diff options
Diffstat (limited to 'drivers/nvme/host/core.c')
| -rw-r--r-- | drivers/nvme/host/core.c | 1170 |
1 files changed, 722 insertions, 448 deletions
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 0a96362912ce..8359d0aa0e44 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -4,6 +4,7 @@ * Copyright (c) 2011-2014, Intel Corporation. */ +#include <linux/async.h> #include <linux/blkdev.h> #include <linux/blk-mq.h> #include <linux/blk-integrity.h> @@ -21,7 +22,7 @@ #include <linux/nvme_ioctl.h> #include <linux/pm_qos.h> #include <linux/ratelimit.h> -#include <asm/unaligned.h> +#include <linux/unaligned.h> #include "nvme.h" #include "fabrics.h" @@ -36,10 +37,13 @@ struct nvme_ns_info { struct nvme_ns_ids ids; u32 nsid; __le32 anagrpid; + u8 pi_offset; bool is_shared; bool is_readonly; bool is_ready; bool is_removed; + bool is_rotational; + bool no_vwc; }; unsigned int admin_timeout = 60; @@ -90,6 +94,17 @@ MODULE_PARM_DESC(apst_secondary_latency_tol_us, "secondary APST latency tolerance in us"); /* + * Older kernels didn't enable protection information if it was at an offset. + * Newer kernels do, so it breaks reads on the upgrade if such formats were + * used in prior kernels since the metadata written did not contain a valid + * checksum. + */ +static bool disable_pi_offsets = false; +module_param(disable_pi_offsets, bool, 0444); +MODULE_PARM_DESC(disable_pi_offsets, + "disable protection information if it has an offset"); + +/* * nvme_wq - hosts nvme related works that are not reset or delete * nvme_reset_wq - hosts nvme reset works * nvme_delete_wq - hosts nvme delete works @@ -110,16 +125,25 @@ struct workqueue_struct *nvme_delete_wq; EXPORT_SYMBOL_GPL(nvme_delete_wq); static LIST_HEAD(nvme_subsystems); -static DEFINE_MUTEX(nvme_subsystems_lock); +DEFINE_MUTEX(nvme_subsystems_lock); static DEFINE_IDA(nvme_instance_ida); static dev_t nvme_ctrl_base_chr_devt; -static struct class *nvme_class; -static struct class *nvme_subsys_class; +static int nvme_class_uevent(const struct device *dev, struct kobj_uevent_env *env); +static const struct class nvme_class = { + .name = "nvme", + .dev_uevent = nvme_class_uevent, +}; + +static const struct class nvme_subsys_class = { + .name = "nvme-subsystem", +}; static DEFINE_IDA(nvme_ns_chr_minor_ida); static dev_t nvme_ns_chr_devt; -static struct class *nvme_ns_chr_class; +static const struct class nvme_ns_chr_class = { + .name = "nvme-generic", +}; static void nvme_put_subsystem(struct nvme_subsystem *subsys); static void nvme_remove_invalid_namespaces(struct nvme_ctrl *ctrl, @@ -252,7 +276,7 @@ void nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl) static blk_status_t nvme_error_status(u16 status) { - switch (status & 0x7ff) { + switch (status & NVME_SCT_SC_MASK) { case NVME_SC_SUCCESS: return BLK_STS_OK; case NVME_SC_CAP_EXCEEDED: @@ -298,7 +322,7 @@ static void nvme_retry_req(struct request *req) u16 crd; /* The mask and shift result must be <= 3 */ - crd = (nvme_req(req)->status & NVME_SC_CRD) >> 11; + crd = (nvme_req(req)->status & NVME_STATUS_CRD) >> 11; if (crd) delay = nvme_req(req)->ctrl->crdt[crd - 1] * 100; @@ -320,10 +344,10 @@ static void nvme_log_error(struct request *req) nvme_sect_to_lba(ns->head, blk_rq_pos(req)), blk_rq_bytes(req) >> ns->head->lba_shift, nvme_get_error_status_str(nr->status), - nr->status >> 8 & 7, /* Status Code Type */ - nr->status & 0xff, /* Status Code */ - nr->status & NVME_SC_MORE ? "MORE " : "", - nr->status & NVME_SC_DNR ? "DNR " : ""); + NVME_SCT(nr->status), /* Status Code Type */ + nr->status & NVME_SC_MASK, /* Status Code */ + nr->status & NVME_STATUS_MORE ? "MORE " : "", + nr->status & NVME_STATUS_DNR ? "DNR " : ""); return; } @@ -332,10 +356,10 @@ static void nvme_log_error(struct request *req) nvme_get_admin_opcode_str(nr->cmd->common.opcode), nr->cmd->common.opcode, nvme_get_error_status_str(nr->status), - nr->status >> 8 & 7, /* Status Code Type */ - nr->status & 0xff, /* Status Code */ - nr->status & NVME_SC_MORE ? "MORE " : "", - nr->status & NVME_SC_DNR ? "DNR " : ""); + NVME_SCT(nr->status), /* Status Code Type */ + nr->status & NVME_SC_MASK, /* Status Code */ + nr->status & NVME_STATUS_MORE ? "MORE " : "", + nr->status & NVME_STATUS_DNR ? "DNR " : ""); } static void nvme_log_err_passthru(struct request *req) @@ -350,10 +374,10 @@ static void nvme_log_err_passthru(struct request *req) nvme_get_admin_opcode_str(nr->cmd->common.opcode), nr->cmd->common.opcode, nvme_get_error_status_str(nr->status), - nr->status >> 8 & 7, /* Status Code Type */ - nr->status & 0xff, /* Status Code */ - nr->status & NVME_SC_MORE ? "MORE " : "", - nr->status & NVME_SC_DNR ? "DNR " : "", + NVME_SCT(nr->status), /* Status Code Type */ + nr->status & NVME_SC_MASK, /* Status Code */ + nr->status & NVME_STATUS_MORE ? "MORE " : "", + nr->status & NVME_STATUS_DNR ? "DNR " : "", nr->cmd->common.cdw10, nr->cmd->common.cdw11, nr->cmd->common.cdw12, @@ -374,14 +398,14 @@ static inline enum nvme_disposition nvme_decide_disposition(struct request *req) if (likely(nvme_req(req)->status == 0)) return COMPLETE; - if ((nvme_req(req)->status & 0x7ff) == NVME_SC_AUTH_REQUIRED) - return AUTHENTICATE; - if (blk_noretry_request(req) || - (nvme_req(req)->status & NVME_SC_DNR) || + (nvme_req(req)->status & NVME_STATUS_DNR) || nvme_req(req)->retries >= nvme_max_retries) return COMPLETE; + if ((nvme_req(req)->status & NVME_SCT_SC_MASK) == NVME_SC_AUTH_REQUIRED) + return AUTHENTICATE; + if (req->cmd_flags & REQ_NVME_MPATH) { if (nvme_is_path_error(nvme_req(req)->status) || blk_queue_dying(req->q)) @@ -405,10 +429,8 @@ static inline void nvme_end_req_zoned(struct request *req) } } -static inline void nvme_end_req(struct request *req) +static inline void __nvme_end_req(struct request *req) { - blk_status_t status = nvme_error_status(nvme_req(req)->status); - if (unlikely(nvme_req(req)->status && !(req->rq_flags & RQF_QUIET))) { if (blk_rq_is_passthrough(req)) nvme_log_err_passthru(req); @@ -419,6 +441,13 @@ static inline void nvme_end_req(struct request *req) nvme_trace_bio_complete(req); if (req->cmd_flags & REQ_NVME_MPATH) nvme_mpath_end_request(req); +} + +void nvme_end_req(struct request *req) +{ + blk_status_t status = nvme_error_status(nvme_req(req)->status); + + __nvme_end_req(req); blk_mq_end_request(req, status); } @@ -467,7 +496,7 @@ void nvme_complete_batch_req(struct request *req) { trace_nvme_complete_rq(req); nvme_cleanup_cmd(req); - nvme_end_req_zoned(req); + __nvme_end_req(req); } EXPORT_SYMBOL_GPL(nvme_complete_batch_req); @@ -535,8 +564,6 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, switch (new_state) { case NVME_CTRL_LIVE: switch (old_state) { - case NVME_CTRL_NEW: - case NVME_CTRL_RESETTING: case NVME_CTRL_CONNECTING: changed = true; fallthrough; @@ -620,27 +647,6 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, EXPORT_SYMBOL_GPL(nvme_change_ctrl_state); /* - * Returns true for sink states that can't ever transition back to live. - */ -static bool nvme_state_terminal(struct nvme_ctrl *ctrl) -{ - switch (nvme_ctrl_state(ctrl)) { - case NVME_CTRL_NEW: - case NVME_CTRL_LIVE: - case NVME_CTRL_RESETTING: - case NVME_CTRL_CONNECTING: - return false; - case NVME_CTRL_DELETING: - case NVME_CTRL_DELETING_NOIO: - case NVME_CTRL_DEAD: - return true; - default: - WARN_ONCE(1, "Unhandled ctrl state:%d", ctrl->state); - return true; - } -} - -/* * Waits for the controller state to be resetting, or returns false if it is * not possible to ever transition to that state. */ @@ -685,7 +691,7 @@ static void nvme_free_ns(struct kref *kref) kfree(ns); } -static inline bool nvme_get_ns(struct nvme_ns *ns) +bool nvme_get_ns(struct nvme_ns *ns) { return kref_get_unless_zero(&ns->kref); } @@ -694,7 +700,7 @@ void nvme_put_ns(struct nvme_ns *ns) { kref_put(&ns->kref, nvme_free_ns); } -EXPORT_SYMBOL_NS_GPL(nvme_put_ns, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_put_ns, "NVME_TARGET_PASSTHRU"); static inline void nvme_clear_nvme_request(struct request *req) { @@ -877,6 +883,12 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *ns, struct request *req, return BLK_STS_OK; } +static void nvme_set_app_tag(struct request *req, struct nvme_command *cmnd) +{ + cmnd->rw.lbat = cpu_to_le16(bio_integrity(req->bio)->app_tag); + cmnd->rw.lbatm = cpu_to_le16(0xffff); +} + static void nvme_set_ref_tag(struct nvme_ns *ns, struct nvme_command *cmnd, struct request *req) { @@ -934,6 +946,36 @@ static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns *ns, return BLK_STS_OK; } +/* + * NVMe does not support a dedicated command to issue an atomic write. A write + * which does adhere to the device atomic limits will silently be executed + * non-atomically. The request issuer should ensure that the write is within + * the queue atomic writes limits, but just validate this in case it is not. + */ +static bool nvme_valid_atomic_write(struct request *req) +{ + struct request_queue *q = req->q; + u32 boundary_bytes = queue_atomic_write_boundary_bytes(q); + + if (blk_rq_bytes(req) > queue_atomic_write_unit_max_bytes(q)) + return false; + + if (boundary_bytes) { + u64 mask = boundary_bytes - 1, imask = ~mask; + u64 start = blk_rq_pos(req) << SECTOR_SHIFT; + u64 end = start + blk_rq_bytes(req) - 1; + + /* If greater then must be crossing a boundary */ + if (blk_rq_bytes(req) > boundary_bytes) + return false; + + if ((start & imask) != (end & imask)) + return false; + } + + return true; +} + static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, struct request *req, struct nvme_command *cmnd, enum nvme_opcode op) @@ -949,6 +991,9 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, if (req->cmd_flags & REQ_RAHEAD) dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH; + if (req->cmd_flags & REQ_ATOMIC && !nvme_valid_atomic_write(req)) + return BLK_STS_INVAL; + cmnd->rw.opcode = op; cmnd->rw.flags = 0; cmnd->rw.nsid = cpu_to_le32(ns->head->ns_id); @@ -960,8 +1005,8 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, cmnd->rw.length = cpu_to_le16((blk_rq_bytes(req) >> ns->head->lba_shift) - 1); cmnd->rw.reftag = 0; - cmnd->rw.apptag = 0; - cmnd->rw.appmask = 0; + cmnd->rw.lbat = 0; + cmnd->rw.lbatm = 0; if (ns->head->ms) { /* @@ -976,18 +1021,17 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, control |= NVME_RW_PRINFO_PRACT; } - switch (ns->head->pi_type) { - case NVME_NS_DPS_PI_TYPE3: + if (bio_integrity_flagged(req->bio, BIP_CHECK_GUARD)) control |= NVME_RW_PRINFO_PRCHK_GUARD; - break; - case NVME_NS_DPS_PI_TYPE1: - case NVME_NS_DPS_PI_TYPE2: - control |= NVME_RW_PRINFO_PRCHK_GUARD | - NVME_RW_PRINFO_PRCHK_REF; + if (bio_integrity_flagged(req->bio, BIP_CHECK_REFTAG)) { + control |= NVME_RW_PRINFO_PRCHK_REF; if (op == nvme_cmd_zone_append) control |= NVME_RW_APPEND_PIREMAP; nvme_set_ref_tag(ns, cmnd, req); - break; + } + if (bio_integrity_flagged(req->bio, BIP_CHECK_APPTAG)) { + control |= NVME_RW_PRINFO_PRCHK_APP; + nvme_set_app_tag(req, cmnd); } } @@ -1005,6 +1049,7 @@ void nvme_cleanup_cmd(struct request *req) clear_bit_unlock(0, &ctrl->discard_page_busy); else kfree(bvec_virt(&req->special_vec)); + req->rq_flags &= ~RQF_SPECIAL_PAYLOAD; } } EXPORT_SYMBOL_GPL(nvme_cleanup_cmd); @@ -1081,7 +1126,7 @@ int nvme_execute_rq(struct request *rq, bool at_head) return nvme_req(rq)->status; return blk_status_to_errno(status); } -EXPORT_SYMBOL_NS_GPL(nvme_execute_rq, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_execute_rq, "NVME_TARGET_PASSTHRU"); /* * Returns 0 on success. If the result is negative, it's a Linux error code; @@ -1161,7 +1206,7 @@ u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opcode) return effects; } -EXPORT_SYMBOL_NS_GPL(nvme_command_effects, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_command_effects, "NVME_TARGET_PASSTHRU"); u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opcode) { @@ -1181,7 +1226,7 @@ u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opcode) } return effects; } -EXPORT_SYMBOL_NS_GPL(nvme_passthru_start, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_passthru_start, "NVME_TARGET_PASSTHRU"); void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u32 effects, struct nvme_command *cmd, int status) @@ -1226,11 +1271,11 @@ void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u32 effects, break; } } -EXPORT_SYMBOL_NS_GPL(nvme_passthru_end, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_passthru_end, "NVME_TARGET_PASSTHRU"); /* * Recommended frequency for KATO commands per NVMe 1.4 section 7.12.1: - * + * * The host should send Keep Alive commands at half of the Keep Alive Timeout * accounting for transport roundtrip times [..]. */ @@ -1267,10 +1312,9 @@ static enum rq_end_io_ret nvme_keep_alive_end_io(struct request *rq, blk_status_t status) { struct nvme_ctrl *ctrl = rq->end_io_data; - unsigned long flags; - bool startka = false; unsigned long rtt = jiffies - (rq->deadline - rq->timeout); unsigned long delay = nvme_keep_alive_work_period(ctrl); + enum nvme_ctrl_state state = nvme_ctrl_state(ctrl); /* * Subtract off the keepalive RTT so nvme_keep_alive_work runs @@ -1295,12 +1339,7 @@ static enum rq_end_io_ret nvme_keep_alive_end_io(struct request *rq, ctrl->ka_last_check_time = jiffies; ctrl->comp_seen = false; - spin_lock_irqsave(&ctrl->lock, flags); - if (ctrl->state == NVME_CTRL_LIVE || - ctrl->state == NVME_CTRL_CONNECTING) - startka = true; - spin_unlock_irqrestore(&ctrl->lock, flags); - if (startka) + if (state == NVME_CTRL_LIVE || state == NVME_CTRL_CONNECTING) queue_delayed_work(nvme_wq, &ctrl->ka_work, delay); return RQ_END_IO_NONE; } @@ -1370,17 +1409,30 @@ static void nvme_update_keep_alive(struct nvme_ctrl *ctrl, nvme_start_keep_alive(ctrl); } -/* - * In NVMe 1.0 the CNS field was just a binary controller or namespace - * flag, thus sending any new CNS opcodes has a big chance of not working. - * Qemu unfortunately had that bug after reporting a 1.1 version compliance - * (but not for any later version). - */ -static bool nvme_ctrl_limited_cns(struct nvme_ctrl *ctrl) +static bool nvme_id_cns_ok(struct nvme_ctrl *ctrl, u8 cns) { - if (ctrl->quirks & NVME_QUIRK_IDENTIFY_CNS) - return ctrl->vs < NVME_VS(1, 2, 0); - return ctrl->vs < NVME_VS(1, 1, 0); + /* + * The CNS field occupies a full byte starting with NVMe 1.2 + */ + if (ctrl->vs >= NVME_VS(1, 2, 0)) + return true; + + /* + * NVMe 1.1 expanded the CNS value to two bits, which means values + * larger than that could get truncated and treated as an incorrect + * value. + * + * Qemu implemented 1.0 behavior for controllers claiming 1.1 + * compliance, so they need to be quirked here. + */ + if (ctrl->vs >= NVME_VS(1, 1, 0) && + !(ctrl->quirks & NVME_QUIRK_IDENTIFY_CNS)) + return cns <= 3; + + /* + * NVMe 1.0 used a single bit for the CNS value. + */ + return cns <= 1; } static int nvme_identify_ctrl(struct nvme_ctrl *dev, struct nvme_id_ctrl **id) @@ -1398,8 +1450,10 @@ static int nvme_identify_ctrl(struct nvme_ctrl *dev, struct nvme_id_ctrl **id) error = nvme_submit_sync_cmd(dev->admin_q, &c, *id, sizeof(struct nvme_id_ctrl)); - if (error) + if (error) { kfree(*id); + *id = NULL; + } return error; } @@ -1528,6 +1582,7 @@ int nvme_identify_ns(struct nvme_ctrl *ctrl, unsigned nsid, if (error) { dev_warn(ctrl->device, "Identify namespace failed (%d)\n", error); kfree(*id); + *id = NULL; } return error; } @@ -1592,6 +1647,8 @@ static int nvme_ns_info_from_id_cs_indep(struct nvme_ctrl *ctrl, info->is_shared = id->nmic & NVME_NS_NMIC_SHARED; info->is_readonly = id->nsattr & NVME_NS_ATTR_RO; info->is_ready = id->nstat & NVME_NSTAT_NRDY; + info->is_rotational = id->nsfeat & NVME_NS_ROTATIONAL; + info->no_vwc = id->nsfeat & NVME_NS_VWC_NOT_PRESENT; } kfree(id); return ret; @@ -1641,7 +1698,13 @@ int nvme_set_queue_count(struct nvme_ctrl *ctrl, int *count) status = nvme_set_features(ctrl, NVME_FEAT_NUM_QUEUES, q_count, NULL, 0, &result); - if (status < 0) + + /* + * It's either a kernel error or the host observed a connection + * lost. In either case it's not possible communicate with the + * controller and thus enter the error code path. + */ + if (status < 0 || status == NVME_SC_HOST_PATH_ERROR) return status; /* @@ -1727,27 +1790,38 @@ int nvme_getgeo(struct block_device *bdev, struct hd_geometry *geo) return 0; } -#ifdef CONFIG_BLK_DEV_INTEGRITY -static void nvme_init_integrity(struct gendisk *disk, - struct nvme_ns_head *head, u32 max_integrity_segments) +static bool nvme_init_integrity(struct nvme_ns_head *head, + struct queue_limits *lim, struct nvme_ns_info *info) { - struct blk_integrity integrity = { }; + struct blk_integrity *bi = &lim->integrity; + + memset(bi, 0, sizeof(*bi)); + + if (!head->ms) + return true; + + /* + * PI can always be supported as we can ask the controller to simply + * insert/strip it, which is not possible for other kinds of metadata. + */ + if (!IS_ENABLED(CONFIG_BLK_DEV_INTEGRITY) || + !(head->features & NVME_NS_METADATA_SUPPORTED)) + return nvme_ns_has_pi(head); switch (head->pi_type) { case NVME_NS_DPS_PI_TYPE3: switch (head->guard_type) { case NVME_NVM_NS_16B_GUARD: - integrity.profile = &t10_pi_type3_crc; - integrity.tag_size = sizeof(u16) + sizeof(u32); - integrity.flags |= BLK_INTEGRITY_DEVICE_CAPABLE; + bi->csum_type = BLK_INTEGRITY_CSUM_CRC; + bi->tag_size = sizeof(u16) + sizeof(u32); + bi->flags |= BLK_INTEGRITY_DEVICE_CAPABLE; break; case NVME_NVM_NS_64B_GUARD: - integrity.profile = &ext_pi_type3_crc64; - integrity.tag_size = sizeof(u16) + 6; - integrity.flags |= BLK_INTEGRITY_DEVICE_CAPABLE; + bi->csum_type = BLK_INTEGRITY_CSUM_CRC64; + bi->tag_size = sizeof(u16) + 6; + bi->flags |= BLK_INTEGRITY_DEVICE_CAPABLE; break; default: - integrity.profile = NULL; break; } break; @@ -1755,73 +1829,48 @@ static void nvme_init_integrity(struct gendisk *disk, case NVME_NS_DPS_PI_TYPE2: switch (head->guard_type) { case NVME_NVM_NS_16B_GUARD: - integrity.profile = &t10_pi_type1_crc; - integrity.tag_size = sizeof(u16); - integrity.flags |= BLK_INTEGRITY_DEVICE_CAPABLE; + bi->csum_type = BLK_INTEGRITY_CSUM_CRC; + bi->tag_size = sizeof(u16); + bi->flags |= BLK_INTEGRITY_DEVICE_CAPABLE | + BLK_INTEGRITY_REF_TAG; break; case NVME_NVM_NS_64B_GUARD: - integrity.profile = &ext_pi_type1_crc64; - integrity.tag_size = sizeof(u16); - integrity.flags |= BLK_INTEGRITY_DEVICE_CAPABLE; + bi->csum_type = BLK_INTEGRITY_CSUM_CRC64; + bi->tag_size = sizeof(u16); + bi->flags |= BLK_INTEGRITY_DEVICE_CAPABLE | + BLK_INTEGRITY_REF_TAG; break; default: - integrity.profile = NULL; break; } break; default: - integrity.profile = NULL; break; } - integrity.tuple_size = head->ms; - blk_integrity_register(disk, &integrity); - blk_queue_max_integrity_segments(disk->queue, max_integrity_segments); -} -#else -static void nvme_init_integrity(struct gendisk *disk, - struct nvme_ns_head *head, u32 max_integrity_segments) -{ + bi->tuple_size = head->ms; + bi->pi_offset = info->pi_offset; + return true; } -#endif /* CONFIG_BLK_DEV_INTEGRITY */ -static void nvme_config_discard(struct nvme_ctrl *ctrl, struct gendisk *disk, - struct nvme_ns_head *head) +static void nvme_config_discard(struct nvme_ns *ns, struct queue_limits *lim) { - struct request_queue *queue = disk->queue; - u32 max_discard_sectors; - - if (ctrl->dmrsl && ctrl->dmrsl <= nvme_sect_to_lba(head, UINT_MAX)) { - max_discard_sectors = nvme_lba_to_sect(head, ctrl->dmrsl); - } else if (ctrl->oncs & NVME_CTRL_ONCS_DSM) { - max_discard_sectors = UINT_MAX; - } else { - blk_queue_max_discard_sectors(queue, 0); - return; - } + struct nvme_ctrl *ctrl = ns->ctrl; - BUILD_BUG_ON(PAGE_SIZE / sizeof(struct nvme_dsm_range) < - NVME_DSM_MAX_RANGES); + if (ctrl->dmrsl && ctrl->dmrsl <= nvme_sect_to_lba(ns->head, UINT_MAX)) + lim->max_hw_discard_sectors = + nvme_lba_to_sect(ns->head, ctrl->dmrsl); + else if (ctrl->oncs & NVME_CTRL_ONCS_DSM) + lim->max_hw_discard_sectors = UINT_MAX; + else + lim->max_hw_discard_sectors = 0; - /* - * If discard is already enabled, don't reset queue limits. - * - * This works around the fact that the block layer can't cope well with - * updating the hardware limits when overridden through sysfs. This is - * harmless because discard limits in NVMe are purely advisory. - */ - if (queue->limits.max_discard_sectors) - return; + lim->discard_granularity = lim->logical_block_size; - blk_queue_max_discard_sectors(queue, max_discard_sectors); if (ctrl->dmrl) - blk_queue_max_discard_segments(queue, ctrl->dmrl); + lim->max_discard_segments = ctrl->dmrl; else - blk_queue_max_discard_segments(queue, NVME_DSM_MAX_RANGES); - queue->limits.discard_granularity = queue_logical_block_size(queue); - - if (ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) - blk_queue_max_write_zeroes_sectors(queue, UINT_MAX); + lim->max_discard_segments = NVME_DSM_MAX_RANGES; } static bool nvme_ns_ids_equal(struct nvme_ns_ids *a, struct nvme_ns_ids *b) @@ -1832,44 +1881,46 @@ static bool nvme_ns_ids_equal(struct nvme_ns_ids *a, struct nvme_ns_ids *b) a->csi == b->csi; } -static int nvme_init_ms(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, - struct nvme_id_ns *id) +static int nvme_identify_ns_nvm(struct nvme_ctrl *ctrl, unsigned int nsid, + struct nvme_id_ns_nvm **nvmp) { - bool first = id->dps & NVME_NS_DPS_PI_FIRST; - unsigned lbaf = nvme_lbaf_index(id->flbas); - struct nvme_command c = { }; + struct nvme_command c = { + .identify.opcode = nvme_admin_identify, + .identify.nsid = cpu_to_le32(nsid), + .identify.cns = NVME_ID_CNS_CS_NS, + .identify.csi = NVME_CSI_NVM, + }; struct nvme_id_ns_nvm *nvm; - int ret = 0; - u32 elbaf; - - head->pi_size = 0; - head->ms = le16_to_cpu(id->lbaf[lbaf].ms); - if (!(ctrl->ctratt & NVME_CTRL_ATTR_ELBAS)) { - head->pi_size = sizeof(struct t10_pi_tuple); - head->guard_type = NVME_NVM_NS_16B_GUARD; - goto set_pi; - } + int ret; nvm = kzalloc(sizeof(*nvm), GFP_KERNEL); if (!nvm) return -ENOMEM; - c.identify.opcode = nvme_admin_identify; - c.identify.nsid = cpu_to_le32(head->ns_id); - c.identify.cns = NVME_ID_CNS_CS_NS; - c.identify.csi = NVME_CSI_NVM; - ret = nvme_submit_sync_cmd(ctrl->admin_q, &c, nvm, sizeof(*nvm)); if (ret) - goto free_data; + kfree(nvm); + else + *nvmp = nvm; + return ret; +} - elbaf = le32_to_cpu(nvm->elbaf[lbaf]); +static void nvme_configure_pi_elbas(struct nvme_ns_head *head, + struct nvme_id_ns *id, struct nvme_id_ns_nvm *nvm) +{ + u32 elbaf = le32_to_cpu(nvm->elbaf[nvme_lbaf_index(id->flbas)]); + u8 guard_type; /* no support for storage tag formats right now */ if (nvme_elbaf_sts(elbaf)) - goto free_data; + return; + + guard_type = nvme_elbaf_guard_type(elbaf); + if ((nvm->pic & NVME_ID_NS_NVM_QPIFS) && + guard_type == NVME_NVM_NS_QTYPE_GUARD) + guard_type = nvme_elbaf_qualified_guard_type(elbaf); - head->guard_type = nvme_elbaf_guard_type(elbaf); + head->guard_type = guard_type; switch (head->guard_type) { case NVME_NVM_NS_64B_GUARD: head->pi_size = sizeof(struct crc64_pi_tuple); @@ -1880,30 +1931,34 @@ static int nvme_init_ms(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, default: break; } - -free_data: - kfree(nvm); -set_pi: - if (head->pi_size && (first || head->ms == head->pi_size)) - head->pi_type = id->dps & NVME_NS_DPS_PI_MASK; - else - head->pi_type = 0; - - return ret; } -static int nvme_configure_metadata(struct nvme_ctrl *ctrl, - struct nvme_ns_head *head, struct nvme_id_ns *id) +static void nvme_configure_metadata(struct nvme_ctrl *ctrl, + struct nvme_ns_head *head, struct nvme_id_ns *id, + struct nvme_id_ns_nvm *nvm, struct nvme_ns_info *info) { - int ret; - - ret = nvme_init_ms(ctrl, head, id); - if (ret) - return ret; - head->features &= ~(NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS); + head->pi_type = 0; + head->pi_size = 0; + head->ms = le16_to_cpu(id->lbaf[nvme_lbaf_index(id->flbas)].ms); if (!head->ms || !(ctrl->ops->flags & NVME_F_METADATA_SUPPORTED)) - return 0; + return; + + if (nvm && (ctrl->ctratt & NVME_CTRL_ATTR_ELBAS)) { + nvme_configure_pi_elbas(head, id, nvm); + } else { + head->pi_size = sizeof(struct t10_pi_tuple); + head->guard_type = NVME_NVM_NS_16B_GUARD; + } + + if (head->pi_size && head->ms >= head->pi_size) + head->pi_type = id->dps & NVME_NS_DPS_PI_MASK; + if (!(id->dps & NVME_NS_DPS_PI_FIRST)) { + if (disable_pi_offsets) + head->pi_type = 0; + else + info->pi_offset = head->ms - head->pi_size; + } if (ctrl->ops->flags & NVME_F_FABRICS) { /* @@ -1912,7 +1967,7 @@ static int nvme_configure_metadata(struct nvme_ctrl *ctrl, * remap the separate metadata buffer from the block layer. */ if (WARN_ON_ONCE(!(id->flbas & NVME_NS_FLBAS_META_EXT))) - return 0; + return; head->features |= NVME_NS_EXT_LBAS; @@ -1939,46 +1994,61 @@ static int nvme_configure_metadata(struct nvme_ctrl *ctrl, else head->features |= NVME_NS_METADATA_SUPPORTED; } - return 0; } -static void nvme_set_queue_limits(struct nvme_ctrl *ctrl, - struct request_queue *q) -{ - bool vwc = ctrl->vwc & NVME_CTRL_VWC_PRESENT; - if (ctrl->max_hw_sectors) { - u32 max_segments = - (ctrl->max_hw_sectors / (NVME_CTRL_PAGE_SIZE >> 9)) + 1; +static void nvme_update_atomic_write_disk_info(struct nvme_ns *ns, + struct nvme_id_ns *id, struct queue_limits *lim, + u32 bs, u32 atomic_bs) +{ + unsigned int boundary = 0; - max_segments = min_not_zero(max_segments, ctrl->max_segments); - blk_queue_max_hw_sectors(q, ctrl->max_hw_sectors); - blk_queue_max_segments(q, min_t(u32, max_segments, USHRT_MAX)); + if (id->nsfeat & NVME_NS_FEAT_ATOMICS && id->nawupf) { + if (le16_to_cpu(id->nabspf)) + boundary = (le16_to_cpu(id->nabspf) + 1) * bs; } - blk_queue_virt_boundary(q, NVME_CTRL_PAGE_SIZE - 1); - blk_queue_dma_alignment(q, 3); - blk_queue_write_cache(q, vwc, vwc); + lim->atomic_write_hw_max = atomic_bs; + lim->atomic_write_hw_boundary = boundary; + lim->atomic_write_hw_unit_min = bs; + lim->atomic_write_hw_unit_max = rounddown_pow_of_two(atomic_bs); + lim->features |= BLK_FEAT_ATOMIC_WRITES; +} + +static u32 nvme_max_drv_segments(struct nvme_ctrl *ctrl) +{ + return ctrl->max_hw_sectors / (NVME_CTRL_PAGE_SIZE >> SECTOR_SHIFT) + 1; +} + +static void nvme_set_ctrl_limits(struct nvme_ctrl *ctrl, + struct queue_limits *lim) +{ + lim->max_hw_sectors = ctrl->max_hw_sectors; + lim->max_segments = min_t(u32, USHRT_MAX, + min_not_zero(nvme_max_drv_segments(ctrl), ctrl->max_segments)); + lim->max_integrity_segments = ctrl->max_integrity_segments; + lim->virt_boundary_mask = NVME_CTRL_PAGE_SIZE - 1; + lim->max_segment_size = UINT_MAX; + lim->dma_alignment = 3; } -static void nvme_update_disk_info(struct nvme_ctrl *ctrl, struct gendisk *disk, - struct nvme_ns_head *head, struct nvme_id_ns *id) +static bool nvme_update_disk_info(struct nvme_ns *ns, struct nvme_id_ns *id, + struct queue_limits *lim) { - sector_t capacity = nvme_lba_to_sect(head, le64_to_cpu(id->nsze)); + struct nvme_ns_head *head = ns->head; u32 bs = 1U << head->lba_shift; u32 atomic_bs, phys_bs, io_opt = 0; + bool valid = true; /* * The block layer can't support LBA sizes larger than the page size * or smaller than a sector size yet, so catch this early and don't * allow block I/O. */ - if (head->lba_shift > PAGE_SHIFT || head->lba_shift < SECTOR_SHIFT) { - capacity = 0; + if (blk_validate_block_size(bs)) { bs = (1 << 9); + valid = false; } - blk_integrity_unregister(disk); - atomic_bs = phys_bs = bs; if (id->nabo == 0) { /* @@ -1989,46 +2059,34 @@ static void nvme_update_disk_info(struct nvme_ctrl *ctrl, struct gendisk *disk, if (id->nsfeat & NVME_NS_FEAT_ATOMICS && id->nawupf) atomic_bs = (1 + le16_to_cpu(id->nawupf)) * bs; else - atomic_bs = (1 + ctrl->subsys->awupf) * bs; + atomic_bs = (1 + ns->ctrl->subsys->awupf) * bs; + + nvme_update_atomic_write_disk_info(ns, id, lim, bs, atomic_bs); } if (id->nsfeat & NVME_NS_FEAT_IO_OPT) { /* NPWG = Namespace Preferred Write Granularity */ phys_bs = bs * (1 + le16_to_cpu(id->npwg)); /* NOWS = Namespace Optimal Write Size */ - io_opt = bs * (1 + le16_to_cpu(id->nows)); + if (id->nows) + io_opt = bs * (1 + le16_to_cpu(id->nows)); } - blk_queue_logical_block_size(disk->queue, bs); /* * Linux filesystems assume writing a single physical block is * an atomic operation. Hence limit the physical block size to the * value of the Atomic Write Unit Power Fail parameter. */ - blk_queue_physical_block_size(disk->queue, min(phys_bs, atomic_bs)); - blk_queue_io_min(disk->queue, phys_bs); - blk_queue_io_opt(disk->queue, io_opt); - - /* - * Register a metadata profile for PI, or the plain non-integrity NVMe - * metadata masquerading as Type 0 if supported, otherwise reject block - * I/O to namespaces with metadata except when the namespace supports - * PI, as it can strip/insert in that case. - */ - if (head->ms) { - if (IS_ENABLED(CONFIG_BLK_DEV_INTEGRITY) && - (head->features & NVME_NS_METADATA_SUPPORTED)) - nvme_init_integrity(disk, head, - ctrl->max_integrity_segments); - else if (!nvme_ns_has_pi(head)) - capacity = 0; - } - - set_capacity_and_notify(disk, capacity); - - nvme_config_discard(ctrl, disk, head); - blk_queue_max_write_zeroes_sectors(disk->queue, - ctrl->max_zeroes_sectors); + lim->logical_block_size = bs; + lim->physical_block_size = min(phys_bs, atomic_bs); + lim->io_min = phys_bs; + lim->io_opt = io_opt; + if ((ns->ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) && + (ns->ctrl->oncs & NVME_CTRL_ONCS_DSM)) + lim->max_write_zeroes_sectors = UINT_MAX; + else + lim->max_write_zeroes_sectors = ns->ctrl->max_zeroes_sectors; + return valid; } static bool nvme_ns_is_readonly(struct nvme_ns *ns, struct nvme_ns_info *info) @@ -2042,7 +2100,8 @@ static inline bool nvme_first_scan(struct gendisk *disk) return !disk_live(disk); } -static void nvme_set_chunk_sectors(struct nvme_ns *ns, struct nvme_id_ns *id) +static void nvme_set_chunk_sectors(struct nvme_ns *ns, struct nvme_id_ns *id, + struct queue_limits *lim) { struct nvme_ctrl *ctrl = ns->ctrl; u32 iob; @@ -2070,38 +2129,39 @@ static void nvme_set_chunk_sectors(struct nvme_ns *ns, struct nvme_id_ns *id) return; } - blk_queue_chunk_sectors(ns->queue, iob); + lim->chunk_sectors = iob; } static int nvme_update_ns_info_generic(struct nvme_ns *ns, struct nvme_ns_info *info) { - blk_mq_freeze_queue(ns->disk->queue); - nvme_set_queue_limits(ns->ctrl, ns->queue); - set_disk_ro(ns->disk, nvme_ns_is_readonly(ns, info)); - blk_mq_unfreeze_queue(ns->disk->queue); + struct queue_limits lim; + unsigned int memflags; + int ret; - if (nvme_ns_head_multipath(ns->head)) { - blk_mq_freeze_queue(ns->head->disk->queue); - set_disk_ro(ns->head->disk, nvme_ns_is_readonly(ns, info)); - nvme_mpath_revalidate_paths(ns); - blk_stack_limits(&ns->head->disk->queue->limits, - &ns->queue->limits, 0); - ns->head->disk->flags |= GENHD_FL_HIDDEN; - blk_mq_unfreeze_queue(ns->head->disk->queue); - } + lim = queue_limits_start_update(ns->disk->queue); + nvme_set_ctrl_limits(ns->ctrl, &lim); - /* Hide the block-interface for these devices */ - ns->disk->flags |= GENHD_FL_HIDDEN; - set_bit(NVME_NS_READY, &ns->flags); + memflags = blk_mq_freeze_queue(ns->disk->queue); + ret = queue_limits_commit_update(ns->disk->queue, &lim); + set_disk_ro(ns->disk, nvme_ns_is_readonly(ns, info)); + blk_mq_unfreeze_queue(ns->disk->queue, memflags); - return 0; + /* Hide the block-interface for these devices */ + if (!ret) + ret = -ENODEV; + return ret; } static int nvme_update_ns_info_block(struct nvme_ns *ns, struct nvme_ns_info *info) { + struct queue_limits lim; + struct nvme_id_ns_nvm *nvm = NULL; + struct nvme_zone_info zi = {}; struct nvme_id_ns *id; + unsigned int memflags; + sector_t capacity; unsigned lbaf; int ret; @@ -2112,31 +2172,64 @@ static int nvme_update_ns_info_block(struct nvme_ns *ns, if (id->ncap == 0) { /* namespace not allocated or attached */ info->is_removed = true; - ret = -ENODEV; - goto error; + ret = -ENXIO; + goto out; } - - blk_mq_freeze_queue(ns->disk->queue); lbaf = nvme_lbaf_index(id->flbas); + + if (ns->ctrl->ctratt & NVME_CTRL_ATTR_ELBAS) { + ret = nvme_identify_ns_nvm(ns->ctrl, info->nsid, &nvm); + if (ret < 0) + goto out; + } + + if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) && + ns->head->ids.csi == NVME_CSI_ZNS) { + ret = nvme_query_zone_info(ns, lbaf, &zi); + if (ret < 0) + goto out; + } + + lim = queue_limits_start_update(ns->disk->queue); + + memflags = blk_mq_freeze_queue(ns->disk->queue); ns->head->lba_shift = id->lbaf[lbaf].ds; ns->head->nuse = le64_to_cpu(id->nuse); - nvme_set_queue_limits(ns->ctrl, ns->queue); + capacity = nvme_lba_to_sect(ns->head, le64_to_cpu(id->nsze)); + nvme_set_ctrl_limits(ns->ctrl, &lim); + nvme_configure_metadata(ns->ctrl, ns->head, id, nvm, info); + nvme_set_chunk_sectors(ns, id, &lim); + if (!nvme_update_disk_info(ns, id, &lim)) + capacity = 0; + nvme_config_discard(ns, &lim); + if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) && + ns->head->ids.csi == NVME_CSI_ZNS) + nvme_update_zone_info(ns, &lim, &zi); - ret = nvme_configure_metadata(ns->ctrl, ns->head, id); - if (ret < 0) { - blk_mq_unfreeze_queue(ns->disk->queue); + if ((ns->ctrl->vwc & NVME_CTRL_VWC_PRESENT) && !info->no_vwc) + lim.features |= BLK_FEAT_WRITE_CACHE | BLK_FEAT_FUA; + else + lim.features &= ~(BLK_FEAT_WRITE_CACHE | BLK_FEAT_FUA); + + if (info->is_rotational) + lim.features |= BLK_FEAT_ROTATIONAL; + + /* + * Register a metadata profile for PI, or the plain non-integrity NVMe + * metadata masquerading as Type 0 if supported, otherwise reject block + * I/O to namespaces with metadata except when the namespace supports + * PI, as it can strip/insert in that case. + */ + if (!nvme_init_integrity(ns->head, &lim, info)) + capacity = 0; + + ret = queue_limits_commit_update(ns->disk->queue, &lim); + if (ret) { + blk_mq_unfreeze_queue(ns->disk->queue, memflags); goto out; } - nvme_set_chunk_sectors(ns, id); - nvme_update_disk_info(ns->ctrl, ns->disk, ns->head, id); - if (ns->head->ids.csi == NVME_CSI_ZNS) { - ret = nvme_update_zone_info(ns, lbaf); - if (ret) { - blk_mq_unfreeze_queue(ns->disk->queue); - goto out; - } - } + set_capacity_and_notify(ns->disk, capacity); /* * Only set the DEAC bit if the device guarantees that reads from @@ -2148,61 +2241,127 @@ static int nvme_update_ns_info_block(struct nvme_ns *ns, ns->head->features |= NVME_NS_DEAC; set_disk_ro(ns->disk, nvme_ns_is_readonly(ns, info)); set_bit(NVME_NS_READY, &ns->flags); - blk_mq_unfreeze_queue(ns->disk->queue); + blk_mq_unfreeze_queue(ns->disk->queue, memflags); if (blk_queue_is_zoned(ns->queue)) { - ret = nvme_revalidate_zones(ns); + ret = blk_revalidate_disk_zones(ns->disk); if (ret && !nvme_first_scan(ns->disk)) goto out; } - if (nvme_ns_head_multipath(ns->head)) { - blk_mq_freeze_queue(ns->head->disk->queue); - nvme_update_disk_info(ns->ctrl, ns->head->disk, ns->head, id); - set_disk_ro(ns->head->disk, nvme_ns_is_readonly(ns, info)); - nvme_mpath_revalidate_paths(ns); - blk_stack_limits(&ns->head->disk->queue->limits, - &ns->queue->limits, 0); - disk_update_readahead(ns->head->disk); - blk_mq_unfreeze_queue(ns->head->disk->queue); - } - ret = 0; out: - /* - * If probing fails due an unsupported feature, hide the block device, - * but still allow other access. - */ - if (ret == -ENODEV) { - ns->disk->flags |= GENHD_FL_HIDDEN; - set_bit(NVME_NS_READY, &ns->flags); - ret = 0; - } - -error: + kfree(nvm); kfree(id); return ret; } static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_ns_info *info) { + bool unsupported = false; + int ret; + switch (info->ids.csi) { case NVME_CSI_ZNS: if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED)) { dev_info(ns->ctrl->device, "block device for nsid %u not supported without CONFIG_BLK_DEV_ZONED\n", info->nsid); - return nvme_update_ns_info_generic(ns, info); + ret = nvme_update_ns_info_generic(ns, info); + break; } - return nvme_update_ns_info_block(ns, info); + ret = nvme_update_ns_info_block(ns, info); + break; case NVME_CSI_NVM: - return nvme_update_ns_info_block(ns, info); + ret = nvme_update_ns_info_block(ns, info); + break; default: dev_info(ns->ctrl->device, "block device for nsid %u not supported (csi %u)\n", info->nsid, info->ids.csi); - return nvme_update_ns_info_generic(ns, info); + ret = nvme_update_ns_info_generic(ns, info); + break; + } + + /* + * If probing fails due an unsupported feature, hide the block device, + * but still allow other access. + */ + if (ret == -ENODEV) { + ns->disk->flags |= GENHD_FL_HIDDEN; + set_bit(NVME_NS_READY, &ns->flags); + unsupported = true; + ret = 0; + } + + if (!ret && nvme_ns_head_multipath(ns->head)) { + struct queue_limits *ns_lim = &ns->disk->queue->limits; + struct queue_limits lim; + unsigned int memflags; + + lim = queue_limits_start_update(ns->head->disk->queue); + memflags = blk_mq_freeze_queue(ns->head->disk->queue); + /* + * queue_limits mixes values that are the hardware limitations + * for bio splitting with what is the device configuration. + * + * For NVMe the device configuration can change after e.g. a + * Format command, and we really want to pick up the new format + * value here. But we must still stack the queue limits to the + * least common denominator for multipathing to split the bios + * properly. + * + * To work around this, we explicitly set the device + * configuration to those that we just queried, but only stack + * the splitting limits in to make sure we still obey possibly + * lower limitations of other controllers. + */ + lim.logical_block_size = ns_lim->logical_block_size; + lim.physical_block_size = ns_lim->physical_block_size; + lim.io_min = ns_lim->io_min; + lim.io_opt = ns_lim->io_opt; + queue_limits_stack_bdev(&lim, ns->disk->part0, 0, + ns->head->disk->disk_name); + if (unsupported) + ns->head->disk->flags |= GENHD_FL_HIDDEN; + else + nvme_init_integrity(ns->head, &lim, info); + ret = queue_limits_commit_update(ns->head->disk->queue, &lim); + + set_capacity_and_notify(ns->head->disk, get_capacity(ns->disk)); + set_disk_ro(ns->head->disk, nvme_ns_is_readonly(ns, info)); + nvme_mpath_revalidate_paths(ns); + + blk_mq_unfreeze_queue(ns->head->disk->queue, memflags); + } + + return ret; +} + +int nvme_ns_get_unique_id(struct nvme_ns *ns, u8 id[16], + enum blk_unique_id type) +{ + struct nvme_ns_ids *ids = &ns->head->ids; + + if (type != BLK_UID_EUI64) + return -EINVAL; + + if (memchr_inv(ids->nguid, 0, sizeof(ids->nguid))) { + memcpy(id, &ids->nguid, sizeof(ids->nguid)); + return sizeof(ids->nguid); + } + if (memchr_inv(ids->eui64, 0, sizeof(ids->eui64))) { + memcpy(id, &ids->eui64, sizeof(ids->eui64)); + return sizeof(ids->eui64); } + + return -EINVAL; +} + +static int nvme_get_unique_id(struct gendisk *disk, u8 id[16], + enum blk_unique_id type) +{ + return nvme_ns_get_unique_id(disk->private_data, id, type); } #ifdef CONFIG_BLK_SED_OPAL @@ -2260,6 +2419,7 @@ const struct block_device_operations nvme_bdev_ops = { .open = nvme_open, .release = nvme_release, .getgeo = nvme_getgeo, + .get_unique_id = nvme_get_unique_id, .report_zones = nvme_report_zones, .pr_ops = &nvme_pr_ops, }; @@ -2342,8 +2502,13 @@ int nvme_enable_ctrl(struct nvme_ctrl *ctrl) else ctrl->ctrl_config = NVME_CC_CSS_NVM; - if (ctrl->cap & NVME_CAP_CRMS_CRWMS && ctrl->cap & NVME_CAP_CRMS_CRIMS) - ctrl->ctrl_config |= NVME_CC_CRIME; + /* + * Setting CRIME results in CSTS.RDY before the media is ready. This + * makes it possible for media related commands to return the error + * NVME_SC_ADMIN_COMMAND_MEDIA_NOT_READY. Until the driver is + * restructured to handle retries, disable CC.CRIME. + */ + ctrl->ctrl_config &= ~NVME_CC_CRIME; ctrl->ctrl_config |= (NVME_CTRL_PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT; ctrl->ctrl_config |= NVME_CC_AMS_RR | NVME_CC_SHN_NONE; @@ -2352,11 +2517,6 @@ int nvme_enable_ctrl(struct nvme_ctrl *ctrl) if (ret) return ret; - /* Flush write to device (required if transport is PCI) */ - ret = ctrl->ops->reg_read32(ctrl, NVME_REG_CC, &ctrl->ctrl_config); - if (ret) - return ret; - /* CAP value may change after initial CC write */ ret = ctrl->ops->reg_read64(ctrl, NVME_REG_CAP, &ctrl->cap); if (ret) @@ -2378,10 +2538,7 @@ int nvme_enable_ctrl(struct nvme_ctrl *ctrl) * devices are known to get this wrong. Use the larger of the * two values. */ - if (ctrl->ctrl_config & NVME_CC_CRIME) - ready_timeout = NVME_CRTO_CRIMT(crto); - else - ready_timeout = NVME_CRTO_CRWMT(crto); + ready_timeout = NVME_CRTO_CRWMT(crto); if (ready_timeout < timeout) dev_warn_once(ctrl->device, "bad crto:%x cap:%llx\n", @@ -2877,7 +3034,7 @@ static int nvme_init_subsystem(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) subsys->awupf = le16_to_cpu(id->awupf); nvme_mpath_default_iopolicy(subsys); - subsys->dev.class = nvme_subsys_class; + subsys->dev.class = &nvme_subsys_class; subsys->dev.release = nvme_release_subsystem; subsys->dev.groups = nvme_subsys_attrs_groups; dev_set_name(&subsys->dev, "nvme-subsys%d", ctrl->instance); @@ -2949,7 +3106,7 @@ int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp, u8 csi, static int nvme_get_effects_log(struct nvme_ctrl *ctrl, u8 csi, struct nvme_effects_log **log) { - struct nvme_effects_log *cel = xa_load(&ctrl->cels, csi); + struct nvme_effects_log *old, *cel = xa_load(&ctrl->cels, csi); int ret; if (cel) @@ -2966,7 +3123,11 @@ static int nvme_get_effects_log(struct nvme_ctrl *ctrl, u8 csi, return ret; } - xa_store(&ctrl->cels, csi, cel, GFP_KERNEL); + old = xa_store(&ctrl->cels, csi, cel, GFP_KERNEL); + if (xa_is_err(old)) { + kfree(cel); + return xa_err(old); + } out: *log = cel; return 0; @@ -3000,7 +3161,7 @@ static int nvme_init_non_mdts_limits(struct nvme_ctrl *ctrl) ctrl->max_zeroes_sectors = 0; if (ctrl->subsys->subtype != NVME_NQN_NVME || - nvme_ctrl_limited_cns(ctrl) || + !nvme_id_cns_ok(ctrl, NVME_ID_CNS_CS_CTRL) || test_bit(NVME_CTRL_SKIP_ID_CNS_CS, &ctrl->flags)) return 0; @@ -3028,6 +3189,25 @@ free_data: return ret; } +static int nvme_init_effects_log(struct nvme_ctrl *ctrl, + u8 csi, struct nvme_effects_log **log) +{ + struct nvme_effects_log *effects, *old; + + effects = kzalloc(sizeof(*effects), GFP_KERNEL); + if (!effects) + return -ENOMEM; + + old = xa_store(&ctrl->cels, csi, effects, GFP_KERNEL); + if (xa_is_err(old)) { + kfree(effects); + return xa_err(old); + } + + *log = effects; + return 0; +} + static void nvme_init_known_nvm_effects(struct nvme_ctrl *ctrl) { struct nvme_effects_log *log = ctrl->effects; @@ -3074,10 +3254,9 @@ static int nvme_init_effects(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) } if (!ctrl->effects) { - ctrl->effects = kzalloc(sizeof(*ctrl->effects), GFP_KERNEL); - if (!ctrl->effects) - return -ENOMEM; - xa_store(&ctrl->cels, NVME_CSI_NVM, ctrl->effects, GFP_KERNEL); + ret = nvme_init_effects_log(ctrl, NVME_CSI_NVM, &ctrl->effects); + if (ret < 0) + return ret; } nvme_init_known_nvm_effects(ctrl); @@ -3117,11 +3296,18 @@ static int nvme_check_ctrl_fabric_info(struct nvme_ctrl *ctrl, struct nvme_id_ct return -EINVAL; } + if (!ctrl->maxcmd) { + dev_warn(ctrl->device, + "Firmware bug: maximum outstanding commands is 0\n"); + ctrl->maxcmd = ctrl->sqsize + 1; + } + return 0; } static int nvme_init_identify(struct nvme_ctrl *ctrl) { + struct queue_limits lim; struct nvme_id_ctrl *id; u32 max_hw_sectors; bool prev_apst_enabled; @@ -3188,7 +3374,12 @@ static int nvme_init_identify(struct nvme_ctrl *ctrl) ctrl->max_hw_sectors = min_not_zero(ctrl->max_hw_sectors, max_hw_sectors); - nvme_set_queue_limits(ctrl, ctrl->admin_q); + lim = queue_limits_start_update(ctrl->admin_q); + nvme_set_ctrl_limits(ctrl, &lim); + ret = queue_limits_commit_update(ctrl->admin_q, &lim); + if (ret) + goto out_free; + ctrl->sgls = le32_to_cpu(id->sgls); ctrl->kas = le16_to_cpu(id->kas); ctrl->max_namespaces = le32_to_cpu(id->mnan); @@ -3206,7 +3397,7 @@ static int nvme_init_identify(struct nvme_ctrl *ctrl) if (ctrl->shutdown_timeout != shutdown_timeout) dev_info(ctrl->device, - "Shutdown timeout set to %u seconds\n", + "D3 entry latency set to %u seconds\n", ctrl->shutdown_timeout); } else ctrl->shutdown_timeout = shutdown_timeout; @@ -3420,7 +3611,7 @@ int nvme_cdev_add(struct cdev *cdev, struct device *cdev_device, if (minor < 0) return minor; cdev_device->devt = MKDEV(MAJOR(nvme_ns_chr_devt), minor); - cdev_device->class = nvme_ns_chr_class; + cdev_device->class = &nvme_ns_chr_class; cdev_device->release = nvme_cdev_rel; device_initialize(cdev_device); cdev_init(cdev, fops); @@ -3493,6 +3684,7 @@ static struct nvme_ns_head *nvme_alloc_ns_head(struct nvme_ctrl *ctrl, head->ns_id = info->nsid; head->ids = info->ids; head->shared = info->is_shared; + head->rotational = info->is_rotational; ratelimit_state_init(&head->rs_nuse, 5 * HZ, 1); ratelimit_set_flags(&head->rs_nuse, RATELIMIT_MSG_ON_RELEASE); kref_init(&head->ref); @@ -3630,7 +3822,7 @@ static int nvme_init_ns_head(struct nvme_ns *ns, struct nvme_ns_info *info) "Found shared namespace %d, but multipathing not supported.\n", info->nsid); dev_warn_once(ctrl->device, - "Support for shared namespaces without CONFIG_NVME_MULTIPATH is deprecated and will be removed in Linux 6.0\n."); + "Support for shared namespaces without CONFIG_NVME_MULTIPATH is deprecated and will be removed in Linux 6.0.\n"); } } @@ -3649,9 +3841,11 @@ out_unlock: struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid) { struct nvme_ns *ns, *ret = NULL; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) { + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) { if (ns->head->ns_id == nsid) { if (!nvme_get_ns(ns)) continue; @@ -3661,10 +3855,10 @@ struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid) if (ns->head->ns_id > nsid) break; } - up_read(&ctrl->namespaces_rwsem); + srcu_read_unlock(&ctrl->srcu, srcu_idx); return ret; } -EXPORT_SYMBOL_NS_GPL(nvme_find_get_ns, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_find_get_ns, "NVME_TARGET_PASSTHRU"); /* * Add the namespace to the controller list while keeping the list ordered. @@ -3675,7 +3869,7 @@ static void nvme_ns_add_to_ctrl_list(struct nvme_ns *ns) list_for_each_entry_reverse(tmp, &ns->ctrl->namespaces, list) { if (tmp->head->ns_id < ns->head->ns_id) { - list_add(&ns->list, &tmp->list); + list_add_rcu(&ns->list, &tmp->list); return; } } @@ -3684,6 +3878,7 @@ static void nvme_ns_add_to_ctrl_list(struct nvme_ns *ns) static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info) { + struct queue_limits lim = { }; struct nvme_ns *ns; struct gendisk *disk; int node = ctrl->numa_node; @@ -3692,7 +3887,13 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info) if (!ns) return; - disk = blk_mq_alloc_disk(ctrl->tagset, ns); + if (ctrl->opts && ctrl->opts->data_digest) + lim.features |= BLK_FEAT_STABLE_WRITES; + if (ctrl->ops->supports_pci_p2pdma && + ctrl->ops->supports_pci_p2pdma(ctrl)) + lim.features |= BLK_FEAT_PCI_P2PDMA; + + disk = blk_mq_alloc_disk(ctrl->tagset, &lim, ns); if (IS_ERR(disk)) goto out_free_ns; disk->fops = &nvme_bdev_ops; @@ -3700,15 +3901,6 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info) ns->disk = disk; ns->queue = disk->queue; - - if (ctrl->opts && ctrl->opts->data_digest) - blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, ns->queue); - - blk_queue_flag_set(QUEUE_FLAG_NONROT, ns->queue); - if (ctrl->ops->supports_pci_p2pdma && - ctrl->ops->supports_pci_p2pdma(ctrl)) - blk_queue_flag_set(QUEUE_FLAG_PCI_P2PDMA, ns->queue); - ns->ctrl = ctrl; kref_init(&ns->kref); @@ -3741,17 +3933,18 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info) if (nvme_update_ns_info(ns, info)) goto out_unlink_ns; - down_write(&ctrl->namespaces_rwsem); + mutex_lock(&ctrl->namespaces_lock); /* * Ensure that no namespaces are added to the ctrl list after the queues * are frozen, thereby avoiding a deadlock between scan and reset. */ if (test_bit(NVME_CTRL_FROZEN, &ctrl->flags)) { - up_write(&ctrl->namespaces_rwsem); + mutex_unlock(&ctrl->namespaces_lock); goto out_unlink_ns; } nvme_ns_add_to_ctrl_list(ns); - up_write(&ctrl->namespaces_rwsem); + mutex_unlock(&ctrl->namespaces_lock); + synchronize_srcu(&ctrl->srcu); nvme_get_ctrl(ctrl); if (device_add_disk(ctrl->device, ns->disk, nvme_ns_attr_groups)) @@ -3774,9 +3967,10 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info) out_cleanup_ns_from_list: nvme_put_ctrl(ctrl); - down_write(&ctrl->namespaces_rwsem); - list_del_init(&ns->list); - up_write(&ctrl->namespaces_rwsem); + mutex_lock(&ctrl->namespaces_lock); + list_del_rcu(&ns->list); + mutex_unlock(&ctrl->namespaces_lock); + synchronize_srcu(&ctrl->srcu); out_unlink_ns: mutex_lock(&ctrl->subsys->lock); list_del_rcu(&ns->siblings); @@ -3826,9 +4020,10 @@ static void nvme_ns_remove(struct nvme_ns *ns) nvme_cdev_del(&ns->cdev, &ns->cdev_device); del_gendisk(ns->disk); - down_write(&ns->ctrl->namespaces_rwsem); - list_del_init(&ns->list); - up_write(&ns->ctrl->namespaces_rwsem); + mutex_lock(&ns->ctrl->namespaces_lock); + list_del_rcu(&ns->list); + mutex_unlock(&ns->ctrl->namespaces_lock); + synchronize_srcu(&ns->ctrl->srcu); if (last_path) nvme_mpath_shutdown_disk(ns->head); @@ -3847,7 +4042,7 @@ static void nvme_ns_remove_by_nsid(struct nvme_ctrl *ctrl, u32 nsid) static void nvme_validate_ns(struct nvme_ns *ns, struct nvme_ns_info *info) { - int ret = NVME_SC_INVALID_NS | NVME_SC_DNR; + int ret = NVME_SC_INVALID_NS | NVME_STATUS_DNR; if (!nvme_ns_ids_equal(&ns->head->ids, &info->ids)) { dev_err(ns->ctrl->device, @@ -3863,7 +4058,7 @@ out: * * TODO: we should probably schedule a delayed retry here. */ - if (ret > 0 && (ret & NVME_SC_DNR)) + if (ret > 0 && (ret & NVME_STATUS_DNR)) nvme_ns_remove(ns); } @@ -3871,7 +4066,7 @@ static void nvme_scan_ns(struct nvme_ctrl *ctrl, unsigned nsid) { struct nvme_ns_info info = { .nsid = nsid }; struct nvme_ns *ns; - int ret; + int ret = 1; if (nvme_identify_ns_descs(ctrl, &info)) return; @@ -3888,9 +4083,10 @@ static void nvme_scan_ns(struct nvme_ctrl *ctrl, unsigned nsid) * set up a namespace. If not fall back to the legacy version. */ if ((ctrl->cap & NVME_CAP_CRMS_CRIMS) || - (info.ids.csi != NVME_CSI_NVM && info.ids.csi != NVME_CSI_ZNS)) + (info.ids.csi != NVME_CSI_NVM && info.ids.csi != NVME_CSI_ZNS) || + ctrl->vs >= NVME_VS(2, 0, 0)) ret = nvme_ns_info_from_id_cs_indep(ctrl, &info); - else + if (ret > 0) ret = nvme_ns_info_from_identify(ctrl, &info); if (info.is_removed) @@ -3912,22 +4108,53 @@ static void nvme_scan_ns(struct nvme_ctrl *ctrl, unsigned nsid) } } +/** + * struct async_scan_info - keeps track of controller & NSIDs to scan + * @ctrl: Controller on which namespaces are being scanned + * @next_nsid: Index of next NSID to scan in ns_list + * @ns_list: Pointer to list of NSIDs to scan + * + * Note: There is a single async_scan_info structure shared by all instances + * of nvme_scan_ns_async() scanning a given controller, so the atomic + * operations on next_nsid are critical to ensure each instance scans a unique + * NSID. + */ +struct async_scan_info { + struct nvme_ctrl *ctrl; + atomic_t next_nsid; + __le32 *ns_list; +}; + +static void nvme_scan_ns_async(void *data, async_cookie_t cookie) +{ + struct async_scan_info *scan_info = data; + int idx; + u32 nsid; + + idx = (u32)atomic_fetch_inc(&scan_info->next_nsid); + nsid = le32_to_cpu(scan_info->ns_list[idx]); + + nvme_scan_ns(scan_info->ctrl, nsid); +} + static void nvme_remove_invalid_namespaces(struct nvme_ctrl *ctrl, unsigned nsid) { struct nvme_ns *ns, *next; LIST_HEAD(rm_list); - down_write(&ctrl->namespaces_rwsem); + mutex_lock(&ctrl->namespaces_lock); list_for_each_entry_safe(ns, next, &ctrl->namespaces, list) { - if (ns->head->ns_id > nsid) - list_move_tail(&ns->list, &rm_list); + if (ns->head->ns_id > nsid) { + list_del_rcu(&ns->list); + synchronize_srcu(&ctrl->srcu); + list_add_tail_rcu(&ns->list, &rm_list); + } } - up_write(&ctrl->namespaces_rwsem); + mutex_unlock(&ctrl->namespaces_lock); list_for_each_entry_safe(ns, next, &rm_list, list) nvme_ns_remove(ns); - } static int nvme_scan_ns_list(struct nvme_ctrl *ctrl) @@ -3936,11 +4163,15 @@ static int nvme_scan_ns_list(struct nvme_ctrl *ctrl) __le32 *ns_list; u32 prev = 0; int ret = 0, i; + ASYNC_DOMAIN(domain); + struct async_scan_info scan_info; ns_list = kzalloc(NVME_IDENTIFY_DATA_SIZE, GFP_KERNEL); if (!ns_list) return -ENOMEM; + scan_info.ctrl = ctrl; + scan_info.ns_list = ns_list; for (;;) { struct nvme_command cmd = { .identify.opcode = nvme_admin_identify, @@ -3956,19 +4187,23 @@ static int nvme_scan_ns_list(struct nvme_ctrl *ctrl) goto free; } + atomic_set(&scan_info.next_nsid, 0); for (i = 0; i < nr_entries; i++) { u32 nsid = le32_to_cpu(ns_list[i]); if (!nsid) /* end of the list? */ goto out; - nvme_scan_ns(ctrl, nsid); + async_schedule_domain(nvme_scan_ns_async, &scan_info, + &domain); while (++prev < nsid) nvme_ns_remove_by_nsid(ctrl, prev); } + async_synchronize_full_domain(&domain); } out: nvme_remove_invalid_namespaces(ctrl, prev); free: + async_synchronize_full_domain(&domain); kfree(ns_list); return ret; } @@ -4044,7 +4279,7 @@ static void nvme_scan_work(struct work_struct *work) } mutex_lock(&ctrl->scan_lock); - if (nvme_ctrl_limited_cns(ctrl)) { + if (!nvme_id_cns_ok(ctrl, NVME_ID_CNS_NS_ACTIVE_LIST)) { nvme_scan_ns_sequential(ctrl); } else { /* @@ -4053,7 +4288,7 @@ static void nvme_scan_work(struct work_struct *work) * they report) but don't actually support it. */ ret = nvme_scan_ns_list(ctrl); - if (ret > 0 && ret & NVME_SC_DNR) + if (ret > 0 && ret & NVME_STATUS_DNR) nvme_scan_ns_sequential(ctrl); } mutex_unlock(&ctrl->scan_lock); @@ -4097,9 +4332,10 @@ void nvme_remove_namespaces(struct nvme_ctrl *ctrl) /* this is a no-op when called from the controller reset handler */ nvme_change_ctrl_state(ctrl, NVME_CTRL_DELETING_NOIO); - down_write(&ctrl->namespaces_rwsem); - list_splice_init(&ctrl->namespaces, &ns_list); - up_write(&ctrl->namespaces_rwsem); + mutex_lock(&ctrl->namespaces_lock); + list_splice_init_rcu(&ctrl->namespaces, &ns_list, synchronize_rcu); + mutex_unlock(&ctrl->namespaces_lock); + synchronize_srcu(&ctrl->srcu); list_for_each_entry_safe(ns, next, &ns_list, list) nvme_ns_remove(ns); @@ -4306,7 +4542,8 @@ static bool nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result) static void nvme_handle_aer_persistent_error(struct nvme_ctrl *ctrl) { - dev_warn(ctrl->device, "resetting controller due to AER\n"); + dev_warn(ctrl->device, + "resetting controller due to persistent internal error\n"); nvme_reset_ctrl(ctrl); } @@ -4353,15 +4590,16 @@ EXPORT_SYMBOL_GPL(nvme_complete_async_event); int nvme_alloc_admin_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set, const struct blk_mq_ops *ops, unsigned int cmd_size) { + struct queue_limits lim = {}; int ret; memset(set, 0, sizeof(*set)); set->ops = ops; set->queue_depth = NVME_AQ_MQ_TAG_DEPTH; if (ctrl->ops->flags & NVME_F_FABRICS) - set->reserved_tags = NVMF_RESERVED_TAGS; + /* Reserved for fabric connect and keep alive */ + set->reserved_tags = 2; set->numa_node = ctrl->numa_node; - set->flags = BLK_MQ_F_NO_SCHED; if (ctrl->ops->flags & NVME_F_BLOCKING) set->flags |= BLK_MQ_F_BLOCKING; set->cmd_size = cmd_size; @@ -4372,14 +4610,14 @@ int nvme_alloc_admin_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set, if (ret) return ret; - ctrl->admin_q = blk_mq_init_queue(set); + ctrl->admin_q = blk_mq_alloc_queue(set, &lim, NULL); if (IS_ERR(ctrl->admin_q)) { ret = PTR_ERR(ctrl->admin_q); goto out_free_tagset; } if (ctrl->ops->flags & NVME_F_FABRICS) { - ctrl->fabrics_q = blk_mq_init_queue(set); + ctrl->fabrics_q = blk_mq_alloc_queue(set, NULL, NULL); if (IS_ERR(ctrl->fabrics_q)) { ret = PTR_ERR(ctrl->fabrics_q); goto out_cleanup_admin_q; @@ -4402,6 +4640,11 @@ EXPORT_SYMBOL_GPL(nvme_alloc_admin_tag_set); void nvme_remove_admin_tag_set(struct nvme_ctrl *ctrl) { + /* + * As we're about to destroy the queue and free tagset + * we can not have keep-alive work running. + */ + nvme_stop_keep_alive(ctrl); blk_mq_destroy_queue(ctrl->admin_q); blk_put_queue(ctrl->admin_q); if (ctrl->ops->flags & NVME_F_FABRICS) { @@ -4428,12 +4671,12 @@ int nvme_alloc_io_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set, if (ctrl->quirks & NVME_QUIRK_SHARED_TAGS) set->reserved_tags = NVME_AQ_DEPTH; else if (ctrl->ops->flags & NVME_F_FABRICS) - set->reserved_tags = NVMF_RESERVED_TAGS; + /* Reserved for fabric connect */ + set->reserved_tags = 1; set->numa_node = ctrl->numa_node; - set->flags = BLK_MQ_F_SHOULD_MERGE; if (ctrl->ops->flags & NVME_F_BLOCKING) set->flags |= BLK_MQ_F_BLOCKING; - set->cmd_size = cmd_size, + set->cmd_size = cmd_size; set->driver_data = ctrl; set->nr_hw_queues = ctrl->queue_count - 1; set->timeout = NVME_IO_TIMEOUT; @@ -4443,13 +4686,15 @@ int nvme_alloc_io_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set, return ret; if (ctrl->ops->flags & NVME_F_FABRICS) { - ctrl->connect_q = blk_mq_init_queue(set); + struct queue_limits lim = { + .features = BLK_FEAT_SKIP_TAGSET_QUIESCE, + }; + + ctrl->connect_q = blk_mq_alloc_queue(set, &lim, NULL); if (IS_ERR(ctrl->connect_q)) { ret = PTR_ERR(ctrl->connect_q); goto out_free_tag_set; } - blk_queue_flag_set(QUEUE_FLAG_SKIP_TAGSET_QUIESCE, - ctrl->connect_q); } ctrl->tagset = set; @@ -4476,7 +4721,6 @@ void nvme_stop_ctrl(struct nvme_ctrl *ctrl) { nvme_mpath_stop(ctrl); nvme_auth_stop(ctrl); - nvme_stop_keep_alive(ctrl); nvme_stop_failfast_work(ctrl); flush_work(&ctrl->async_event_work); cancel_work_sync(&ctrl->fw_act_work); @@ -4512,6 +4756,7 @@ EXPORT_SYMBOL_GPL(nvme_start_ctrl); void nvme_uninit_ctrl(struct nvme_ctrl *ctrl) { + nvme_stop_keep_alive(ctrl); nvme_hwmon_exit(ctrl); nvme_fault_inject_fini(&ctrl->fault_inject); dev_pm_qos_hide_latency_tolerance(ctrl->device); @@ -4541,9 +4786,9 @@ static void nvme_free_ctrl(struct device *dev) if (!subsys || ctrl->instance != subsys->instance) ida_free(&nvme_instance_ida, ctrl->instance); - key_put(ctrl->tls_key); nvme_free_cels(ctrl); nvme_mpath_uninit(ctrl); + cleanup_srcu_struct(&ctrl->srcu); nvme_auth_stop(ctrl); nvme_auth_free(ctrl); __free_page(ctrl->discard_page); @@ -4566,6 +4811,9 @@ static void nvme_free_ctrl(struct device *dev) * Initialize a NVMe controller structures. This needs to be called during * earliest initialization so that we have the initialized structured around * during probing. + * + * On success, the caller must use the nvme_put_ctrl() to release this when + * needed, which also invokes the ops->free_ctrl() callback. */ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, const struct nvme_ctrl_ops *ops, unsigned long quirks) @@ -4576,10 +4824,15 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, ctrl->passthru_err_log_enabled = false; clear_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags); spin_lock_init(&ctrl->lock); + mutex_init(&ctrl->namespaces_lock); + + ret = init_srcu_struct(&ctrl->srcu); + if (ret) + return ret; + mutex_init(&ctrl->scan_lock); INIT_LIST_HEAD(&ctrl->namespaces); xa_init(&ctrl->cels); - init_rwsem(&ctrl->namespaces_rwsem); ctrl->dev = dev; ctrl->ops = ops; ctrl->quirks = quirks; @@ -4609,11 +4862,17 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, goto out; ctrl->instance = ret; + ret = nvme_auth_init_ctrl(ctrl); + if (ret) + goto out_release_instance; + + nvme_mpath_init_ctrl(ctrl); + device_initialize(&ctrl->ctrl_device); ctrl->device = &ctrl->ctrl_device; ctrl->device->devt = MKDEV(MAJOR(nvme_ctrl_base_chr_devt), ctrl->instance); - ctrl->device->class = nvme_class; + ctrl->device->class = &nvme_class; ctrl->device->parent = ctrl->dev; if (ops->dev_attr_groups) ctrl->device->groups = ops->dev_attr_groups; @@ -4621,16 +4880,36 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, ctrl->device->groups = nvme_dev_attr_groups; ctrl->device->release = nvme_free_ctrl; dev_set_drvdata(ctrl->device, ctrl); + + return ret; + +out_release_instance: + ida_free(&nvme_instance_ida, ctrl->instance); +out: + if (ctrl->discard_page) + __free_page(ctrl->discard_page); + cleanup_srcu_struct(&ctrl->srcu); + return ret; +} +EXPORT_SYMBOL_GPL(nvme_init_ctrl); + +/* + * On success, returns with an elevated controller reference and caller must + * use nvme_uninit_ctrl() to properly free resources associated with the ctrl. + */ +int nvme_add_ctrl(struct nvme_ctrl *ctrl) +{ + int ret; + ret = dev_set_name(ctrl->device, "nvme%d", ctrl->instance); if (ret) - goto out_release_instance; + return ret; - nvme_get_ctrl(ctrl); cdev_init(&ctrl->cdev, &nvme_dev_fops); - ctrl->cdev.owner = ops->module; + ctrl->cdev.owner = ctrl->ops->module; ret = cdev_device_add(&ctrl->cdev, ctrl->device); if (ret) - goto out_free_name; + return ret; /* * Initialize latency tolerance controls. The sysfs files won't @@ -4641,48 +4920,36 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, min(default_ps_max_latency_us, (unsigned long)S32_MAX)); nvme_fault_inject_init(&ctrl->fault_inject, dev_name(ctrl->device)); - nvme_mpath_init_ctrl(ctrl); - ret = nvme_auth_init_ctrl(ctrl); - if (ret) - goto out_free_cdev; + nvme_get_ctrl(ctrl); return 0; -out_free_cdev: - nvme_fault_inject_fini(&ctrl->fault_inject); - dev_pm_qos_hide_latency_tolerance(ctrl->device); - cdev_device_del(&ctrl->cdev, ctrl->device); -out_free_name: - nvme_put_ctrl(ctrl); - kfree_const(ctrl->device->kobj.name); -out_release_instance: - ida_free(&nvme_instance_ida, ctrl->instance); -out: - if (ctrl->discard_page) - __free_page(ctrl->discard_page); - return ret; } -EXPORT_SYMBOL_GPL(nvme_init_ctrl); +EXPORT_SYMBOL_GPL(nvme_add_ctrl); /* let I/O to all namespaces fail in preparation for surprise removal */ void nvme_mark_namespaces_dead(struct nvme_ctrl *ctrl) { struct nvme_ns *ns; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) blk_mark_disk_dead(ns->disk); - up_read(&ctrl->namespaces_rwsem); + srcu_read_unlock(&ctrl->srcu, srcu_idx); } EXPORT_SYMBOL_GPL(nvme_mark_namespaces_dead); void nvme_unfreeze(struct nvme_ctrl *ctrl) { struct nvme_ns *ns; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) - blk_mq_unfreeze_queue(ns->queue); - up_read(&ctrl->namespaces_rwsem); + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) + blk_mq_unfreeze_queue_non_owner(ns->queue); + srcu_read_unlock(&ctrl->srcu, srcu_idx); clear_bit(NVME_CTRL_FROZEN, &ctrl->flags); } EXPORT_SYMBOL_GPL(nvme_unfreeze); @@ -4690,14 +4957,16 @@ EXPORT_SYMBOL_GPL(nvme_unfreeze); int nvme_wait_freeze_timeout(struct nvme_ctrl *ctrl, long timeout) { struct nvme_ns *ns; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) { + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) { timeout = blk_mq_freeze_queue_wait_timeout(ns->queue, timeout); if (timeout <= 0) break; } - up_read(&ctrl->namespaces_rwsem); + srcu_read_unlock(&ctrl->srcu, srcu_idx); return timeout; } EXPORT_SYMBOL_GPL(nvme_wait_freeze_timeout); @@ -4705,23 +4974,32 @@ EXPORT_SYMBOL_GPL(nvme_wait_freeze_timeout); void nvme_wait_freeze(struct nvme_ctrl *ctrl) { struct nvme_ns *ns; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) blk_mq_freeze_queue_wait(ns->queue); - up_read(&ctrl->namespaces_rwsem); + srcu_read_unlock(&ctrl->srcu, srcu_idx); } EXPORT_SYMBOL_GPL(nvme_wait_freeze); void nvme_start_freeze(struct nvme_ctrl *ctrl) { struct nvme_ns *ns; + int srcu_idx; set_bit(NVME_CTRL_FROZEN, &ctrl->flags); - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) - blk_freeze_queue_start(ns->queue); - up_read(&ctrl->namespaces_rwsem); + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) + /* + * Typical non_owner use case is from pci driver, in which + * start_freeze is called from timeout work function, but + * unfreeze is done in reset work context + */ + blk_freeze_queue_start_non_owner(ns->queue); + srcu_read_unlock(&ctrl->srcu, srcu_idx); } EXPORT_SYMBOL_GPL(nvme_start_freeze); @@ -4764,11 +5042,13 @@ EXPORT_SYMBOL_GPL(nvme_unquiesce_admin_queue); void nvme_sync_io_queues(struct nvme_ctrl *ctrl) { struct nvme_ns *ns; + int srcu_idx; - down_read(&ctrl->namespaces_rwsem); - list_for_each_entry(ns, &ctrl->namespaces, list) + srcu_idx = srcu_read_lock(&ctrl->srcu); + list_for_each_entry_srcu(ns, &ctrl->namespaces, list, + srcu_read_lock_held(&ctrl->srcu)) blk_sync_queue(ns->queue); - up_read(&ctrl->namespaces_rwsem); + srcu_read_unlock(&ctrl->srcu, srcu_idx); } EXPORT_SYMBOL_GPL(nvme_sync_io_queues); @@ -4786,7 +5066,7 @@ struct nvme_ctrl *nvme_ctrl_from_file(struct file *file) return NULL; return file->private_data; } -EXPORT_SYMBOL_NS_GPL(nvme_ctrl_from_file, NVME_TARGET_PASSTHRU); +EXPORT_SYMBOL_NS_GPL(nvme_ctrl_from_file, "NVME_TARGET_PASSTHRU"); /* * Check we didn't inadvertently grow the command structure sizes: @@ -4814,6 +5094,8 @@ static inline void _nvme_check_size(void) BUILD_BUG_ON(sizeof(struct nvme_id_ctrl_nvm) != NVME_IDENTIFY_DATA_SIZE); BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64); BUILD_BUG_ON(sizeof(struct nvme_smart_log) != 512); + BUILD_BUG_ON(sizeof(struct nvme_endurance_group_log) != 512); + BUILD_BUG_ON(sizeof(struct nvme_rotational_media_log) != 512); BUILD_BUG_ON(sizeof(struct nvme_dbbuf) != 64); BUILD_BUG_ON(sizeof(struct nvme_directive_cmd) != 64); BUILD_BUG_ON(sizeof(struct nvme_feat_host_behavior) != 512); @@ -4822,22 +5104,20 @@ static inline void _nvme_check_size(void) static int __init nvme_core_init(void) { + unsigned int wq_flags = WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS; int result = -ENOMEM; _nvme_check_size(); - nvme_wq = alloc_workqueue("nvme-wq", - WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + nvme_wq = alloc_workqueue("nvme-wq", wq_flags, 0); if (!nvme_wq) goto out; - nvme_reset_wq = alloc_workqueue("nvme-reset-wq", - WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + nvme_reset_wq = alloc_workqueue("nvme-reset-wq", wq_flags, 0); if (!nvme_reset_wq) goto destroy_wq; - nvme_delete_wq = alloc_workqueue("nvme-delete-wq", - WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + nvme_delete_wq = alloc_workqueue("nvme-delete-wq", wq_flags, 0); if (!nvme_delete_wq) goto destroy_reset_wq; @@ -4846,42 +5126,36 @@ static int __init nvme_core_init(void) if (result < 0) goto destroy_delete_wq; - nvme_class = class_create("nvme"); - if (IS_ERR(nvme_class)) { - result = PTR_ERR(nvme_class); + result = class_register(&nvme_class); + if (result) goto unregister_chrdev; - } - nvme_class->dev_uevent = nvme_class_uevent; - nvme_subsys_class = class_create("nvme-subsystem"); - if (IS_ERR(nvme_subsys_class)) { - result = PTR_ERR(nvme_subsys_class); + result = class_register(&nvme_subsys_class); + if (result) goto destroy_class; - } result = alloc_chrdev_region(&nvme_ns_chr_devt, 0, NVME_MINORS, "nvme-generic"); if (result < 0) goto destroy_subsys_class; - nvme_ns_chr_class = class_create("nvme-generic"); - if (IS_ERR(nvme_ns_chr_class)) { - result = PTR_ERR(nvme_ns_chr_class); + result = class_register(&nvme_ns_chr_class); + if (result) goto unregister_generic_ns; - } + result = nvme_init_auth(); if (result) goto destroy_ns_chr; return 0; destroy_ns_chr: - class_destroy(nvme_ns_chr_class); + class_unregister(&nvme_ns_chr_class); unregister_generic_ns: unregister_chrdev_region(nvme_ns_chr_devt, NVME_MINORS); destroy_subsys_class: - class_destroy(nvme_subsys_class); + class_unregister(&nvme_subsys_class); destroy_class: - class_destroy(nvme_class); + class_unregister(&nvme_class); unregister_chrdev: unregister_chrdev_region(nvme_ctrl_base_chr_devt, NVME_MINORS); destroy_delete_wq: @@ -4897,9 +5171,9 @@ out: static void __exit nvme_core_exit(void) { nvme_exit_auth(); - class_destroy(nvme_ns_chr_class); - class_destroy(nvme_subsys_class); - class_destroy(nvme_class); + class_unregister(&nvme_ns_chr_class); + class_unregister(&nvme_subsys_class); + class_unregister(&nvme_class); unregister_chrdev_region(nvme_ns_chr_devt, NVME_MINORS); unregister_chrdev_region(nvme_ctrl_base_chr_devt, NVME_MINORS); destroy_workqueue(nvme_delete_wq); |