[mirror_ubuntu-bionic-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_entry_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_cached umem_tree;
	struct list_head umem_list;
	int numem;
};

enum vhost_uaddr_type {
	VHOST_ADDR_DESC = 0,
	VHOST_ADDR_AVAIL = 1,
	VHOST_ADDR_USED = 2,
	VHOST_NUM_ADDRS = 3,
};

/* 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;
	const struct vhost_umem_node *meta_iotlb[VHOST_NUM_ADDRS];
	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
	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>
	8	#include <linux/poll.h>
	9	#include <linux/file.h>
	10	#include <linux/uio.h>
	11	#include <linux/virtio_config.h>
	12	#include <linux/virtio_ring.h>
	13	#include <linux/atomic.h>
	14
	15	struct vhost_work;
	16	typedef void (vhost_work_fn_t)(struct vhost_work work);
	17
	18	#define VHOST_WORK_QUEUED 1
	19	struct vhost_work {
	20	struct llist_node node;
	21	vhost_work_fn_t fn;
	22	wait_queue_head_t done;
	23	int flushing;
	24	unsigned queue_seq;
	25	unsigned done_seq;
	26	unsigned long flags;
	27	};
	28
	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_entry_t wait;
	35	struct vhost_work work;
	36	unsigned long mask;
	37	struct vhost_dev *dev;
	38	};
	39
	40	void vhost_work_init(struct vhost_work *work, vhost_work_fn_t fn);
	41	void vhost_work_queue(struct vhost_dev dev, struct vhost_work work);
	42	bool vhost_has_work(struct vhost_dev *dev);
	43
	44	void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
	45	unsigned long mask, struct vhost_dev *dev);
	46	int vhost_poll_start(struct vhost_poll poll, struct file file);
	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);
	50	void vhost_work_flush(struct vhost_dev dev, struct vhost_work work);
	51	long vhost_vring_ioctl(struct vhost_dev d, int ioctl, void __user argp);
	52
	53	struct vhost_log {
	54	u64 addr;
	55	u64 len;
	56	};
	57
	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;
	68	__u32 perm;
	69	__u32 flags_padding;
	70	__u64 __subtree_last;
	71	};
	72
	73	struct vhost_umem {
	74	struct rb_root_cached umem_tree;
	75	struct list_head umem_list;
	76	int numem;
	77	};
	78
	79	enum vhost_uaddr_type {
	80	VHOST_ADDR_DESC = 0,
	81	VHOST_ADDR_AVAIL = 1,
	82	VHOST_ADDR_USED = 2,
	83	VHOST_NUM_ADDRS = 3,
	84	};
	85
	86	/* The virtqueue structure describes a queue attached to a device. */
	87	struct vhost_virtqueue {
	88	struct vhost_dev *dev;
	89
	90	/* The actual ring of buffers. */
	91	struct mutex mutex;
	92	unsigned int num;
	93	struct vring_desc __user *desc;
	94	struct vring_avail __user *avail;
	95	struct vring_used __user *used;
	96	const struct vhost_umem_node *meta_iotlb[VHOST_NUM_ADDRS];
	97	struct file *kick;
	98	struct file *call;
	99	struct file *error;
	100	struct eventfd_ctx *call_ctx;
	101	struct eventfd_ctx *error_ctx;
	102	struct eventfd_ctx *log_ctx;
	103
	104	struct vhost_poll poll;
	105
	106	/* The routine to call when the Guest pings us, or timeout. */
	107	vhost_work_fn_t handle_kick;
	108
	109	/* Last available index we saw. */
	110	u16 last_avail_idx;
	111
	112	/* Caches available index value from user. */
	113	u16 avail_idx;
	114
	115	/* Last index we used. */
	116	u16 last_used_idx;
	117
	118	/* Used flags */
	119	u16 used_flags;
	120
	121	/* Last used index value we have signalled on */
	122	u16 signalled_used;
	123
	124	/* Last used index value we have signalled on */
	125	bool signalled_used_valid;
	126
	127	/* Log writes to used structure. */
	128	bool log_used;
	129	u64 log_addr;
	130
	131	struct iovec iov[UIO_MAXIOV];
	132	struct iovec iotlb_iov[64];
	133	struct iovec *indirect;
	134	struct vring_used_elem *heads;
	135	/* Protected by virtqueue mutex. */
	136	struct vhost_umem *umem;
	137	struct vhost_umem *iotlb;
	138	void *private_data;
	139	u64 acked_features;
	140	/* Log write descriptors */
	141	void __user *log_base;
	142	struct vhost_log *log;
	143
	144	/* Ring endianness. Defaults to legacy native endianness.
	145	* Set to true when starting a modern virtio device. */
	146	bool is_le;
	147	#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
	148	/* Ring endianness requested by userspace for cross-endian support. */
	149	bool user_be;
	150	#endif
	151	u32 busyloop_timeout;
	152	};
	153
	154	struct vhost_msg_node {
	155	struct vhost_msg msg;
	156	struct vhost_virtqueue *vq;
	157	struct list_head node;
	158	};
	159
	160	struct vhost_dev {
	161	struct mm_struct *mm;
	162	struct mutex mutex;
	163	struct vhost_virtqueue **vqs;
	164	int nvqs;
	165	struct file *log_file;
	166	struct eventfd_ctx *log_ctx;
	167	struct llist_head work_list;
	168	struct task_struct *worker;
	169	struct vhost_umem *umem;
	170	struct vhost_umem *iotlb;
	171	spinlock_t iotlb_lock;
	172	struct list_head read_list;
	173	struct list_head pending_list;
	174	wait_queue_head_t wait;
	175	};
	176
	177	void vhost_dev_init(struct vhost_dev , struct vhost_virtqueue *vqs, int nvqs);
	178	long vhost_dev_set_owner(struct vhost_dev *dev);
	179	bool vhost_dev_has_owner(struct vhost_dev *dev);
	180	long vhost_dev_check_owner(struct vhost_dev *);
	181	struct vhost_umem *vhost_dev_reset_owner_prepare(void);
	182	void vhost_dev_reset_owner(struct vhost_dev , struct vhost_umem );
	183	void vhost_dev_cleanup(struct vhost_dev *, bool locked);
	184	void vhost_dev_stop(struct vhost_dev *);
	185	long vhost_dev_ioctl(struct vhost_dev , unsigned int ioctl, void __user argp);
	186	long vhost_vring_ioctl(struct vhost_dev d, int ioctl, void __user argp);
	187	int vhost_vq_access_ok(struct vhost_virtqueue *vq);
	188	int vhost_log_access_ok(struct vhost_dev *);
	189
	190	int vhost_get_vq_desc(struct vhost_virtqueue *,
	191	struct iovec iov[], unsigned int iov_count,
	192	unsigned int out_num, unsigned int in_num,
	193	struct vhost_log log, unsigned int log_num);
	194	void vhost_discard_vq_desc(struct vhost_virtqueue *, int n);
	195
	196	int vhost_vq_init_access(struct vhost_virtqueue *);
	197	int vhost_add_used(struct vhost_virtqueue *, unsigned int head, int len);
	198	int vhost_add_used_n(struct vhost_virtqueue , struct vring_used_elem heads,
	199	unsigned count);
	200	void vhost_add_used_and_signal(struct vhost_dev , struct vhost_virtqueue ,
	201	unsigned int id, int len);
	202	void vhost_add_used_and_signal_n(struct vhost_dev , struct vhost_virtqueue ,
	203	struct vring_used_elem *heads, unsigned count);
	204	void vhost_signal(struct vhost_dev , struct vhost_virtqueue );
	205	void vhost_disable_notify(struct vhost_dev , struct vhost_virtqueue );
	206	bool vhost_vq_avail_empty(struct vhost_dev , struct vhost_virtqueue );
	207	bool vhost_enable_notify(struct vhost_dev , struct vhost_virtqueue );
	208
	209	int vhost_log_write(struct vhost_virtqueue vq, struct vhost_log log,
	210	unsigned int log_num, u64 len);
	211	int vq_iotlb_prefetch(struct vhost_virtqueue *vq);
	212
	213	struct vhost_msg_node vhost_new_msg(struct vhost_virtqueue vq, int type);
	214	void vhost_enqueue_msg(struct vhost_dev *dev,
	215	struct list_head *head,
	216	struct vhost_msg_node *node);
	217	struct vhost_msg_node vhost_dequeue_msg(struct vhost_dev dev,
	218	struct list_head *head);
	219	unsigned int vhost_chr_poll(struct file file, struct vhost_dev dev,
	220	poll_table *wait);
	221	ssize_t vhost_chr_read_iter(struct vhost_dev dev, struct iov_iter to,
	222	int noblock);
	223	ssize_t vhost_chr_write_iter(struct vhost_dev *dev,
	224	struct iov_iter *from);
	225	int vhost_init_device_iotlb(struct vhost_dev *d, bool enabled);
	226
	227	#define vq_err(vq, fmt, ...) do { \
	228	pr_debug(pr_fmt(fmt), ##__VA_ARGS__); \
	229	if ((vq)->error_ctx) \
	230	eventfd_signal((vq)->error_ctx, 1);\
	231	} while (0)
	232
	233	enum {
	234	VHOST_FEATURES = (1ULL << VIRTIO_F_NOTIFY_ON_EMPTY) \|
	235	(1ULL << VIRTIO_RING_F_INDIRECT_DESC) \|
	236	(1ULL << VIRTIO_RING_F_EVENT_IDX) \|
	237	(1ULL << VHOST_F_LOG_ALL) \|
	238	(1ULL << VIRTIO_F_ANY_LAYOUT) \|
	239	(1ULL << VIRTIO_F_VERSION_1)
	240	};
	241
	242	static inline bool vhost_has_feature(struct vhost_virtqueue *vq, int bit)
	243	{
	244	return vq->acked_features & (1ULL << bit);
	245	}
	246
	247	#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
	248	static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
	249	{
	250	return vq->is_le;
	251	}
	252	#else
	253	static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq)
	254	{
	255	return virtio_legacy_is_little_endian() \|\| vq->is_le;
	256	}
	257	#endif
	258
	259	/* Memory accessors */
	260	static inline u16 vhost16_to_cpu(struct vhost_virtqueue *vq, __virtio16 val)
	261	{
	262	return __virtio16_to_cpu(vhost_is_little_endian(vq), val);
	263	}
	264
	265	static inline __virtio16 cpu_to_vhost16(struct vhost_virtqueue *vq, u16 val)
	266	{
	267	return __cpu_to_virtio16(vhost_is_little_endian(vq), val);
	268	}
	269
	270	static inline u32 vhost32_to_cpu(struct vhost_virtqueue *vq, __virtio32 val)
	271	{
	272	return __virtio32_to_cpu(vhost_is_little_endian(vq), val);
	273	}
	274
	275	static inline __virtio32 cpu_to_vhost32(struct vhost_virtqueue *vq, u32 val)
	276	{
	277	return __cpu_to_virtio32(vhost_is_little_endian(vq), val);
	278	}
	279
	280	static inline u64 vhost64_to_cpu(struct vhost_virtqueue *vq, __virtio64 val)
	281	{
	282	return __virtio64_to_cpu(vhost_is_little_endian(vq), val);
	283	}
	284
	285	static inline __virtio64 cpu_to_vhost64(struct vhost_virtqueue *vq, u64 val)
	286	{
	287	return __cpu_to_virtio64(vhost_is_little_endian(vq), val);
	288	}
	289	#endif