Commit d86c4d8e authored by Christoph Hellwig's avatar Christoph Hellwig
Browse files

nvme: move reset workqueue handling to common code 



This moves the nvme_reset function from the PCIe driver to common code,
renaming it to nvme_reset_ctrl in the process.  Additionally a new
helper nvme_reset_ctrl_sync is added for the case where we want to
wait for the reset.  To facilitate that the reset_work work structure is
move to the common nvme_ctrl structure and the ->reset_ctrl method is
removed.  For now the drivers initialize the reset_work with their own
callback, but longer term we should move to callouts for specific
parts of the reset process and move even more code to the core.

Signed-off-by: default avatarChristoph Hellwig <hch@lst.de>
Reviewed-by: default avatarSagi Grimberg <sagi@grimberg.me>
parent 0350815a

drivers/nvme/host/core.c

+23 −3
Original line number Diff line number Diff line
@@ -73,6 +73,26 @@ static DEFINE_SPINLOCK(dev_list_lock); 


static struct class *nvme_class;



int nvme_reset_ctrl(struct nvme_ctrl *ctrl)

{

	if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_RESETTING))

		return -EBUSY;

	if (!queue_work(nvme_wq, &ctrl->reset_work))

		return -EBUSY;

	return 0;

}

EXPORT_SYMBOL_GPL(nvme_reset_ctrl);



static int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl)

{

	int ret;



	ret = nvme_reset_ctrl(ctrl);

	if (!ret)

		flush_work(&ctrl->reset_work);

	return ret;

}



static blk_status_t nvme_error_status(struct request *req)

{

	switch (nvme_req(req)->status & 0x7ff) {
@@ -604,7 +624,7 @@ static void nvme_keep_alive_work(struct work_struct *work) 
	if (nvme_keep_alive(ctrl)) {

		/* allocation failure, reset the controller */

		dev_err(ctrl->device, "keep-alive failed\n");

		ctrl->ops->reset_ctrl(ctrl);

		nvme_reset_ctrl_sync(ctrl);

		return;

	}

}
@@ -1821,7 +1841,7 @@ static long nvme_dev_ioctl(struct file *file, unsigned int cmd, 
		return nvme_dev_user_cmd(ctrl, argp);

	case NVME_IOCTL_RESET:

		dev_warn(ctrl->device, "resetting controller\n");

		return ctrl->ops->reset_ctrl(ctrl);

		return nvme_reset_ctrl_sync(ctrl);

	case NVME_IOCTL_SUBSYS_RESET:

		return nvme_reset_subsystem(ctrl);

	case NVME_IOCTL_RESCAN:
@@ -1847,7 +1867,7 @@ static ssize_t nvme_sysfs_reset(struct device *dev, 
	struct nvme_ctrl *ctrl = dev_get_drvdata(dev);

	int ret;



	ret = ctrl->ops->reset_ctrl(ctrl);

	ret = nvme_reset_ctrl_sync(ctrl);

	if (ret < 0)

		return ret;

	return count;

drivers/nvme/host/fc.c

+4 −32
Original line number Diff line number Diff line
@@ -161,7 +161,6 @@ struct nvme_fc_ctrl { 
	struct blk_mq_tag_set	tag_set;



	struct work_struct	delete_work;

	struct work_struct	reset_work;

	struct delayed_work	connect_work;



	struct kref		ref;
@@ -1764,10 +1763,7 @@ nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg) 
		return;

	}



	if (!queue_work(nvme_wq, &ctrl->reset_work))

		dev_err(ctrl->ctrl.device,

			"NVME-FC{%d}: error_recovery: Failed to schedule "

			"reset work\n", ctrl->cnum);

	nvme_reset_ctrl(&ctrl->ctrl);

}



static enum blk_eh_timer_return
@@ -2517,7 +2513,7 @@ nvme_fc_delete_ctrl_work(struct work_struct *work) 
	struct nvme_fc_ctrl *ctrl =

		container_of(work, struct nvme_fc_ctrl, delete_work);



	cancel_work_sync(&ctrl->reset_work);

	cancel_work_sync(&ctrl->ctrl.reset_work);

	cancel_delayed_work_sync(&ctrl->connect_work);



	/*
@@ -2611,7 +2607,7 @@ static void 
nvme_fc_reset_ctrl_work(struct work_struct *work)

{

	struct nvme_fc_ctrl *ctrl =

			container_of(work, struct nvme_fc_ctrl, reset_work);

		container_of(work, struct nvme_fc_ctrl, ctrl.reset_work);

	int ret;



	/* will block will waiting for io to terminate */
@@ -2625,29 +2621,6 @@ nvme_fc_reset_ctrl_work(struct work_struct *work) 
			"NVME-FC{%d}: controller reset complete\n", ctrl->cnum);

}



/*

 * called by the nvme core layer, for sysfs interface that requests

 * a reset of the nvme controller

 */

static int

nvme_fc_reset_nvme_ctrl(struct nvme_ctrl *nctrl)

{

	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);



	dev_info(ctrl->ctrl.device,

		"NVME-FC{%d}: admin requested controller reset\n", ctrl->cnum);



	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_RESETTING))

		return -EBUSY;



	if (!queue_work(nvme_wq, &ctrl->reset_work))

		return -EBUSY;



	flush_work(&ctrl->reset_work);



	return 0;

}



static const struct nvme_ctrl_ops nvme_fc_ctrl_ops = {

	.name			= "fc",

	.module			= THIS_MODULE,
@@ -2655,7 +2628,6 @@ static const struct nvme_ctrl_ops nvme_fc_ctrl_ops = { 
	.reg_read32		= nvmf_reg_read32,

	.reg_read64		= nvmf_reg_read64,

	.reg_write32		= nvmf_reg_write32,

	.reset_ctrl		= nvme_fc_reset_nvme_ctrl,

	.free_ctrl		= nvme_fc_nvme_ctrl_freed,

	.submit_async_event	= nvme_fc_submit_async_event,

	.delete_ctrl		= nvme_fc_del_nvme_ctrl,
@@ -2730,7 +2702,7 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts, 
	kref_init(&ctrl->ref);



	INIT_WORK(&ctrl->delete_work, nvme_fc_delete_ctrl_work);

	INIT_WORK(&ctrl->reset_work, nvme_fc_reset_ctrl_work);

	INIT_WORK(&ctrl->ctrl.reset_work, nvme_fc_reset_ctrl_work);

	INIT_DELAYED_WORK(&ctrl->connect_work, nvme_fc_connect_ctrl_work);

	spin_lock_init(&ctrl->lock);

drivers/nvme/host/nvme.h

+2 −1
Original line number Diff line number Diff line
@@ -130,6 +130,7 @@ struct nvme_ctrl { 
	struct device *device;	/* char device */

	struct list_head node;

	struct ida ns_ida;

	struct work_struct reset_work;



	struct opal_dev *opal_dev;
@@ -218,7 +219,6 @@ struct nvme_ctrl_ops { 
	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);

	int (*reset_ctrl)(struct nvme_ctrl *ctrl);

	void (*free_ctrl)(struct nvme_ctrl *ctrl);

	void (*submit_async_event)(struct nvme_ctrl *ctrl, int aer_idx);

	int (*delete_ctrl)(struct nvme_ctrl *ctrl);
@@ -325,6 +325,7 @@ int nvme_set_features(struct nvme_ctrl *dev, unsigned fid, unsigned dword11, 
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);



struct sg_io_hdr;

drivers/nvme/host/pci.c

+11 −34
Original line number Diff line number Diff line
@@ -61,7 +61,6 @@ MODULE_PARM_DESC(max_host_mem_size_mb, 
struct nvme_dev;

struct nvme_queue;



static int nvme_reset(struct nvme_dev *dev);

static void nvme_process_cq(struct nvme_queue *nvmeq);

static void nvme_dev_disable(struct nvme_dev *dev, bool shutdown);
@@ -83,7 +82,6 @@ struct nvme_dev { 
	u32 db_stride;

	void __iomem *bar;

	unsigned long bar_mapped_size;

	struct work_struct reset_work;

	struct work_struct remove_work;

	struct mutex shutdown_lock;

	bool subsystem;
@@ -983,7 +981,7 @@ static enum blk_eh_timer_return nvme_timeout(struct request *req, bool reserved) 
	if (nvme_should_reset(dev, csts)) {

		nvme_warn_reset(dev, csts);

		nvme_dev_disable(dev, false);

		nvme_reset(dev);

		nvme_reset_ctrl(&dev->ctrl);

		return BLK_EH_HANDLED;

	}
@@ -1022,7 +1020,7 @@ static enum blk_eh_timer_return nvme_timeout(struct request *req, bool reserved) 
			 "I/O %d QID %d timeout, reset controller\n",

			 req->tag, nvmeq->qid);

		nvme_dev_disable(dev, false);

		nvme_reset(dev);

		nvme_reset_ctrl(&dev->ctrl);



		/*

		 * Mark the request as handled, since the inline shutdown
@@ -2055,7 +2053,8 @@ static void nvme_remove_dead_ctrl(struct nvme_dev *dev, int status) 


static void nvme_reset_work(struct work_struct *work)

{

	struct nvme_dev *dev = container_of(work, struct nvme_dev, reset_work);

	struct nvme_dev *dev =

		container_of(work, struct nvme_dev, ctrl.reset_work);

	bool was_suspend = !!(dev->ctrl.ctrl_config & NVME_CC_SHN_NORMAL);

	int result = -ENODEV;
@@ -2159,17 +2158,6 @@ static void nvme_remove_dead_ctrl_work(struct work_struct *work) 
	nvme_put_ctrl(&dev->ctrl);

}



static int nvme_reset(struct nvme_dev *dev)

{

	if (!dev->ctrl.admin_q || blk_queue_dying(dev->ctrl.admin_q))

		return -ENODEV;

	if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING))

		return -EBUSY;

	if (!queue_work(nvme_wq, &dev->reset_work))

		return -EBUSY;

	return 0;

}



static int nvme_pci_reg_read32(struct nvme_ctrl *ctrl, u32 off, u32 *val)

{

	*val = readl(to_nvme_dev(ctrl)->bar + off);
@@ -2188,16 +2176,6 @@ static int nvme_pci_reg_read64(struct nvme_ctrl *ctrl, u32 off, u64 *val) 
	return 0;

}



static int nvme_pci_reset_ctrl(struct nvme_ctrl *ctrl)

{

	struct nvme_dev *dev = to_nvme_dev(ctrl);

	int ret = nvme_reset(dev);



	if (!ret)

		flush_work(&dev->reset_work);

	return ret;

}



static const struct nvme_ctrl_ops nvme_pci_ctrl_ops = {

	.name			= "pcie",

	.module			= THIS_MODULE,
@@ -2205,7 +2183,6 @@ static const struct nvme_ctrl_ops nvme_pci_ctrl_ops = { 
	.reg_read32		= nvme_pci_reg_read32,

	.reg_write32		= nvme_pci_reg_write32,

	.reg_read64		= nvme_pci_reg_read64,

	.reset_ctrl		= nvme_pci_reset_ctrl,

	.free_ctrl		= nvme_pci_free_ctrl,

	.submit_async_event	= nvme_pci_submit_async_event,

};
@@ -2271,7 +2248,7 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) 
	if (result)

		goto free;



	INIT_WORK(&dev->reset_work, nvme_reset_work);

	INIT_WORK(&dev->ctrl.reset_work, nvme_reset_work);

	INIT_WORK(&dev->remove_work, nvme_remove_dead_ctrl_work);

	mutex_init(&dev->shutdown_lock);

	init_completion(&dev->ioq_wait);
@@ -2290,7 +2267,7 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) 
	nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING);

	dev_info(dev->ctrl.device, "pci function %s\n", dev_name(&pdev->dev));



	queue_work(nvme_wq, &dev->reset_work);

	queue_work(nvme_wq, &dev->ctrl.reset_work);

	return 0;



 release_pools:
@@ -2311,7 +2288,7 @@ static void nvme_reset_notify(struct pci_dev *pdev, bool prepare) 
	if (prepare)

		nvme_dev_disable(dev, false);

	else

		nvme_reset(dev);

		nvme_reset_ctrl(&dev->ctrl);

}



static void nvme_shutdown(struct pci_dev *pdev)
@@ -2331,7 +2308,7 @@ static void nvme_remove(struct pci_dev *pdev) 


	nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_DELETING);



	cancel_work_sync(&dev->reset_work);

	cancel_work_sync(&dev->ctrl.reset_work);

	pci_set_drvdata(pdev, NULL);



	if (!pci_device_is_present(pdev)) {
@@ -2339,7 +2316,7 @@ static void nvme_remove(struct pci_dev *pdev) 
		nvme_dev_disable(dev, false);

	}



	flush_work(&dev->reset_work);

	flush_work(&dev->ctrl.reset_work);

	nvme_uninit_ctrl(&dev->ctrl);

	nvme_dev_disable(dev, true);

	nvme_free_host_mem(dev);
@@ -2383,7 +2360,7 @@ static int nvme_resume(struct device *dev) 
	struct pci_dev *pdev = to_pci_dev(dev);

	struct nvme_dev *ndev = pci_get_drvdata(pdev);



	nvme_reset(ndev);

	nvme_reset_ctrl(&ndev->ctrl);

	return 0;

}

#endif
@@ -2422,7 +2399,7 @@ static pci_ers_result_t nvme_slot_reset(struct pci_dev *pdev) 


	dev_info(dev->ctrl.device, "restart after slot reset\n");

	pci_restore_state(pdev);

	nvme_reset(dev);

	nvme_reset_ctrl(&dev->ctrl);

	return PCI_ERS_RESULT_RECOVERED;

}

drivers/nvme/host/rdma.c

+3 −20
Original line number Diff line number Diff line
@@ -108,7 +108,6 @@ struct nvme_rdma_ctrl { 
	/* other member variables */

	struct blk_mq_tag_set	tag_set;

	struct work_struct	delete_work;

	struct work_struct	reset_work;

	struct work_struct	err_work;



	struct nvme_rdma_qe	async_event_sqe;
@@ -1703,8 +1702,8 @@ static void nvme_rdma_remove_ctrl_work(struct work_struct *work) 


static void nvme_rdma_reset_ctrl_work(struct work_struct *work)

{

	struct nvme_rdma_ctrl *ctrl = container_of(work,

					struct nvme_rdma_ctrl, reset_work);

	struct nvme_rdma_ctrl *ctrl =

		container_of(work, struct nvme_rdma_ctrl, ctrl.reset_work);

	int ret;

	bool changed;
@@ -1748,21 +1747,6 @@ del_dead_ctrl: 
	WARN_ON(!queue_work(nvme_wq, &ctrl->delete_work));

}



static int nvme_rdma_reset_ctrl(struct nvme_ctrl *nctrl)

{

	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);



	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_RESETTING))

		return -EBUSY;



	if (!queue_work(nvme_wq, &ctrl->reset_work))

		return -EBUSY;



	flush_work(&ctrl->reset_work);



	return 0;

}



static const struct nvme_ctrl_ops nvme_rdma_ctrl_ops = {

	.name			= "rdma",

	.module			= THIS_MODULE,
@@ -1770,7 +1754,6 @@ static const struct nvme_ctrl_ops nvme_rdma_ctrl_ops = { 
	.reg_read32		= nvmf_reg_read32,

	.reg_read64		= nvmf_reg_read64,

	.reg_write32		= nvmf_reg_write32,

	.reset_ctrl		= nvme_rdma_reset_ctrl,

	.free_ctrl		= nvme_rdma_free_ctrl,

	.submit_async_event	= nvme_rdma_submit_async_event,

	.delete_ctrl		= nvme_rdma_del_ctrl,
@@ -1879,7 +1862,7 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev, 
			nvme_rdma_reconnect_ctrl_work);

	INIT_WORK(&ctrl->err_work, nvme_rdma_error_recovery_work);

	INIT_WORK(&ctrl->delete_work, nvme_rdma_del_ctrl_work);

	INIT_WORK(&ctrl->reset_work, nvme_rdma_reset_ctrl_work);

	INIT_WORK(&ctrl->ctrl.reset_work, nvme_rdma_reset_ctrl_work);



	ctrl->queue_count = opts->nr_io_queues + 1; /* +1 for admin queue */

	ctrl->ctrl.sqsize = opts->queue_size - 1;

CRA Git | Maintained and supported by SUSTech CRA and CCSE