lock->owner != SPINLOCK_OWNER_INIT)
-                       owner = lock->owner;
-               printk(KERN_EMERG "BUG: spinlock %s on CPU#%d, %s/%d\n",
-                       msg, raw_smp_processor_id(),
-                       current->comm, current->pid);
-               printk(KERN_EMERG " lock: %p, .magic: %08x, .owner: %s/%d, "
-                               ".owner_cpu: %d\n",
-                       lock, lock->magic,
-                       owner ? owner->comm : "<none>",
-                       owner ? owner->pid : -1,
-                       lock->owner_cpu);
-               dump_stack();
-#ifdef CONFIG_SMP
-               /*
-                * We cannot continue on SMP:
-                */
-//             panic("bad locking");
-#endif
-       }
+       if (!debug_locks_off())
+               return;
+
+       if (lock->owner && lock->owner != SPINLOCK_OWNER_INIT)
+               owner = lock->owner;
+       printk(KERN_EMERG "BUG: spinlock %s on CPU#%d, %s/%d\n",
+               msg, raw_smp_processor_id(),
+               current->comm, current->pid);
+       printk(KERN_EMERG " lock: %p, .magic: %08x, .owner: %s/%d, "
+                       ".owner_cpu: %d\n",
+               lock, lock->magic,
+               owner ? owner->comm : "<none>",
+               owner ? owner->pid : -1,
+               lock->owner_cpu);
+       dump_stack();
 }
 
 #define SPIN_BUG_ON(cond, lock, msg) if (unlikely(cond)) spin_bug(lock, msg)
 
-static inline void debug_spin_lock_before(spinlock_t *lock)
+static inline void
+debug_spin_lock_before(spinlock_t *lock)
 {
        SPIN_BUG_ON(lock->magic != SPINLOCK_MAGIC, lock, "bad magic");
        SPIN_BUG_ON(lock->owner == current, lock, "recursion");
@@ -118,20 +115,13 @@ void _raw_spin_unlock(spinlock_t *lock)
 
 static void rwlock_bug(rwlock_t *lock, const char *msg)
 {
-       static long print_once = 1;
-
-       if (xchg(&print_once, 0)) {
-               printk(KERN_EMERG "BUG: rwlock %s on CPU#%d, %s/%d, %p\n",
-                       msg, raw_smp_processor_id(), current->comm,
-                       current->pid, lock);
-               dump_stack();
-#ifdef CONFIG_SMP
-               /*
-                * We cannot continue on SMP:
-                */
-               panic("bad locking");
-#endif
-       }
+       if (!debug_locks_off())
+               return;
+
+       printk(KERN_EMERG "BUG: rwlock %s on CPU#%d, %s/%d, %p\n",
+               msg, raw_smp_processor_id(), current->comm,
+               current->pid, lock);
+       dump_stack();
 }
 
 #define RWLOCK_BUG_ON(cond, lock, msg) if (unlikely(cond)) rwlock_bug(lock, msg)
index 35f8553..7b45079 100644 (file)
@@ -330,6 +330,8 @@ void __vunmap(void *addr, int deallocate_pages)
                return;
        }
 
+       debug_check_no_locks_freed(addr, area->size);
+
        if (deallocate_pages) {
                int i;