Commit c786427f authored by Linus Torvalds's avatar Linus Torvalds
Browse files

Merge tag 'for-linus-20180217' of git://git.kernel.dk/linux-block 

Pull block fixes from Jens Axboe:

 - NVMe pull request from Keith, with fixes all over the map for nvme.
   From various folks.

 - Classic polling fix, that avoids a latency issue where we still end
   up waiting for an interrupt in some cases. From Nitesh Shetty.

 - Comment typo fix from Minwoo Im.

* tag 'for-linus-20180217' of git://git.kernel.dk/linux-block:
  block: fix a typo in comment of BLK_MQ_POLL_STATS_BKTS
  nvme-rdma: fix sysfs invoked reset_ctrl error flow
  nvmet: Change return code of discard command if not supported
  nvme-pci: Fix timeouts in connecting state
  nvme-pci: Remap CMB SQ entries on every controller reset
  nvme: fix the deadlock in nvme_update_formats
  blk: optimization for classic polling
  nvme: Don't use a stack buffer for keep-alive command
  nvme_fc: cleanup io completion
  nvme_fc: correct abort race condition on resets
  nvme: Fix discard buffer overrun
  nvme: delete NVME_CTRL_LIVE --> NVME_CTRL_CONNECTING transition
  nvme-rdma: use NVME_CTRL_CONNECTING state to mark init process
  nvme: rename NVME_CTRL_RECONNECTING state to NVME_CTRL_CONNECTING
parents fa2139ef 096392e0

block/blk-mq.c

+1 −0
Original line number Diff line number Diff line
@@ -3164,6 +3164,7 @@ static bool __blk_mq_poll(struct blk_mq_hw_ctx *hctx, struct request *rq) 
		cpu_relax();

	}



	__set_current_state(TASK_RUNNING);

	return false;

}

drivers/nvme/host/core.c

+27 −18
Original line number Diff line number Diff line
@@ -120,8 +120,12 @@ int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl) 
	int ret;



	ret = nvme_reset_ctrl(ctrl);

	if (!ret)

	if (!ret) {

		flush_work(&ctrl->reset_work);

		if (ctrl->state != NVME_CTRL_LIVE)

			ret = -ENETRESET;

	}



	return ret;

}

EXPORT_SYMBOL_GPL(nvme_reset_ctrl_sync);
@@ -265,7 +269,7 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, 
	switch (new_state) {

	case NVME_CTRL_ADMIN_ONLY:

		switch (old_state) {

		case NVME_CTRL_RECONNECTING:

		case NVME_CTRL_CONNECTING:

			changed = true;

			/* FALLTHRU */

		default:
@@ -276,7 +280,7 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, 
		switch (old_state) {

		case NVME_CTRL_NEW:

		case NVME_CTRL_RESETTING:

		case NVME_CTRL_RECONNECTING:

		case NVME_CTRL_CONNECTING:

			changed = true;

			/* FALLTHRU */

		default:
@@ -294,9 +298,9 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, 
			break;

		}

		break;

	case NVME_CTRL_RECONNECTING:

	case NVME_CTRL_CONNECTING:

		switch (old_state) {

		case NVME_CTRL_LIVE:

		case NVME_CTRL_NEW:

		case NVME_CTRL_RESETTING:

			changed = true;

			/* FALLTHRU */
@@ -309,7 +313,7 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, 
		case NVME_CTRL_LIVE:

		case NVME_CTRL_ADMIN_ONLY:

		case NVME_CTRL_RESETTING:

		case NVME_CTRL_RECONNECTING:

		case NVME_CTRL_CONNECTING:

			changed = true;

			/* FALLTHRU */

		default:
@@ -518,9 +522,11 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *ns, struct request *req, 
		u64 slba = nvme_block_nr(ns, bio->bi_iter.bi_sector);

		u32 nlb = bio->bi_iter.bi_size >> ns->lba_shift;



		if (n < segments) {

			range[n].cattr = cpu_to_le32(0);

			range[n].nlb = cpu_to_le32(nlb);

			range[n].slba = cpu_to_le64(slba);

		}

		n++;

	}
@@ -794,13 +800,9 @@ static void nvme_keep_alive_end_io(struct request *rq, blk_status_t status) 


static int nvme_keep_alive(struct nvme_ctrl *ctrl)

{

	struct nvme_command c;

	struct request *rq;



	memset(&c, 0, sizeof(c));

	c.common.opcode = nvme_admin_keep_alive;



	rq = nvme_alloc_request(ctrl->admin_q, &c, BLK_MQ_REQ_RESERVED,

	rq = nvme_alloc_request(ctrl->admin_q, &ctrl->ka_cmd, BLK_MQ_REQ_RESERVED,

			NVME_QID_ANY);

	if (IS_ERR(rq))

		return PTR_ERR(rq);
@@ -832,6 +834,8 @@ void nvme_start_keep_alive(struct nvme_ctrl *ctrl) 
		return;



	INIT_DELAYED_WORK(&ctrl->ka_work, nvme_keep_alive_work);

	memset(&ctrl->ka_cmd, 0, sizeof(ctrl->ka_cmd));

	ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive;

	schedule_delayed_work(&ctrl->ka_work, ctrl->kato * HZ);

}

EXPORT_SYMBOL_GPL(nvme_start_keep_alive);
@@ -1117,14 +1121,19 @@ static u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, 


static void nvme_update_formats(struct nvme_ctrl *ctrl)

{

	struct nvme_ns *ns;

	struct nvme_ns *ns, *next;

	LIST_HEAD(rm_list);



	mutex_lock(&ctrl->namespaces_mutex);

	list_for_each_entry(ns, &ctrl->namespaces, list) {

		if (ns->disk && nvme_revalidate_disk(ns->disk))

			nvme_ns_remove(ns);

		if (ns->disk && nvme_revalidate_disk(ns->disk)) {

			list_move_tail(&ns->list, &rm_list);

		}

	}

	mutex_unlock(&ctrl->namespaces_mutex);



	list_for_each_entry_safe(ns, next, &rm_list, list)

		nvme_ns_remove(ns);

}



static void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects)
@@ -2687,7 +2696,7 @@ static ssize_t nvme_sysfs_show_state(struct device *dev, 
		[NVME_CTRL_LIVE]	= "live",

		[NVME_CTRL_ADMIN_ONLY]	= "only-admin",

		[NVME_CTRL_RESETTING]	= "resetting",

		[NVME_CTRL_RECONNECTING]= "reconnecting",

		[NVME_CTRL_CONNECTING]	= "connecting",

		[NVME_CTRL_DELETING]	= "deleting",

		[NVME_CTRL_DEAD]	= "dead",

	};

drivers/nvme/host/fabrics.h

+5 −4
Original line number Diff line number Diff line
@@ -171,13 +171,14 @@ static inline blk_status_t nvmf_check_init_req(struct nvme_ctrl *ctrl, 
	    cmd->common.opcode != nvme_fabrics_command ||

	    cmd->fabrics.fctype != nvme_fabrics_type_connect) {

		/*

		 * Reconnecting state means transport disruption, which can take

		 * a long time and even might fail permanently, fail fast to

		 * give upper layers a chance to failover.

		 * Connecting state means transport disruption or initial

		 * establishment, which can take a long time and even might

		 * fail permanently, fail fast to give upper layers a chance

		 * to failover.

		 * Deleting state means that the ctrl will never accept commands

		 * again, fail it permanently.

		 */

		if (ctrl->state == NVME_CTRL_RECONNECTING ||

		if (ctrl->state == NVME_CTRL_CONNECTING ||

		    ctrl->state == NVME_CTRL_DELETING) {

			nvme_req(rq)->status = NVME_SC_ABORT_REQ;

			return BLK_STS_IOERR;

drivers/nvme/host/fc.c

+36 −121
Original line number Diff line number Diff line
@@ -55,9 +55,7 @@ struct nvme_fc_queue { 


enum nvme_fcop_flags {

	FCOP_FLAGS_TERMIO	= (1 << 0),

	FCOP_FLAGS_RELEASED	= (1 << 1),

	FCOP_FLAGS_COMPLETE	= (1 << 2),

	FCOP_FLAGS_AEN		= (1 << 3),

	FCOP_FLAGS_AEN		= (1 << 1),

};



struct nvmefc_ls_req_op {
@@ -532,7 +530,7 @@ nvme_fc_resume_controller(struct nvme_fc_ctrl *ctrl) 
{

	switch (ctrl->ctrl.state) {

	case NVME_CTRL_NEW:

	case NVME_CTRL_RECONNECTING:

	case NVME_CTRL_CONNECTING:

		/*

		 * As all reconnects were suppressed, schedule a

		 * connect.
@@ -777,7 +775,7 @@ nvme_fc_ctrl_connectivity_loss(struct nvme_fc_ctrl *ctrl) 
		}

		break;



	case NVME_CTRL_RECONNECTING:

	case NVME_CTRL_CONNECTING:

		/*

		 * The association has already been terminated and the

		 * controller is attempting reconnects.  No need to do anything
@@ -1470,7 +1468,6 @@ nvme_fc_xmt_disconnect_assoc(struct nvme_fc_ctrl *ctrl) 


/* *********************** NVME Ctrl Routines **************************** */



static void __nvme_fc_final_op_cleanup(struct request *rq);

static void nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg);



static int
@@ -1512,13 +1509,19 @@ nvme_fc_exit_request(struct blk_mq_tag_set *set, struct request *rq, 
static int

__nvme_fc_abort_op(struct nvme_fc_ctrl *ctrl, struct nvme_fc_fcp_op *op)

{

	int state;

	unsigned long flags;

	int opstate;



	spin_lock_irqsave(&ctrl->lock, flags);

	opstate = atomic_xchg(&op->state, FCPOP_STATE_ABORTED);

	if (opstate != FCPOP_STATE_ACTIVE)

		atomic_set(&op->state, opstate);

	else if (ctrl->flags & FCCTRL_TERMIO)

		ctrl->iocnt++;

	spin_unlock_irqrestore(&ctrl->lock, flags);



	state = atomic_xchg(&op->state, FCPOP_STATE_ABORTED);

	if (state != FCPOP_STATE_ACTIVE) {

		atomic_set(&op->state, state);

	if (opstate != FCPOP_STATE_ACTIVE)

		return -ECANCELED;

	}



	ctrl->lport->ops->fcp_abort(&ctrl->lport->localport,

					&ctrl->rport->remoteport,
@@ -1532,60 +1535,26 @@ static void 
nvme_fc_abort_aen_ops(struct nvme_fc_ctrl *ctrl)

{

	struct nvme_fc_fcp_op *aen_op = ctrl->aen_ops;

	unsigned long flags;

	int i, ret;



	for (i = 0; i < NVME_NR_AEN_COMMANDS; i++, aen_op++) {

		if (atomic_read(&aen_op->state) != FCPOP_STATE_ACTIVE)

			continue;



		spin_lock_irqsave(&ctrl->lock, flags);

		if (ctrl->flags & FCCTRL_TERMIO) {

			ctrl->iocnt++;

			aen_op->flags |= FCOP_FLAGS_TERMIO;

		}

		spin_unlock_irqrestore(&ctrl->lock, flags);



		ret = __nvme_fc_abort_op(ctrl, aen_op);

		if (ret) {

			/*

			 * if __nvme_fc_abort_op failed the io wasn't

			 * active. Thus this call path is running in

			 * parallel to the io complete. Treat as non-error.

			 */

	int i;



			/* back out the flags/counters */

			spin_lock_irqsave(&ctrl->lock, flags);

			if (ctrl->flags & FCCTRL_TERMIO)

				ctrl->iocnt--;

			aen_op->flags &= ~FCOP_FLAGS_TERMIO;

			spin_unlock_irqrestore(&ctrl->lock, flags);

			return;

		}

	}

	for (i = 0; i < NVME_NR_AEN_COMMANDS; i++, aen_op++)

		__nvme_fc_abort_op(ctrl, aen_op);

}



static inline int

static inline void

__nvme_fc_fcpop_chk_teardowns(struct nvme_fc_ctrl *ctrl,

		struct nvme_fc_fcp_op *op)

		struct nvme_fc_fcp_op *op, int opstate)

{

	unsigned long flags;

	bool complete_rq = false;



	if (opstate == FCPOP_STATE_ABORTED) {

		spin_lock_irqsave(&ctrl->lock, flags);

	if (unlikely(op->flags & FCOP_FLAGS_TERMIO)) {

		if (ctrl->flags & FCCTRL_TERMIO) {

			if (!--ctrl->iocnt)

				wake_up(&ctrl->ioabort_wait);

		}

	}

	if (op->flags & FCOP_FLAGS_RELEASED)

		complete_rq = true;

	else

		op->flags |= FCOP_FLAGS_COMPLETE;

		spin_unlock_irqrestore(&ctrl->lock, flags);



	return complete_rq;

	}

}



static void
@@ -1601,6 +1570,7 @@ nvme_fc_fcpio_done(struct nvmefc_fcp_req *req) 
	__le16 status = cpu_to_le16(NVME_SC_SUCCESS << 1);

	union nvme_result result;

	bool terminate_assoc = true;

	int opstate;



	/*

	 * WARNING:
@@ -1639,11 +1609,12 @@ nvme_fc_fcpio_done(struct nvmefc_fcp_req *req) 
	 * association to be terminated.

	 */



	opstate = atomic_xchg(&op->state, FCPOP_STATE_COMPLETE);



	fc_dma_sync_single_for_cpu(ctrl->lport->dev, op->fcp_req.rspdma,

				sizeof(op->rsp_iu), DMA_FROM_DEVICE);



	if (atomic_read(&op->state) == FCPOP_STATE_ABORTED ||

			op->flags & FCOP_FLAGS_TERMIO)

	if (opstate == FCPOP_STATE_ABORTED)

		status = cpu_to_le16(NVME_SC_ABORT_REQ << 1);

	else if (freq->status)

		status = cpu_to_le16(NVME_SC_INTERNAL << 1);
@@ -1708,7 +1679,7 @@ nvme_fc_fcpio_done(struct nvmefc_fcp_req *req) 
done:

	if (op->flags & FCOP_FLAGS_AEN) {

		nvme_complete_async_event(&queue->ctrl->ctrl, status, &result);

		__nvme_fc_fcpop_chk_teardowns(ctrl, op);

		__nvme_fc_fcpop_chk_teardowns(ctrl, op, opstate);

		atomic_set(&op->state, FCPOP_STATE_IDLE);

		op->flags = FCOP_FLAGS_AEN;	/* clear other flags */

		nvme_fc_ctrl_put(ctrl);
@@ -1722,12 +1693,10 @@ done: 
	if (status &&

	    (blk_queue_dying(rq->q) ||

	     ctrl->ctrl.state == NVME_CTRL_NEW ||

	     ctrl->ctrl.state == NVME_CTRL_RECONNECTING))

	     ctrl->ctrl.state == NVME_CTRL_CONNECTING))

		status |= cpu_to_le16(NVME_SC_DNR << 1);



	if (__nvme_fc_fcpop_chk_teardowns(ctrl, op))

		__nvme_fc_final_op_cleanup(rq);

	else

	__nvme_fc_fcpop_chk_teardowns(ctrl, op, opstate);

	nvme_end_request(rq, status, result);



check_error:
@@ -2415,46 +2384,16 @@ nvme_fc_submit_async_event(struct nvme_ctrl *arg) 
}



static void

__nvme_fc_final_op_cleanup(struct request *rq)

nvme_fc_complete_rq(struct request *rq)

{

	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(rq);

	struct nvme_fc_ctrl *ctrl = op->ctrl;



	atomic_set(&op->state, FCPOP_STATE_IDLE);

	op->flags &= ~(FCOP_FLAGS_TERMIO | FCOP_FLAGS_RELEASED |

			FCOP_FLAGS_COMPLETE);



	nvme_fc_unmap_data(ctrl, rq, op);

	nvme_complete_rq(rq);

	nvme_fc_ctrl_put(ctrl);



}



static void

nvme_fc_complete_rq(struct request *rq)

{

	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(rq);

	struct nvme_fc_ctrl *ctrl = op->ctrl;

	unsigned long flags;

	bool completed = false;



	/*

	 * the core layer, on controller resets after calling

	 * nvme_shutdown_ctrl(), calls complete_rq without our

	 * calling blk_mq_complete_request(), thus there may still

	 * be live i/o outstanding with the LLDD. Means transport has

	 * to track complete calls vs fcpio_done calls to know what

	 * path to take on completes and dones.

	 */

	spin_lock_irqsave(&ctrl->lock, flags);

	if (op->flags & FCOP_FLAGS_COMPLETE)

		completed = true;

	else

		op->flags |= FCOP_FLAGS_RELEASED;

	spin_unlock_irqrestore(&ctrl->lock, flags);



	if (completed)

		__nvme_fc_final_op_cleanup(rq);

}



/*
@@ -2476,35 +2415,11 @@ nvme_fc_terminate_exchange(struct request *req, void *data, bool reserved) 
	struct nvme_ctrl *nctrl = data;

	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);

	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(req);

	unsigned long flags;

	int status;



	if (!blk_mq_request_started(req))

		return;



	spin_lock_irqsave(&ctrl->lock, flags);

	if (ctrl->flags & FCCTRL_TERMIO) {

		ctrl->iocnt++;

		op->flags |= FCOP_FLAGS_TERMIO;

	}

	spin_unlock_irqrestore(&ctrl->lock, flags);



	status = __nvme_fc_abort_op(ctrl, op);

	if (status) {

		/*

		 * if __nvme_fc_abort_op failed the io wasn't

		 * active. Thus this call path is running in

		 * parallel to the io complete. Treat as non-error.

		 */



		/* back out the flags/counters */

		spin_lock_irqsave(&ctrl->lock, flags);

		if (ctrl->flags & FCCTRL_TERMIO)

			ctrl->iocnt--;

		op->flags &= ~FCOP_FLAGS_TERMIO;

		spin_unlock_irqrestore(&ctrl->lock, flags);

		return;

	}

	__nvme_fc_abort_op(ctrl, op);

}
@@ -2943,7 +2858,7 @@ nvme_fc_reconnect_or_delete(struct nvme_fc_ctrl *ctrl, int status) 
	unsigned long recon_delay = ctrl->ctrl.opts->reconnect_delay * HZ;

	bool recon = true;



	if (ctrl->ctrl.state != NVME_CTRL_RECONNECTING)

	if (ctrl->ctrl.state != NVME_CTRL_CONNECTING)

		return;



	if (portptr->port_state == FC_OBJSTATE_ONLINE)
@@ -2991,10 +2906,10 @@ nvme_fc_reset_ctrl_work(struct work_struct *work) 
	/* will block will waiting for io to terminate */

	nvme_fc_delete_association(ctrl);



	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_RECONNECTING)) {

	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING)) {

		dev_err(ctrl->ctrl.device,

			"NVME-FC{%d}: error_recovery: Couldn't change state "

			"to RECONNECTING\n", ctrl->cnum);

			"to CONNECTING\n", ctrl->cnum);

		return;

	}
@@ -3195,7 +3110,7 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts, 
	 * transport errors (frame drop, LS failure) inherently must kill

	 * the association. The transport is coded so that any command used

	 * to create the association (prior to a LIVE state transition

	 * while NEW or RECONNECTING) will fail if it completes in error or

	 * while NEW or CONNECTING) will fail if it completes in error or

	 * times out.

	 *

	 * As such: as the connect request was mostly likely due to a

drivers/nvme/host/nvme.h

+2 −1
Original line number Diff line number Diff line
@@ -123,7 +123,7 @@ enum nvme_ctrl_state { 
	NVME_CTRL_LIVE,

	NVME_CTRL_ADMIN_ONLY,    /* Only admin queue live */

	NVME_CTRL_RESETTING,

	NVME_CTRL_RECONNECTING,

	NVME_CTRL_CONNECTING,

	NVME_CTRL_DELETING,

	NVME_CTRL_DEAD,

};
@@ -183,6 +183,7 @@ struct nvme_ctrl { 
	struct work_struct scan_work;

	struct work_struct async_event_work;

	struct delayed_work ka_work;

	struct nvme_command ka_cmd;

	struct work_struct fw_act_work;



	/* Power saving configuration */

CRA Git | Maintained and supported by SUSTech CRA and CCSE