[linux-2.6-block.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>

/*
 * 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))
#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,
};

/*
 * Default protocol if the frontend doesn't specify one.
 */
#ifdef CONFIG_X86
#  define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_X86_32
#else
#  define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_NATIVE
#endif

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_PER_PAGE		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;
	atomic_t		inflight;
	/* 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;

	struct work_struct	free_work;
	/* 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];
	struct gnttab_unmap_grant_ref unmap[MAX_INDIRECT_SEGMENTS];
	struct page                   *unmap_pages[MAX_INDIRECT_SEGMENTS];
	struct gntab_unmap_queue_data gnttab_unmap_data;
};


#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))	\
			schedule_work(&(_b)->free_work);\
	} 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);
void xen_blkbk_free_caches(struct xen_blkif *blkif);

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);
void xen_blkbk_unmap_purged_grants(struct work_struct *work);

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
402b27f9 RPM	47	/*
	48	* This is the maximum number of segments that would be allowed in indirect
	49	* requests. This value will also be passed to the frontend.
	50	*/
	51	#define MAX_INDIRECT_SEGMENTS 256
	52
	53	#define SEGS_PER_INDIRECT_FRAME \
80bfa2f6	54	(PAGE_SIZE/sizeof(struct blkif_request_segment))
402b27f9 RPM	55	#define MAX_INDIRECT_PAGES \
	56	((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	57	#define INDIRECT_PAGES(_segs) \
	58	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	59
452a6b2b KRW	60	/* Not a real protocol. Used to generate ring structs which contain
	61	* the elements common to all protocols only. This way we get a
	62	* compiler-checkable way to use common struct elements, so we can
	63	* avoid using switch(protocol) in a number of places. */
	64	struct blkif_common_request {
	65	char dummy;
	66	};
	67	struct blkif_common_response {
	68	char dummy;
	69	};
	70
b3cb0d6a	71	struct blkif_x86_32_request_rw {
97e36834 KRW	72	uint8_t nr_segments; /* number of segments */
	73	blkif_vdev_t handle; /* only for read/write requests */
	74	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a LD	75	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	76	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	77	} __attribute__((__packed__));
b3cb0d6a LD	78
b3cb0d6a LD	79	struct blkif_x86_32_request_discard {
5ea42986	80	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	81	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
97e36834 KRW	82	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a	83	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	84	uint64_t nr_sectors;
97e36834 KRW	85	} __attribute__((__packed__));
b3cb0d6a	86
0e367ae4 DV	87	struct blkif_x86_32_request_other {
	88	uint8_t _pad1;
	89	blkif_vdev_t _pad2;
	90	uint64_t id; /* private guest value, echoed in resp */
	91	} __attribute__((__packed__));
	92
402b27f9 RPM	93	struct blkif_x86_32_request_indirect {
	94	uint8_t indirect_op;
	95	uint16_t nr_segments;
	96	uint64_t id;
	97	blkif_sector_t sector_number;
	98	blkif_vdev_t handle;
	99	uint16_t _pad1;
	100	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	101	/*
	102	* The maximum number of indirect segments (and pages) that will
	103	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	104	* is also exported to the guest (via xenstore
	105	* feature-max-indirect-segments entry), so the frontend knows how
	106	* many indirect segments the backend supports.
	107	*/
	108	uint64_t _pad2; /* make it 64 byte aligned */
	109	} __attribute__((__packed__));
	110
452a6b2b KRW	111	struct blkif_x86_32_request {
452a6b2b KRW	112	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	113	union {
	114	struct blkif_x86_32_request_rw rw;
	115	struct blkif_x86_32_request_discard discard;
0e367ae4	116	struct blkif_x86_32_request_other other;
402b27f9	117	struct blkif_x86_32_request_indirect indirect;
b3cb0d6a	118	} u;
97e36834 KRW	119	} __attribute__((__packed__));
	120
	121	/* i386 protocol version */
	122	#pragma pack(push, 4)
452a6b2b KRW	123	struct blkif_x86_32_response {
	124	uint64_t id; /* copied from request */
	125	uint8_t operation; /* copied from request */
	126	int16_t status; /* BLKIF_RSP_??? */
	127	};
452a6b2b	128	#pragma pack(pop)
452a6b2b	129	/* x86_64 protocol version */
b3cb0d6a LD	130
b3cb0d6a LD	131	struct blkif_x86_64_request_rw {
97e36834 KRW	132	uint8_t nr_segments; /* number of segments */
	133	blkif_vdev_t handle; /* only for read/write requests */
	134	uint32_t _pad1; /* offsetof(blkif_reqest..,u.rw.id)==8 */
	135	uint64_t id;
b3cb0d6a LD	136	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	137	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	138	} __attribute__((__packed__));
b3cb0d6a LD	139
b3cb0d6a LD	140	struct blkif_x86_64_request_discard {
5ea42986	141	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	142	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
	143	uint32_t _pad2; /* offsetof(blkif_..,u.discard.id)==8 */
	144	uint64_t id;
b3cb0d6a	145	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	146	uint64_t nr_sectors;
97e36834 KRW	147	} __attribute__((__packed__));
b3cb0d6a	148
0e367ae4 DV	149	struct blkif_x86_64_request_other {
	150	uint8_t _pad1;
	151	blkif_vdev_t _pad2;
	152	uint32_t _pad3; /* offsetof(blkif_..,u.discard.id)==8 */
	153	uint64_t id; /* private guest value, echoed in resp */
	154	} __attribute__((__packed__));
	155
402b27f9 RPM	156	struct blkif_x86_64_request_indirect {
	157	uint8_t indirect_op;
	158	uint16_t nr_segments;
	159	uint32_t _pad1; /* offsetof(blkif_..,u.indirect.id)==8 */
	160	uint64_t id;
	161	blkif_sector_t sector_number;
	162	blkif_vdev_t handle;
	163	uint16_t _pad2;
	164	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	165	/*
	166	* The maximum number of indirect segments (and pages) that will
	167	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	168	* is also exported to the guest (via xenstore
	169	* feature-max-indirect-segments entry), so the frontend knows how
	170	* many indirect segments the backend supports.
	171	*/
	172	uint32_t _pad3; /* make it 64 byte aligned */
	173	} __attribute__((__packed__));
	174
452a6b2b KRW	175	struct blkif_x86_64_request {
452a6b2b KRW	176	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	177	union {
	178	struct blkif_x86_64_request_rw rw;
	179	struct blkif_x86_64_request_discard discard;
0e367ae4	180	struct blkif_x86_64_request_other other;
402b27f9	181	struct blkif_x86_64_request_indirect indirect;
b3cb0d6a	182	} u;
97e36834 KRW	183	} __attribute__((__packed__));
97e36834 KRW	184
452a6b2b KRW	185	struct blkif_x86_64_response {
	186	uint64_t __attribute__((__aligned__(8))) id;
	187	uint8_t operation; /* copied from request */
	188	int16_t status; /* BLKIF_RSP_??? */
	189	};
452a6b2b KRW	190
	191	DEFINE_RING_TYPES(blkif_common, struct blkif_common_request,
	192	struct blkif_common_response);
	193	DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request,
	194	struct blkif_x86_32_response);
	195	DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request,
	196	struct blkif_x86_64_response);
	197
	198	union blkif_back_rings {
	199	struct blkif_back_ring native;
	200	struct blkif_common_back_ring common;
	201	struct blkif_x86_32_back_ring x86_32;
	202	struct blkif_x86_64_back_ring x86_64;
	203	};
	204
	205	enum blkif_protocol {
	206	BLKIF_PROTOCOL_NATIVE = 1,
	207	BLKIF_PROTOCOL_X86_32 = 2,
	208	BLKIF_PROTOCOL_X86_64 = 3,
	209	};
	210
b042a3ca DV	211	/*
	212	* Default protocol if the frontend doesn't specify one.
	213	*/
	214	#ifdef CONFIG_X86
	215	# define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_X86_32
	216	#else
	217	# define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_NATIVE
	218	#endif
	219
3d814731	220	struct xen_vbd {
01f37f2d KRW	221	/* What the domain refers to this vbd as. */
	222	blkif_vdev_t handle;
	223	/* Non-zero -> read-only */
	224	unsigned char readonly;
	225	/* VDISK_xxx */
	226	unsigned char type;
	227	/* phys device that this vbd maps to. */
	228	u32 pdevice;
	229	struct block_device *bdev;
	230	/* Cached size parameter. */
	231	sector_t size;
1f999572 OC	232	unsigned int flush_support:1;
1f999572 OC	233	unsigned int discard_secure:1;
0a8704a5 RPM	234	unsigned int feature_gnt_persistent:1;
0a8704a5 RPM	235	unsigned int overflow_max_grants:1;
4d05a28d KRW	236	};
	237
	238	struct backend_info;
	239
3f3aad5e RPM	240	/* Number of available flags */
	241	#define PERSISTENT_GNT_FLAGS_SIZE 2
	242	/* This persistent grant is currently in use */
	243	#define PERSISTENT_GNT_ACTIVE 0
	244	/*
	245	* This persistent grant has been used, this flag is set when we remove the
	246	* PERSISTENT_GNT_ACTIVE, to know that this grant has been used recently.
	247	*/
	248	#define PERSISTENT_GNT_WAS_ACTIVE 1
0a8704a5	249
bf0720c4	250	/* Number of requests that we can fit in a ring */
69b91ede	251	#define XEN_BLKIF_REQS_PER_PAGE 32
bf0720c4	252
0a8704a5 RPM	253	struct persistent_gnt {
	254	struct page *page;
	255	grant_ref_t gnt;
	256	grant_handle_t handle;
3f3aad5e	257	DECLARE_BITMAP(flags, PERSISTENT_GNT_FLAGS_SIZE);
0a8704a5	258	struct rb_node node;
3f3aad5e	259	struct list_head remove_node;
0a8704a5 RPM	260	};
0a8704a5 RPM	261
30fd1502	262	struct xen_blkif {
4d05a28d	263	/* Unique identifier for this interface. */
01f37f2d KRW	264	domid_t domid;
01f37f2d KRW	265	unsigned int handle;
4d05a28d	266	/* Physical parameters of the comms window. */
01f37f2d	267	unsigned int irq;
4d05a28d	268	/* Comms information. */
01f37f2d KRW	269	enum blkif_protocol blk_protocol;
01f37f2d KRW	270	union blkif_back_rings blk_rings;
2d073846	271	void *blk_ring;
4d05a28d	272	/* The VBD attached to this interface. */
3d814731	273	struct xen_vbd vbd;
4d05a28d	274	/* Back pointer to the backend_info. */
01f37f2d	275	struct backend_info *be;
4d05a28d	276	/* Private fields. */
01f37f2d KRW	277	spinlock_t blk_ring_lock;
01f37f2d KRW	278	atomic_t refcnt;
4d05a28d	279
01f37f2d	280	wait_queue_head_t wq;
29bde093 KRW	281	/* for barrier (drain) requests */
	282	struct completion drain_complete;
	283	atomic_t drain;
c05f3e3c	284	atomic_t inflight;
a1397fa3	285	/* One thread per one blkif. */
01f37f2d KRW	286	struct task_struct *xenblkd;
01f37f2d KRW	287	unsigned int waiting_reqs;
4d05a28d	288
0a8704a5 RPM	289	/* tree to store persistent grants */
	290	struct rb_root persistent_gnts;
	291	unsigned int persistent_gnt_c;
3f3aad5e RPM	292	atomic_t persistent_gnt_in_use;
	293	unsigned long next_lru;
	294
	295	/* used by the kworker that offload work from the persistent purge */
	296	struct list_head persistent_purge_list;
	297	struct work_struct persistent_purge_work;
0a8704a5	298
c6cc142d RPM	299	/* buffer of free pages to map grant refs */
	300	spinlock_t free_pages_lock;
	301	int free_pages_num;
	302	struct list_head free_pages;
	303
bf0720c4 RPM	304	/* List of all 'pending_req' available */
	305	struct list_head pending_free;
	306	/* And its spinlock. */
	307	spinlock_t pending_free_lock;
	308	wait_queue_head_t pending_free_wq;
	309
4d05a28d	310	/* statistics */
01f37f2d	311	unsigned long st_print;
986cacbd ZK	312	unsigned long long st_rd_req;
	313	unsigned long long st_wr_req;
	314	unsigned long long st_oo_req;
	315	unsigned long long st_f_req;
	316	unsigned long long st_ds_req;
	317	unsigned long long st_rd_sect;
	318	unsigned long long st_wr_sect;
01f37f2d	319
814d04e7	320	struct work_struct free_work;
8e3f8755 KRW	321	/* Thread shutdown wait queue. */
8e3f8755 KRW	322	wait_queue_head_t shutdown_wq;
5489377c	323	};
4d05a28d	324
402b27f9 RPM	325	struct seg_buf {
	326	unsigned long offset;
	327	unsigned int nsec;
	328	};
	329
bb642e83 RPM	330	struct grant_page {
	331	struct page *page;
	332	struct persistent_gnt *persistent_gnt;
	333	grant_handle_t handle;
	334	grant_ref_t gref;
	335	};
	336
bf0720c4 RPM	337	/*
	338	* Each outstanding request that we've passed to the lower device layers has a
	339	* 'pending_req' allocated to it. Each buffer_head that completes decrements
	340	* the pendcnt towards zero. When it hits zero, the specified domain has a
	341	* response queued for it, with the saved 'id' passed back.
	342	*/
	343	struct pending_req {
	344	struct xen_blkif *blkif;
	345	u64 id;
	346	int nr_pages;
	347	atomic_t pendcnt;
	348	unsigned short operation;
	349	int status;
	350	struct list_head free_list;
bb642e83	351	struct grant_page *segments[MAX_INDIRECT_SEGMENTS];
402b27f9	352	/* Indirect descriptors */
bb642e83	353	struct grant_page *indirect_pages[MAX_INDIRECT_PAGES];
402b27f9 RPM	354	struct seg_buf seg[MAX_INDIRECT_SEGMENTS];
402b27f9 RPM	355	struct bio *biolist[MAX_INDIRECT_SEGMENTS];
c43cf3ea JH	356	struct gnttab_unmap_grant_ref unmap[MAX_INDIRECT_SEGMENTS];
	357	struct page *unmap_pages[MAX_INDIRECT_SEGMENTS];
	358	struct gntab_unmap_queue_data gnttab_unmap_data;
bf0720c4 RPM	359	};
bf0720c4 RPM	360
42c7841d KRW	361
	362	#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
	363	(_v)->bdev->bd_part->nr_sects : \
	364	get_capacity((_v)->bdev->bd_disk))
4d05a28d	365
8b6bf747 KRW	366	#define xen_blkif_get(_b) (atomic_inc(&(_b)->refcnt))
8b6bf747 KRW	367	#define xen_blkif_put(_b) \
4d05a28d KRW	368	do { \
4d05a28d KRW	369	if (atomic_dec_and_test(&(_b)->refcnt)) \
814d04e7	370	schedule_work(&(_b)->free_work);\
4d05a28d KRW	371	} while (0)
4d05a28d KRW	372
4d05a28d	373	struct phys_req {
01f37f2d	374	unsigned short dev;
b3cb0d6a	375	blkif_sector_t nr_sects;
01f37f2d KRW	376	struct block_device *bdev;
01f37f2d KRW	377	blkif_sector_t sector_number;
4d05a28d	378	};
8b6bf747	379	int xen_blkif_interface_init(void);
4d05a28d	380
8b6bf747	381	int xen_blkif_xenbus_init(void);
4d05a28d	382
8b6bf747 KRW	383	irqreturn_t xen_blkif_be_int(int irq, void *dev_id);
8b6bf747 KRW	384	int xen_blkif_schedule(void *arg);
3f3aad5e	385	int xen_blkif_purge_persistent(void *arg);
ef753411	386	void xen_blkbk_free_caches(struct xen_blkif *blkif);
4d05a28d	387
24f567f9 KRW	388	int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt,
24f567f9 KRW	389	struct backend_info *be, int state);
4d05a28d	390
29bde093 KRW	391	int xen_blkbk_barrier(struct xenbus_transaction xbt,
29bde093 KRW	392	struct backend_info *be, int state);
8b6bf747	393	struct xenbus_device xen_blkbk_xenbus(struct backend_info be);
abb97b8c	394	void xen_blkbk_unmap_purged_grants(struct work_struct *work);
98e036a3	395
b0f80127	396	static inline void blkif_get_x86_32_req(struct blkif_request *dst,
68c88dd7 KRW	397	struct blkif_x86_32_request *src)
68c88dd7 KRW	398	{
402b27f9	399	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	400	dst->operation = src->operation;
b3cb0d6a LD	401	switch (src->operation) {
	402	case BLKIF_OP_READ:
	403	case BLKIF_OP_WRITE:
	404	case BLKIF_OP_WRITE_BARRIER:
	405	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	406	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	407	dst->u.rw.handle = src->u.rw.handle;
	408	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	409	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	410	barrier();
97e36834 KRW	411	if (n > dst->u.rw.nr_segments)
97e36834 KRW	412	n = dst->u.rw.nr_segments;
b3cb0d6a LD	413	for (i = 0; i < n; i++)
	414	dst->u.rw.seg[i] = src->u.rw.seg[i];
	415	break;
	416	case BLKIF_OP_DISCARD:
5ea42986	417	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	418	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	419	dst->u.discard.sector_number = src->u.discard.sector_number;
	420	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	421	break;
402b27f9 RPM	422	case BLKIF_OP_INDIRECT:
	423	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	424	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	425	dst->u.indirect.handle = src->u.indirect.handle;
	426	dst->u.indirect.id = src->u.indirect.id;
	427	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	428	barrier();
	429	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	430	for (i = 0; i < j; i++)
	431	dst->u.indirect.indirect_grefs[i] =
	432	src->u.indirect.indirect_grefs[i];
	433	break;
b3cb0d6a	434	default:
0e367ae4 DV	435	/*
	436	* Don't know how to translate this op. Only get the
	437	* ID so failure can be reported to the frontend.
	438	*/
	439	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	440	break;
b3cb0d6a LD	441	}
68c88dd7 KRW	442	}
68c88dd7 KRW	443
b0f80127	444	static inline void blkif_get_x86_64_req(struct blkif_request *dst,
68c88dd7 KRW	445	struct blkif_x86_64_request *src)
68c88dd7 KRW	446	{
402b27f9	447	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	448	dst->operation = src->operation;
b3cb0d6a LD	449	switch (src->operation) {
	450	case BLKIF_OP_READ:
	451	case BLKIF_OP_WRITE:
	452	case BLKIF_OP_WRITE_BARRIER:
	453	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	454	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	455	dst->u.rw.handle = src->u.rw.handle;
	456	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	457	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	458	barrier();
97e36834 KRW	459	if (n > dst->u.rw.nr_segments)
97e36834 KRW	460	n = dst->u.rw.nr_segments;
b3cb0d6a LD	461	for (i = 0; i < n; i++)
	462	dst->u.rw.seg[i] = src->u.rw.seg[i];
	463	break;
	464	case BLKIF_OP_DISCARD:
5ea42986	465	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	466	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	467	dst->u.discard.sector_number = src->u.discard.sector_number;
	468	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	469	break;
402b27f9 RPM	470	case BLKIF_OP_INDIRECT:
	471	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	472	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	473	dst->u.indirect.handle = src->u.indirect.handle;
	474	dst->u.indirect.id = src->u.indirect.id;
	475	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	476	barrier();
	477	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	478	for (i = 0; i < j; i++)
	479	dst->u.indirect.indirect_grefs[i] =
	480	src->u.indirect.indirect_grefs[i];
	481	break;
b3cb0d6a	482	default:
0e367ae4 DV	483	/*
	484	* Don't know how to translate this op. Only get the
	485	* ID so failure can be reported to the frontend.
	486	*/
	487	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	488	break;
b3cb0d6a LD	489	}
68c88dd7 KRW	490	}
68c88dd7 KRW	491
5a577e38	492	#endif /* __XEN_BLKIF__BACKEND__COMMON_H__ */