[net-next-2.6.git] / drivers / vhost / vhost.h

#ifndef _VHOST_H
#define _VHOST_H

#include <linux/eventfd.h>
#include <linux/vhost.h>
#include <linux/mm.h>
#include <linux/mutex.h>
#include <linux/poll.h>
#include <linux/file.h>
#include <linux/skbuff.h>
#include <linux/uio.h>
#include <linux/virtio_config.h>
#include <linux/virtio_ring.h>
#include <asm/atomic.h>

struct vhost_device;

enum {
	/* Enough place for all fragments, head, and virtio net header. */
	VHOST_NET_MAX_SG = MAX_SKB_FRAGS + 2,
};

struct vhost_work;
typedef void (*vhost_work_fn_t)(struct vhost_work *work);

struct vhost_work {
	struct list_head	  node;
	vhost_work_fn_t		  fn;
	wait_queue_head_t	  done;
	int			  flushing;
	unsigned		  queue_seq;
	unsigned		  done_seq;
};

/* Poll a file (eventfd or socket) */
/* Note: there's nothing vhost specific about this structure. */
struct vhost_poll {
	poll_table                table;
	wait_queue_head_t        *wqh;
	wait_queue_t              wait;
	struct vhost_work	  work;
	unsigned long		  mask;
	struct vhost_dev	 *dev;
};

void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
		     unsigned long mask, struct vhost_dev *dev);
void vhost_poll_start(struct vhost_poll *poll, struct file *file);
void vhost_poll_stop(struct vhost_poll *poll);
void vhost_poll_flush(struct vhost_poll *poll);
void vhost_poll_queue(struct vhost_poll *poll);

struct vhost_log {
	u64 addr;
	u64 len;
};

/* The virtqueue structure describes a queue attached to a device. */
struct vhost_virtqueue {
	struct vhost_dev *dev;

	/* The actual ring of buffers. */
	struct mutex mutex;
	unsigned int num;
	struct vring_desc __user *desc;
	struct vring_avail __user *avail;
	struct vring_used __user *used;
	struct file *kick;
	struct file *call;
	struct file *error;
	struct eventfd_ctx *call_ctx;
	struct eventfd_ctx *error_ctx;
	struct eventfd_ctx *log_ctx;

	struct vhost_poll poll;

	/* The routine to call when the Guest pings us, or timeout. */
	vhost_work_fn_t handle_kick;

	/* Last available index we saw. */
	u16 last_avail_idx;

	/* Caches available index value from user. */
	u16 avail_idx;

	/* Last index we used. */
	u16 last_used_idx;

	/* Used flags */
	u16 used_flags;

	/* Log writes to used structure. */
	bool log_used;
	u64 log_addr;

	struct iovec indirect[VHOST_NET_MAX_SG];
	struct iovec iov[VHOST_NET_MAX_SG];
	struct iovec hdr[VHOST_NET_MAX_SG];
	size_t vhost_hlen;
	size_t sock_hlen;
	struct vring_used_elem heads[VHOST_NET_MAX_SG];
	/* We use a kind of RCU to access private pointer.
	 * All readers access it from worker, which makes it possible to
	 * flush the vhost_work instead of synchronize_rcu. Therefore readers do
	 * not need to call rcu_read_lock/rcu_read_unlock: the beginning of
	 * vhost_work execution acts instead of rcu_read_lock() and the end of
	 * vhost_work execution acts instead of rcu_read_lock().
	 * Writers use virtqueue mutex. */
	void *private_data;
	/* Log write descriptors */
	void __user *log_base;
	struct vhost_log log[VHOST_NET_MAX_SG];
};

struct vhost_dev {
	/* Readers use RCU to access memory table pointer
	 * log base pointer and features.
	 * Writers use mutex below.*/
	struct vhost_memory *memory;
	struct mm_struct *mm;
	struct mutex mutex;
	unsigned acked_features;
	struct vhost_virtqueue *vqs;
	int nvqs;
	struct file *log_file;
	struct eventfd_ctx *log_ctx;
	spinlock_t work_lock;
	struct list_head work_list;
	struct task_struct *worker;
};

long vhost_dev_init(struct vhost_dev *, struct vhost_virtqueue *vqs, int nvqs);
long vhost_dev_check_owner(struct vhost_dev *);
long vhost_dev_reset_owner(struct vhost_dev *);
void vhost_dev_cleanup(struct vhost_dev *);
long vhost_dev_ioctl(struct vhost_dev *, unsigned int ioctl, unsigned long arg);
int vhost_vq_access_ok(struct vhost_virtqueue *vq);
int vhost_log_access_ok(struct vhost_dev *);

int vhost_get_vq_desc(struct vhost_dev *, struct vhost_virtqueue *,
		      struct iovec iov[], unsigned int iov_count,
		      unsigned int *out_num, unsigned int *in_num,
		      struct vhost_log *log, unsigned int *log_num);
void vhost_discard_vq_desc(struct vhost_virtqueue *, int n);

int vhost_add_used(struct vhost_virtqueue *, unsigned int head, int len);
int vhost_add_used_n(struct vhost_virtqueue *, struct vring_used_elem *heads,
		     unsigned count);
void vhost_add_used_and_signal(struct vhost_dev *, struct vhost_virtqueue *,
			       unsigned int id, int len);
void vhost_add_used_and_signal_n(struct vhost_dev *, struct vhost_virtqueue *,
			       struct vring_used_elem *heads, unsigned count);
void vhost_signal(struct vhost_dev *, struct vhost_virtqueue *);
void vhost_disable_notify(struct vhost_virtqueue *);
bool vhost_enable_notify(struct vhost_virtqueue *);

int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log,
		    unsigned int log_num, u64 len);

#define vq_err(vq, fmt, ...) do {                                  \
		pr_debug(pr_fmt(fmt), ##__VA_ARGS__);       \
		if ((vq)->error_ctx)                               \
				eventfd_signal((vq)->error_ctx, 1);\
	} while (0)

enum {
	VHOST_FEATURES = (1 << VIRTIO_F_NOTIFY_ON_EMPTY) |
			 (1 << VIRTIO_RING_F_INDIRECT_DESC) |
			 (1 << VHOST_F_LOG_ALL) |
			 (1 << VHOST_NET_F_VIRTIO_NET_HDR) |
			 (1 << VIRTIO_NET_F_MRG_RXBUF),
};

static inline int vhost_has_feature(struct vhost_dev *dev, int bit)
{
	unsigned acked_features = rcu_dereference(dev->acked_features);
	return acked_features & (1 << bit);
}

#endif
Commit	Line	Data
3a4d5c94 MT	1	#ifndef _VHOST_H
	2	#define _VHOST_H
	3
	4	#include <linux/eventfd.h>
	5	#include <linux/vhost.h>
	6	#include <linux/mm.h>
	7	#include <linux/mutex.h>
3a4d5c94 MT	8	#include <linux/poll.h>
	9	#include <linux/file.h>
	10	#include <linux/skbuff.h>
	11	#include <linux/uio.h>
	12	#include <linux/virtio_config.h>
	13	#include <linux/virtio_ring.h>
c23f3445	14	#include <asm/atomic.h>
3a4d5c94 MT	15
	16	struct vhost_device;
	17
	18	enum {
	19	/* Enough place for all fragments, head, and virtio net header. */
	20	VHOST_NET_MAX_SG = MAX_SKB_FRAGS + 2,
	21	};
	22
c23f3445 TH	23	struct vhost_work;
	24	typedef void (vhost_work_fn_t)(struct vhost_work work);
	25
	26	struct vhost_work {
	27	struct list_head node;
	28	vhost_work_fn_t fn;
	29	wait_queue_head_t done;
	30	int flushing;
	31	unsigned queue_seq;
	32	unsigned done_seq;
	33	};
	34
3a4d5c94 MT	35	/* Poll a file (eventfd or socket) */
	36	/* Note: there's nothing vhost specific about this structure. */
	37	struct vhost_poll {
	38	poll_table table;
	39	wait_queue_head_t *wqh;
	40	wait_queue_t wait;
c23f3445	41	struct vhost_work work;
3a4d5c94	42	unsigned long mask;
c23f3445	43	struct vhost_dev *dev;
3a4d5c94 MT	44	};
3a4d5c94 MT	45
c23f3445 TH	46	void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
c23f3445 TH	47	unsigned long mask, struct vhost_dev *dev);
3a4d5c94 MT	48	void vhost_poll_start(struct vhost_poll poll, struct file file);
	49	void vhost_poll_stop(struct vhost_poll *poll);
	50	void vhost_poll_flush(struct vhost_poll *poll);
	51	void vhost_poll_queue(struct vhost_poll *poll);
	52
	53	struct vhost_log {
	54	u64 addr;
	55	u64 len;
	56	};
	57
	58	/* The virtqueue structure describes a queue attached to a device. */
	59	struct vhost_virtqueue {
	60	struct vhost_dev *dev;
	61
	62	/* The actual ring of buffers. */
	63	struct mutex mutex;
	64	unsigned int num;
	65	struct vring_desc __user *desc;
	66	struct vring_avail __user *avail;
	67	struct vring_used __user *used;
	68	struct file *kick;
	69	struct file *call;
	70	struct file *error;
	71	struct eventfd_ctx *call_ctx;
	72	struct eventfd_ctx *error_ctx;
	73	struct eventfd_ctx *log_ctx;
	74
	75	struct vhost_poll poll;
	76
	77	/* The routine to call when the Guest pings us, or timeout. */
c23f3445	78	vhost_work_fn_t handle_kick;
3a4d5c94 MT	79
	80	/* Last available index we saw. */
	81	u16 last_avail_idx;
	82
	83	/* Caches available index value from user. */
	84	u16 avail_idx;
	85
	86	/* Last index we used. */
	87	u16 last_used_idx;
	88
	89	/* Used flags */
	90	u16 used_flags;
	91
	92	/* Log writes to used structure. */
	93	bool log_used;
	94	u64 log_addr;
	95
	96	struct iovec indirect[VHOST_NET_MAX_SG];
	97	struct iovec iov[VHOST_NET_MAX_SG];
	98	struct iovec hdr[VHOST_NET_MAX_SG];
8dd014ad DS	99	size_t vhost_hlen;
	100	size_t sock_hlen;
	101	struct vring_used_elem heads[VHOST_NET_MAX_SG];
3a4d5c94	102	/* We use a kind of RCU to access private pointer.
c23f3445 TH	103	* All readers access it from worker, which makes it possible to
c23f3445 TH	104	* flush the vhost_work instead of synchronize_rcu. Therefore readers do
3a4d5c94	105	* not need to call rcu_read_lock/rcu_read_unlock: the beginning of
c23f3445 TH	106	* vhost_work execution acts instead of rcu_read_lock() and the end of
c23f3445 TH	107	* vhost_work execution acts instead of rcu_read_lock().
3a4d5c94 MT	108	* Writers use virtqueue mutex. */
	109	void *private_data;
	110	/* Log write descriptors */
	111	void __user *log_base;
	112	struct vhost_log log[VHOST_NET_MAX_SG];
	113	};
	114
	115	struct vhost_dev {
	116	/* Readers use RCU to access memory table pointer
	117	* log base pointer and features.
	118	* Writers use mutex below.*/
	119	struct vhost_memory *memory;
	120	struct mm_struct *mm;
	121	struct mutex mutex;
	122	unsigned acked_features;
	123	struct vhost_virtqueue *vqs;
	124	int nvqs;
	125	struct file *log_file;
	126	struct eventfd_ctx *log_ctx;
c23f3445 TH	127	spinlock_t work_lock;
	128	struct list_head work_list;
	129	struct task_struct *worker;
3a4d5c94 MT	130	};
	131
	132	long vhost_dev_init(struct vhost_dev , struct vhost_virtqueue vqs, int nvqs);
	133	long vhost_dev_check_owner(struct vhost_dev *);
	134	long vhost_dev_reset_owner(struct vhost_dev *);
	135	void vhost_dev_cleanup(struct vhost_dev *);
	136	long vhost_dev_ioctl(struct vhost_dev *, unsigned int ioctl, unsigned long arg);
	137	int vhost_vq_access_ok(struct vhost_virtqueue *vq);
	138	int vhost_log_access_ok(struct vhost_dev *);
	139
d5675bd2 MT	140	int vhost_get_vq_desc(struct vhost_dev , struct vhost_virtqueue ,
	141	struct iovec iov[], unsigned int iov_count,
	142	unsigned int out_num, unsigned int in_num,
	143	struct vhost_log log, unsigned int log_num);
8dd014ad	144	void vhost_discard_vq_desc(struct vhost_virtqueue *, int n);
3a4d5c94 MT	145
3a4d5c94 MT	146	int vhost_add_used(struct vhost_virtqueue *, unsigned int head, int len);
8dd014ad DS	147	int vhost_add_used_n(struct vhost_virtqueue , struct vring_used_elem heads,
8dd014ad DS	148	unsigned count);
3a4d5c94	149	void vhost_add_used_and_signal(struct vhost_dev , struct vhost_virtqueue ,
8dd014ad DS	150	unsigned int id, int len);
	151	void vhost_add_used_and_signal_n(struct vhost_dev , struct vhost_virtqueue ,
	152	struct vring_used_elem *heads, unsigned count);
	153	void vhost_signal(struct vhost_dev , struct vhost_virtqueue );
3a4d5c94 MT	154	void vhost_disable_notify(struct vhost_virtqueue *);
	155	bool vhost_enable_notify(struct vhost_virtqueue *);
	156
	157	int vhost_log_write(struct vhost_virtqueue vq, struct vhost_log log,
	158	unsigned int log_num, u64 len);
	159
3a4d5c94 MT	160	#define vq_err(vq, fmt, ...) do { \
	161	pr_debug(pr_fmt(fmt), ##__VA_ARGS__); \
	162	if ((vq)->error_ctx) \
	163	eventfd_signal((vq)->error_ctx, 1);\
	164	} while (0)
	165
	166	enum {
	167	VHOST_FEATURES = (1 << VIRTIO_F_NOTIFY_ON_EMPTY) \|
	168	(1 << VIRTIO_RING_F_INDIRECT_DESC) \|
	169	(1 << VHOST_F_LOG_ALL) \|
8dd014ad DS	170	(1 << VHOST_NET_F_VIRTIO_NET_HDR) \|
8dd014ad DS	171	(1 << VIRTIO_NET_F_MRG_RXBUF),
3a4d5c94 MT	172	};
	173
	174	static inline int vhost_has_feature(struct vhost_dev *dev, int bit)
	175	{
	176	unsigned acked_features = rcu_dereference(dev->acked_features);
	177	return acked_features & (1 << bit);
	178	}
	179
	180	#endif