[mirror_ubuntu-bionic-kernel.git] / drivers / nvme / host / nvme.h

/*
 * Copyright (c) 2011-2014, Intel Corporation.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 */

#ifndef _NVME_H
#define _NVME_H

#include <linux/nvme.h>
#include <linux/cdev.h>
#include <linux/pci.h>
#include <linux/kref.h>
#include <linux/blk-mq.h>
#include <linux/lightnvm.h>
#include <linux/sed-opal.h>

extern unsigned int nvme_io_timeout;
#define NVME_IO_TIMEOUT	(nvme_io_timeout * HZ)

extern unsigned int admin_timeout;
#define ADMIN_TIMEOUT	(admin_timeout * HZ)

#define NVME_DEFAULT_KATO	5
#define NVME_KATO_GRACE		10

extern struct workqueue_struct *nvme_wq;

enum {
	NVME_NS_LBA		= 0,
	NVME_NS_LIGHTNVM	= 1,
};

/*
 * List of workarounds for devices that required behavior not specified in
 * the standard.
 */
enum nvme_quirks {
	/*
	 * Prefers I/O aligned to a stripe size specified in a vendor
	 * specific Identify field.
	 */
	NVME_QUIRK_STRIPE_SIZE			= (1 << 0),

	/*
	 * The controller doesn't handle Identify value others than 0 or 1
	 * correctly.
	 */
	NVME_QUIRK_IDENTIFY_CNS			= (1 << 1),

	/*
	 * The controller deterministically returns O's on reads to
	 * logical blocks that deallocate was called on.
	 */
	NVME_QUIRK_DEALLOCATE_ZEROES		= (1 << 2),

	/*
	 * The controller needs a delay before starts checking the device
	 * readiness, which is done by reading the NVME_CSTS_RDY bit.
	 */
	NVME_QUIRK_DELAY_BEFORE_CHK_RDY		= (1 << 3),

	/*
	 * APST should not be used.
	 */
	NVME_QUIRK_NO_APST			= (1 << 4),

	/*
	 * The deepest sleep state should not be used.
	 */
	NVME_QUIRK_NO_DEEPEST_PS		= (1 << 5),

	/*
	 * Supports the LighNVM command set if indicated in vs[1].
	 */
	NVME_QUIRK_LIGHTNVM			= (1 << 6),

	/*
	 * Set MEDIUM priority on SQ creation
	 */
	NVME_QUIRK_MEDIUM_PRIO_SQ		= (1 << 7),
};

/*
 * Common request structure for NVMe passthrough.  All drivers must have
 * this structure as the first member of their request-private data.
 */
struct nvme_request {
	struct nvme_command	*cmd;
	union nvme_result	result;
	u8			retries;
	u8			flags;
	u16			status;
};

/*
 * Mark a bio as coming in through the mpath node.
 */
#define REQ_NVME_MPATH		REQ_DRV

enum {
	NVME_REQ_CANCELLED		= (1 << 0),
};

static inline struct nvme_request *nvme_req(struct request *req)
{
	return blk_mq_rq_to_pdu(req);
}

/* The below value is the specific amount of delay needed before checking
 * readiness in case of the PCI_DEVICE(0x1c58, 0x0003), which needs the
 * NVME_QUIRK_DELAY_BEFORE_CHK_RDY quirk enabled. The value (in ms) was
 * found empirically.
 */
#define NVME_QUIRK_DELAY_AMOUNT		2300

enum nvme_ctrl_state {
	NVME_CTRL_NEW,
	NVME_CTRL_LIVE,
	NVME_CTRL_RESETTING,
	NVME_CTRL_RECONNECTING,
	NVME_CTRL_DELETING,
	NVME_CTRL_DEAD,
};

struct nvme_ctrl {
	enum nvme_ctrl_state state;
	bool identified;
	spinlock_t lock;
	const struct nvme_ctrl_ops *ops;
	struct request_queue *admin_q;
	struct request_queue *connect_q;
	struct device *dev;
	int instance;
	struct blk_mq_tag_set *tagset;
	struct blk_mq_tag_set *admin_tagset;
	struct list_head namespaces;
	struct mutex namespaces_mutex;
	struct device ctrl_device;
	struct device *device;	/* char device */
	struct cdev cdev;
	struct work_struct reset_work;
	struct work_struct delete_work;

	struct nvme_subsystem *subsys;
	struct list_head subsys_entry;

	struct opal_dev *opal_dev;

	char name[12];
	u16 cntlid;

	u32 ctrl_config;
	u16 mtfa;
	u32 queue_count;

	u64 cap;
	u32 page_size;
	u32 max_hw_sectors;
	u16 oncs;
	u16 oacs;
	u16 nssa;
	u16 nr_streams;
	atomic_t abort_limit;
	u8 vwc;
	u32 vs;
	u32 sgls;
	u16 kas;
	u8 npss;
	u8 apsta;
	u32 aen_result;
	unsigned int shutdown_timeout;
	unsigned int kato;
	bool subsystem;
	unsigned long quirks;
	struct nvme_id_power_state psd[32];
	struct nvme_effects_log *effects;
	struct work_struct scan_work;
	struct work_struct async_event_work;
	struct delayed_work ka_work;
	struct work_struct fw_act_work;

	/* Power saving configuration */
	u64 ps_max_latency_us;
	bool apst_enabled;

	/* PCIe only: */
	u32 hmpre;
	u32 hmmin;
	u32 hmminds;
	u16 hmmaxd;

	/* Fabrics only */
	u16 sqsize;
	u32 ioccsz;
	u32 iorcsz;
	u16 icdoff;
	u16 maxcmd;
	int nr_reconnects;
	struct nvmf_ctrl_options *opts;
};

struct nvme_subsystem {
	int			instance;
	struct device		dev;
	/*
	 * Because we unregister the device on the last put we need
	 * a separate refcount.
	 */
	struct kref		ref;
	struct list_head	entry;
	struct mutex		lock;
	struct list_head	ctrls;
	struct list_head	nsheads;
	char			subnqn[NVMF_NQN_SIZE];
	char			serial[20];
	char			model[40];
	char			firmware_rev[8];
	u8			cmic;
	u16			vendor_id;
	struct ida		ns_ida;
};

/*
 * Container structure for uniqueue namespace identifiers.
 */
struct nvme_ns_ids {
	u8	eui64[8];
	u8	nguid[16];
	uuid_t	uuid;
};

/*
 * Anchor structure for namespaces.  There is one for each namespace in a
 * NVMe subsystem that any of our controllers can see, and the namespace
 * structure for each controller is chained of it.  For private namespaces
 * there is a 1:1 relation to our namespace structures, that is ->list
 * only ever has a single entry for private namespaces.
 */
struct nvme_ns_head {
#ifdef CONFIG_NVME_MULTIPATH
	struct gendisk		*disk;
	struct nvme_ns __rcu	*current_path;
	struct bio_list		requeue_list;
	spinlock_t		requeue_lock;
	struct work_struct	requeue_work;
#endif
	struct list_head	list;
	struct srcu_struct      srcu;
	struct nvme_subsystem	*subsys;
	unsigned		ns_id;
	struct nvme_ns_ids	ids;
	struct list_head	entry;
	struct kref		ref;
	int			instance;
};

struct nvme_ns {
	struct list_head list;

	struct nvme_ctrl *ctrl;
	struct request_queue *queue;
	struct gendisk *disk;
	struct list_head siblings;
	struct nvm_dev *ndev;
	struct kref kref;
	struct nvme_ns_head *head;

	int lba_shift;
	u16 ms;
	u16 sgs;
	u32 sws;
	bool ext;
	u8 pi_type;
	unsigned long flags;
#define NVME_NS_REMOVING 0
#define NVME_NS_DEAD     1
	u16 noiob;
};

struct nvme_ctrl_ops {
	const char *name;
	struct module *module;
	unsigned int flags;
#define NVME_F_FABRICS			(1 << 0)
#define NVME_F_METADATA_SUPPORTED	(1 << 1)
	int (*reg_read32)(struct nvme_ctrl *ctrl, u32 off, u32 *val);
	int (*reg_write32)(struct nvme_ctrl *ctrl, u32 off, u32 val);
	int (*reg_read64)(struct nvme_ctrl *ctrl, u32 off, u64 *val);
	void (*free_ctrl)(struct nvme_ctrl *ctrl);
	void (*submit_async_event)(struct nvme_ctrl *ctrl);
	void (*delete_ctrl)(struct nvme_ctrl *ctrl);
	int (*get_address)(struct nvme_ctrl *ctrl, char *buf, int size);
	int (*reinit_request)(void *data, struct request *rq);
};

static inline bool nvme_ctrl_ready(struct nvme_ctrl *ctrl)
{
	u32 val = 0;

	if (ctrl->ops->reg_read32(ctrl, NVME_REG_CSTS, &val))
		return false;
	return val & NVME_CSTS_RDY;
}

static inline int nvme_reset_subsystem(struct nvme_ctrl *ctrl)
{
	if (!ctrl->subsystem)
		return -ENOTTY;
	return ctrl->ops->reg_write32(ctrl, NVME_REG_NSSR, 0x4E564D65);
}

static inline u64 nvme_block_nr(struct nvme_ns *ns, sector_t sector)
{
	return (sector >> (ns->lba_shift - 9));
}

static inline void nvme_cleanup_cmd(struct request *req)
{
	if (req->rq_flags & RQF_SPECIAL_PAYLOAD) {
		kfree(page_address(req->special_vec.bv_page) +
		      req->special_vec.bv_offset);
	}
}

static inline void nvme_end_request(struct request *req, __le16 status,
		union nvme_result result)
{
	struct nvme_request *rq = nvme_req(req);

	rq->status = le16_to_cpu(status) >> 1;
	rq->result = result;
	blk_mq_complete_request(req);
}

static inline void nvme_get_ctrl(struct nvme_ctrl *ctrl)
{
	get_device(ctrl->device);
}

static inline void nvme_put_ctrl(struct nvme_ctrl *ctrl)
{
	put_device(ctrl->device);
}

void nvme_complete_rq(struct request *req);
void nvme_cancel_request(struct request *req, void *data, bool reserved);
bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
		enum nvme_ctrl_state new_state);
int nvme_disable_ctrl(struct nvme_ctrl *ctrl, u64 cap);
int nvme_enable_ctrl(struct nvme_ctrl *ctrl, u64 cap);
int nvme_shutdown_ctrl(struct nvme_ctrl *ctrl);
int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev,
		const struct nvme_ctrl_ops *ops, unsigned long quirks);
void nvme_uninit_ctrl(struct nvme_ctrl *ctrl);
void nvme_start_ctrl(struct nvme_ctrl *ctrl);
void nvme_stop_ctrl(struct nvme_ctrl *ctrl);
void nvme_put_ctrl(struct nvme_ctrl *ctrl);
int nvme_init_identify(struct nvme_ctrl *ctrl);

void nvme_queue_scan(struct nvme_ctrl *ctrl);
void nvme_remove_namespaces(struct nvme_ctrl *ctrl);

int nvme_sec_submit(void *data, u16 spsp, u8 secp, void *buffer, size_t len,
		bool send);

void nvme_complete_async_event(struct nvme_ctrl *ctrl, __le16 status,
		union nvme_result *res);

void nvme_stop_queues(struct nvme_ctrl *ctrl);
void nvme_start_queues(struct nvme_ctrl *ctrl);
void nvme_kill_queues(struct nvme_ctrl *ctrl);
void nvme_unfreeze(struct nvme_ctrl *ctrl);
void nvme_wait_freeze(struct nvme_ctrl *ctrl);
void nvme_wait_freeze_timeout(struct nvme_ctrl *ctrl, long timeout);
void nvme_start_freeze(struct nvme_ctrl *ctrl);
int nvme_reinit_tagset(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set);

#define NVME_QID_ANY -1
struct request *nvme_alloc_request(struct request_queue *q,
		struct nvme_command *cmd, blk_mq_req_flags_t flags, int qid);
blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
		struct nvme_command *cmd);
int nvme_submit_sync_cmd(struct request_queue *q, struct nvme_command *cmd,
		void *buf, unsigned bufflen);
int __nvme_submit_sync_cmd(struct request_queue *q, struct nvme_command *cmd,
		union nvme_result *result, void *buffer, unsigned bufflen,
		unsigned timeout, int qid, int at_head,
		blk_mq_req_flags_t flags);
int nvme_set_queue_count(struct nvme_ctrl *ctrl, int *count);
void nvme_start_keep_alive(struct nvme_ctrl *ctrl);
void nvme_stop_keep_alive(struct nvme_ctrl *ctrl);
int nvme_reset_ctrl(struct nvme_ctrl *ctrl);
int nvme_delete_ctrl(struct nvme_ctrl *ctrl);
int nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl);

extern const struct attribute_group nvme_ns_id_attr_group;
extern const struct block_device_operations nvme_ns_head_ops;

#ifdef CONFIG_NVME_MULTIPATH
void nvme_failover_req(struct request *req);
bool nvme_req_needs_failover(struct request *req);
void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl);
int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,struct nvme_ns_head *head);
void nvme_mpath_add_disk(struct nvme_ns_head *head);
void nvme_mpath_remove_disk(struct nvme_ns_head *head);

static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
{
	struct nvme_ns_head *head = ns->head;

	if (head && ns == srcu_dereference(head->current_path, &head->srcu))
		rcu_assign_pointer(head->current_path, NULL);
}
struct nvme_ns *nvme_find_path(struct nvme_ns_head *head);

static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
{
	struct nvme_ns_head *head = ns->head;

	if (head->disk && list_empty(&head->list))
		kblockd_schedule_work(&head->requeue_work);
}

#else
static inline void nvme_failover_req(struct request *req)
{
}
static inline bool nvme_req_needs_failover(struct request *req)
{
	return false;
}
static inline void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
{
}
static inline int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,
		struct nvme_ns_head *head)
{
	return 0;
}
static inline void nvme_mpath_add_disk(struct nvme_ns_head *head)
{
}
static inline void nvme_mpath_remove_disk(struct nvme_ns_head *head)
{
}
static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
{
}
static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
{
}
#endif /* CONFIG_NVME_MULTIPATH */

#ifdef CONFIG_NVM
int nvme_nvm_register(struct nvme_ns *ns, char *disk_name, int node);
void nvme_nvm_unregister(struct nvme_ns *ns);
int nvme_nvm_register_sysfs(struct nvme_ns *ns);
void nvme_nvm_unregister_sysfs(struct nvme_ns *ns);
int nvme_nvm_ioctl(struct nvme_ns *ns, unsigned int cmd, unsigned long arg);
#else
static inline int nvme_nvm_register(struct nvme_ns *ns, char *disk_name,
				    int node)
{
	return 0;
}

static inline void nvme_nvm_unregister(struct nvme_ns *ns) {};
static inline int nvme_nvm_register_sysfs(struct nvme_ns *ns)
{
	return 0;
}
static inline void nvme_nvm_unregister_sysfs(struct nvme_ns *ns) {};
static inline int nvme_nvm_ioctl(struct nvme_ns *ns, unsigned int cmd,
							unsigned long arg)
{
	return -ENOTTY;
}
#endif /* CONFIG_NVM */

static inline struct nvme_ns *nvme_get_ns_from_dev(struct device *dev)
{
	return dev_to_disk(dev)->private_data;
}

int __init nvme_core_init(void);
void nvme_core_exit(void);

#endif /* _NVME_H */
Commit	Line	Data
f11bb3e2 CH	1	/*
	2	* Copyright (c) 2011-2014, Intel Corporation.
	3	*
	4	* This program is free software; you can redistribute it and/or modify it
	5	* under the terms and conditions of the GNU General Public License,
	6	* version 2, as published by the Free Software Foundation.
	7	*
	8	* This program is distributed in the hope it will be useful, but WITHOUT
	9	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
	10	* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
	11	* more details.
	12	*/
	13
	14	#ifndef _NVME_H
	15	#define _NVME_H
	16
	17	#include <linux/nvme.h>
a6a5149b	18	#include <linux/cdev.h>
f11bb3e2 CH	19	#include <linux/pci.h>
	20	#include <linux/kref.h>
	21	#include <linux/blk-mq.h>
b0b4e09c	22	#include <linux/lightnvm.h>
a98e58e5	23	#include <linux/sed-opal.h>
f11bb3e2	24
8ae4e447	25	extern unsigned int nvme_io_timeout;
f11bb3e2 CH	26	#define NVME_IO_TIMEOUT (nvme_io_timeout * HZ)
f11bb3e2 CH	27
8ae4e447	28	extern unsigned int admin_timeout;
21d34711 CH	29	#define ADMIN_TIMEOUT (admin_timeout * HZ)
21d34711 CH	30
038bd4cb SG	31	#define NVME_DEFAULT_KATO 5
	32	#define NVME_KATO_GRACE 10
	33
9a6327d2 SG	34	extern struct workqueue_struct *nvme_wq;
9a6327d2 SG	35
ca064085 MB	36	enum {
	37	NVME_NS_LBA = 0,
	38	NVME_NS_LIGHTNVM = 1,
	39	};
	40
f11bb3e2	41	/*
106198ed CH	42	* List of workarounds for devices that required behavior not specified in
106198ed CH	43	* the standard.
f11bb3e2	44	*/
106198ed CH	45	enum nvme_quirks {
	46	/*
	47	* Prefers I/O aligned to a stripe size specified in a vendor
	48	* specific Identify field.
	49	*/
	50	NVME_QUIRK_STRIPE_SIZE = (1 << 0),
540c801c KB	51
	52	/*
	53	* The controller doesn't handle Identify value others than 0 or 1
	54	* correctly.
	55	*/
	56	NVME_QUIRK_IDENTIFY_CNS = (1 << 1),
08095e70 KB	57
08095e70 KB	58	/*
e850fd16 CH	59	* The controller deterministically returns O's on reads to
e850fd16 CH	60	* logical blocks that deallocate was called on.
08095e70	61	*/
e850fd16	62	NVME_QUIRK_DEALLOCATE_ZEROES = (1 << 2),
54adc010 GP	63
	64	/*
	65	* The controller needs a delay before starts checking the device
	66	* readiness, which is done by reading the NVME_CSTS_RDY bit.
	67	*/
	68	NVME_QUIRK_DELAY_BEFORE_CHK_RDY = (1 << 3),
c5552fde AL	69
	70	/*
	71	* APST should not be used.
	72	*/
	73	NVME_QUIRK_NO_APST = (1 << 4),
ff5350a8 AL	74
	75	/*
	76	* The deepest sleep state should not be used.
	77	*/
	78	NVME_QUIRK_NO_DEEPEST_PS = (1 << 5),
608cc4b1 CH	79
	80	/*
	81	* Supports the LighNVM command set if indicated in vs[1].
	82	*/
	83	NVME_QUIRK_LIGHTNVM = (1 << 6),
5750cb1c JA	84
	85	/*
	86	* Set MEDIUM priority on SQ creation
	87	*/
	88	NVME_QUIRK_MEDIUM_PRIO_SQ = (1 << 7),
106198ed CH	89	};
106198ed CH	90
d49187e9 CH	91	/*
	92	* Common request structure for NVMe passthrough. All drivers must have
	93	* this structure as the first member of their request-private data.
	94	*/
	95	struct nvme_request {
	96	struct nvme_command *cmd;
	97	union nvme_result result;
44e44b29	98	u8 retries;
27fa9bc5 CH	99	u8 flags;
	100	u16 status;
	101	};
	102
32acab31 CH	103	/*
	104	* Mark a bio as coming in through the mpath node.
	105	*/
	106	#define REQ_NVME_MPATH REQ_DRV
	107
27fa9bc5 CH	108	enum {
27fa9bc5 CH	109	NVME_REQ_CANCELLED = (1 << 0),
d49187e9 CH	110	};
	111
	112	static inline struct nvme_request nvme_req(struct request req)
	113	{
	114	return blk_mq_rq_to_pdu(req);
	115	}
	116
54adc010 GP	117	/* The below value is the specific amount of delay needed before checking
	118	* readiness in case of the PCI_DEVICE(0x1c58, 0x0003), which needs the
	119	* NVME_QUIRK_DELAY_BEFORE_CHK_RDY quirk enabled. The value (in ms) was
	120	* found empirically.
	121	*/
8c97eecc	122	#define NVME_QUIRK_DELAY_AMOUNT 2300
54adc010	123
bb8d261e CH	124	enum nvme_ctrl_state {
	125	NVME_CTRL_NEW,
	126	NVME_CTRL_LIVE,
	127	NVME_CTRL_RESETTING,
def61eca	128	NVME_CTRL_RECONNECTING,
bb8d261e	129	NVME_CTRL_DELETING,
0ff9d4e1	130	NVME_CTRL_DEAD,
bb8d261e CH	131	};
bb8d261e CH	132
1c63dc66	133	struct nvme_ctrl {
bb8d261e	134	enum nvme_ctrl_state state;
bd4da3ab	135	bool identified;
bb8d261e	136	spinlock_t lock;
1c63dc66	137	const struct nvme_ctrl_ops *ops;
f11bb3e2	138	struct request_queue *admin_q;
07bfcd09	139	struct request_queue *connect_q;
f11bb3e2	140	struct device *dev;
f11bb3e2	141	int instance;
5bae7f73	142	struct blk_mq_tag_set *tagset;
34b6c231	143	struct blk_mq_tag_set *admin_tagset;
f11bb3e2	144	struct list_head namespaces;
69d3b8ac	145	struct mutex namespaces_mutex;
d22524a4	146	struct device ctrl_device;
5bae7f73	147	struct device device; / char device */
a6a5149b	148	struct cdev cdev;
d86c4d8e	149	struct work_struct reset_work;
c5017e85	150	struct work_struct delete_work;
1c63dc66	151
ab9e00cc CH	152	struct nvme_subsystem *subsys;
	153	struct list_head subsys_entry;
	154
4f1244c8	155	struct opal_dev *opal_dev;
a98e58e5	156
f11bb3e2	157	char name[12];
76e3914a	158	u16 cntlid;
5fd4ce1b CH	159
5fd4ce1b CH	160	u32 ctrl_config;
b6dccf7f	161	u16 mtfa;
d858e5f0	162	u32 queue_count;
5fd4ce1b	163
20d0dfe6	164	u64 cap;
5fd4ce1b	165	u32 page_size;
f11bb3e2	166	u32 max_hw_sectors;
f11bb3e2	167	u16 oncs;
8a9ae523	168	u16 oacs;
f5d11840 JA	169	u16 nssa;
f5d11840 JA	170	u16 nr_streams;
6bf25d16	171	atomic_t abort_limit;
f11bb3e2	172	u8 vwc;
f3ca80fc	173	u32 vs;
07bfcd09	174	u32 sgls;
038bd4cb	175	u16 kas;
c5552fde AL	176	u8 npss;
c5552fde AL	177	u8 apsta;
e3d7874d	178	u32 aen_result;
07fbd32a	179	unsigned int shutdown_timeout;
038bd4cb	180	unsigned int kato;
f3ca80fc	181	bool subsystem;
106198ed	182	unsigned long quirks;
c5552fde	183	struct nvme_id_power_state psd[32];
84fef62d	184	struct nvme_effects_log *effects;
5955be21	185	struct work_struct scan_work;
f866fc42	186	struct work_struct async_event_work;
038bd4cb	187	struct delayed_work ka_work;
b6dccf7f	188	struct work_struct fw_act_work;
07bfcd09	189
c5552fde AL	190	/* Power saving configuration */
c5552fde AL	191	u64 ps_max_latency_us;
76a5af84	192	bool apst_enabled;
c5552fde	193
044a9df1	194	/* PCIe only: */
fe6d53c9 CH	195	u32 hmpre;
fe6d53c9 CH	196	u32 hmmin;
044a9df1 CH	197	u32 hmminds;
044a9df1 CH	198	u16 hmmaxd;
fe6d53c9	199
07bfcd09 CH	200	/* Fabrics only */
	201	u16 sqsize;
	202	u32 ioccsz;
	203	u32 iorcsz;
	204	u16 icdoff;
	205	u16 maxcmd;
fdf9dfa8	206	int nr_reconnects;
07bfcd09	207	struct nvmf_ctrl_options *opts;
f11bb3e2 CH	208	};
f11bb3e2 CH	209
ab9e00cc CH	210	struct nvme_subsystem {
	211	int instance;
	212	struct device dev;
	213	/*
	214	* Because we unregister the device on the last put we need
	215	* a separate refcount.
	216	*/
	217	struct kref ref;
	218	struct list_head entry;
	219	struct mutex lock;
	220	struct list_head ctrls;
ed754e5d	221	struct list_head nsheads;
ab9e00cc CH	222	char subnqn[NVMF_NQN_SIZE];
	223	char serial[20];
	224	char model[40];
	225	char firmware_rev[8];
	226	u8 cmic;
	227	u16 vendor_id;
ed754e5d	228	struct ida ns_ida;
ab9e00cc CH	229	};
ab9e00cc CH	230
002fab04 CH	231	/*
	232	* Container structure for uniqueue namespace identifiers.
	233	*/
	234	struct nvme_ns_ids {
	235	u8 eui64[8];
	236	u8 nguid[16];
	237	uuid_t uuid;
	238	};
	239
ed754e5d CH	240	/*
	241	* Anchor structure for namespaces. There is one for each namespace in a
	242	* NVMe subsystem that any of our controllers can see, and the namespace
	243	* structure for each controller is chained of it. For private namespaces
	244	* there is a 1:1 relation to our namespace structures, that is ->list
	245	* only ever has a single entry for private namespaces.
	246	*/
	247	struct nvme_ns_head {
32acab31 CH	248	#ifdef CONFIG_NVME_MULTIPATH
	249	struct gendisk *disk;
	250	struct nvme_ns __rcu *current_path;
	251	struct bio_list requeue_list;
	252	spinlock_t requeue_lock;
	253	struct work_struct requeue_work;
	254	#endif
ed754e5d CH	255	struct list_head list;
	256	struct srcu_struct srcu;
	257	struct nvme_subsystem *subsys;
	258	unsigned ns_id;
	259	struct nvme_ns_ids ids;
	260	struct list_head entry;
	261	struct kref ref;
	262	int instance;
	263	};
	264
f11bb3e2 CH	265	struct nvme_ns {
	266	struct list_head list;
	267
1c63dc66	268	struct nvme_ctrl *ctrl;
f11bb3e2 CH	269	struct request_queue *queue;
f11bb3e2 CH	270	struct gendisk *disk;
ed754e5d	271	struct list_head siblings;
b0b4e09c	272	struct nvm_dev *ndev;
f11bb3e2	273	struct kref kref;
ed754e5d	274	struct nvme_ns_head *head;
f11bb3e2	275
f11bb3e2 CH	276	int lba_shift;
f11bb3e2 CH	277	u16 ms;
f5d11840 JA	278	u16 sgs;
f5d11840 JA	279	u32 sws;
f11bb3e2 CH	280	bool ext;
f11bb3e2 CH	281	u8 pi_type;
646017a6	282	unsigned long flags;
646017a6	283	#define NVME_NS_REMOVING 0
69d9a99c	284	#define NVME_NS_DEAD 1
57eeaf8e	285	u16 noiob;
f11bb3e2 CH	286	};
f11bb3e2 CH	287
1c63dc66	288	struct nvme_ctrl_ops {
1a353d85	289	const char *name;
e439bb12	290	struct module *module;
d3d5b87d CH	291	unsigned int flags;
d3d5b87d CH	292	#define NVME_F_FABRICS (1 << 0)
c81bfba9	293	#define NVME_F_METADATA_SUPPORTED (1 << 1)
1c63dc66	294	int (reg_read32)(struct nvme_ctrl ctrl, u32 off, u32 *val);
5fd4ce1b	295	int (reg_write32)(struct nvme_ctrl ctrl, u32 off, u32 val);
7fd8930f	296	int (reg_read64)(struct nvme_ctrl ctrl, u32 off, u64 *val);
1673f1f0	297	void (free_ctrl)(struct nvme_ctrl ctrl);
ad22c355	298	void (submit_async_event)(struct nvme_ctrl ctrl);
c5017e85	299	void (delete_ctrl)(struct nvme_ctrl ctrl);
1a353d85	300	int (get_address)(struct nvme_ctrl ctrl, char *buf, int size);
31b84460	301	int (reinit_request)(void data, struct request *rq);
f11bb3e2 CH	302	};
f11bb3e2 CH	303
1c63dc66 CH	304	static inline bool nvme_ctrl_ready(struct nvme_ctrl *ctrl)
	305	{
	306	u32 val = 0;
	307
	308	if (ctrl->ops->reg_read32(ctrl, NVME_REG_CSTS, &val))
	309	return false;
	310	return val & NVME_CSTS_RDY;
	311	}
	312
f3ca80fc CH	313	static inline int nvme_reset_subsystem(struct nvme_ctrl *ctrl)
	314	{
	315	if (!ctrl->subsystem)
	316	return -ENOTTY;
	317	return ctrl->ops->reg_write32(ctrl, NVME_REG_NSSR, 0x4E564D65);
	318	}
	319
f11bb3e2 CH	320	static inline u64 nvme_block_nr(struct nvme_ns *ns, sector_t sector)
	321	{
	322	return (sector >> (ns->lba_shift - 9));
	323	}
	324
6904242d ML	325	static inline void nvme_cleanup_cmd(struct request *req)
6904242d ML	326	{
f9d03f96 CH	327	if (req->rq_flags & RQF_SPECIAL_PAYLOAD) {
	328	kfree(page_address(req->special_vec.bv_page) +
	329	req->special_vec.bv_offset);
	330	}
6904242d ML	331	}
6904242d ML	332
27fa9bc5 CH	333	static inline void nvme_end_request(struct request *req, __le16 status,
27fa9bc5 CH	334	union nvme_result result)
15a190f7	335	{
27fa9bc5	336	struct nvme_request *rq = nvme_req(req);
15a190f7	337
27fa9bc5 CH	338	rq->status = le16_to_cpu(status) >> 1;
27fa9bc5 CH	339	rq->result = result;
08e0029a	340	blk_mq_complete_request(req);
7688faa6 CH	341	}
7688faa6 CH	342
d22524a4 CH	343	static inline void nvme_get_ctrl(struct nvme_ctrl *ctrl)
	344	{
	345	get_device(ctrl->device);
	346	}
	347
	348	static inline void nvme_put_ctrl(struct nvme_ctrl *ctrl)
	349	{
	350	put_device(ctrl->device);
	351	}
	352
77f02a7a	353	void nvme_complete_rq(struct request *req);
c55a2fd4	354	void nvme_cancel_request(struct request req, void data, bool reserved);
bb8d261e CH	355	bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
bb8d261e CH	356	enum nvme_ctrl_state new_state);
5fd4ce1b CH	357	int nvme_disable_ctrl(struct nvme_ctrl *ctrl, u64 cap);
	358	int nvme_enable_ctrl(struct nvme_ctrl *ctrl, u64 cap);
	359	int nvme_shutdown_ctrl(struct nvme_ctrl *ctrl);
f3ca80fc CH	360	int nvme_init_ctrl(struct nvme_ctrl ctrl, struct device dev,
f3ca80fc CH	361	const struct nvme_ctrl_ops *ops, unsigned long quirks);
53029b04	362	void nvme_uninit_ctrl(struct nvme_ctrl *ctrl);
d09f2b45 SG	363	void nvme_start_ctrl(struct nvme_ctrl *ctrl);
d09f2b45 SG	364	void nvme_stop_ctrl(struct nvme_ctrl *ctrl);
1673f1f0	365	void nvme_put_ctrl(struct nvme_ctrl *ctrl);
7fd8930f	366	int nvme_init_identify(struct nvme_ctrl *ctrl);
5bae7f73	367
5955be21	368	void nvme_queue_scan(struct nvme_ctrl *ctrl);
5bae7f73	369	void nvme_remove_namespaces(struct nvme_ctrl *ctrl);
1673f1f0	370
4f1244c8 CH	371	int nvme_sec_submit(void data, u16 spsp, u8 secp, void buffer, size_t len,
4f1244c8 CH	372	bool send);
a98e58e5	373
7bf58533 CH	374	void nvme_complete_async_event(struct nvme_ctrl *ctrl, __le16 status,
7bf58533 CH	375	union nvme_result *res);
f866fc42	376
25646264 KB	377	void nvme_stop_queues(struct nvme_ctrl *ctrl);
25646264 KB	378	void nvme_start_queues(struct nvme_ctrl *ctrl);
69d9a99c	379	void nvme_kill_queues(struct nvme_ctrl *ctrl);
302ad8cc KB	380	void nvme_unfreeze(struct nvme_ctrl *ctrl);
	381	void nvme_wait_freeze(struct nvme_ctrl *ctrl);
	382	void nvme_wait_freeze_timeout(struct nvme_ctrl *ctrl, long timeout);
	383	void nvme_start_freeze(struct nvme_ctrl *ctrl);
31b84460	384	int nvme_reinit_tagset(struct nvme_ctrl ctrl, struct blk_mq_tag_set set);
363c9aac	385
eb71f435	386	#define NVME_QID_ANY -1
4160982e	387	struct request nvme_alloc_request(struct request_queue q,
9a95e4ef	388	struct nvme_command *cmd, blk_mq_req_flags_t flags, int qid);
fc17b653	389	blk_status_t nvme_setup_cmd(struct nvme_ns ns, struct request req,
8093f7ca	390	struct nvme_command *cmd);
f11bb3e2 CH	391	int nvme_submit_sync_cmd(struct request_queue q, struct nvme_command cmd,
	392	void *buf, unsigned bufflen);
	393	int __nvme_submit_sync_cmd(struct request_queue q, struct nvme_command cmd,
d49187e9	394	union nvme_result result, void buffer, unsigned bufflen,
9a95e4ef BVA	395	unsigned timeout, int qid, int at_head,
9a95e4ef BVA	396	blk_mq_req_flags_t flags);
9a0be7ab	397	int nvme_set_queue_count(struct nvme_ctrl ctrl, int count);
038bd4cb SG	398	void nvme_start_keep_alive(struct nvme_ctrl *ctrl);
038bd4cb SG	399	void nvme_stop_keep_alive(struct nvme_ctrl *ctrl);
d86c4d8e	400	int nvme_reset_ctrl(struct nvme_ctrl *ctrl);
c5017e85 CH	401	int nvme_delete_ctrl(struct nvme_ctrl *ctrl);
c5017e85 CH	402	int nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl);
f11bb3e2	403
5b85b826	404	extern const struct attribute_group nvme_ns_id_attr_group;
32acab31 CH	405	extern const struct block_device_operations nvme_ns_head_ops;
	406
	407	#ifdef CONFIG_NVME_MULTIPATH
	408	void nvme_failover_req(struct request *req);
	409	bool nvme_req_needs_failover(struct request *req);
	410	void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl);
	411	int nvme_mpath_alloc_disk(struct nvme_ctrl ctrl,struct nvme_ns_head head);
	412	void nvme_mpath_add_disk(struct nvme_ns_head *head);
	413	void nvme_mpath_remove_disk(struct nvme_ns_head *head);
	414
	415	static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
	416	{
	417	struct nvme_ns_head *head = ns->head;
	418
	419	if (head && ns == srcu_dereference(head->current_path, &head->srcu))
	420	rcu_assign_pointer(head->current_path, NULL);
	421	}
	422	struct nvme_ns nvme_find_path(struct nvme_ns_head head);
479a322f SG	423
	424	static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
	425	{
	426	struct nvme_ns_head *head = ns->head;
	427
	428	if (head->disk && list_empty(&head->list))
	429	kblockd_schedule_work(&head->requeue_work);
	430	}
	431
32acab31 CH	432	#else
	433	static inline void nvme_failover_req(struct request *req)
	434	{
	435	}
	436	static inline bool nvme_req_needs_failover(struct request *req)
	437	{
	438	return false;
	439	}
	440	static inline void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
	441	{
	442	}
	443	static inline int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,
	444	struct nvme_ns_head *head)
	445	{
	446	return 0;
	447	}
	448	static inline void nvme_mpath_add_disk(struct nvme_ns_head *head)
	449	{
	450	}
	451	static inline void nvme_mpath_remove_disk(struct nvme_ns_head *head)
	452	{
	453	}
	454	static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
479a322f SG	455	{
	456	}
	457	static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
32acab31 CH	458	{
	459	}
	460	#endif /* CONFIG_NVME_MULTIPATH */
	461
c4699e70	462	#ifdef CONFIG_NVM
3dc87dd0	463	int nvme_nvm_register(struct nvme_ns ns, char disk_name, int node);
b0b4e09c	464	void nvme_nvm_unregister(struct nvme_ns *ns);
3dc87dd0 MB	465	int nvme_nvm_register_sysfs(struct nvme_ns *ns);
3dc87dd0 MB	466	void nvme_nvm_unregister_sysfs(struct nvme_ns *ns);
84d4add7	467	int nvme_nvm_ioctl(struct nvme_ns *ns, unsigned int cmd, unsigned long arg);
c4699e70	468	#else
b0b4e09c	469	static inline int nvme_nvm_register(struct nvme_ns ns, char disk_name,
3dc87dd0	470	int node)
c4699e70 KB	471	{
	472	return 0;
	473	}
	474
b0b4e09c	475	static inline void nvme_nvm_unregister(struct nvme_ns *ns) {};
3dc87dd0 MB	476	static inline int nvme_nvm_register_sysfs(struct nvme_ns *ns)
	477	{
	478	return 0;
	479	}
	480	static inline void nvme_nvm_unregister_sysfs(struct nvme_ns *ns) {};
84d4add7 MB	481	static inline int nvme_nvm_ioctl(struct nvme_ns *ns, unsigned int cmd,
	482	unsigned long arg)
	483	{
	484	return -ENOTTY;
	485	}
3dc87dd0 MB	486	#endif /* CONFIG_NVM */
3dc87dd0 MB	487
40267efd SL	488	static inline struct nvme_ns nvme_get_ns_from_dev(struct device dev)
	489	{
	490	return dev_to_disk(dev)->private_data;
	491	}
ca064085	492
5bae7f73 CH	493	int __init nvme_core_init(void);
	494	void nvme_core_exit(void);
	495
f11bb3e2	496	#endif /* _NVME_H */