[PATCH 169/196] Kobject: auto-cleanup on final unref
From: Greg Kroah-Hartman
Date: Fri Jan 25 2008 - 03:11:47 EST
From: Kay Sievers <kay.sievers@xxxxxxxx>
We save the current state in the object itself, so we can do proper
cleanup when the last reference is dropped.
If the initial reference is dropped, the object will be removed from
sysfs if needed, if an "add" event was sent, "remove" will be send, and
the allocated resources are released.
This allows us to clean up some driver core usage as well as allowing us
to do other such changes to the rest of the kernel.
Signed-off-by: Kay Sievers <kay.sievers@xxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxx>
---
drivers/base/core.c | 32 ++-------
include/linux/kobject.h | 5 ++
lib/kobject.c | 170 ++++++++++++++++++++++++++--------------------
lib/kobject_uevent.c | 11 +++
4 files changed, 119 insertions(+), 99 deletions(-)
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 675a719..d5d542d 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -576,8 +576,8 @@ static struct kobject *get_device_parent(struct device *dev,
/*
* If we have no parent, we live in "virtual".
- * Class-devices with a bus-device as parent, live
- * in a class-directory to prevent namespace collisions.
+ * Class-devices with a non class-device as parent, live
+ * in a "glue" directory to prevent namespace collisions.
*/
if (parent == NULL)
parent_kobj = virtual_device_parent(dev);
@@ -607,8 +607,7 @@ static struct kobject *get_device_parent(struct device *dev,
kobject_put(k);
return NULL;
}
- /* Do not emit a uevent, as it's not needed for this
- * "class glue" directory. */
+ /* do not emit an uevent for this simple "glue" directory */
return k;
}
@@ -619,30 +618,13 @@ static struct kobject *get_device_parent(struct device *dev,
static void cleanup_device_parent(struct device *dev)
{
- struct device *d;
- int other = 0;
+ struct kobject *glue_dir = dev->kobj.parent;
- if (!dev->class)
- return;
-
- /* see if we live in a parent class directory */
- if (dev->kobj.parent->kset != &dev->class->class_dirs)
+ /* see if we live in a "glue" directory */
+ if (!dev->class || glue_dir->kset != &dev->class->class_dirs)
return;
- /* if we are the last child of our class, delete the directory */
- down(&dev->class->sem);
- list_for_each_entry(d, &dev->class->devices, node) {
- if (d == dev)
- continue;
- if (d->kobj.parent == dev->kobj.parent) {
- other = 1;
- break;
- }
- }
- if (!other)
- kobject_del(dev->kobj.parent);
- kobject_put(dev->kobj.parent);
- up(&dev->class->sem);
+ kobject_put(glue_dir);
}
#endif
diff --git a/include/linux/kobject.h b/include/linux/kobject.h
index 63967da..be03ce8 100644
--- a/include/linux/kobject.h
+++ b/include/linux/kobject.h
@@ -68,6 +68,11 @@ struct kobject {
struct kset * kset;
struct kobj_type * ktype;
struct sysfs_dirent * sd;
+ unsigned int state_initialized:1;
+ unsigned int state_name_set:1;
+ unsigned int state_in_sysfs:1;
+ unsigned int state_add_uevent_sent:1;
+ unsigned int state_remove_uevent_sent:1;
};
extern int kobject_set_name(struct kobject *, const char *, ...)
diff --git a/lib/kobject.c b/lib/kobject.c
index c321f19..4fce5ca 100644
--- a/lib/kobject.c
+++ b/lib/kobject.c
@@ -124,85 +124,74 @@ char *kobject_get_path(struct kobject *kobj, gfp_t gfp_mask)
}
EXPORT_SYMBOL_GPL(kobject_get_path);
-static void kobject_init_internal(struct kobject * kobj)
+/* add the kobject to its kset's list */
+static void kobj_kset_join(struct kobject *kobj)
{
- if (!kobj)
+ if (!kobj->kset)
return;
- kref_init(&kobj->kref);
- INIT_LIST_HEAD(&kobj->entry);
+
+ kset_get(kobj->kset);
+ spin_lock(&kobj->kset->list_lock);
+ list_add_tail(&kobj->entry, &kobj->kset->list);
+ spin_unlock(&kobj->kset->list_lock);
}
+/* remove the kobject from its kset's list */
+static void kobj_kset_leave(struct kobject *kobj)
+{
+ if (!kobj->kset)
+ return;
-/**
- * unlink - remove kobject from kset list.
- * @kobj: kobject.
- *
- * Remove the kobject from the kset list and decrement
- * its parent's refcount.
- * This is separated out, so we can use it in both
- * kobject_del() and kobject_add_internal() on error.
- */
+ spin_lock(&kobj->kset->list_lock);
+ list_del_init(&kobj->entry);
+ spin_unlock(&kobj->kset->list_lock);
+ kset_put(kobj->kset);
+}
-static void unlink(struct kobject * kobj)
+static void kobject_init_internal(struct kobject * kobj)
{
- struct kobject *parent = kobj->parent;
-
- if (kobj->kset) {
- spin_lock(&kobj->kset->list_lock);
- list_del_init(&kobj->entry);
- spin_unlock(&kobj->kset->list_lock);
- }
- kobj->parent = NULL;
- kobject_put(kobj);
- kobject_put(parent);
+ if (!kobj)
+ return;
+ kref_init(&kobj->kref);
+ INIT_LIST_HEAD(&kobj->entry);
}
+
static int kobject_add_internal(struct kobject *kobj)
{
int error = 0;
struct kobject * parent;
- if (!(kobj = kobject_get(kobj)))
+ if (!kobj)
return -ENOENT;
- if (!kobj->k_name)
- kobject_set_name(kobj, "NO_NAME");
- if (!*kobj->k_name) {
- pr_debug("kobject (%p) attempted to be registered with no "
+
+ if (!kobj->k_name || !kobj->k_name[0]) {
+ pr_debug("kobject: (%p): attempted to be registered with empty "
"name!\n", kobj);
WARN_ON(1);
- kobject_put(kobj);
return -EINVAL;
}
- parent = kobject_get(kobj->parent);
- pr_debug("kobject: '%s' (%p): %s: parent: '%s', set: '%s'\n",
- kobject_name(kobj), kobj, __FUNCTION__,
- parent ? kobject_name(parent) : "<NULL>",
- kobj->kset ? kobject_name(&kobj->kset->kobj) : "<NULL>" );
+ parent = kobject_get(kobj->parent);
+ /* join kset if set, use it as parent if we do not already have one */
if (kobj->kset) {
- kobj->kset = kset_get(kobj->kset);
-
- if (!parent) {
+ if (!parent)
parent = kobject_get(&kobj->kset->kobj);
- /*
- * If the kset is our parent, get a second
- * reference, we drop both the kset and the
- * parent ref on cleanup
- */
- kobject_get(parent);
- }
-
- spin_lock(&kobj->kset->list_lock);
- list_add_tail(&kobj->entry, &kobj->kset->list);
- spin_unlock(&kobj->kset->list_lock);
+ kobj_kset_join(kobj);
kobj->parent = parent;
}
+ pr_debug("kobject: '%s' (%p): %s: parent: '%s', set: '%s'\n",
+ kobject_name(kobj), kobj, __FUNCTION__,
+ parent ? kobject_name(parent) : "<NULL>",
+ kobj->kset ? kobject_name(&kobj->kset->kobj) : "<NULL>" );
+
error = create_dir(kobj);
if (error) {
- /* unlink does the kobject_put() for us */
- unlink(kobj);
+ kobj_kset_leave(kobj);
+ kobject_put(parent);
+ kobj->parent = NULL;
/* be noisy on error issues */
if (error == -EEXIST)
@@ -214,7 +203,8 @@ static int kobject_add_internal(struct kobject *kobj)
printk(KERN_ERR "%s failed for %s (%d)\n",
__FUNCTION__, kobject_name(kobj), error);
dump_stack();
- }
+ } else
+ kobj->state_in_sysfs = 1;
return error;
}
@@ -238,11 +228,13 @@ static int kobject_set_name_vargs(struct kobject *kobj, const char *fmt,
if (!name)
return -ENOMEM;
+
/* Free the old name, if necessary. */
kfree(kobj->k_name);
/* Now, set the new name */
kobj->k_name = name;
+ kobj->state_name_set = 1;
return 0;
}
@@ -293,20 +285,25 @@ void kobject_init(struct kobject *kobj, struct kobj_type *ktype)
err_str = "must have a ktype to be initialized properly!\n";
goto error;
}
- if (atomic_read(&kobj->kref.refcount)) {
+ if (kobj->state_initialized) {
/* do not error out as sometimes we can recover */
- printk(KERN_ERR "kobject: reference count is already set, "
- "something is seriously wrong.\n");
+ printk(KERN_ERR "kobject (%p): tried to init an initialized "
+ "object, something is seriously wrong.\n", kobj);
dump_stack();
}
kref_init(&kobj->kref);
INIT_LIST_HEAD(&kobj->entry);
kobj->ktype = ktype;
+ kobj->state_name_set = 0;
+ kobj->state_in_sysfs = 0;
+ kobj->state_add_uevent_sent = 0;
+ kobj->state_remove_uevent_sent = 0;
+ kobj->state_initialized = 1;
return;
error:
- printk(KERN_ERR "kobject: %s\n", err_str);
+ printk(KERN_ERR "kobject (%p): %s\n", kobj, err_str);
dump_stack();
}
EXPORT_SYMBOL(kobject_init);
@@ -345,17 +342,10 @@ static int kobject_add_varg(struct kobject *kobj, struct kobject *parent,
*
* If this function returns an error, kobject_put() must be called to
* properly clean up the memory associated with the object.
- *
- * If the function is successful, the only way to properly clean up the
- * memory is with a call to kobject_del(), in which case, a call to
- * kobject_put() is not necessary (kobject_del() does the final
- * kobject_put() to call the release function in the ktype's release
- * pointer.)
- *
* Under no instance should the kobject that is passed to this function
* be directly freed with a call to kfree(), that can leak memory.
*
- * Note, no uevent will be created with this call, the caller should set
+ * Note, no "add" uevent will be created with this call, the caller should set
* up all of the necessary sysfs files for the object and then call
* kobject_uevent() with the UEVENT_ADD parameter to ensure that
* userspace is properly notified of this kobject's creation.
@@ -369,6 +359,13 @@ int kobject_add(struct kobject *kobj, struct kobject *parent,
if (!kobj)
return -EINVAL;
+ if (!kobj->state_initialized) {
+ printk(KERN_ERR "kobject '%s' (%p): tried to add an "
+ "uninitialized object, something is seriously wrong.\n",
+ kobject_name(kobj), kobj);
+ dump_stack();
+ return -EINVAL;
+ }
va_start(args, fmt);
retval = kobject_add_varg(kobj, parent, fmt, args);
va_end(args);
@@ -527,8 +524,12 @@ void kobject_del(struct kobject * kobj)
{
if (!kobj)
return;
+
sysfs_remove_dir(kobj);
- unlink(kobj);
+ kobj->state_in_sysfs = 0;
+ kobj_kset_leave(kobj);
+ kobject_put(kobj->parent);
+ kobj->parent = NULL;
}
/**
@@ -565,21 +566,43 @@ struct kobject * kobject_get(struct kobject * kobj)
*/
static void kobject_cleanup(struct kobject *kobj)
{
- struct kobj_type * t = get_ktype(kobj);
- struct kset * s = kobj->kset;
+ struct kobj_type *t = get_ktype(kobj);
const char *name = kobj->k_name;
+ int name_set = kobj->state_name_set;
pr_debug("kobject: '%s' (%p): %s\n",
kobject_name(kobj), kobj, __FUNCTION__);
+
+ if (t && !t->release)
+ pr_debug("kobject: '%s' (%p): does not have a release() "
+ "function, it is broken and must be fixed.\n",
+ kobject_name(kobj), kobj);
+
+ /* send "remove" if the caller did not do it but sent "add" */
+ if (kobj->state_add_uevent_sent && !kobj->state_remove_uevent_sent) {
+ pr_debug("kobject: '%s' (%p): auto cleanup 'remove' event\n",
+ kobject_name(kobj), kobj);
+ kobject_uevent(kobj, KOBJ_REMOVE);
+ }
+
+ /* remove from sysfs if the caller did not do it */
+ if (kobj->state_in_sysfs) {
+ pr_debug("kobject: '%s' (%p): auto cleanup kobject_del\n",
+ kobject_name(kobj), kobj);
+ kobject_del(kobj);
+ }
+
if (t && t->release) {
+ pr_debug("kobject: '%s' (%p): calling ktype release\n",
+ kobject_name(kobj), kobj);
t->release(kobj);
- /* If we have a release function, we can guess that this was
- * not a statically allocated kobject, so we should be safe to
- * free the name */
+ }
+
+ /* free name if we allocated it */
+ if (name_set && name) {
+ pr_debug("kobject: '%s': free name\n", name);
kfree(name);
}
- if (s)
- kset_put(s);
}
static void kobject_release(struct kref *kref)
@@ -601,8 +624,7 @@ void kobject_put(struct kobject * kobj)
static void dynamic_kobj_release(struct kobject *kobj)
{
- pr_debug("kobject: '%s' (%p): %s\n",
- kobject_name(kobj), kobj, __FUNCTION__);
+ pr_debug("kobject: (%p): %s\n", kobj, __FUNCTION__);
kfree(kobj);
}
diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
index 51dc4d2..b021e67 100644
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -180,6 +180,17 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action,
}
}
+ /*
+ * Mark "add" and "remove" events in the object to ensure proper
+ * events to userspace during automatic cleanup. If the object did
+ * send an "add" event, "remove" will automatically generated by
+ * the core, if not already done by the caller.
+ */
+ if (action == KOBJ_ADD)
+ kobj->state_add_uevent_sent = 1;
+ else if (action == KOBJ_REMOVE)
+ kobj->state_remove_uevent_sent = 1;
+
/* we will send an event, so request a new sequence number */
spin_lock(&sequence_lock);
seq = ++uevent_seqnum;
--
1.5.3.8
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/