[deliverable/linux.git] / drivers / block / xen-blkback / common.h

/*
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License version 2
 * as published by the Free Software Foundation; or, when distributed
 * separately from the Linux kernel or incorporated into other
 * software packages, subject to the following license:
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this source file (the "Software"), to deal in the Software without
 * restriction, including without limitation the rights to use, copy, modify,
 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
 * and to permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

#ifndef __XEN_BLKIF__BACKEND__COMMON_H__
#define __XEN_BLKIF__BACKEND__COMMON_H__

#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
#include <linux/vmalloc.h>
#include <linux/wait.h>
#include <linux/io.h>
#include <linux/rbtree.h>
#include <asm/setup.h>
#include <asm/pgalloc.h>
#include <asm/hypervisor.h>
#include <xen/grant_table.h>
#include <xen/xenbus.h>
#include <xen/interface/io/ring.h>
#include <xen/interface/io/blkif.h>
#include <xen/interface/io/protocols.h>

#define DRV_PFX "xen-blkback:"
#define DPRINTK(fmt, args...)				\
	pr_debug(DRV_PFX "(%s:%d) " fmt ".\n",		\
		 __func__, __LINE__, ##args)


/*
 * This is the maximum number of segments that would be allowed in indirect
 * requests. This value will also be passed to the frontend.
 */
#define MAX_INDIRECT_SEGMENTS 256

#define SEGS_PER_INDIRECT_FRAME \
	(PAGE_SIZE/sizeof(struct blkif_request_segment_aligned))
#define MAX_INDIRECT_PAGES \
	((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
#define INDIRECT_PAGES(_segs) \
	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)

/* Not a real protocol.  Used to generate ring structs which contain
 * the elements common to all protocols only.  This way we get a
 * compiler-checkable way to use common struct elements, so we can
 * avoid using switch(protocol) in a number of places.  */
struct blkif_common_request {
	char dummy;
};
struct blkif_common_response {
	char dummy;
};

struct blkif_x86_32_request_rw {
	uint8_t        nr_segments;  /* number of segments                   */
	blkif_vdev_t   handle;       /* only for read/write requests         */
	uint64_t       id;           /* private guest value, echoed in resp  */
	blkif_sector_t sector_number;/* start sector idx on disk (r/w only)  */
	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
} __attribute__((__packed__));

struct blkif_x86_32_request_discard {
	uint8_t        flag;         /* BLKIF_DISCARD_SECURE or zero         */
	blkif_vdev_t   _pad1;        /* was "handle" for read/write requests */
	uint64_t       id;           /* private guest value, echoed in resp  */
	blkif_sector_t sector_number;/* start sector idx on disk (r/w only)  */
	uint64_t       nr_sectors;
} __attribute__((__packed__));

struct blkif_x86_32_request_other {
	uint8_t        _pad1;
	blkif_vdev_t   _pad2;
	uint64_t       id;           /* private guest value, echoed in resp  */
} __attribute__((__packed__));

struct blkif_x86_32_request_indirect {
	uint8_t        indirect_op;
	uint16_t       nr_segments;
	uint64_t       id;
	blkif_sector_t sector_number;
	blkif_vdev_t   handle;
	uint16_t       _pad1;
	grant_ref_t    indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	/*
	 * The maximum number of indirect segments (and pages) that will
	 * be used is determined by MAX_INDIRECT_SEGMENTS, this value
	 * is also exported to the guest (via xenstore
	 * feature-max-indirect-segments entry), so the frontend knows how
	 * many indirect segments the backend supports.
	 */
	uint64_t       _pad2;        /* make it 64 byte aligned */
} __attribute__((__packed__));

struct blkif_x86_32_request {
	uint8_t        operation;    /* BLKIF_OP_???                         */
	union {
		struct blkif_x86_32_request_rw rw;
		struct blkif_x86_32_request_discard discard;
		struct blkif_x86_32_request_other other;
		struct blkif_x86_32_request_indirect indirect;
	} u;
} __attribute__((__packed__));

/* i386 protocol version */
#pragma pack(push, 4)
struct blkif_x86_32_response {
	uint64_t        id;              /* copied from request */
	uint8_t         operation;       /* copied from request */
	int16_t         status;          /* BLKIF_RSP_???       */
};
#pragma pack(pop)
/* x86_64 protocol version */

struct blkif_x86_64_request_rw {
	uint8_t        nr_segments;  /* number of segments                   */
	blkif_vdev_t   handle;       /* only for read/write requests         */
	uint32_t       _pad1;        /* offsetof(blkif_reqest..,u.rw.id)==8  */
	uint64_t       id;
	blkif_sector_t sector_number;/* start sector idx on disk (r/w only)  */
	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
} __attribute__((__packed__));

struct blkif_x86_64_request_discard {
	uint8_t        flag;         /* BLKIF_DISCARD_SECURE or zero         */
	blkif_vdev_t   _pad1;        /* was "handle" for read/write requests */
        uint32_t       _pad2;        /* offsetof(blkif_..,u.discard.id)==8   */
	uint64_t       id;
	blkif_sector_t sector_number;/* start sector idx on disk (r/w only)  */
	uint64_t       nr_sectors;
} __attribute__((__packed__));

struct blkif_x86_64_request_other {
	uint8_t        _pad1;
	blkif_vdev_t   _pad2;
	uint32_t       _pad3;        /* offsetof(blkif_..,u.discard.id)==8   */
	uint64_t       id;           /* private guest value, echoed in resp  */
} __attribute__((__packed__));

struct blkif_x86_64_request_indirect {
	uint8_t        indirect_op;
	uint16_t       nr_segments;
	uint32_t       _pad1;        /* offsetof(blkif_..,u.indirect.id)==8   */
	uint64_t       id;
	blkif_sector_t sector_number;
	blkif_vdev_t   handle;
	uint16_t       _pad2;
	grant_ref_t    indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	/*
	 * The maximum number of indirect segments (and pages) that will
	 * be used is determined by MAX_INDIRECT_SEGMENTS, this value
	 * is also exported to the guest (via xenstore
	 * feature-max-indirect-segments entry), so the frontend knows how
	 * many indirect segments the backend supports.
	 */
	uint32_t       _pad3;        /* make it 64 byte aligned */
} __attribute__((__packed__));

struct blkif_x86_64_request {
	uint8_t        operation;    /* BLKIF_OP_???                         */
	union {
		struct blkif_x86_64_request_rw rw;
		struct blkif_x86_64_request_discard discard;
		struct blkif_x86_64_request_other other;
		struct blkif_x86_64_request_indirect indirect;
	} u;
} __attribute__((__packed__));

struct blkif_x86_64_response {
	uint64_t       __attribute__((__aligned__(8))) id;
	uint8_t         operation;       /* copied from request */
	int16_t         status;          /* BLKIF_RSP_???       */
};

DEFINE_RING_TYPES(blkif_common, struct blkif_common_request,
		  struct blkif_common_response);
DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request,
		  struct blkif_x86_32_response);
DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request,
		  struct blkif_x86_64_response);

union blkif_back_rings {
	struct blkif_back_ring        native;
	struct blkif_common_back_ring common;
	struct blkif_x86_32_back_ring x86_32;
	struct blkif_x86_64_back_ring x86_64;
};

enum blkif_protocol {
	BLKIF_PROTOCOL_NATIVE = 1,
	BLKIF_PROTOCOL_X86_32 = 2,
	BLKIF_PROTOCOL_X86_64 = 3,
};

struct xen_vbd {
	/* What the domain refers to this vbd as. */
	blkif_vdev_t		handle;
	/* Non-zero -> read-only */
	unsigned char		readonly;
	/* VDISK_xxx */
	unsigned char		type;
	/* phys device that this vbd maps to. */
	u32			pdevice;
	struct block_device	*bdev;
	/* Cached size parameter. */
	sector_t		size;
	unsigned int		flush_support:1;
	unsigned int		discard_secure:1;
	unsigned int		feature_gnt_persistent:1;
	unsigned int		overflow_max_grants:1;
};

struct backend_info;

/* Number of available flags */
#define PERSISTENT_GNT_FLAGS_SIZE	2
/* This persistent grant is currently in use */
#define PERSISTENT_GNT_ACTIVE		0
/*
 * This persistent grant has been used, this flag is set when we remove the
 * PERSISTENT_GNT_ACTIVE, to know that this grant has been used recently.
 */
#define PERSISTENT_GNT_WAS_ACTIVE	1

/* Number of requests that we can fit in a ring */
#define XEN_BLKIF_REQS			32

struct persistent_gnt {
	struct page *page;
	grant_ref_t gnt;
	grant_handle_t handle;
	DECLARE_BITMAP(flags, PERSISTENT_GNT_FLAGS_SIZE);
	struct rb_node node;
	struct list_head remove_node;
};

struct xen_blkif {
	/* Unique identifier for this interface. */
	domid_t			domid;
	unsigned int		handle;
	/* Physical parameters of the comms window. */
	unsigned int		irq;
	/* Comms information. */
	enum blkif_protocol	blk_protocol;
	union blkif_back_rings	blk_rings;
	void			*blk_ring;
	/* The VBD attached to this interface. */
	struct xen_vbd		vbd;
	/* Back pointer to the backend_info. */
	struct backend_info	*be;
	/* Private fields. */
	spinlock_t		blk_ring_lock;
	atomic_t		refcnt;

	wait_queue_head_t	wq;
	/* for barrier (drain) requests */
	struct completion	drain_complete;
	atomic_t		drain;
	/* One thread per one blkif. */
	struct task_struct	*xenblkd;
	unsigned int		waiting_reqs;

	/* tree to store persistent grants */
	struct rb_root		persistent_gnts;
	unsigned int		persistent_gnt_c;
	atomic_t		persistent_gnt_in_use;
	unsigned long           next_lru;

	/* used by the kworker that offload work from the persistent purge */
	struct list_head	persistent_purge_list;
	struct work_struct	persistent_purge_work;

	/* buffer of free pages to map grant refs */
	spinlock_t		free_pages_lock;
	int			free_pages_num;
	struct list_head	free_pages;

	/* List of all 'pending_req' available */
	struct list_head	pending_free;
	/* And its spinlock. */
	spinlock_t		pending_free_lock;
	wait_queue_head_t	pending_free_wq;

	/* statistics */
	unsigned long		st_print;
	unsigned long long			st_rd_req;
	unsigned long long			st_wr_req;
	unsigned long long			st_oo_req;
	unsigned long long			st_f_req;
	unsigned long long			st_ds_req;
	unsigned long long			st_rd_sect;
	unsigned long long			st_wr_sect;

	wait_queue_head_t	waiting_to_free;
	/* Thread shutdown wait queue. */
	wait_queue_head_t	shutdown_wq;
};

struct seg_buf {
	unsigned long offset;
	unsigned int nsec;
};

struct grant_page {
	struct page 		*page;
	struct persistent_gnt	*persistent_gnt;
	grant_handle_t		handle;
	grant_ref_t		gref;
};

/*
 * Each outstanding request that we've passed to the lower device layers has a
 * 'pending_req' allocated to it. Each buffer_head that completes decrements
 * the pendcnt towards zero. When it hits zero, the specified domain has a
 * response queued for it, with the saved 'id' passed back.
 */
struct pending_req {
	struct xen_blkif	*blkif;
	u64			id;
	int			nr_pages;
	atomic_t		pendcnt;
	unsigned short		operation;
	int			status;
	struct list_head	free_list;
	struct grant_page	*segments[MAX_INDIRECT_SEGMENTS];
	/* Indirect descriptors */
	struct grant_page	*indirect_pages[MAX_INDIRECT_PAGES];
	struct seg_buf		seg[MAX_INDIRECT_SEGMENTS];
	struct bio		*biolist[MAX_INDIRECT_SEGMENTS];
};


#define vbd_sz(_v)	((_v)->bdev->bd_part ? \
			 (_v)->bdev->bd_part->nr_sects : \
			  get_capacity((_v)->bdev->bd_disk))

#define xen_blkif_get(_b) (atomic_inc(&(_b)->refcnt))
#define xen_blkif_put(_b)				\
	do {						\
		if (atomic_dec_and_test(&(_b)->refcnt))	\
			wake_up(&(_b)->waiting_to_free);\
	} while (0)

struct phys_req {
	unsigned short		dev;
	blkif_sector_t		nr_sects;
	struct block_device	*bdev;
	blkif_sector_t		sector_number;
};
int xen_blkif_interface_init(void);

int xen_blkif_xenbus_init(void);

irqreturn_t xen_blkif_be_int(int irq, void *dev_id);
int xen_blkif_schedule(void *arg);
int xen_blkif_purge_persistent(void *arg);

int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt,
			      struct backend_info *be, int state);

int xen_blkbk_barrier(struct xenbus_transaction xbt,
		      struct backend_info *be, int state);
struct xenbus_device *xen_blkbk_xenbus(struct backend_info *be);

static inline void blkif_get_x86_32_req(struct blkif_request *dst,
					struct blkif_x86_32_request *src)
{
	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
	dst->operation = src->operation;
	switch (src->operation) {
	case BLKIF_OP_READ:
	case BLKIF_OP_WRITE:
	case BLKIF_OP_WRITE_BARRIER:
	case BLKIF_OP_FLUSH_DISKCACHE:
		dst->u.rw.nr_segments = src->u.rw.nr_segments;
		dst->u.rw.handle = src->u.rw.handle;
		dst->u.rw.id = src->u.rw.id;
		dst->u.rw.sector_number = src->u.rw.sector_number;
		barrier();
		if (n > dst->u.rw.nr_segments)
			n = dst->u.rw.nr_segments;
		for (i = 0; i < n; i++)
			dst->u.rw.seg[i] = src->u.rw.seg[i];
		break;
	case BLKIF_OP_DISCARD:
		dst->u.discard.flag = src->u.discard.flag;
		dst->u.discard.id = src->u.discard.id;
		dst->u.discard.sector_number = src->u.discard.sector_number;
		dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
		break;
	case BLKIF_OP_INDIRECT:
		dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
		dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
		dst->u.indirect.handle = src->u.indirect.handle;
		dst->u.indirect.id = src->u.indirect.id;
		dst->u.indirect.sector_number = src->u.indirect.sector_number;
		barrier();
		j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
		for (i = 0; i < j; i++)
			dst->u.indirect.indirect_grefs[i] =
				src->u.indirect.indirect_grefs[i];
		break;
	default:
		/*
		 * Don't know how to translate this op. Only get the
		 * ID so failure can be reported to the frontend.
		 */
		dst->u.other.id = src->u.other.id;
		break;
	}
}

static inline void blkif_get_x86_64_req(struct blkif_request *dst,
					struct blkif_x86_64_request *src)
{
	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
	dst->operation = src->operation;
	switch (src->operation) {
	case BLKIF_OP_READ:
	case BLKIF_OP_WRITE:
	case BLKIF_OP_WRITE_BARRIER:
	case BLKIF_OP_FLUSH_DISKCACHE:
		dst->u.rw.nr_segments = src->u.rw.nr_segments;
		dst->u.rw.handle = src->u.rw.handle;
		dst->u.rw.id = src->u.rw.id;
		dst->u.rw.sector_number = src->u.rw.sector_number;
		barrier();
		if (n > dst->u.rw.nr_segments)
			n = dst->u.rw.nr_segments;
		for (i = 0; i < n; i++)
			dst->u.rw.seg[i] = src->u.rw.seg[i];
		break;
	case BLKIF_OP_DISCARD:
		dst->u.discard.flag = src->u.discard.flag;
		dst->u.discard.id = src->u.discard.id;
		dst->u.discard.sector_number = src->u.discard.sector_number;
		dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
		break;
	case BLKIF_OP_INDIRECT:
		dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
		dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
		dst->u.indirect.handle = src->u.indirect.handle;
		dst->u.indirect.id = src->u.indirect.id;
		dst->u.indirect.sector_number = src->u.indirect.sector_number;
		barrier();
		j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
		for (i = 0; i < j; i++)
			dst->u.indirect.indirect_grefs[i] =
				src->u.indirect.indirect_grefs[i];
		break;
	default:
		/*
		 * Don't know how to translate this op. Only get the
		 * ID so failure can be reported to the frontend.
		 */
		dst->u.other.id = src->u.other.id;
		break;
	}
}

#endif /* __XEN_BLKIF__BACKEND__COMMON_H__ */
Commit	Line	Data
4d05a28d KRW	1	/*
	2	* This program is free software; you can redistribute it and/or
	3	* modify it under the terms of the GNU General Public License version 2
	4	* as published by the Free Software Foundation; or, when distributed
	5	* separately from the Linux kernel or incorporated into other
	6	* software packages, subject to the following license:
	7	*
	8	* Permission is hereby granted, free of charge, to any person obtaining a copy
	9	* of this source file (the "Software"), to deal in the Software without
	10	* restriction, including without limitation the rights to use, copy, modify,
	11	* merge, publish, distribute, sublicense, and/or sell copies of the Software,
	12	* and to permit persons to whom the Software is furnished to do so, subject to
	13	* the following conditions:
	14	*
	15	* The above copyright notice and this permission notice shall be included in
	16	* all copies or substantial portions of the Software.
	17	*
	18	* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
	19	* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
	20	* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
	21	* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
	22	* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
	23	* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
	24	* IN THE SOFTWARE.
	25	*/
	26
5a577e38 KRW	27	#ifndef __XEN_BLKIF__BACKEND__COMMON_H__
5a577e38 KRW	28	#define __XEN_BLKIF__BACKEND__COMMON_H__
4d05a28d	29
4d05a28d KRW	30	#include <linux/module.h>
	31	#include <linux/interrupt.h>
	32	#include <linux/slab.h>
	33	#include <linux/blkdev.h>
	34	#include <linux/vmalloc.h>
	35	#include <linux/wait.h>
5489377c	36	#include <linux/io.h>
0a8704a5	37	#include <linux/rbtree.h>
4d05a28d KRW	38	#include <asm/setup.h>
4d05a28d KRW	39	#include <asm/pgalloc.h>
4d05a28d	40	#include <asm/hypervisor.h>
88122933	41	#include <xen/grant_table.h>
4d05a28d	42	#include <xen/xenbus.h>
452a6b2b KRW	43	#include <xen/interface/io/ring.h>
	44	#include <xen/interface/io/blkif.h>
	45	#include <xen/interface/io/protocols.h>
4d05a28d	46
22b20f2d	47	#define DRV_PFX "xen-blkback:"
1afbd730	48	#define DPRINTK(fmt, args...) \
1bc05b0a	49	pr_debug(DRV_PFX "(%s:%d) " fmt ".\n", \
1afbd730	50	__func__, __LINE__, ##args)
4d05a28d	51
452a6b2b	52
402b27f9 RPM	53	/*
	54	* This is the maximum number of segments that would be allowed in indirect
	55	* requests. This value will also be passed to the frontend.
	56	*/
	57	#define MAX_INDIRECT_SEGMENTS 256
	58
	59	#define SEGS_PER_INDIRECT_FRAME \
	60	(PAGE_SIZE/sizeof(struct blkif_request_segment_aligned))
	61	#define MAX_INDIRECT_PAGES \
	62	((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	63	#define INDIRECT_PAGES(_segs) \
	64	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	65
452a6b2b KRW	66	/* Not a real protocol. Used to generate ring structs which contain
	67	* the elements common to all protocols only. This way we get a
	68	* compiler-checkable way to use common struct elements, so we can
	69	* avoid using switch(protocol) in a number of places. */
	70	struct blkif_common_request {
	71	char dummy;
	72	};
	73	struct blkif_common_response {
	74	char dummy;
	75	};
	76
b3cb0d6a	77	struct blkif_x86_32_request_rw {
97e36834 KRW	78	uint8_t nr_segments; /* number of segments */
	79	blkif_vdev_t handle; /* only for read/write requests */
	80	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a LD	81	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	82	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	83	} __attribute__((__packed__));
b3cb0d6a LD	84
b3cb0d6a LD	85	struct blkif_x86_32_request_discard {
5ea42986	86	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	87	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
97e36834 KRW	88	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a	89	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	90	uint64_t nr_sectors;
97e36834 KRW	91	} __attribute__((__packed__));
b3cb0d6a	92
0e367ae4 DV	93	struct blkif_x86_32_request_other {
	94	uint8_t _pad1;
	95	blkif_vdev_t _pad2;
	96	uint64_t id; /* private guest value, echoed in resp */
	97	} __attribute__((__packed__));
	98
402b27f9 RPM	99	struct blkif_x86_32_request_indirect {
	100	uint8_t indirect_op;
	101	uint16_t nr_segments;
	102	uint64_t id;
	103	blkif_sector_t sector_number;
	104	blkif_vdev_t handle;
	105	uint16_t _pad1;
	106	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	107	/*
	108	* The maximum number of indirect segments (and pages) that will
	109	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	110	* is also exported to the guest (via xenstore
	111	* feature-max-indirect-segments entry), so the frontend knows how
	112	* many indirect segments the backend supports.
	113	*/
	114	uint64_t _pad2; /* make it 64 byte aligned */
	115	} __attribute__((__packed__));
	116
452a6b2b KRW	117	struct blkif_x86_32_request {
452a6b2b KRW	118	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	119	union {
	120	struct blkif_x86_32_request_rw rw;
	121	struct blkif_x86_32_request_discard discard;
0e367ae4	122	struct blkif_x86_32_request_other other;
402b27f9	123	struct blkif_x86_32_request_indirect indirect;
b3cb0d6a	124	} u;
97e36834 KRW	125	} __attribute__((__packed__));
	126
	127	/* i386 protocol version */
	128	#pragma pack(push, 4)
452a6b2b KRW	129	struct blkif_x86_32_response {
	130	uint64_t id; /* copied from request */
	131	uint8_t operation; /* copied from request */
	132	int16_t status; /* BLKIF_RSP_??? */
	133	};
452a6b2b	134	#pragma pack(pop)
452a6b2b	135	/* x86_64 protocol version */
b3cb0d6a LD	136
b3cb0d6a LD	137	struct blkif_x86_64_request_rw {
97e36834 KRW	138	uint8_t nr_segments; /* number of segments */
	139	blkif_vdev_t handle; /* only for read/write requests */
	140	uint32_t _pad1; /* offsetof(blkif_reqest..,u.rw.id)==8 */
	141	uint64_t id;
b3cb0d6a LD	142	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	143	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	144	} __attribute__((__packed__));
b3cb0d6a LD	145
b3cb0d6a LD	146	struct blkif_x86_64_request_discard {
5ea42986	147	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	148	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
	149	uint32_t _pad2; /* offsetof(blkif_..,u.discard.id)==8 */
	150	uint64_t id;
b3cb0d6a	151	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	152	uint64_t nr_sectors;
97e36834 KRW	153	} __attribute__((__packed__));
b3cb0d6a	154
0e367ae4 DV	155	struct blkif_x86_64_request_other {
	156	uint8_t _pad1;
	157	blkif_vdev_t _pad2;
	158	uint32_t _pad3; /* offsetof(blkif_..,u.discard.id)==8 */
	159	uint64_t id; /* private guest value, echoed in resp */
	160	} __attribute__((__packed__));
	161
402b27f9 RPM	162	struct blkif_x86_64_request_indirect {
	163	uint8_t indirect_op;
	164	uint16_t nr_segments;
	165	uint32_t _pad1; /* offsetof(blkif_..,u.indirect.id)==8 */
	166	uint64_t id;
	167	blkif_sector_t sector_number;
	168	blkif_vdev_t handle;
	169	uint16_t _pad2;
	170	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	171	/*
	172	* The maximum number of indirect segments (and pages) that will
	173	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	174	* is also exported to the guest (via xenstore
	175	* feature-max-indirect-segments entry), so the frontend knows how
	176	* many indirect segments the backend supports.
	177	*/
	178	uint32_t _pad3; /* make it 64 byte aligned */
	179	} __attribute__((__packed__));
	180
452a6b2b KRW	181	struct blkif_x86_64_request {
452a6b2b KRW	182	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	183	union {
	184	struct blkif_x86_64_request_rw rw;
	185	struct blkif_x86_64_request_discard discard;
0e367ae4	186	struct blkif_x86_64_request_other other;
402b27f9	187	struct blkif_x86_64_request_indirect indirect;
b3cb0d6a	188	} u;
97e36834 KRW	189	} __attribute__((__packed__));
97e36834 KRW	190
452a6b2b KRW	191	struct blkif_x86_64_response {
	192	uint64_t __attribute__((__aligned__(8))) id;
	193	uint8_t operation; /* copied from request */
	194	int16_t status; /* BLKIF_RSP_??? */
	195	};
452a6b2b KRW	196
	197	DEFINE_RING_TYPES(blkif_common, struct blkif_common_request,
	198	struct blkif_common_response);
	199	DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request,
	200	struct blkif_x86_32_response);
	201	DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request,
	202	struct blkif_x86_64_response);
	203
	204	union blkif_back_rings {
	205	struct blkif_back_ring native;
	206	struct blkif_common_back_ring common;
	207	struct blkif_x86_32_back_ring x86_32;
	208	struct blkif_x86_64_back_ring x86_64;
	209	};
	210
	211	enum blkif_protocol {
	212	BLKIF_PROTOCOL_NATIVE = 1,
	213	BLKIF_PROTOCOL_X86_32 = 2,
	214	BLKIF_PROTOCOL_X86_64 = 3,
	215	};
	216
3d814731	217	struct xen_vbd {
01f37f2d KRW	218	/* What the domain refers to this vbd as. */
	219	blkif_vdev_t handle;
	220	/* Non-zero -> read-only */
	221	unsigned char readonly;
	222	/* VDISK_xxx */
	223	unsigned char type;
	224	/* phys device that this vbd maps to. */
	225	u32 pdevice;
	226	struct block_device *bdev;
	227	/* Cached size parameter. */
	228	sector_t size;
1f999572 OC	229	unsigned int flush_support:1;
1f999572 OC	230	unsigned int discard_secure:1;
0a8704a5 RPM	231	unsigned int feature_gnt_persistent:1;
0a8704a5 RPM	232	unsigned int overflow_max_grants:1;
4d05a28d KRW	233	};
	234
	235	struct backend_info;
	236
3f3aad5e RPM	237	/* Number of available flags */
	238	#define PERSISTENT_GNT_FLAGS_SIZE 2
	239	/* This persistent grant is currently in use */
	240	#define PERSISTENT_GNT_ACTIVE 0
	241	/*
	242	* This persistent grant has been used, this flag is set when we remove the
	243	* PERSISTENT_GNT_ACTIVE, to know that this grant has been used recently.
	244	*/
	245	#define PERSISTENT_GNT_WAS_ACTIVE 1
0a8704a5	246
bf0720c4 RPM	247	/* Number of requests that we can fit in a ring */
	248	#define XEN_BLKIF_REQS 32
	249
0a8704a5 RPM	250	struct persistent_gnt {
	251	struct page *page;
	252	grant_ref_t gnt;
	253	grant_handle_t handle;
3f3aad5e	254	DECLARE_BITMAP(flags, PERSISTENT_GNT_FLAGS_SIZE);
0a8704a5	255	struct rb_node node;
3f3aad5e	256	struct list_head remove_node;
0a8704a5 RPM	257	};
0a8704a5 RPM	258
30fd1502	259	struct xen_blkif {
4d05a28d	260	/* Unique identifier for this interface. */
01f37f2d KRW	261	domid_t domid;
01f37f2d KRW	262	unsigned int handle;
4d05a28d	263	/* Physical parameters of the comms window. */
01f37f2d	264	unsigned int irq;
4d05a28d	265	/* Comms information. */
01f37f2d KRW	266	enum blkif_protocol blk_protocol;
01f37f2d KRW	267	union blkif_back_rings blk_rings;
2d073846	268	void *blk_ring;
4d05a28d	269	/* The VBD attached to this interface. */
3d814731	270	struct xen_vbd vbd;
4d05a28d	271	/* Back pointer to the backend_info. */
01f37f2d	272	struct backend_info *be;
4d05a28d	273	/* Private fields. */
01f37f2d KRW	274	spinlock_t blk_ring_lock;
01f37f2d KRW	275	atomic_t refcnt;
4d05a28d	276
01f37f2d	277	wait_queue_head_t wq;
29bde093 KRW	278	/* for barrier (drain) requests */
	279	struct completion drain_complete;
	280	atomic_t drain;
a1397fa3	281	/* One thread per one blkif. */
01f37f2d KRW	282	struct task_struct *xenblkd;
01f37f2d KRW	283	unsigned int waiting_reqs;
4d05a28d	284
0a8704a5 RPM	285	/* tree to store persistent grants */
	286	struct rb_root persistent_gnts;
	287	unsigned int persistent_gnt_c;
3f3aad5e RPM	288	atomic_t persistent_gnt_in_use;
	289	unsigned long next_lru;
	290
	291	/* used by the kworker that offload work from the persistent purge */
	292	struct list_head persistent_purge_list;
	293	struct work_struct persistent_purge_work;
0a8704a5	294
c6cc142d RPM	295	/* buffer of free pages to map grant refs */
	296	spinlock_t free_pages_lock;
	297	int free_pages_num;
	298	struct list_head free_pages;
	299
bf0720c4 RPM	300	/* List of all 'pending_req' available */
	301	struct list_head pending_free;
	302	/* And its spinlock. */
	303	spinlock_t pending_free_lock;
	304	wait_queue_head_t pending_free_wq;
	305
4d05a28d	306	/* statistics */
01f37f2d	307	unsigned long st_print;
986cacbd ZK	308	unsigned long long st_rd_req;
	309	unsigned long long st_wr_req;
	310	unsigned long long st_oo_req;
	311	unsigned long long st_f_req;
	312	unsigned long long st_ds_req;
	313	unsigned long long st_rd_sect;
	314	unsigned long long st_wr_sect;
01f37f2d KRW	315
01f37f2d KRW	316	wait_queue_head_t waiting_to_free;
8e3f8755 KRW	317	/* Thread shutdown wait queue. */
8e3f8755 KRW	318	wait_queue_head_t shutdown_wq;
5489377c	319	};
4d05a28d	320
402b27f9 RPM	321	struct seg_buf {
	322	unsigned long offset;
	323	unsigned int nsec;
	324	};
	325
bb642e83 RPM	326	struct grant_page {
	327	struct page *page;
	328	struct persistent_gnt *persistent_gnt;
	329	grant_handle_t handle;
	330	grant_ref_t gref;
	331	};
	332
bf0720c4 RPM	333	/*
	334	* Each outstanding request that we've passed to the lower device layers has a
	335	* 'pending_req' allocated to it. Each buffer_head that completes decrements
	336	* the pendcnt towards zero. When it hits zero, the specified domain has a
	337	* response queued for it, with the saved 'id' passed back.
	338	*/
	339	struct pending_req {
	340	struct xen_blkif *blkif;
	341	u64 id;
	342	int nr_pages;
	343	atomic_t pendcnt;
	344	unsigned short operation;
	345	int status;
	346	struct list_head free_list;
bb642e83	347	struct grant_page *segments[MAX_INDIRECT_SEGMENTS];
402b27f9	348	/* Indirect descriptors */
bb642e83	349	struct grant_page *indirect_pages[MAX_INDIRECT_PAGES];
402b27f9 RPM	350	struct seg_buf seg[MAX_INDIRECT_SEGMENTS];
402b27f9 RPM	351	struct bio *biolist[MAX_INDIRECT_SEGMENTS];
bf0720c4 RPM	352	};
bf0720c4 RPM	353
42c7841d KRW	354
	355	#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
	356	(_v)->bdev->bd_part->nr_sects : \
	357	get_capacity((_v)->bdev->bd_disk))
4d05a28d	358
8b6bf747 KRW	359	#define xen_blkif_get(_b) (atomic_inc(&(_b)->refcnt))
8b6bf747 KRW	360	#define xen_blkif_put(_b) \
4d05a28d KRW	361	do { \
	362	if (atomic_dec_and_test(&(_b)->refcnt)) \
	363	wake_up(&(_b)->waiting_to_free);\
	364	} while (0)
	365
4d05a28d	366	struct phys_req {
01f37f2d	367	unsigned short dev;
b3cb0d6a	368	blkif_sector_t nr_sects;
01f37f2d KRW	369	struct block_device *bdev;
01f37f2d KRW	370	blkif_sector_t sector_number;
4d05a28d	371	};
8b6bf747	372	int xen_blkif_interface_init(void);
4d05a28d	373
8b6bf747	374	int xen_blkif_xenbus_init(void);
4d05a28d	375
8b6bf747 KRW	376	irqreturn_t xen_blkif_be_int(int irq, void *dev_id);
8b6bf747 KRW	377	int xen_blkif_schedule(void *arg);
3f3aad5e	378	int xen_blkif_purge_persistent(void *arg);
4d05a28d	379
24f567f9 KRW	380	int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt,
24f567f9 KRW	381	struct backend_info *be, int state);
4d05a28d	382
29bde093 KRW	383	int xen_blkbk_barrier(struct xenbus_transaction xbt,
29bde093 KRW	384	struct backend_info *be, int state);
8b6bf747	385	struct xenbus_device xen_blkbk_xenbus(struct backend_info be);
98e036a3	386
b0f80127	387	static inline void blkif_get_x86_32_req(struct blkif_request *dst,
68c88dd7 KRW	388	struct blkif_x86_32_request *src)
68c88dd7 KRW	389	{
402b27f9	390	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	391	dst->operation = src->operation;
b3cb0d6a LD	392	switch (src->operation) {
	393	case BLKIF_OP_READ:
	394	case BLKIF_OP_WRITE:
	395	case BLKIF_OP_WRITE_BARRIER:
	396	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	397	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	398	dst->u.rw.handle = src->u.rw.handle;
	399	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	400	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	401	barrier();
97e36834 KRW	402	if (n > dst->u.rw.nr_segments)
97e36834 KRW	403	n = dst->u.rw.nr_segments;
b3cb0d6a LD	404	for (i = 0; i < n; i++)
	405	dst->u.rw.seg[i] = src->u.rw.seg[i];
	406	break;
	407	case BLKIF_OP_DISCARD:
5ea42986	408	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	409	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	410	dst->u.discard.sector_number = src->u.discard.sector_number;
	411	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	412	break;
402b27f9 RPM	413	case BLKIF_OP_INDIRECT:
	414	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	415	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	416	dst->u.indirect.handle = src->u.indirect.handle;
	417	dst->u.indirect.id = src->u.indirect.id;
	418	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	419	barrier();
	420	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	421	for (i = 0; i < j; i++)
	422	dst->u.indirect.indirect_grefs[i] =
	423	src->u.indirect.indirect_grefs[i];
	424	break;
b3cb0d6a	425	default:
0e367ae4 DV	426	/*
	427	* Don't know how to translate this op. Only get the
	428	* ID so failure can be reported to the frontend.
	429	*/
	430	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	431	break;
b3cb0d6a LD	432	}
68c88dd7 KRW	433	}
68c88dd7 KRW	434
b0f80127	435	static inline void blkif_get_x86_64_req(struct blkif_request *dst,
68c88dd7 KRW	436	struct blkif_x86_64_request *src)
68c88dd7 KRW	437	{
402b27f9	438	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	439	dst->operation = src->operation;
b3cb0d6a LD	440	switch (src->operation) {
	441	case BLKIF_OP_READ:
	442	case BLKIF_OP_WRITE:
	443	case BLKIF_OP_WRITE_BARRIER:
	444	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	445	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	446	dst->u.rw.handle = src->u.rw.handle;
	447	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	448	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	449	barrier();
97e36834 KRW	450	if (n > dst->u.rw.nr_segments)
97e36834 KRW	451	n = dst->u.rw.nr_segments;
b3cb0d6a LD	452	for (i = 0; i < n; i++)
	453	dst->u.rw.seg[i] = src->u.rw.seg[i];
	454	break;
	455	case BLKIF_OP_DISCARD:
5ea42986	456	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	457	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	458	dst->u.discard.sector_number = src->u.discard.sector_number;
	459	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	460	break;
402b27f9 RPM	461	case BLKIF_OP_INDIRECT:
	462	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	463	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	464	dst->u.indirect.handle = src->u.indirect.handle;
	465	dst->u.indirect.id = src->u.indirect.id;
	466	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	467	barrier();
	468	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	469	for (i = 0; i < j; i++)
	470	dst->u.indirect.indirect_grefs[i] =
	471	src->u.indirect.indirect_grefs[i];
	472	break;
b3cb0d6a	473	default:
0e367ae4 DV	474	/*
	475	* Don't know how to translate this op. Only get the
	476	* ID so failure can be reported to the frontend.
	477	*/
	478	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	479	break;
b3cb0d6a LD	480	}
68c88dd7 KRW	481	}
68c88dd7 KRW	482
5a577e38	483	#endif /* __XEN_BLKIF__BACKEND__COMMON_H__ */