On Wed, Apr 11, 2007 at 06:26:42PM +0900, Tejun Heo wrote:
> Hello,
> 
> Cornelia Huck wrote:
> > =======================================================
> > [ INFO: possible circular locking dependency detected ]
> > 2.6.21-rc6-ge666c753-dirty #54
> > -------------------------------------------------------
> > kslowcrw/64 is trying to acquire lock:
> >  (&sch->reg_mutex){--..}, at: [<00000000004233b2>] mutex_lock+0x3e/0x4c
> > 
> > but task is already holding lock:
> >  (&sd->s_active){----}, at: [<0000000000209578>] remove_dir+0xec/0x144
> > 
> > which lock already depends on the new lock.
> 
> This is probably because s_active has different write locked and
> unlocked by different threads but it doesn't tell lockdep about it.
> I'll read lockdep docs and update it.

Does this patch fix the problem?

Index: work/fs/sysfs/dir.c
===================================================================
--- work.orig/fs/sysfs/dir.c    2007-04-11 19:12:02.000000000 +0900
+++ work/fs/sysfs/dir.c 2007-04-11 19:12:12.000000000 +0900
@@ -26,7 +26,8 @@ void release_sysfs_dirent(struct sysfs_d
         * locked.  If @sd is cursor for directory walk or being
         * released prematurely, s_active has no reader or writer.
         */
-       down_write_trylock(&sd->s_active);
+       if (!down_write_trylock(&sd->s_active))
+               rwsem_acquire(&sd->s_active.dep_map, 0, 0, _RET_IP_);
        up_write(&sd->s_active);
 
        if (sd->s_type & SYSFS_KOBJ_LINK)
Index: work/fs/sysfs/sysfs.h
===================================================================
--- work.orig/fs/sysfs/sysfs.h  2007-04-11 19:12:02.000000000 +0900
+++ work/fs/sysfs/sysfs.h       2007-04-11 19:12:03.000000000 +0900
@@ -171,6 +171,7 @@ static inline void sysfs_put_active_two(
 static inline void sysfs_deactivate(struct sysfs_dirent *sd)
 {
        down_write(&sd->s_active);
+       rwsem_release(&sd->s_active.dep_map, 1, _RET_IP_);
 }
 
 static inline int sysfs_is_shadowed_inode(struct inode *inode)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Reply via email to