[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>

extern unsigned int xen_blkif_max_ring_order;
/*
 * 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;
	unsigned int nr_ring_pages;
};

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_segs;
	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
86839c56	47	extern unsigned int xen_blkif_max_ring_order;
402b27f9 RPM	48	/*
	49	* This is the maximum number of segments that would be allowed in indirect
	50	* requests. This value will also be passed to the frontend.
	51	*/
	52	#define MAX_INDIRECT_SEGMENTS 256
	53
	54	#define SEGS_PER_INDIRECT_FRAME \
80bfa2f6	55	(PAGE_SIZE/sizeof(struct blkif_request_segment))
402b27f9 RPM	56	#define MAX_INDIRECT_PAGES \
	57	((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	58	#define INDIRECT_PAGES(_segs) \
	59	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
	60
452a6b2b KRW	61	/* Not a real protocol. Used to generate ring structs which contain
	62	* the elements common to all protocols only. This way we get a
	63	* compiler-checkable way to use common struct elements, so we can
	64	* avoid using switch(protocol) in a number of places. */
	65	struct blkif_common_request {
	66	char dummy;
	67	};
	68	struct blkif_common_response {
	69	char dummy;
	70	};
	71
b3cb0d6a	72	struct blkif_x86_32_request_rw {
97e36834 KRW	73	uint8_t nr_segments; /* number of segments */
	74	blkif_vdev_t handle; /* only for read/write requests */
	75	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a LD	76	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	77	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	78	} __attribute__((__packed__));
b3cb0d6a LD	79
b3cb0d6a LD	80	struct blkif_x86_32_request_discard {
5ea42986	81	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	82	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
97e36834 KRW	83	uint64_t id; /* private guest value, echoed in resp */
b3cb0d6a	84	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	85	uint64_t nr_sectors;
97e36834 KRW	86	} __attribute__((__packed__));
b3cb0d6a	87
0e367ae4 DV	88	struct blkif_x86_32_request_other {
	89	uint8_t _pad1;
	90	blkif_vdev_t _pad2;
	91	uint64_t id; /* private guest value, echoed in resp */
	92	} __attribute__((__packed__));
	93
402b27f9 RPM	94	struct blkif_x86_32_request_indirect {
	95	uint8_t indirect_op;
	96	uint16_t nr_segments;
	97	uint64_t id;
	98	blkif_sector_t sector_number;
	99	blkif_vdev_t handle;
	100	uint16_t _pad1;
	101	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	102	/*
	103	* The maximum number of indirect segments (and pages) that will
	104	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	105	* is also exported to the guest (via xenstore
	106	* feature-max-indirect-segments entry), so the frontend knows how
	107	* many indirect segments the backend supports.
	108	*/
	109	uint64_t _pad2; /* make it 64 byte aligned */
	110	} __attribute__((__packed__));
	111
452a6b2b KRW	112	struct blkif_x86_32_request {
452a6b2b KRW	113	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	114	union {
	115	struct blkif_x86_32_request_rw rw;
	116	struct blkif_x86_32_request_discard discard;
0e367ae4	117	struct blkif_x86_32_request_other other;
402b27f9	118	struct blkif_x86_32_request_indirect indirect;
b3cb0d6a	119	} u;
97e36834 KRW	120	} __attribute__((__packed__));
	121
	122	/* i386 protocol version */
	123	#pragma pack(push, 4)
452a6b2b KRW	124	struct blkif_x86_32_response {
	125	uint64_t id; /* copied from request */
	126	uint8_t operation; /* copied from request */
	127	int16_t status; /* BLKIF_RSP_??? */
	128	};
452a6b2b	129	#pragma pack(pop)
452a6b2b	130	/* x86_64 protocol version */
b3cb0d6a LD	131
b3cb0d6a LD	132	struct blkif_x86_64_request_rw {
97e36834 KRW	133	uint8_t nr_segments; /* number of segments */
	134	blkif_vdev_t handle; /* only for read/write requests */
	135	uint32_t _pad1; /* offsetof(blkif_reqest..,u.rw.id)==8 */
	136	uint64_t id;
b3cb0d6a LD	137	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
b3cb0d6a LD	138	struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
97e36834	139	} __attribute__((__packed__));
b3cb0d6a LD	140
b3cb0d6a LD	141	struct blkif_x86_64_request_discard {
5ea42986	142	uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
97e36834 KRW	143	blkif_vdev_t _pad1; /* was "handle" for read/write requests */
	144	uint32_t _pad2; /* offsetof(blkif_..,u.discard.id)==8 */
	145	uint64_t id;
b3cb0d6a	146	blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
97e36834 KRW	147	uint64_t nr_sectors;
97e36834 KRW	148	} __attribute__((__packed__));
b3cb0d6a	149
0e367ae4 DV	150	struct blkif_x86_64_request_other {
	151	uint8_t _pad1;
	152	blkif_vdev_t _pad2;
	153	uint32_t _pad3; /* offsetof(blkif_..,u.discard.id)==8 */
	154	uint64_t id; /* private guest value, echoed in resp */
	155	} __attribute__((__packed__));
	156
402b27f9 RPM	157	struct blkif_x86_64_request_indirect {
	158	uint8_t indirect_op;
	159	uint16_t nr_segments;
	160	uint32_t _pad1; /* offsetof(blkif_..,u.indirect.id)==8 */
	161	uint64_t id;
	162	blkif_sector_t sector_number;
	163	blkif_vdev_t handle;
	164	uint16_t _pad2;
	165	grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
	166	/*
	167	* The maximum number of indirect segments (and pages) that will
	168	* be used is determined by MAX_INDIRECT_SEGMENTS, this value
	169	* is also exported to the guest (via xenstore
	170	* feature-max-indirect-segments entry), so the frontend knows how
	171	* many indirect segments the backend supports.
	172	*/
	173	uint32_t _pad3; /* make it 64 byte aligned */
	174	} __attribute__((__packed__));
	175
452a6b2b KRW	176	struct blkif_x86_64_request {
452a6b2b KRW	177	uint8_t operation; /* BLKIF_OP_??? */
b3cb0d6a LD	178	union {
	179	struct blkif_x86_64_request_rw rw;
	180	struct blkif_x86_64_request_discard discard;
0e367ae4	181	struct blkif_x86_64_request_other other;
402b27f9	182	struct blkif_x86_64_request_indirect indirect;
b3cb0d6a	183	} u;
97e36834 KRW	184	} __attribute__((__packed__));
97e36834 KRW	185
452a6b2b KRW	186	struct blkif_x86_64_response {
	187	uint64_t __attribute__((__aligned__(8))) id;
	188	uint8_t operation; /* copied from request */
	189	int16_t status; /* BLKIF_RSP_??? */
	190	};
452a6b2b KRW	191
	192	DEFINE_RING_TYPES(blkif_common, struct blkif_common_request,
	193	struct blkif_common_response);
	194	DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request,
	195	struct blkif_x86_32_response);
	196	DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request,
	197	struct blkif_x86_64_response);
	198
	199	union blkif_back_rings {
	200	struct blkif_back_ring native;
	201	struct blkif_common_back_ring common;
	202	struct blkif_x86_32_back_ring x86_32;
	203	struct blkif_x86_64_back_ring x86_64;
	204	};
	205
	206	enum blkif_protocol {
	207	BLKIF_PROTOCOL_NATIVE = 1,
	208	BLKIF_PROTOCOL_X86_32 = 2,
	209	BLKIF_PROTOCOL_X86_64 = 3,
	210	};
	211
b042a3ca DV	212	/*
	213	* Default protocol if the frontend doesn't specify one.
	214	*/
	215	#ifdef CONFIG_X86
	216	# define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_X86_32
	217	#else
	218	# define BLKIF_PROTOCOL_DEFAULT BLKIF_PROTOCOL_NATIVE
	219	#endif
	220
3d814731	221	struct xen_vbd {
01f37f2d KRW	222	/* What the domain refers to this vbd as. */
	223	blkif_vdev_t handle;
	224	/* Non-zero -> read-only */
	225	unsigned char readonly;
	226	/* VDISK_xxx */
	227	unsigned char type;
	228	/* phys device that this vbd maps to. */
	229	u32 pdevice;
	230	struct block_device *bdev;
	231	/* Cached size parameter. */
	232	sector_t size;
1f999572 OC	233	unsigned int flush_support:1;
1f999572 OC	234	unsigned int discard_secure:1;
0a8704a5 RPM	235	unsigned int feature_gnt_persistent:1;
0a8704a5 RPM	236	unsigned int overflow_max_grants:1;
4d05a28d KRW	237	};
	238
	239	struct backend_info;
	240
3f3aad5e RPM	241	/* Number of available flags */
	242	#define PERSISTENT_GNT_FLAGS_SIZE 2
	243	/* This persistent grant is currently in use */
	244	#define PERSISTENT_GNT_ACTIVE 0
	245	/*
	246	* This persistent grant has been used, this flag is set when we remove the
	247	* PERSISTENT_GNT_ACTIVE, to know that this grant has been used recently.
	248	*/
	249	#define PERSISTENT_GNT_WAS_ACTIVE 1
0a8704a5	250
bf0720c4	251	/* Number of requests that we can fit in a ring */
69b91ede	252	#define XEN_BLKIF_REQS_PER_PAGE 32
bf0720c4	253
0a8704a5 RPM	254	struct persistent_gnt {
	255	struct page *page;
	256	grant_ref_t gnt;
	257	grant_handle_t handle;
3f3aad5e	258	DECLARE_BITMAP(flags, PERSISTENT_GNT_FLAGS_SIZE);
0a8704a5	259	struct rb_node node;
3f3aad5e	260	struct list_head remove_node;
0a8704a5 RPM	261	};
0a8704a5 RPM	262
30fd1502	263	struct xen_blkif {
4d05a28d	264	/* Unique identifier for this interface. */
01f37f2d KRW	265	domid_t domid;
01f37f2d KRW	266	unsigned int handle;
4d05a28d	267	/* Physical parameters of the comms window. */
01f37f2d	268	unsigned int irq;
4d05a28d	269	/* Comms information. */
01f37f2d KRW	270	enum blkif_protocol blk_protocol;
01f37f2d KRW	271	union blkif_back_rings blk_rings;
2d073846	272	void *blk_ring;
4d05a28d	273	/* The VBD attached to this interface. */
3d814731	274	struct xen_vbd vbd;
4d05a28d	275	/* Back pointer to the backend_info. */
01f37f2d	276	struct backend_info *be;
4d05a28d	277	/* Private fields. */
01f37f2d KRW	278	spinlock_t blk_ring_lock;
01f37f2d KRW	279	atomic_t refcnt;
4d05a28d	280
01f37f2d	281	wait_queue_head_t wq;
29bde093 KRW	282	/* for barrier (drain) requests */
	283	struct completion drain_complete;
	284	atomic_t drain;
c05f3e3c	285	atomic_t inflight;
a1397fa3	286	/* One thread per one blkif. */
01f37f2d KRW	287	struct task_struct *xenblkd;
01f37f2d KRW	288	unsigned int waiting_reqs;
4d05a28d	289
0a8704a5 RPM	290	/* tree to store persistent grants */
	291	struct rb_root persistent_gnts;
	292	unsigned int persistent_gnt_c;
3f3aad5e RPM	293	atomic_t persistent_gnt_in_use;
	294	unsigned long next_lru;
	295
	296	/* used by the kworker that offload work from the persistent purge */
	297	struct list_head persistent_purge_list;
	298	struct work_struct persistent_purge_work;
0a8704a5	299
c6cc142d RPM	300	/* buffer of free pages to map grant refs */
	301	spinlock_t free_pages_lock;
	302	int free_pages_num;
	303	struct list_head free_pages;
	304
bf0720c4 RPM	305	/* List of all 'pending_req' available */
	306	struct list_head pending_free;
	307	/* And its spinlock. */
	308	spinlock_t pending_free_lock;
	309	wait_queue_head_t pending_free_wq;
	310
4d05a28d	311	/* statistics */
01f37f2d	312	unsigned long st_print;
986cacbd ZK	313	unsigned long long st_rd_req;
	314	unsigned long long st_wr_req;
	315	unsigned long long st_oo_req;
	316	unsigned long long st_f_req;
	317	unsigned long long st_ds_req;
	318	unsigned long long st_rd_sect;
	319	unsigned long long st_wr_sect;
01f37f2d	320
814d04e7	321	struct work_struct free_work;
8e3f8755 KRW	322	/* Thread shutdown wait queue. */
8e3f8755 KRW	323	wait_queue_head_t shutdown_wq;
86839c56	324	unsigned int nr_ring_pages;
5489377c	325	};
4d05a28d	326
402b27f9 RPM	327	struct seg_buf {
	328	unsigned long offset;
	329	unsigned int nsec;
	330	};
	331
bb642e83 RPM	332	struct grant_page {
	333	struct page *page;
	334	struct persistent_gnt *persistent_gnt;
	335	grant_handle_t handle;
	336	grant_ref_t gref;
	337	};
	338
bf0720c4 RPM	339	/*
	340	* Each outstanding request that we've passed to the lower device layers has a
	341	* 'pending_req' allocated to it. Each buffer_head that completes decrements
	342	* the pendcnt towards zero. When it hits zero, the specified domain has a
	343	* response queued for it, with the saved 'id' passed back.
	344	*/
	345	struct pending_req {
	346	struct xen_blkif *blkif;
	347	u64 id;
6684fa1c	348	int nr_segs;
bf0720c4 RPM	349	atomic_t pendcnt;
	350	unsigned short operation;
	351	int status;
	352	struct list_head free_list;
bb642e83	353	struct grant_page *segments[MAX_INDIRECT_SEGMENTS];
402b27f9	354	/* Indirect descriptors */
bb642e83	355	struct grant_page *indirect_pages[MAX_INDIRECT_PAGES];
402b27f9 RPM	356	struct seg_buf seg[MAX_INDIRECT_SEGMENTS];
402b27f9 RPM	357	struct bio *biolist[MAX_INDIRECT_SEGMENTS];
c43cf3ea JH	358	struct gnttab_unmap_grant_ref unmap[MAX_INDIRECT_SEGMENTS];
	359	struct page *unmap_pages[MAX_INDIRECT_SEGMENTS];
	360	struct gntab_unmap_queue_data gnttab_unmap_data;
bf0720c4 RPM	361	};
bf0720c4 RPM	362
42c7841d KRW	363
	364	#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
	365	(_v)->bdev->bd_part->nr_sects : \
	366	get_capacity((_v)->bdev->bd_disk))
4d05a28d	367
8b6bf747 KRW	368	#define xen_blkif_get(_b) (atomic_inc(&(_b)->refcnt))
8b6bf747 KRW	369	#define xen_blkif_put(_b) \
4d05a28d KRW	370	do { \
4d05a28d KRW	371	if (atomic_dec_and_test(&(_b)->refcnt)) \
814d04e7	372	schedule_work(&(_b)->free_work);\
4d05a28d KRW	373	} while (0)
4d05a28d KRW	374
4d05a28d	375	struct phys_req {
01f37f2d	376	unsigned short dev;
b3cb0d6a	377	blkif_sector_t nr_sects;
01f37f2d KRW	378	struct block_device *bdev;
01f37f2d KRW	379	blkif_sector_t sector_number;
4d05a28d	380	};
8b6bf747	381	int xen_blkif_interface_init(void);
4d05a28d	382
8b6bf747	383	int xen_blkif_xenbus_init(void);
4d05a28d	384
8b6bf747 KRW	385	irqreturn_t xen_blkif_be_int(int irq, void *dev_id);
8b6bf747 KRW	386	int xen_blkif_schedule(void *arg);
3f3aad5e	387	int xen_blkif_purge_persistent(void *arg);
ef753411	388	void xen_blkbk_free_caches(struct xen_blkif *blkif);
4d05a28d	389
24f567f9 KRW	390	int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt,
24f567f9 KRW	391	struct backend_info *be, int state);
4d05a28d	392
29bde093 KRW	393	int xen_blkbk_barrier(struct xenbus_transaction xbt,
29bde093 KRW	394	struct backend_info *be, int state);
8b6bf747	395	struct xenbus_device xen_blkbk_xenbus(struct backend_info be);
abb97b8c	396	void xen_blkbk_unmap_purged_grants(struct work_struct *work);
98e036a3	397
b0f80127	398	static inline void blkif_get_x86_32_req(struct blkif_request *dst,
68c88dd7 KRW	399	struct blkif_x86_32_request *src)
68c88dd7 KRW	400	{
402b27f9	401	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	402	dst->operation = src->operation;
b3cb0d6a LD	403	switch (src->operation) {
	404	case BLKIF_OP_READ:
	405	case BLKIF_OP_WRITE:
	406	case BLKIF_OP_WRITE_BARRIER:
	407	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	408	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	409	dst->u.rw.handle = src->u.rw.handle;
	410	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	411	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	412	barrier();
97e36834 KRW	413	if (n > dst->u.rw.nr_segments)
97e36834 KRW	414	n = dst->u.rw.nr_segments;
b3cb0d6a LD	415	for (i = 0; i < n; i++)
	416	dst->u.rw.seg[i] = src->u.rw.seg[i];
	417	break;
	418	case BLKIF_OP_DISCARD:
5ea42986	419	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	420	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	421	dst->u.discard.sector_number = src->u.discard.sector_number;
	422	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	423	break;
402b27f9 RPM	424	case BLKIF_OP_INDIRECT:
	425	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	426	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	427	dst->u.indirect.handle = src->u.indirect.handle;
	428	dst->u.indirect.id = src->u.indirect.id;
	429	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	430	barrier();
	431	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	432	for (i = 0; i < j; i++)
	433	dst->u.indirect.indirect_grefs[i] =
	434	src->u.indirect.indirect_grefs[i];
	435	break;
b3cb0d6a	436	default:
0e367ae4 DV	437	/*
	438	* Don't know how to translate this op. Only get the
	439	* ID so failure can be reported to the frontend.
	440	*/
	441	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	442	break;
b3cb0d6a LD	443	}
68c88dd7 KRW	444	}
68c88dd7 KRW	445
b0f80127	446	static inline void blkif_get_x86_64_req(struct blkif_request *dst,
68c88dd7 KRW	447	struct blkif_x86_64_request *src)
68c88dd7 KRW	448	{
402b27f9	449	int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
68c88dd7	450	dst->operation = src->operation;
b3cb0d6a LD	451	switch (src->operation) {
	452	case BLKIF_OP_READ:
	453	case BLKIF_OP_WRITE:
	454	case BLKIF_OP_WRITE_BARRIER:
	455	case BLKIF_OP_FLUSH_DISKCACHE:
97e36834 KRW	456	dst->u.rw.nr_segments = src->u.rw.nr_segments;
	457	dst->u.rw.handle = src->u.rw.handle;
	458	dst->u.rw.id = src->u.rw.id;
b3cb0d6a LD	459	dst->u.rw.sector_number = src->u.rw.sector_number;
b3cb0d6a LD	460	barrier();
97e36834 KRW	461	if (n > dst->u.rw.nr_segments)
97e36834 KRW	462	n = dst->u.rw.nr_segments;
b3cb0d6a LD	463	for (i = 0; i < n; i++)
	464	dst->u.rw.seg[i] = src->u.rw.seg[i];
	465	break;
	466	case BLKIF_OP_DISCARD:
5ea42986	467	dst->u.discard.flag = src->u.discard.flag;
8c9ce606	468	dst->u.discard.id = src->u.discard.id;
b3cb0d6a LD	469	dst->u.discard.sector_number = src->u.discard.sector_number;
	470	dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
	471	break;
402b27f9 RPM	472	case BLKIF_OP_INDIRECT:
	473	dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
	474	dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
	475	dst->u.indirect.handle = src->u.indirect.handle;
	476	dst->u.indirect.id = src->u.indirect.id;
	477	dst->u.indirect.sector_number = src->u.indirect.sector_number;
	478	barrier();
	479	j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
	480	for (i = 0; i < j; i++)
	481	dst->u.indirect.indirect_grefs[i] =
	482	src->u.indirect.indirect_grefs[i];
	483	break;
b3cb0d6a	484	default:
0e367ae4 DV	485	/*
	486	* Don't know how to translate this op. Only get the
	487	* ID so failure can be reported to the frontend.
	488	*/
	489	dst->u.other.id = src->u.other.id;
b3cb0d6a LD	490	break;
b3cb0d6a LD	491	}
68c88dd7 KRW	492	}
68c88dd7 KRW	493
5a577e38	494	#endif /* __XEN_BLKIF__BACKEND__COMMON_H__ */