linux內核部件分析（六）——設備驅動模型之device

     linux的設備驅動模型，是建立在sysfs和kobject之上的，由總線、設備、驅動、類所組成的關係結構。從本節開始，我們將對linux這一設備驅動模型進行深入分析。

     頭文件是include/linux/device.h，實現在drivers/base目錄中。本節要分析的，是其中的設備，主要在core.c中。

struct device {
	struct device		*parent;

	struct device_private	*p;

	struct kobject kobj;
	const char		*init_name; /* initial name of the device */
	struct device_type	*type;

	struct semaphore	sem;	/* semaphore to synchronize calls to
					 * its driver.
					 */

	struct bus_type	*bus;		/* type of bus device is on */
	struct device_driver *driver;	/* which driver has allocated this
					   device */
	void		*platform_data;	/* Platform specific data, device
					   core doesn't touch it */
	struct dev_pm_info	power;

#ifdef CONFIG_NUMA
	int		numa_node;	/* NUMA node this device is close to */
#endif
	u64		*dma_mask;	/* dma mask (if dma'able device) */
	u64		coherent_dma_mask;/* Like dma_mask, but for
					     alloc_coherent mappings as
					     not all hardware supports
					     64 bit addresses for consistent
					     allocations such descriptors. */

	struct device_dma_parameters *dma_parms;

	struct list_head	dma_pools;	/* dma pools (if dma'ble) */

	struct dma_coherent_mem	*dma_mem; /* internal for coherent mem
					     override */
	/* arch specific additions */
	struct dev_archdata	archdata;

	dev_t			devt;	/* dev_t, creates the sysfs "dev" */

	spinlock_t		devres_lock;
	struct list_head	devres_head;

	struct klist_node	knode_class;
	struct class		*class;
	const struct attribute_group **groups;	/* optional groups */

	void	(*release)(struct device *dev);
};

先來分析下struct device的結構變量。首先是指向父節點的指針parent，kobj是內嵌在device中的kobject，用於把它聯繫到sysfs中。bus是對設備所在總線的指針，driver是對設備所用驅動的指針。還有DMA需要的數據，表示設備號的devt，表示設備資源的devres_head和保護它的devres_lock。指向類的指針class，knode_class是被連入class鏈表時所用的klist節點。group是設備的屬性集合。release應該是設備釋放時調用的函數。

struct device_private {
	struct klist klist_children;
	struct klist_node knode_parent;
	struct klist_node knode_driver;
	struct klist_node knode_bus;
	void *driver_data;
	struct device *device;
};
#define to_device_private_parent(obj)	\
	container_of(obj, struct device_private, knode_parent)
#define to_device_private_driver(obj)	\
	container_of(obj, struct device_private, knode_driver)
#define to_device_private_bus(obj)	\
	container_of(obj, struct device_private, knode_bus)

struct device中有一部分不願意讓外界看到，所以做出struct device_private結構，包括了設備驅動模型內部的鏈接。klist_children是子設備的鏈表，knode_parent是連入父設備的klist_children時所用的節點，knode_driver是連入驅動的設備鏈表所用的節點，knode_bus是連入總線的設備鏈表時所用的節點。driver_data用於在設備結構中存放相關的驅動信息，也許是驅動專門爲設備建立的結構實例。device則是指向struct device_private所屬的device。

下面還有一些宏，to_device_private_parent()是從父設備的klist_children上節點，獲得相應的device_private。to_device_private_driver()是從驅動的設備鏈表上節點，獲得對應的device_private。to_device_private_bus()是從總線的設備鏈表上節點，獲得對應的device_private。

或許會奇怪，爲什麼knode_class沒有被移入struct device_private，或許有外部模塊需要用到它。

/*
 * The type of device, "struct device" is embedded in. A class
 * or bus can contain devices of different types
 * like "partitions" and "disks", "mouse" and "event".
 * This identifies the device type and carries type-specific
 * information, equivalent to the kobj_type of a kobject.
 * If "name" is specified, the uevent will contain it in
 * the DEVTYPE variable.
 */
struct device_type {
	const char *name;
	const struct attribute_group **groups;
	int (*uevent)(struct device *dev, struct kobj_uevent_env *env);
	char *(*devnode)(struct device *dev, mode_t *mode);
	void (*release)(struct device *dev);

	const struct dev_pm_ops *pm;
};

device竟然有device_type，類似於與kobject相對的kobj_type，之後我們再看它怎麼用。

/* interface for exporting device attributes */
struct device_attribute {
	struct attribute	attr;
	ssize_t (*show)(struct device *dev, struct device_attribute *attr,
			char *buf);
	ssize_t (*store)(struct device *dev, struct device_attribute *attr,
			 const char *buf, size_t count);
};

#define DEVICE_ATTR(_name, _mode, _show, _store) \
struct device_attribute dev_attr_##_name = __ATTR(_name, _mode, _show, _store)

這個device_attribute顯然就是device對struct attribute的封裝，新加的show()、store()函數都是以與設備相關的結構調用的。

至於device中其它的archdata、dma、devres，都是作爲設備特有的，我們現在主要關心設備驅動模型的建立，這些會盡量忽略。

 

下面就來看看device的實現，這主要在core.c中。

int __init devices_init(void)
{
	devices_kset = kset_create_and_add("devices", &device_uevent_ops, NULL);
	if (!devices_kset)
		return -ENOMEM;
	dev_kobj = kobject_create_and_add("dev", NULL);
	if (!dev_kobj)
		goto dev_kobj_err;
	sysfs_dev_block_kobj = kobject_create_and_add("block", dev_kobj);
	if (!sysfs_dev_block_kobj)
		goto block_kobj_err;
	sysfs_dev_char_kobj = kobject_create_and_add("char", dev_kobj);
	if (!sysfs_dev_char_kobj)
		goto char_kobj_err;

	return 0;

 char_kobj_err:
	kobject_put(sysfs_dev_block_kobj);
 block_kobj_err:
	kobject_put(dev_kobj);
 dev_kobj_err:
	kset_unregister(devices_kset);
	return -ENOMEM;
}

這是在設備驅動模型初始化時調用的device部分初始的函數devices_init()。它乾的事情我們都很熟悉，就是建立sysfs中的devices目錄，和dev目錄。還在dev目錄下又建立了block和char兩個子目錄。因爲dev目錄只打算存放輔助的設備號，所以沒必要使用kset。

static ssize_t dev_attr_show(struct kobject *kobj, struct attribute *attr,
			     char *buf)
{
	struct device_attribute *dev_attr = to_dev_attr(attr);
	struct device *dev = to_dev(kobj);
	ssize_t ret = -EIO;

	if (dev_attr->show)
		ret = dev_attr->show(dev, dev_attr, buf);
	if (ret >= (ssize_t)PAGE_SIZE) {
		print_symbol("dev_attr_show: %s returned bad count\n",
				(unsigned long)dev_attr->show);
	}
	return ret;
}

static ssize_t dev_attr_store(struct kobject *kobj, struct attribute *attr,
			      const char *buf, size_t count)
{
	struct device_attribute *dev_attr = to_dev_attr(attr);
	struct device *dev = to_dev(kobj);
	ssize_t ret = -EIO;

	if (dev_attr->store)
		ret = dev_attr->store(dev, dev_attr, buf, count);
	return ret;
}

static struct sysfs_ops dev_sysfs_ops = {
	.show	= dev_attr_show,
	.store	= dev_attr_store,
};

看到這裏是不是很熟悉，dev_sysfs_ops就是device準備註冊到sysfs中的操作函數。dev_attr_show()和dev_attr_store()都會再調用與屬性相關的函數。

static void device_release(struct kobject *kobj)
{
	struct device *dev = to_dev(kobj);
	struct device_private *p = dev->p;

	if (dev->release)
		dev->release(dev);
	else if (dev->type && dev->type->release)
		dev->type->release(dev);
	else if (dev->class && dev->class->dev_release)
		dev->class->dev_release(dev);
	else
		WARN(1, KERN_ERR "Device '%s' does not have a release() "
			"function, it is broken and must be fixed.\n",
			dev_name(dev));
	kfree(p);
}

static struct kobj_type device_ktype = {
	.release	= device_release,
	.sysfs_ops	= &dev_sysfs_ops,
};

使用的release函數是device_release。在釋放device時，會依次調用device結構中定義的release函數，device_type中定義的release函數，device所屬的class中所定義的release函數，最後會吧device_private結構釋放掉。

 

static int dev_uevent_filter(struct kset *kset, struct kobject *kobj)
{
	struct kobj_type *ktype = get_ktype(kobj);

	if (ktype == &device_ktype) {
		struct device *dev = to_dev(kobj);
		if (dev->bus)
			return 1;
		if (dev->class)
			return 1;
	}
	return 0;
}

static const char *dev_uevent_name(struct kset *kset, struct kobject *kobj)
{
	struct device *dev = to_dev(kobj);

	if (dev->bus)
		return dev->bus->name;
	if (dev->class)
		return dev->class->name;
	return NULL;
}

static int dev_uevent(struct kset *kset, struct kobject *kobj,
		      struct kobj_uevent_env *env)
{
	struct device *dev = to_dev(kobj);
	int retval = 0;

	/* add device node properties if present */
	if (MAJOR(dev->devt)) {
		const char *tmp;
		const char *name;
		mode_t mode = 0;

		add_uevent_var(env, "MAJOR=%u", MAJOR(dev->devt));
		add_uevent_var(env, "MINOR=%u", MINOR(dev->devt));
		name = device_get_devnode(dev, &mode, &tmp);
		if (name) {
			add_uevent_var(env, "DEVNAME=%s", name);
			kfree(tmp);
			if (mode)
				add_uevent_var(env, "DEVMODE=%#o", mode & 0777);
		}
	}

	if (dev->type && dev->type->name)
		add_uevent_var(env, "DEVTYPE=%s", dev->type->name);

	if (dev->driver)
		add_uevent_var(env, "DRIVER=%s", dev->driver->name);

#ifdef CONFIG_SYSFS_DEPRECATED
	if (dev->class) {
		struct device *parent = dev->parent;

		/* find first bus device in parent chain */
		while (parent && !parent->bus)
			parent = parent->parent;
		if (parent && parent->bus) {
			const char *path;

			path = kobject_get_path(&parent->kobj, GFP_KERNEL);
			if (path) {
				add_uevent_var(env, "PHYSDEVPATH=%s", path);
				kfree(path);
			}

			add_uevent_var(env, "PHYSDEVBUS=%s", parent->bus->name);

			if (parent->driver)
				add_uevent_var(env, "PHYSDEVDRIVER=%s",
					       parent->driver->name);
		}
	} else if (dev->bus) {
		add_uevent_var(env, "PHYSDEVBUS=%s", dev->bus->name);

		if (dev->driver)
			add_uevent_var(env, "PHYSDEVDRIVER=%s",
				       dev->driver->name);
	}
#endif

	/* have the bus specific function add its stuff */
	if (dev->bus && dev->bus->uevent) {
		retval = dev->bus->uevent(dev, env);
		if (retval)
			pr_debug("device: '%s': %s: bus uevent() returned %d\n",
				 dev_name(dev), __func__, retval);
	}

	/* have the class specific function add its stuff */
	if (dev->class && dev->class->dev_uevent) {
		retval = dev->class->dev_uevent(dev, env);
		if (retval)
			pr_debug("device: '%s': %s: class uevent() "
				 "returned %d\n", dev_name(dev),
				 __func__, retval);
	}

	/* have the device type specific fuction add its stuff */
	if (dev->type && dev->type->uevent) {
		retval = dev->type->uevent(dev, env);
		if (retval)
			pr_debug("device: '%s': %s: dev_type uevent() "
				 "returned %d\n", dev_name(dev),
				 __func__, retval);
	}

	return retval;
}

static struct kset_uevent_ops device_uevent_ops = {
	.filter =	dev_uevent_filter,
	.name =		dev_uevent_name,
	.uevent =	dev_uevent,
};

前面在講到kset時，我們並未關注其中的kset_event_ops結構變量。但這裏device既然用到了，我們就對其中的三個函數做簡單介紹。kset_uevent_ops中的函數是用於管理kset內部kobject的uevent操作。其中filter函數用於阻止一個kobject向用戶空間發送uevent，返回值爲0表示阻止。這裏dev_uevent_filter()檢查device所屬的bus或者class是否存在，如果都不存在，也就沒有發送uevent的必要了。name函數是用於覆蓋kset發送給用戶空間的名稱。這裏dev_uevent_name()選擇使用bus或者class的名稱。uevent()函數是在uevent將被髮送到用戶空間之前調用的，用於向uevent中增加新的環境變量。dev_uevent()的實現很熱鬧，向uevent中添加了各種環境變量。

static ssize_t show_uevent(struct device *dev, struct device_attribute *attr,
			   char *buf)
{
	struct kobject *top_kobj;
	struct kset *kset;
	struct kobj_uevent_env *env = NULL;
	int i;
	size_t count = 0;
	int retval;

	/* search the kset, the device belongs to */
	top_kobj = &dev->kobj;
	while (!top_kobj->kset && top_kobj->parent)
		top_kobj = top_kobj->parent;
	if (!top_kobj->kset)
		goto out;

	kset = top_kobj->kset;
	if (!kset->uevent_ops || !kset->uevent_ops->uevent)
		goto out;

	/* respect filter */
	if (kset->uevent_ops && kset->uevent_ops->filter)
		if (!kset->uevent_ops->filter(kset, &dev->kobj))
			goto out;

	env = kzalloc(sizeof(struct kobj_uevent_env), GFP_KERNEL);
	if (!env)
		return -ENOMEM;

	/* let the kset specific function add its keys */
	retval = kset->uevent_ops->uevent(kset, &dev->kobj, env);
	if (retval)
		goto out;

	/* copy keys to file */
	for (i = 0; i < env->envp_idx; i++)
		count += sprintf(&buf[count], "%s\n", env->envp[i]);
out:
	kfree(env);
	return count;
}

static ssize_t store_uevent(struct device *dev, struct device_attribute *attr,
			    const char *buf, size_t count)
{
	enum kobject_action action;

	if (kobject_action_type(buf, count, &action) == 0) {
		kobject_uevent(&dev->kobj, action);
		goto out;
	}

	dev_err(dev, "uevent: unsupported action-string; this will "
		     "be ignored in a future kernel version\n");
	kobject_uevent(&dev->kobj, KOBJ_ADD);
out:
	return count;
}

static struct device_attribute uevent_attr =
	__ATTR(uevent, S_IRUGO | S_IWUSR, show_uevent, store_uevent);

device不僅在kset中添加了對uevent的管理，而且還把uevent信息做成設備的一個屬性uevent。其中show_event()是顯示uevent中環境變量的，store_uevent()是發送uevent的。

static int device_add_attributes(struct device *dev,
				 struct device_attribute *attrs)
{
	int error = 0;
	int i;

	if (attrs) {
		for (i = 0; attr_name(attrs[i]); i++) {
			error = device_create_file(dev, &attrs[i]);
			if (error)
				break;
		}
		if (error)
			while (--i >= 0)
				device_remove_file(dev, &attrs[i]);
	}
	return error;
}

static void device_remove_attributes(struct device *dev,
				     struct device_attribute *attrs)
{
	int i;

	if (attrs)
		for (i = 0; attr_name(attrs[i]); i++)
			device_remove_file(dev, &attrs[i]);
}

static int device_add_groups(struct device *dev,
			     const struct attribute_group **groups)
{
	int error = 0;
	int i;

	if (groups) {
		for (i = 0; groups[i]; i++) {
			error = sysfs_create_group(&dev->kobj, groups[i]);
			if (error) {
				while (--i >= 0)
					sysfs_remove_group(&dev->kobj,
							   groups[i]);
				break;
			}
		}
	}
	return error;
}

static void device_remove_groups(struct device *dev,
				 const struct attribute_group **groups)
{
	int i;

	if (groups)
		for (i = 0; groups[i]; i++)
			sysfs_remove_group(&dev->kobj, groups[i]);
}

以上四個內部函數是用來向device中添加或刪除屬性與屬性集合的。

device_add_attributes、device_remove_attributes、device_add_groups、device_remove_groups，都是直接通過sysfs提供的API實現。

static int device_add_attrs(struct device *dev)
{
	struct class *class = dev->class;
	struct device_type *type = dev->type;
	int error;

	if (class) {
		error = device_add_attributes(dev, class->dev_attrs);
		if (error)
			return error;
	}

	if (type) {
		error = device_add_groups(dev, type->groups);
		if (error)
			goto err_remove_class_attrs;
	}

	error = device_add_groups(dev, dev->groups);
	if (error)
		goto err_remove_type_groups;

	return 0;

 err_remove_type_groups:
	if (type)
		device_remove_groups(dev, type->groups);
 err_remove_class_attrs:
	if (class)
		device_remove_attributes(dev, class->dev_attrs);

	return error;
}

static void device_remove_attrs(struct device *dev)
{
	struct class *class = dev->class;
	struct device_type *type = dev->type;

	device_remove_groups(dev, dev->groups);

	if (type)
		device_remove_groups(dev, type->groups);

	if (class)
		device_remove_attributes(dev, class->dev_attrs);
}

device_add_attrs()實際負責device中的屬性添加。也是幾個部分的集合，包括class中的dev_attrs，device_type中的groups，還有device本身的groups。

device_remove_attrs()則負責對應的device屬性刪除工作。

 

#define print_dev_t(buffer, dev)					\
	sprintf((buffer), "%u:%u\n", MAJOR(dev), MINOR(dev))

static ssize_t show_dev(struct device *dev, struct device_attribute *attr,
			char *buf)
{
	return print_dev_t(buf, dev->devt);
}

static struct device_attribute devt_attr =
	__ATTR(dev, S_IRUGO, show_dev, NULL);

這裏又定義了一個名爲dev的屬性，就是顯示設備的設備號。

/**
 * device_create_file - create sysfs attribute file for device.
 * @dev: device.
 * @attr: device attribute descriptor.
 */
int device_create_file(struct device *dev, struct device_attribute *attr)
{
	int error = 0;
	if (dev)
		error = sysfs_create_file(&dev->kobj, &attr->attr);
	return error;
}

/**
 * device_remove_file - remove sysfs attribute file.
 * @dev: device.
 * @attr: device attribute descriptor.
 */
void device_remove_file(struct device *dev, struct device_attribute *attr)
{
	if (dev)
		sysfs_remove_file(&dev->kobj, &attr->attr);
}

/**
 * device_create_bin_file - create sysfs binary attribute file for device.
 * @dev: device.
 * @attr: device binary attribute descriptor.
 */
int device_create_bin_file(struct device *dev, struct bin_attribute *attr)
{
	int error = -EINVAL;
	if (dev)
		error = sysfs_create_bin_file(&dev->kobj, attr);
	return error;
}

/**
 * device_remove_bin_file - remove sysfs binary attribute file
 * @dev: device.
 * @attr: device binary attribute descriptor.
 */
void device_remove_bin_file(struct device *dev, struct bin_attribute *attr)
{
	if (dev)
		sysfs_remove_bin_file(&dev->kobj, attr);
}

int device_schedule_callback_owner(struct device *dev,
		void (*func)(struct device *), struct module *owner)
{
	return sysfs_schedule_callback(&dev->kobj,
			(void (*)(void *)) func, dev, owner);
}

這裏的五個函數，也是對sysfs提供的API的簡單封裝。

device_create_file()和device_remove_file()提供直接的屬性文件管理方法。

 device_create_bin_file()和device_remove_bin_file()則是提供設備管理二進制文件的方法。

device_schedule_callback_owner()也是簡單地將func加入工作隊列。

 

static void klist_children_get(struct klist_node *n)
{
	struct device_private *p = to_device_private_parent(n);
	struct device *dev = p->device;

	get_device(dev);
}

static void klist_children_put(struct klist_node *n)
{
	struct device_private *p = to_device_private_parent(n);
	struct device *dev = p->device;

	put_device(dev);
}

如果之前認真看過klist的實現，應該知道，klist_children_get()和klist_children_put()就是在設備掛入和刪除父設備的klist_children鏈表時調用的函數。在父設備klist_children鏈表上的指針，相當於對device的一個引用計數。

struct device *get_device(struct device *dev)
{
	return dev ? to_dev(kobject_get(&dev->kobj)) : NULL;
}

/**
 * put_device - decrement reference count.
 * @dev: device in question.
 */
void put_device(struct device *dev)
{
	/* might_sleep(); */
	if (dev)
		kobject_put(&dev->kobj);
}

device中的引用計數，完全交給內嵌的kobject來做。如果引用計數降爲零，自然是調用之前說到的包含甚廣的device_release函數。

void device_initialize(struct device *dev)
{
	dev->kobj.kset = devices_kset;
	kobject_init(&dev->kobj, &device_ktype);
	INIT_LIST_HEAD(&dev->dma_pools);
	init_MUTEX(&dev->sem);
	spin_lock_init(&dev->devres_lock);
	INIT_LIST_HEAD(&dev->devres_head);
	device_init_wakeup(dev, 0);
	device_pm_init(dev);
	set_dev_node(dev, -1);
}

device_initialize()就是device結構的初始化函數，它把device中能初始化的部分全初始化。它的界限在其中kobj的位置與device在設備驅動模型中的位置，這些必須由外部設置。可以看到，調用kobject_init()時，object的kobj_type選擇了device_ktype，其中主要是sysops的兩個函數，還有device_release函數。

static struct kobject *virtual_device_parent(struct device *dev)
{
	static struct kobject *virtual_dir = NULL;

	if (!virtual_dir)
		virtual_dir = kobject_create_and_add("virtual",
						     &devices_kset->kobj);

	return virtual_dir;
}

static struct kobject *get_device_parent(struct device *dev,
					 struct device *parent)
{
	int retval;

	if (dev->class) {
		struct kobject *kobj = NULL;
		struct kobject *parent_kobj;
		struct kobject *k;

		/*
		 * If we have no parent, we live in "virtual".
		 * 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);
		else if (parent->class)
			return &parent->kobj;
		else
			parent_kobj = &parent->kobj;

		/* find our class-directory at the parent and reference it */
		spin_lock(&dev->class->p->class_dirs.list_lock);
		list_for_each_entry(k, &dev->class->p->class_dirs.list, entry)
			if (k->parent == parent_kobj) {
				kobj = kobject_get(k);
				break;
			}
		spin_unlock(&dev->class->p->class_dirs.list_lock);
		if (kobj)
			return kobj;

		/* or create a new class-directory at the parent device */
		k = kobject_create();
		if (!k)
			return NULL;
		k->kset = &dev->class->p->class_dirs;
		retval = kobject_add(k, parent_kobj, "%s", dev->class->name);
		if (retval < 0) {
			kobject_put(k);
			return NULL;
		}
		/* do not emit an uevent for this simple "glue" directory */
		return k;
	}

	if (parent)
		return &parent->kobj;
	return NULL;
}

這裏的get_device_parent()就是獲取父節點的kobject，但也並非就如此簡單。get_device_parent()的返回值直接決定了device將被掛在哪個目錄下。到底該掛在哪，是由dev->class、dev->parent、dev->parent->class等因素綜合決定的。我們看get_device_parent()中是如何判斷的。如果dev->class爲空，表示一切隨父設備，有parent則返回parent->kobj，沒有則返回NULL。如果有dev->class呢，情況就比較複雜了，也許device有着與parent不同的class，也許device還沒有一個parent，等等。我們看具體的情況。如果parent不爲空，而且存在parent->class，則還放在parent目錄下。不然，要麼parent不存在，要麼parent沒有class，很難直接將有class的device放在parent下面。目前的解決方法很簡單，在parent與device之間，再加一層表示class的目錄。如果parent都沒有，那就把/sys/devices/virtual當做parent。class->p->class_dirs就是專門存放這種中間kobject的kset。思路理清後，再結合實際的sysfs，代碼就很容易看懂了。

static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)
{
	/* see if we live in a "glue" directory */
	if (!glue_dir || !dev->class ||
	    glue_dir->kset != &dev->class->p->class_dirs)
		return;

	kobject_put(glue_dir);
}

static void cleanup_device_parent(struct device *dev)
{
	cleanup_glue_dir(dev, dev->kobj.parent);
}

cleanup_device_parent()是取消對parent引用時調用的函數，看起來只針對這種glue形式的目錄起作用。

static void setup_parent(struct device *dev, struct device *parent)
{
	struct kobject *kobj;
	kobj = get_device_parent(dev, parent);
	if (kobj)
		dev->kobj.parent = kobj;
}

setup_parent()就是調用get_device_parent()獲得應該存放的父目錄kobj，並把dev->kobj.parent設爲它。

static int device_add_class_symlinks(struct device *dev)
{
	int error;

	if (!dev->class)
		return 0;

	error = sysfs_create_link(&dev->kobj,
				  &dev->class->p->class_subsys.kobj,
				  "subsystem");
	if (error)
		goto out;
	/* link in the class directory pointing to the device */
	error = sysfs_create_link(&dev->class->p->class_subsys.kobj,
				  &dev->kobj, dev_name(dev));
	if (error)
		goto out_subsys;

	if (dev->parent && device_is_not_partition(dev)) {
		error = sysfs_create_link(&dev->kobj, &dev->parent->kobj,
					  "device");
		if (error)
			goto out_busid;
	}
	return 0;

out_busid:
	sysfs_remove_link(&dev->class->p->class_subsys.kobj, dev_name(dev));
out_subsys:
	sysfs_remove_link(&dev->kobj, "subsystem");
out:
	return error;
}

device_add_class_symlinks()在device和class直接添加一些軟鏈接。在device目錄下創建指向class的subsystem文件，在class目錄下創建指向device的同名文件。如果device有父設備，而且device不是塊設備分區，則在device目錄下建立一個指向父設備的device鏈接文件。這一點在usb設備和usb接口間很常見。

static void device_remove_class_symlinks(struct device *dev)
{
	if (!dev->class)
		return;

#ifdef CONFIG_SYSFS_DEPRECATED
	if (dev->parent && device_is_not_partition(dev)) {
		char *class_name;

		class_name = make_class_name(dev->class->name, &dev->kobj);
		if (class_name) {
			sysfs_remove_link(&dev->parent->kobj, class_name);
			kfree(class_name);
		}
		sysfs_remove_link(&dev->kobj, "device");
	}

	if (dev->kobj.parent != &dev->class->p->class_subsys.kobj &&
	    device_is_not_partition(dev))
		sysfs_remove_link(&dev->class->p->class_subsys.kobj,
				  dev_name(dev));
#else
	if (dev->parent && device_is_not_partition(dev))
		sysfs_remove_link(&dev->kobj, "device");

	sysfs_remove_link(&dev->class->p->class_subsys.kobj, dev_name(dev));
#endif

	sysfs_remove_link(&dev->kobj, "subsystem");
}

device_remove_class_symlinks()刪除device和class之間的軟鏈接。

static inline const char *dev_name(const struct device *dev)
{
	return kobject_name(&dev->kobj);
}

int dev_set_name(struct device *dev, const char *fmt, ...)
{
	va_list vargs;
	int err;

	va_start(vargs, fmt);
	err = kobject_set_name_vargs(&dev->kobj, fmt, vargs);
	va_end(vargs);
	return err;
}

dev_name()獲得設備名稱，dev_set_name()設置設備名稱。但這裏的dev_set_name()只能在設備未註冊前使用。device的名稱其實是完全靠dev->kobj管理的。

static struct kobject *device_to_dev_kobj(struct device *dev)
{
	struct kobject *kobj;

	if (dev->class)
		kobj = dev->class->dev_kobj;
	else
		kobj = sysfs_dev_char_kobj;

	return kobj;
}

device_to_dev_kobj()爲dev選擇合適的/sys/dev下的kobject，或者是塊設備，或者是字符設備，或者沒有。

#define format_dev_t(buffer, dev)					\
	({								\
		sprintf(buffer, "%u:%u", MAJOR(dev), MINOR(dev));	\
		buffer;							\
	})

static int device_create_sys_dev_entry(struct device *dev)
{
	struct kobject *kobj = device_to_dev_kobj(dev);
	int error = 0;
	char devt_str[15];

	if (kobj) {
		format_dev_t(devt_str, dev->devt);
		error = sysfs_create_link(kobj, &dev->kobj, devt_str);
	}

	return error;
}

static void device_remove_sys_dev_entry(struct device *dev)
{
	struct kobject *kobj = device_to_dev_kobj(dev);
	char devt_str[15];

	if (kobj) {
		format_dev_t(devt_str, dev->devt);
		sysfs_remove_link(kobj, devt_str);
	}
}

device_create_sys_dev_entry()是在/sys/dev相應的目錄下建立對設備的軟鏈接。先是通過device_to_dev_kobj()獲得父節點的kobj，然後調用sysfs_create_link()建立軟鏈接。

device_remove_sys_dev_entry()與其操作正相反，刪除在/sys/dev下建立的軟鏈接。

int device_private_init(struct device *dev)
{
	dev->p = kzalloc(sizeof(*dev->p), GFP_KERNEL);
	if (!dev->p)
		return -ENOMEM;
	dev->p->device = dev;
	klist_init(&dev->p->klist_children, klist_children_get,
		   klist_children_put);
	return 0;
}

device_private_init()分配並初始化dev->p。至於空間的釋放，是等到釋放設備時調用的device_release()中。

 

之前的函數比較散亂，或許找不出一個整體的印象。但下面馬上就要看到重要的部分了，因爲代碼終於攢到了爆發的程度！

/**
 * device_register - register a device with the system.
 * @dev: pointer to the device structure
 *
 * This happens in two clean steps - initialize the device
 * and add it to the system. The two steps can be called
 * separately, but this is the easiest and most common.
 * I.e. you should only call the two helpers separately if
 * have a clearly defined need to use and refcount the device
 * before it is added to the hierarchy.
 *
 * NOTE: _Never_ directly free @dev after calling this function, even
 * if it returned an error! Always use put_device() to give up the
 * reference initialized in this function instead.
 */
int device_register(struct device *dev)
{
	device_initialize(dev);
	return device_add(dev);
}

device_register()是提供給外界註冊設備的接口。它先是調用device_initialize()初始化dev結構，然後調用device_add()將其加入系統中。但要注意，在調用device_register()註冊dev之前，有一些dev結構變量是需要自行設置的。這其中有指明設備位置的struct device *parent,struct bus_type *bus, struct class *class，有指明設備屬性的 const char *init_name, struct device_type *type, const struct attribute_group **groups, void (*release)(struct device *dev), dev_t devt，等等。不同設備的使用方法不同，我們留待之後再具體分析。device_initialize()我們已經看過，下面重點看看device_add()是如何實現的。

int device_add(struct device *dev)
{
	struct device *parent = NULL;
	struct class_interface *class_intf;
	int error = -EINVAL;

	dev = get_device(dev);
	if (!dev)
		goto done;

	if (!dev->p) {
		error = device_private_init(dev);
		if (error)
			goto done;
	}

	/*
	 * for statically allocated devices, which should all be converted
	 * some day, we need to initialize the name. We prevent reading back
	 * the name, and force the use of dev_name()
	 */
	if (dev->init_name) {
		dev_set_name(dev, "%s", dev->init_name);
		dev->init_name = NULL;
	}

	if (!dev_name(dev))
		goto name_error;

	pr_debug("device: '%s': %s\n", dev_name(dev), __func__);

	parent = get_device(dev->parent);
	setup_parent(dev, parent);

	/* use parent numa_node */
	if (parent)
		set_dev_node(dev, dev_to_node(parent));

	/* first, register with generic layer. */
	/* we require the name to be set before, and pass NULL */
	error = kobject_add(&dev->kobj, dev->kobj.parent, NULL);
	if (error)
		goto Error;

	/* notify platform of device entry */
	if (platform_notify)
		platform_notify(dev);

	error = device_create_file(dev, &uevent_attr);
	if (error)
		goto attrError;

	if (MAJOR(dev->devt)) {
		error = device_create_file(dev, &devt_attr);
		if (error)
			goto ueventattrError;

		error = device_create_sys_dev_entry(dev);
		if (error)
			goto devtattrError;

		devtmpfs_create_node(dev);
	}

	error = device_add_class_symlinks(dev);
	if (error)
		goto SymlinkError;
	error = device_add_attrs(dev);
	if (error)
		goto AttrsError;
	error = bus_add_device(dev);
	if (error)
		goto BusError;
	error = dpm_sysfs_add(dev);
	if (error)
		goto DPMError;
	device_pm_add(dev);

	/* Notify clients of device addition.  This call must come
	 * after dpm_sysf_add() and before kobject_uevent().
	 */
	if (dev->bus)
		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
					     BUS_NOTIFY_ADD_DEVICE, dev);

	kobject_uevent(&dev->kobj, KOBJ_ADD);
	bus_probe_device(dev);
	if (parent)
		klist_add_tail(&dev->p->knode_parent,
			       &parent->p->klist_children);

	if (dev->class) {
		mutex_lock(&dev->class->p->class_mutex);
		/* tie the class to the device */
		klist_add_tail(&dev->knode_class,
			       &dev->class->p->class_devices);

		/* notify any interfaces that the device is here */
		list_for_each_entry(class_intf,
				    &dev->class->p->class_interfaces, node)
			if (class_intf->add_dev)
				class_intf->add_dev(dev, class_intf);
		mutex_unlock(&dev->class->p->class_mutex);
	}
done:
	put_device(dev);
	return error;
 DPMError:
	bus_remove_device(dev);
 BusError:
	device_remove_attrs(dev);
 AttrsError:
	device_remove_class_symlinks(dev);
 SymlinkError:
	if (MAJOR(dev->devt))
		device_remove_sys_dev_entry(dev);
 devtattrError:
	if (MAJOR(dev->devt))
		device_remove_file(dev, &devt_attr);
 ueventattrError:
	device_remove_file(dev, &uevent_attr);
 attrError:
	kobject_uevent(&dev->kobj, KOBJ_REMOVE);
	kobject_del(&dev->kobj);
 Error:
	cleanup_device_parent(dev);
	if (parent)
		put_device(parent);
name_error:
	kfree(dev->p);
	dev->p = NULL;
	goto done;
}

device_add()將dev加入設備驅動模型。它先是調用get_device(dev)增加dev的引用計數，然後調用device_private_init()分配和初始化dev->p，調用dev_set_name()設置dev名字。然後是準備將dev加入sysfs，先是用get_device(parent)增加對parent的引用計數(無論是直接掛在parent下還是通過一個類層掛在parent下都要增加parent的引用計數），然後調用setup_parent()找到實際要加入的父kobject，通過kobject_add()加入其下。然後是添加屬性和屬性集合的操作，調用device_create_file()添加uevent屬性，調用device_add_attrs()添加device/type/class預定義的屬性與屬性集合。如果dev有被分配設備號，再用device_create_file()添加dev屬性，並用device_create_sys_dev_entry()在/sys/dev下添加相應的軟鏈接，最後調用devtmpfs_create_node()在/dev下創建相應的設備文件。然後調用device_add_class_symlinks()添加dev與class間的軟鏈接，調用bus_add_device()添加dev與bus間的軟鏈接，並將dev掛入bus的設備鏈表。調用dpm_sysfs_add()增加dev下的power屬性集合，調用device_pm_add()將dev加入dpm_list鏈表。

調用kobject_uevent()發佈KOBJ_ADD消息，調用bus_probe_device()爲dev尋找合適的驅動。如果有parent節點，把dev->p->knode_parent掛入parent->p->klist_children鏈表。如果dev有所屬的class，將dev->knode_class掛在class->p->class_devices上，並調用可能的類設備接口的add_dev()方法。可能對於直接在bus上的設備來說，自然可以調用bus_probe_device()查找驅動，而不與總線直接接觸的設備，則要靠class來發現驅動，這裏的class_interface中的add_dev()方法，就是一個絕好的機會。最後會調用put_device(dev)釋放在函數開頭增加的引用計數。

device_add()要做的事很多，但想想每件事都在情理之中。device是設備驅動模型的基本元素，在class、bus、dev、devices中都有它的身影。device_add()要適應各種類型的設備註冊，自然會越來越複雜。可以說文件開頭定義的內部函數，差不多都是爲了這裏服務的。

 

void device_unregister(struct device *dev)
{
	pr_debug("device: '%s': %s\n", dev_name(dev), __func__);
	device_del(dev);
	put_device(dev);
}

有註冊自然又註銷。device_unregister()就是用於將dev從系統中註銷，並釋放創建時產生的引用計數。

void device_del(struct device *dev)
{
	struct device *parent = dev->parent;
	struct class_interface *class_intf;

	/* Notify clients of device removal.  This call must come
	 * before dpm_sysfs_remove().
	 */
	if (dev->bus)
		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
					     BUS_NOTIFY_DEL_DEVICE, dev);
	device_pm_remove(dev);
	dpm_sysfs_remove(dev);
	if (parent)
		klist_del(&dev->p->knode_parent);
	if (MAJOR(dev->devt)) {
		devtmpfs_delete_node(dev);
		device_remove_sys_dev_entry(dev);
		device_remove_file(dev, &devt_attr);
	}
	if (dev->class) {
		device_remove_class_symlinks(dev);

		mutex_lock(&dev->class->p->class_mutex);
		/* notify any interfaces that the device is now gone */
		list_for_each_entry(class_intf,
				    &dev->class->p->class_interfaces, node)
			if (class_intf->remove_dev)
				class_intf->remove_dev(dev, class_intf);
		/* remove the device from the class list */
		klist_del(&dev->knode_class);
		mutex_unlock(&dev->class->p->class_mutex);
	}
	device_remove_file(dev, &uevent_attr);
	device_remove_attrs(dev);
	bus_remove_device(dev);

	/*
	 * Some platform devices are driven without driver attached
	 * and managed resources may have been acquired.  Make sure
	 * all resources are released.
	 */
	devres_release_all(dev);

	/* Notify the platform of the removal, in case they
	 * need to do anything...
	 */
	if (platform_notify_remove)
		platform_notify_remove(dev);
	kobject_uevent(&dev->kobj, KOBJ_REMOVE);
	cleanup_device_parent(dev);
	kobject_del(&dev->kobj);
	put_device(parent);
}

device_del()是與device_add()相對的函數，進行實際的將dev從系統中脫離的工作。這其中既有將dev從設備驅動模型各種鏈表中脫離的工作，又有將dev從sysfs的各個角落刪除的工作。大致流程與dev_add()相對，就不一一介紹。

 

爆發結束，下面來看一些比較輕鬆的函數。

/**
 * device_get_devnode - path of device node file
 * @dev: device
 * @mode: returned file access mode
 * @tmp: possibly allocated string
 *
 * Return the relative path of a possible device node.
 * Non-default names may need to allocate a memory to compose
 * a name. This memory is returned in tmp and needs to be
 * freed by the caller.
 */
const char *device_get_devnode(struct device *dev,
			       mode_t *mode, const char **tmp)
{
	char *s;

	*tmp = NULL;

	/* the device type may provide a specific name */
	if (dev->type && dev->type->devnode)
		*tmp = dev->type->devnode(dev, mode);
	if (*tmp)
		return *tmp;

	/* the class may provide a specific name */
	if (dev->class && dev->class->devnode)
		*tmp = dev->class->devnode(dev, mode);
	if (*tmp)
		return *tmp;

	/* return name without allocation, tmp == NULL */
	if (strchr(dev_name(dev), '!') == NULL)
		return dev_name(dev);

	/* replace '!' in the name with '/' */
	*tmp = kstrdup(dev_name(dev), GFP_KERNEL);
	if (!*tmp)
		return NULL;
	while ((s = strchr(*tmp, '!')))
		s[0] = '/';
	return *tmp;
}

device_get_devnode()返回設備的路徑名。不過似乎可以由device_type或者class定義一些獨特的返回名稱。

static struct device *next_device(struct klist_iter *i)
{
	struct klist_node *n = klist_next(i);
	struct device *dev = NULL;
	struct device_private *p;

	if (n) {
		p = to_device_private_parent(n);
		dev = p->device;
	}
	return dev;
}

int device_for_each_child(struct device *parent, void *data,
			  int (*fn)(struct device *dev, void *data))
{
	struct klist_iter i;
	struct device *child;
	int error = 0;

	if (!parent->p)
		return 0;

	klist_iter_init(&parent->p->klist_children, &i);
	while ((child = next_device(&i)) && !error)
		error = fn(child, data);
	klist_iter_exit(&i);
	return error;
}

struct device *device_find_child(struct device *parent, void *data,
				 int (*match)(struct device *dev, void *data))
{
	struct klist_iter i;
	struct device *child;

	if (!parent)
		return NULL;

	klist_iter_init(&parent->p->klist_children, &i);
	while ((child = next_device(&i)))
		if (match(child, data) && get_device(child))
			break;
	klist_iter_exit(&i);
	return child;
}

device_for_each_child()對dev下的每個子device，都調用一遍特定的處理函數。

device_find_child()則是查找dev下特點的子device，查找使用特定的match函數。

這兩個遍歷過程都使用了klist特有的遍歷函數，支持遍歷過程中的節點刪除等功能。next_device()則是爲了遍歷方便封裝的一個內部函數。

 

下面本該是root_device註冊相關的代碼。但經過檢查，linux內核中使用到的root_device很少見，而且在sysfs中也未能找到一個實際的例子。所以root_device即使還未被棄用，也並非主流，我們將其跳過。

與kobject和kset類似，device也爲我們提供了快速device創建方法，下面就看看吧。

static void device_create_release(struct device *dev)
{
	pr_debug("device: '%s': %s\n", dev_name(dev), __func__);
	kfree(dev);
}

struct device *device_create_vargs(struct class *class, struct device *parent,
				   dev_t devt, void *drvdata, const char *fmt,
				   va_list args)
{
	struct device *dev = NULL;
	int retval = -ENODEV;

	if (class == NULL || IS_ERR(class))
		goto error;

	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
	if (!dev) {
		retval = -ENOMEM;
		goto error;
	}

	dev->devt = devt;
	dev->class = class;
	dev->parent = parent;
	dev->release = device_create_release;
	dev_set_drvdata(dev, drvdata);

	retval = kobject_set_name_vargs(&dev->kobj, fmt, args);
	if (retval)
		goto error;

	retval = device_register(dev);
	if (retval)
		goto error;

	return dev;

error:
	put_device(dev);
	return ERR_PTR(retval);
}

struct device *device_create(struct class *class, struct device *parent,
			     dev_t devt, void *drvdata, const char *fmt, ...)
{
	va_list vargs;
	struct device *dev;

	va_start(vargs, fmt);
	dev = device_create_vargs(class, parent, devt, drvdata, fmt, vargs);
	va_end(vargs);
	return dev;
}

這裏的device_create()提供了一個快速的dev創建註冊方法。只是中間沒有提供設置device_type的方法，或許是這樣的device已經夠特立獨行了，不需要搞出一類來。

 

static int __match_devt(struct device *dev, void *data)
{
	dev_t *devt = data;

	return dev->devt == *devt;
}

void device_destroy(struct class *class, dev_t devt)
{
	struct device *dev;

	dev = class_find_device(class, NULL, &devt, __match_devt);
	if (dev) {
		put_device(dev);
		device_unregister(dev);
	}
}

device_destroy()就是與device_create()相對的註銷函數。至於這裏爲什麼會多一個put_device(dev)，也很簡單，因爲在class_find_device()找到dev時，調用了get_device()。

struct device *class_find_device(struct class *class, struct device *start,
				 void *data,
				 int (*match)(struct device *, void *))
{
	struct class_dev_iter iter;
	struct device *dev;

	if (!class)
		return NULL;
	if (!class->p) {
		WARN(1, "%s called for class '%s' before it was initialized",
		     __func__, class->name);
		return NULL;
	}

	class_dev_iter_init(&iter, class, start, NULL);
	while ((dev = class_dev_iter_next(&iter))) {
		if (match(dev, data)) {
			get_device(dev);
			break;
		}
	}
	class_dev_iter_exit(&iter);

	return dev;
}

class_find_device()本來是class.c中的內容，其實現也於之前將的遍歷dev->p->klist_children類似，無非是在klist提供的遍歷方法上加以封裝。但我們這裏列出class_find_device()的實現與使用它的device_destroy()，卻是爲了更好地分析這個調用流程中dev是如何被保護的。它實際上是經歷了三個保護手段：首先在class_dev_iter_next()->klist_next()中，是受到struct klist中 spinlock_t k_lock保護的。在找到下一點並解鎖之前，就增加了struct klist_node中的struct kref n_ref引用計數。在當前的next()調用完，到下一個next()調用之前，都是受這個增加的引用計數保護的。再看class_find_device()中，使用get_device(dev)增加了dev本身的引用計數保護(當然也要追溯到kobj->kref中)，這是第三種保護。知道device_destroy()中主動調用put_device(dev)纔去除了這種保護。

本來對dev的保護，應該完全是由dev中的引用計數完成的。但實際上這種保護很多時候是間接完成的。例如這裏的klist中的自旋鎖，klist_node中的引用計數，都不過是爲了保持class的設備鏈表中對dev的引用計數不消失，這是一種間接保護的手段，保證了這中間即使外界主動釋放class設備鏈表對dev的引用計數，dev仍然不會被實際註銷。這種曲折的聯繫，才真正發揮了引用計數的作用，構成設備驅動模型獨特的魅力。

int device_rename(struct device *dev, char *new_name)
{
	char *old_device_name = NULL;
	int error;

	dev = get_device(dev);
	if (!dev)
		return -EINVAL;

	pr_debug("device: '%s': %s: renaming to '%s'\n", dev_name(dev),
		 __func__, new_name);
old_device_name = kstrdup(dev_name(dev), GFP_KERNEL);
	if (!old_device_name) {
		error = -ENOMEM;
		goto out;
	}

	error = kobject_rename(&dev->kobj, new_name);
	if (error)
		goto out;
if (dev->class) {
		error = sysfs_create_link_nowarn(&dev->class->p->class_subsys.kobj,
						 &dev->kobj, dev_name(dev));
		if (error)
			goto out;
		sysfs_remove_link(&dev->class->p->class_subsys.kobj,
				  old_device_name);
	}
out:
	put_device(dev);

	kfree(old_device_name);

	return error;
}

device_rename()是供設備註冊後改變名稱用的，除了改變/sys/devices下地名稱，還改變了/sys/class下地軟鏈接名稱。前者很自然，但後者卻很難想到。即使簡單的地方，經過重重調試，我們也會驚訝於linux的心細如髮。

static int device_move_class_links(struct device *dev,
				   struct device *old_parent,
				   struct device *new_parent)
{
	int error = 0;
	if (old_parent)
		sysfs_remove_link(&dev->kobj, "device");
	if (new_parent)
		error = sysfs_create_link(&dev->kobj, &new_parent->kobj,
					  "device");
	return error;
#endif
}

device_move_class_links()只是一個內部函數，後面還有操縱它的那隻手。這裏的device_move_class_links顯得很名不副實，並沒用操作class中軟鏈接的舉動。這很正常，因爲在sysfs中軟鏈接是針對kobject來說的，所以即使位置變掉了，軟鏈接還是很很準確地定位。

/**
 * device_move - moves a device to a new parent
 * @dev: the pointer to the struct device to be moved
 * @new_parent: the new parent of the device (can by NULL)
 * @dpm_order: how to reorder the dpm_list
 */
int device_move(struct device *dev, struct device *new_parent,
		enum dpm_order dpm_order)
{
	int error;
	struct device *old_parent;
	struct kobject *new_parent_kobj;

	dev = get_device(dev);
	if (!dev)
		return -EINVAL;

	device_pm_lock();
	new_parent = get_device(new_parent);
	new_parent_kobj = get_device_parent(dev, new_parent);

	pr_debug("device: '%s': %s: moving to '%s'\n", dev_name(dev),
		 __func__, new_parent ? dev_name(new_parent) : "<NULL>");
	error = kobject_move(&dev->kobj, new_parent_kobj);
	if (error) {
		cleanup_glue_dir(dev, new_parent_kobj);
		put_device(new_parent);
		goto out;
	}
	old_parent = dev->parent;
	dev->parent = new_parent;
	if (old_parent)
		klist_remove(&dev->p->knode_parent);
	if (new_parent) {
		klist_add_tail(&dev->p->knode_parent,
			       &new_parent->p->klist_children);
		set_dev_node(dev, dev_to_node(new_parent));
	}

	if (!dev->class)
		goto out_put;
	error = device_move_class_links(dev, old_parent, new_parent);
	if (error) {
		/* We ignore errors on cleanup since we're hosed anyway... */
		device_move_class_links(dev, new_parent, old_parent);
		if (!kobject_move(&dev->kobj, &old_parent->kobj)) {
			if (new_parent)
				klist_remove(&dev->p->knode_parent);
			dev->parent = old_parent;
			if (old_parent) {
				klist_add_tail(&dev->p->knode_parent,
					       &old_parent->p->klist_children);
				set_dev_node(dev, dev_to_node(old_parent));
			}
		}
		cleanup_glue_dir(dev, new_parent_kobj);
		put_device(new_parent);
		goto out;
	}
	switch (dpm_order) {
	case DPM_ORDER_NONE:
		break;
	case DPM_ORDER_DEV_AFTER_PARENT:
		device_pm_move_after(dev, new_parent);
		break;
	case DPM_ORDER_PARENT_BEFORE_DEV:
		device_pm_move_before(new_parent, dev);
		break;
	case DPM_ORDER_DEV_LAST:
		device_pm_move_last(dev);
		break;
	}
out_put:
	put_device(old_parent);
out:
	device_pm_unlock();
	put_device(dev);
	return error;
}

device_move()就是將dev移到一個新的parent下。但也有可能這個parent是空的。大部分操作圍繞在引用計數上，get_device()，put_device()。而且換了新的parent，到底要加到sysfs中哪個目錄下，還要再調用get_device_parent()研究一下。主要的操作就是kobject_move()和device_move_class_links()。因爲在sysfs中軟鏈接是針對kobject來說的，所以即使位置變掉了，軟鏈接還是很很準確地定位，所以在/sys/dev、/sys/bus、/sys/class中的軟鏈接都不用變，這實在是sysfs的一大優勢。除此之外，device_move()還涉及到電源管理的問題，device移動影響到dev在dpm_list上的位置，我們對此不瞭解，先忽略之。

void device_shutdown(void)
{
	struct device *dev, *devn;

	list_for_each_entry_safe_reverse(dev, devn, &devices_kset->list,
				kobj.entry) {
		if (dev->bus && dev->bus->shutdown) {
			dev_dbg(dev, "shutdown\n");
			dev->bus->shutdown(dev);
		} else if (dev->driver && dev->driver->shutdown) {
			dev_dbg(dev, "shutdown\n");
			dev->driver->shutdown(dev);
		}
	}
	kobject_put(sysfs_dev_char_kobj);
	kobject_put(sysfs_dev_block_kobj);
	kobject_put(dev_kobj);
	async_synchronize_full();
}

這個device_shutdown()是在系統關閉時才調用的。它動用了很少使用的devices_kset，從而可以遍歷到每個註冊到sysfs上的設備，調用相應的總線或驅動定義的shutdown()函數。提起這個，還是在device_initialize()中將dev->kobj->kset統一設爲devices_kset的。原來設備雖然有不同的parent，但kset還是一樣的。這樣我們就能理解/sys/devices下的頂層設備目錄是怎麼來的，因爲沒用parent，就在調用kobject_add()時將kset->kobj當成了parent，所以會直接掛在頂層目錄下。這樣的目錄大致有pci0000:00、virtual等等。

看完了core.c，我有種明白機器人也是由零件組成的的感覺。linux設備驅動模型的大門已經打開了四分之一。隨着分析的深入，我們大概也會越來越明白linux的良苦用心。