[mirror_ubuntu-zesty-kernel.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/uio.h>
#include <linux/virtio_config.h>
#include <linux/virtio_ring.h>
#include <linux/atomic.h>

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

#define VHOST_WORK_QUEUED 1
struct vhost_work {
	struct llist_node	  node;
	vhost_work_fn_t		  fn;
	wait_queue_head_t	  done;
	int			  flushing;
	unsigned		  queue_seq;
	unsigned		  done_seq;
	unsigned long		  flags;
};

/* 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_work_init(struct vhost_work *work, vhost_work_fn_t fn);
void vhost_work_queue(struct vhost_dev *dev, struct vhost_work *work);
bool vhost_has_work(struct vhost_dev *dev);

void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
		     unsigned long mask, struct vhost_dev *dev);
int 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);
void vhost_work_flush(struct vhost_dev *dev, struct vhost_work *work);
long vhost_vring_ioctl(struct vhost_dev *d, int ioctl, void __user *argp);

struct vhost_log {
	u64 addr;
	u64 len;
};

#define START(node) ((node)->start)
#define LAST(node) ((node)->last)

struct vhost_umem_node {
	struct rb_node rb;
	struct list_head link;
	__u64 start;
	__u64 last;
	__u64 size;
	__u64 userspace_addr;
	__u32 perm;
	__u32 flags_padding;
	__u64 __subtree_last;
};

struct vhost_umem {
	struct rb_root umem_tree;
	struct list_head umem_list;
	int numem;
};

/* 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;

	/* Last used index value we have signalled on */
	u16 signalled_used;

	/* Last used index value we have signalled on */
	bool signalled_used_valid;

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

	struct iovec iov[UIO_MAXIOV];
	struct iovec iotlb_iov[64];
	struct iovec *indirect;
	struct vring_used_elem *heads;
	/* Protected by virtqueue mutex. */
	struct vhost_umem *umem;
	struct vhost_umem *iotlb;
	void *private_data;
	u64 acked_features;
	/* Log write descriptors */
	void __user *log_base;
	struct vhost_log *log;

	/* Ring endianness. Defaults to legacy native endianness.
	 * Set to true when starting a modern virtio device. */
	bool is_le;
#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
	/* Ring endianness requested by userspace for cross-endian support. */
	bool user_be;
#endif
	u32 busyloop_timeout;
};

struct vhost_msg_node {
  struct vhost_msg msg;
  struct vhost_virtqueue *vq;
  struct list_head node;
};

struct vhost_dev {
	struct mm_struct *mm;
	struct mutex mutex;
	struct vhost_virtqueue **vqs;
	int nvqs;
	struct file *log_file;
	struct eventfd_ctx *log_ctx;
	struct llist_head work_list;
	struct task_struct *worker;
	struct vhost_umem *umem;
	struct vhost_umem *iotlb;
	spinlock_t iotlb_lock;
	struct list_head read_list;
	struct list_head pending_list;
	wait_queue_head_t wait;
};

void vhost_dev_init(struct vhost_dev *, struct vhost_virtqueue **vqs, int nvqs);
long vhost_dev_set_owner(struct vhost_dev *dev);
bool vhost_dev_has_owner(struct vhost_dev *dev);
long vhost_dev_check_owner(struct vhost_dev *);
struct vhost_umem *vhost_dev_reset_owner_prepare(void);
void vhost_dev_reset_owner(struct vhost_dev *, struct vhost_umem *);
void vhost_dev_cleanup(struct vhost_dev *, bool locked);
void vhost_dev_stop(struct vhost_dev *);
long vhost_dev_ioctl(struct vhost_dev *, unsigned int ioctl, void __user *argp);
long vhost_vring_ioctl(struct vhost_dev *d, int ioctl, void __user *argp);
int vhost_vq_access_ok(struct vhost_virtqueue *vq);
int vhost_log_access_ok(struct vhost_dev *);

int vhost_get_vq_desc(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_vq_init_access(struct vhost_virtqueue *);
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_dev *, struct vhost_virtqueue *);
bool vhost_vq_avail_empty(struct vhost_dev *, struct vhost_virtqueue *);
bool vhost_enable_notify(struct vhost_dev *, struct vhost_virtqueue *);

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

struct vhost_msg_node *vhost_new_msg(struct vhost_virtqueue *vq, int type);
void vhost_enqueue_msg(struct vhost_dev *dev,
		       struct list_head *head,
		       struct vhost_msg_node *node);
struct vhost_msg_node *vhost_dequeue_msg(struct vhost_dev *dev,
					 struct list_head *head);
unsigned int vhost_chr_poll(struct file *file, struct vhost_dev *dev,
			    poll_table *wait);
ssize_t vhost_chr_read_iter(struct vhost_dev *dev, struct iov_iter *to,
			    int noblock);
ssize_t vhost_chr_write_iter(struct vhost_dev *dev,
			     struct iov_iter *from);
int vhost_init_device_iotlb(struct vhost_dev *d, bool enabled);

#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 = (1ULL << VIRTIO_F_NOTIFY_ON_EMPTY) |
			 (1ULL << VIRTIO_RING_F_INDIRECT_DESC) |
			 (1ULL << VIRTIO_RING_F_EVENT_IDX) |
			 (1ULL << VHOST_F_LOG_ALL) |
			 (1ULL << VIRTIO_F_ANY_LAYOUT) |
			 (1ULL << VIRTIO_F_VERSION_1)
};

static inline bool vhost_has_feature(struct vhost_virtqueue *vq, int bit)
{
	return vq->acked_features & (1ULL << bit);
}

#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
{
	return vq->is_le;
}
#else
static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
{
	return virtio_legacy_is_little_endian() || vq->is_le;
}
#endif

/* Memory accessors */
static inline u16 vhost16_to_cpu(struct vhost_virtqueue *vq, __virtio16 val)
{
	return __virtio16_to_cpu(vhost_is_little_endian(vq), val);
}

static inline __virtio16 cpu_to_vhost16(struct vhost_virtqueue *vq, u16 val)
{
	return __cpu_to_virtio16(vhost_is_little_endian(vq), val);
}

static inline u32 vhost32_to_cpu(struct vhost_virtqueue *vq, __virtio32 val)
{
	return __virtio32_to_cpu(vhost_is_little_endian(vq), val);
}

static inline __virtio32 cpu_to_vhost32(struct vhost_virtqueue *vq, u32 val)
{
	return __cpu_to_virtio32(vhost_is_little_endian(vq), val);
}

static inline u64 vhost64_to_cpu(struct vhost_virtqueue *vq, __virtio64 val)
{
	return __virtio64_to_cpu(vhost_is_little_endian(vq), val);
}

static inline __virtio64 cpu_to_vhost64(struct vhost_virtqueue *vq, u64 val)
{
	return __cpu_to_virtio64(vhost_is_little_endian(vq), val);
}
#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>
3a4d5c94 MT	9	#include <linux/file.h>
3a4d5c94 MT	10	#include <linux/uio.h>
	11	#include <linux/virtio_config.h>
	12	#include <linux/virtio_ring.h>
60063497	13	#include <linux/atomic.h>
3a4d5c94	14
c23f3445 TH	15	struct vhost_work;
	16	typedef void (vhost_work_fn_t)(struct vhost_work work);
	17
04b96e55	18	#define VHOST_WORK_QUEUED 1
c23f3445	19	struct vhost_work {
04b96e55	20	struct llist_node node;
c23f3445 TH	21	vhost_work_fn_t fn;
	22	wait_queue_head_t done;
	23	int flushing;
	24	unsigned queue_seq;
	25	unsigned done_seq;
04b96e55	26	unsigned long flags;
c23f3445 TH	27	};
c23f3445 TH	28
3a4d5c94 MT	29	/* Poll a file (eventfd or socket) */
	30	/* Note: there's nothing vhost specific about this structure. */
	31	struct vhost_poll {
	32	poll_table table;
	33	wait_queue_head_t *wqh;
	34	wait_queue_t wait;
c23f3445	35	struct vhost_work work;
3a4d5c94	36	unsigned long mask;
c23f3445	37	struct vhost_dev *dev;
3a4d5c94 MT	38	};
3a4d5c94 MT	39
163049ae SH	40	void vhost_work_init(struct vhost_work *work, vhost_work_fn_t fn);
163049ae SH	41	void vhost_work_queue(struct vhost_dev dev, struct vhost_work work);
526d3e7f	42	bool vhost_has_work(struct vhost_dev *dev);
163049ae	43
c23f3445 TH	44	void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
c23f3445 TH	45	unsigned long mask, struct vhost_dev *dev);
2b8b328b	46	int vhost_poll_start(struct vhost_poll poll, struct file file);
3a4d5c94 MT	47	void vhost_poll_stop(struct vhost_poll *poll);
	48	void vhost_poll_flush(struct vhost_poll *poll);
	49	void vhost_poll_queue(struct vhost_poll *poll);
6ac1afbf AH	50	void vhost_work_flush(struct vhost_dev dev, struct vhost_work work);
6ac1afbf AH	51	long vhost_vring_ioctl(struct vhost_dev d, int ioctl, void __user argp);
3a4d5c94 MT	52
	53	struct vhost_log {
	54	u64 addr;
	55	u64 len;
	56	};
	57
a9709d68 JW	58	#define START(node) ((node)->start)
	59	#define LAST(node) ((node)->last)
	60
	61	struct vhost_umem_node {
	62	struct rb_node rb;
	63	struct list_head link;
	64	__u64 start;
	65	__u64 last;
	66	__u64 size;
	67	__u64 userspace_addr;
6b1e6cc7 JW	68	__u32 perm;
6b1e6cc7 JW	69	__u32 flags_padding;
a9709d68 JW	70	__u64 __subtree_last;
	71	};
	72
	73	struct vhost_umem {
	74	struct rb_root umem_tree;
	75	struct list_head umem_list;
6b1e6cc7	76	int numem;
a9709d68 JW	77	};
a9709d68 JW	78
3a4d5c94 MT	79	/* The virtqueue structure describes a queue attached to a device. */
	80	struct vhost_virtqueue {
	81	struct vhost_dev *dev;
	82
	83	/* The actual ring of buffers. */
	84	struct mutex mutex;
	85	unsigned int num;
	86	struct vring_desc __user *desc;
	87	struct vring_avail __user *avail;
	88	struct vring_used __user *used;
	89	struct file *kick;
	90	struct file *call;
	91	struct file *error;
	92	struct eventfd_ctx *call_ctx;
	93	struct eventfd_ctx *error_ctx;
	94	struct eventfd_ctx *log_ctx;
	95
	96	struct vhost_poll poll;
	97
	98	/* The routine to call when the Guest pings us, or timeout. */
c23f3445	99	vhost_work_fn_t handle_kick;
3a4d5c94 MT	100
	101	/* Last available index we saw. */
	102	u16 last_avail_idx;
	103
	104	/* Caches available index value from user. */
	105	u16 avail_idx;
	106
	107	/* Last index we used. */
	108	u16 last_used_idx;
	109
	110	/* Used flags */
	111	u16 used_flags;
	112
8ea8cf89 MT	113	/* Last used index value we have signalled on */
	114	u16 signalled_used;
	115
	116	/* Last used index value we have signalled on */
	117	bool signalled_used_valid;
	118
3a4d5c94 MT	119	/* Log writes to used structure. */
	120	bool log_used;
	121	u64 log_addr;
	122
e0e9b406	123	struct iovec iov[UIO_MAXIOV];
6b1e6cc7	124	struct iovec iotlb_iov[64];
e0e9b406	125	struct iovec *indirect;
e0e9b406	126	struct vring_used_elem *heads;
22fa90c7	127	/* Protected by virtqueue mutex. */
a9709d68	128	struct vhost_umem *umem;
6b1e6cc7	129	struct vhost_umem *iotlb;
22fa90c7	130	void *private_data;
bd82752a	131	u64 acked_features;
3a4d5c94 MT	132	/* Log write descriptors */
3a4d5c94 MT	133	void __user *log_base;
e0e9b406	134	struct vhost_log *log;
2751c988 GK	135
	136	/* Ring endianness. Defaults to legacy native endianness.
	137	* Set to true when starting a modern virtio device. */
	138	bool is_le;
	139	#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
	140	/* Ring endianness requested by userspace for cross-endian support. */
	141	bool user_be;
	142	#endif
03088137	143	u32 busyloop_timeout;
3a4d5c94 MT	144	};
3a4d5c94 MT	145
6b1e6cc7 JW	146	struct vhost_msg_node {
	147	struct vhost_msg msg;
	148	struct vhost_virtqueue *vq;
	149	struct list_head node;
	150	};
	151
3a4d5c94	152	struct vhost_dev {
3a4d5c94 MT	153	struct mm_struct *mm;
3a4d5c94 MT	154	struct mutex mutex;
3ab2e420	155	struct vhost_virtqueue **vqs;
3a4d5c94 MT	156	int nvqs;
	157	struct file *log_file;
	158	struct eventfd_ctx *log_ctx;
04b96e55	159	struct llist_head work_list;
c23f3445	160	struct task_struct *worker;
a9709d68	161	struct vhost_umem *umem;
6b1e6cc7 JW	162	struct vhost_umem *iotlb;
	163	spinlock_t iotlb_lock;
	164	struct list_head read_list;
	165	struct list_head pending_list;
	166	wait_queue_head_t wait;
3a4d5c94 MT	167	};
3a4d5c94 MT	168
59566b6e	169	void vhost_dev_init(struct vhost_dev , struct vhost_virtqueue *vqs, int nvqs);
54db63c2	170	long vhost_dev_set_owner(struct vhost_dev *dev);
05c05351	171	bool vhost_dev_has_owner(struct vhost_dev *dev);
3a4d5c94	172	long vhost_dev_check_owner(struct vhost_dev *);
a9709d68 JW	173	struct vhost_umem *vhost_dev_reset_owner_prepare(void);
a9709d68 JW	174	void vhost_dev_reset_owner(struct vhost_dev , struct vhost_umem );
ea5d4046	175	void vhost_dev_cleanup(struct vhost_dev *, bool locked);
b211616d	176	void vhost_dev_stop(struct vhost_dev *);
935cdee7 MT	177	long vhost_dev_ioctl(struct vhost_dev , unsigned int ioctl, void __user argp);
935cdee7 MT	178	long vhost_vring_ioctl(struct vhost_dev d, int ioctl, void __user argp);
3a4d5c94 MT	179	int vhost_vq_access_ok(struct vhost_virtqueue *vq);
	180	int vhost_log_access_ok(struct vhost_dev *);
	181
47283bef	182	int vhost_get_vq_desc(struct vhost_virtqueue *,
d5675bd2 MT	183	struct iovec iov[], unsigned int iov_count,
	184	unsigned int out_num, unsigned int in_num,
	185	struct vhost_log log, unsigned int log_num);
8dd014ad	186	void vhost_discard_vq_desc(struct vhost_virtqueue *, int n);
3a4d5c94	187
80f7d030	188	int vhost_vq_init_access(struct vhost_virtqueue *);
3a4d5c94	189	int vhost_add_used(struct vhost_virtqueue *, unsigned int head, int len);
8dd014ad DS	190	int vhost_add_used_n(struct vhost_virtqueue , struct vring_used_elem heads,
8dd014ad DS	191	unsigned count);
3a4d5c94	192	void vhost_add_used_and_signal(struct vhost_dev , struct vhost_virtqueue ,
8dd014ad DS	193	unsigned int id, int len);
	194	void vhost_add_used_and_signal_n(struct vhost_dev , struct vhost_virtqueue ,
	195	struct vring_used_elem *heads, unsigned count);
	196	void vhost_signal(struct vhost_dev , struct vhost_virtqueue );
8ea8cf89	197	void vhost_disable_notify(struct vhost_dev , struct vhost_virtqueue );
d4a60603	198	bool vhost_vq_avail_empty(struct vhost_dev , struct vhost_virtqueue );
8ea8cf89	199	bool vhost_enable_notify(struct vhost_dev , struct vhost_virtqueue );
3a4d5c94 MT	200
	201	int vhost_log_write(struct vhost_virtqueue vq, struct vhost_log log,
	202	unsigned int log_num, u64 len);
6b1e6cc7 JW	203	int vq_iotlb_prefetch(struct vhost_virtqueue *vq);
	204
	205	struct vhost_msg_node vhost_new_msg(struct vhost_virtqueue vq, int type);
	206	void vhost_enqueue_msg(struct vhost_dev *dev,
	207	struct list_head *head,
	208	struct vhost_msg_node *node);
	209	struct vhost_msg_node vhost_dequeue_msg(struct vhost_dev dev,
	210	struct list_head *head);
	211	unsigned int vhost_chr_poll(struct file file, struct vhost_dev dev,
	212	poll_table *wait);
	213	ssize_t vhost_chr_read_iter(struct vhost_dev dev, struct iov_iter to,
	214	int noblock);
	215	ssize_t vhost_chr_write_iter(struct vhost_dev *dev,
	216	struct iov_iter *from);
	217	int vhost_init_device_iotlb(struct vhost_dev *d, bool enabled);
3a4d5c94	218
3a4d5c94 MT	219	#define vq_err(vq, fmt, ...) do { \
	220	pr_debug(pr_fmt(fmt), ##__VA_ARGS__); \
	221	if ((vq)->error_ctx) \
	222	eventfd_signal((vq)->error_ctx, 1);\
	223	} while (0)
	224
	225	enum {
8ea8cf89 MT	226	VHOST_FEATURES = (1ULL << VIRTIO_F_NOTIFY_ON_EMPTY) \|
	227	(1ULL << VIRTIO_RING_F_INDIRECT_DESC) \|
	228	(1ULL << VIRTIO_RING_F_EVENT_IDX) \|
4e9fa50c MT	229	(1ULL << VHOST_F_LOG_ALL) \|
	230	(1ULL << VIRTIO_F_ANY_LAYOUT) \|
	231	(1ULL << VIRTIO_F_VERSION_1)
3a4d5c94 MT	232	};
3a4d5c94 MT	233
bd82752a	234	static inline bool vhost_has_feature(struct vhost_virtqueue *vq, int bit)
3a4d5c94	235	{
bd82752a	236	return vq->acked_features & (1ULL << bit);
3a4d5c94	237	}
e05fd12b	238
e407f39a	239	#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
ab27c07f GK	240	static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
ab27c07f GK	241	{
2751c988	242	return vq->is_le;
ab27c07f	243	}
e407f39a MT	244	#else
	245	static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
	246	{
	247	return virtio_legacy_is_little_endian() \|\| vq->is_le;
	248	}
	249	#endif
ab27c07f	250
e05fd12b MT	251	/* Memory accessors */
	252	static inline u16 vhost16_to_cpu(struct vhost_virtqueue *vq, __virtio16 val)
	253	{
ab27c07f	254	return __virtio16_to_cpu(vhost_is_little_endian(vq), val);
e05fd12b MT	255	}
	256
	257	static inline __virtio16 cpu_to_vhost16(struct vhost_virtqueue *vq, u16 val)
	258	{
ab27c07f	259	return __cpu_to_virtio16(vhost_is_little_endian(vq), val);
e05fd12b MT	260	}
	261
	262	static inline u32 vhost32_to_cpu(struct vhost_virtqueue *vq, __virtio32 val)
	263	{
ab27c07f	264	return __virtio32_to_cpu(vhost_is_little_endian(vq), val);
e05fd12b MT	265	}
	266
	267	static inline __virtio32 cpu_to_vhost32(struct vhost_virtqueue *vq, u32 val)
	268	{
ab27c07f	269	return __cpu_to_virtio32(vhost_is_little_endian(vq), val);
e05fd12b MT	270	}
	271
	272	static inline u64 vhost64_to_cpu(struct vhost_virtqueue *vq, __virtio64 val)
	273	{
ab27c07f	274	return __virtio64_to_cpu(vhost_is_little_endian(vq), val);
e05fd12b MT	275	}
	276
	277	static inline __virtio64 cpu_to_vhost64(struct vhost_virtqueue *vq, u64 val)
	278	{
ab27c07f	279	return __cpu_to_virtio64(vhost_is_little_endian(vq), val);
e05fd12b	280	}
3a4d5c94	281	#endif