ib_srp.c

linux内核源码

/*
 * Target ports are added by writing
 *
 *     id_ext=<SRP ID ext>,ioc_guid=<SRP IOC GUID>,dgid=<dest GID>,
 *     pkey=<P_Key>,service_id=<service ID>
 *
 * to the add_target sysfs attribute.
 */
enum {
	SRP_OPT_ERR		= 0,
	SRP_OPT_ID_EXT		= 1 << 0,
	SRP_OPT_IOC_GUID	= 1 << 1,
	SRP_OPT_DGID		= 1 << 2,
	SRP_OPT_PKEY		= 1 << 3,
	SRP_OPT_SERVICE_ID	= 1 << 4,
	SRP_OPT_MAX_SECT	= 1 << 5,
	SRP_OPT_MAX_CMD_PER_LUN	= 1 << 6,
	SRP_OPT_IO_CLASS	= 1 << 7,
	SRP_OPT_INITIATOR_EXT	= 1 << 8,
	SRP_OPT_ALL		= (SRP_OPT_ID_EXT	|
				   SRP_OPT_IOC_GUID	|
				   SRP_OPT_DGID		|
				   SRP_OPT_PKEY		|
				   SRP_OPT_SERVICE_ID),
};

static match_table_t srp_opt_tokens = {
	{ SRP_OPT_ID_EXT,		"id_ext=%s" 		},
	{ SRP_OPT_IOC_GUID,		"ioc_guid=%s" 		},
	{ SRP_OPT_DGID,			"dgid=%s" 		},
	{ SRP_OPT_PKEY,			"pkey=%x" 		},
	{ SRP_OPT_SERVICE_ID,		"service_id=%s"		},
	{ SRP_OPT_MAX_SECT,		"max_sect=%d" 		},
	{ SRP_OPT_MAX_CMD_PER_LUN,	"max_cmd_per_lun=%d" 	},
	{ SRP_OPT_IO_CLASS,		"io_class=%x"		},
	{ SRP_OPT_INITIATOR_EXT,	"initiator_ext=%s"	},
	{ SRP_OPT_ERR,			NULL 			}
};

static int srp_parse_options(const char *buf, struct srp_target_port *target)
{
	char *options, *sep_opt;
	char *p;
	char dgid[3];
	substring_t args[MAX_OPT_ARGS];
	int opt_mask = 0;
	int token;
	int ret = -EINVAL;
	int i;

	options = kstrdup(buf, GFP_KERNEL);
	if (!options)
		return -ENOMEM;

	sep_opt = options;
	while ((p = strsep(&sep_opt, ",")) != NULL) {
		if (!*p)
			continue;

		token = match_token(p, srp_opt_tokens, args);
		opt_mask |= token;

		switch (token) {
		case SRP_OPT_ID_EXT:
			p = match_strdup(args);
			if (!p) {
				ret = -ENOMEM;
				goto out;
			}
			target->id_ext = cpu_to_be64(simple_strtoull(p, NULL, 16));
			kfree(p);
			break;

		case SRP_OPT_IOC_GUID:
			p = match_strdup(args);
			if (!p) {
				ret = -ENOMEM;
				goto out;
			}
			target->ioc_guid = cpu_to_be64(simple_strtoull(p, NULL, 16));
			kfree(p);
			break;

		case SRP_OPT_DGID:
			p = match_strdup(args);
			if (!p) {
				ret = -ENOMEM;
				goto out;
			}
			if (strlen(p) != 32) {
				printk(KERN_WARNING PFX "bad dest GID parameter '%s'\n", p);
				kfree(p);
				goto out;
			}

			for (i = 0; i < 16; ++i) {
				strlcpy(dgid, p + i * 2, 3);
				target->path.dgid.raw[i] = simple_strtoul(dgid, NULL, 16);
			}
			kfree(p);
			memcpy(target->orig_dgid, target->path.dgid.raw, 16);
			break;

		case SRP_OPT_PKEY:
			if (match_hex(args, &token)) {
				printk(KERN_WARNING PFX "bad P_Key parameter '%s'\n", p);
				goto out;
			}
			target->path.pkey = cpu_to_be16(token);
			break;

		case SRP_OPT_SERVICE_ID:
			p = match_strdup(args);
			if (!p) {
				ret = -ENOMEM;
				goto out;
			}
			target->service_id = cpu_to_be64(simple_strtoull(p, NULL, 16));
			target->path.service_id = target->service_id;
			kfree(p);
			break;

		case SRP_OPT_MAX_SECT:
			if (match_int(args, &token)) {
				printk(KERN_WARNING PFX "bad max sect parameter '%s'\n", p);
				goto out;
			}
			target->scsi_host->max_sectors = token;
			break;

		case SRP_OPT_MAX_CMD_PER_LUN:
			if (match_int(args, &token)) {
				printk(KERN_WARNING PFX "bad max cmd_per_lun parameter '%s'\n", p);
				goto out;
			}
			target->scsi_host->cmd_per_lun = min(token, SRP_SQ_SIZE);
			break;

		case SRP_OPT_IO_CLASS:
			if (match_hex(args, &token)) {
				printk(KERN_WARNING PFX "bad  IO class parameter '%s' \n", p);
				goto out;
			}
			if (token != SRP_REV10_IB_IO_CLASS &&
			    token != SRP_REV16A_IB_IO_CLASS) {
				printk(KERN_WARNING PFX "unknown IO class parameter value"
				       " %x specified (use %x or %x).\n",
				       token, SRP_REV10_IB_IO_CLASS, SRP_REV16A_IB_IO_CLASS);
				goto out;
			}
			target->io_class = token;
			break;

		case SRP_OPT_INITIATOR_EXT:
			p = match_strdup(args);
			if (!p) {
				ret = -ENOMEM;
				goto out;
			}
			target->initiator_ext = cpu_to_be64(simple_strtoull(p, NULL, 16));
			kfree(p);
			break;

		default:
			printk(KERN_WARNING PFX "unknown parameter or missing value "
			       "'%s' in target creation request\n", p);
			goto out;
		}
	}

	if ((opt_mask & SRP_OPT_ALL) == SRP_OPT_ALL)
		ret = 0;
	else
		for (i = 0; i < ARRAY_SIZE(srp_opt_tokens); ++i)
			if ((srp_opt_tokens[i].token & SRP_OPT_ALL) &&
			    !(srp_opt_tokens[i].token & opt_mask))
				printk(KERN_WARNING PFX "target creation request is "
				       "missing parameter '%s'\n",
				       srp_opt_tokens[i].pattern);

out:
	kfree(options);
	return ret;
}

static ssize_t srp_create_target(struct class_device *class_dev,
				 const char *buf, size_t count)
{
	struct srp_host *host =
		container_of(class_dev, struct srp_host, class_dev);
	struct Scsi_Host *target_host;
	struct srp_target_port *target;
	int ret;
	int i;

	target_host = scsi_host_alloc(&srp_template,
				      sizeof (struct srp_target_port));
	if (!target_host)
		return -ENOMEM;

	target_host->transportt = ib_srp_transport_template;
	target_host->max_lun     = SRP_MAX_LUN;
	target_host->max_cmd_len = sizeof ((struct srp_cmd *) (void *) 0L)->cdb;

	target = host_to_target(target_host);

	target->io_class   = SRP_REV16A_IB_IO_CLASS;
	target->scsi_host  = target_host;
	target->srp_host   = host;

	INIT_LIST_HEAD(&target->free_reqs);
	INIT_LIST_HEAD(&target->req_queue);
	for (i = 0; i < SRP_SQ_SIZE; ++i) {
		target->req_ring[i].index = i;
		list_add_tail(&target->req_ring[i].list, &target->free_reqs);
	}

	ret = srp_parse_options(buf, target);
	if (ret)
		goto err;

	ib_get_cached_gid(host->dev->dev, host->port, 0, &target->path.sgid);

	printk(KERN_DEBUG PFX "new target: id_ext %016llx ioc_guid %016llx pkey %04x "
	       "service_id %016llx dgid %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x\n",
	       (unsigned long long) be64_to_cpu(target->id_ext),
	       (unsigned long long) be64_to_cpu(target->ioc_guid),
	       be16_to_cpu(target->path.pkey),
	       (unsigned long long) be64_to_cpu(target->service_id),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[0]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[2]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[4]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[6]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[8]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[10]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[12]),
	       (int) be16_to_cpu(*(__be16 *) &target->path.dgid.raw[14]));

	ret = srp_create_target_ib(target);
	if (ret)
		goto err;

	target->cm_id = ib_create_cm_id(host->dev->dev, srp_cm_handler, target);
	if (IS_ERR(target->cm_id)) {
		ret = PTR_ERR(target->cm_id);
		goto err_free;
	}

	target->qp_in_error = 0;
	ret = srp_connect_target(target);
	if (ret) {
		printk(KERN_ERR PFX "Connection failed\n");
		goto err_cm_id;
	}

	ret = srp_add_target(host, target);
	if (ret)
		goto err_disconnect;

	return count;

err_disconnect:
	srp_disconnect_target(target);

err_cm_id:
	ib_destroy_cm_id(target->cm_id);

err_free:
	srp_free_target_ib(target);

err:
	scsi_host_put(target_host);

	return ret;
}

static CLASS_DEVICE_ATTR(add_target, S_IWUSR, NULL, srp_create_target);

static ssize_t show_ibdev(struct class_device *class_dev, char *buf)
{
	struct srp_host *host =
		container_of(class_dev, struct srp_host, class_dev);

	return sprintf(buf, "%s\n", host->dev->dev->name);
}

static CLASS_DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);

static ssize_t show_port(struct class_device *class_dev, char *buf)
{
	struct srp_host *host =
		container_of(class_dev, struct srp_host, class_dev);

	return sprintf(buf, "%d\n", host->port);
}

static CLASS_DEVICE_ATTR(port, S_IRUGO, show_port, NULL);

static struct srp_host *srp_add_port(struct srp_device *device, u8 port)
{
	struct srp_host *host;

	host = kzalloc(sizeof *host, GFP_KERNEL);
	if (!host)
		return NULL;

	INIT_LIST_HEAD(&host->target_list);
	spin_lock_init(&host->target_lock);
	init_completion(&host->released);
	host->dev  = device;
	host->port = port;

	host->class_dev.class = &srp_class;
	host->class_dev.dev   = device->dev->dma_device;
	snprintf(host->class_dev.class_id, BUS_ID_SIZE, "srp-%s-%d",
		 device->dev->name, port);

	if (class_device_register(&host->class_dev))
		goto free_host;
	if (class_device_create_file(&host->class_dev, &class_device_attr_add_target))
		goto err_class;
	if (class_device_create_file(&host->class_dev, &class_device_attr_ibdev))
		goto err_class;
	if (class_device_create_file(&host->class_dev, &class_device_attr_port))
		goto err_class;

	return host;

err_class:
	class_device_unregister(&host->class_dev);

free_host:
	kfree(host);

	return NULL;
}

static void srp_add_one(struct ib_device *device)
{
	struct srp_device *srp_dev;
	struct ib_device_attr *dev_attr;
	struct ib_fmr_pool_param fmr_param;
	struct srp_host *host;
	int s, e, p;

	dev_attr = kmalloc(sizeof *dev_attr, GFP_KERNEL);
	if (!dev_attr)
		return;

	if (ib_query_device(device, dev_attr)) {
		printk(KERN_WARNING PFX "Query device failed for %s\n",
		       device->name);
		goto free_attr;
	}

	srp_dev = kmalloc(sizeof *srp_dev, GFP_KERNEL);
	if (!srp_dev)
		goto free_attr;

	/*
	 * Use the smallest page size supported by the HCA, down to a
	 * minimum of 512 bytes (which is the smallest sector that a
	 * SCSI command will ever carry).
	 */
	srp_dev->fmr_page_shift = max(9, ffs(dev_attr->page_size_cap) - 1);
	srp_dev->fmr_page_size  = 1 << srp_dev->fmr_page_shift;
	srp_dev->fmr_page_mask  = ~((u64) srp_dev->fmr_page_size - 1);

	INIT_LIST_HEAD(&srp_dev->dev_list);

	srp_dev->dev = device;
	srp_dev->pd  = ib_alloc_pd(device);
	if (IS_ERR(srp_dev->pd))
		goto free_dev;

	srp_dev->mr = ib_get_dma_mr(srp_dev->pd,
				    IB_ACCESS_LOCAL_WRITE |
				    IB_ACCESS_REMOTE_READ |
				    IB_ACCESS_REMOTE_WRITE);
	if (IS_ERR(srp_dev->mr))
		goto err_pd;

	memset(&fmr_param, 0, sizeof fmr_param);
	fmr_param.pool_size	    = SRP_FMR_POOL_SIZE;
	fmr_param.dirty_watermark   = SRP_FMR_DIRTY_SIZE;
	fmr_param.cache		    = 1;
	fmr_param.max_pages_per_fmr = SRP_FMR_SIZE;
	fmr_param.page_shift	    = srp_dev->fmr_page_shift;
	fmr_param.access	    = (IB_ACCESS_LOCAL_WRITE |
				       IB_ACCESS_REMOTE_WRITE |
				       IB_ACCESS_REMOTE_READ);

	srp_dev->fmr_pool = ib_create_fmr_pool(srp_dev->pd, &fmr_param);
	if (IS_ERR(srp_dev->fmr_pool))
		srp_dev->fmr_pool = NULL;

	if (device->node_type == RDMA_NODE_IB_SWITCH) {
		s = 0;
		e = 0;
	} else {
		s = 1;
		e = device->phys_port_cnt;
	}

	for (p = s; p <= e; ++p) {
		host = srp_add_port(srp_dev, p);
		if (host)
			list_add_tail(&host->list, &srp_dev->dev_list);
	}

	ib_set_client_data(device, &srp_client, srp_dev);

	goto free_attr;

err_pd:
	ib_dealloc_pd(srp_dev->pd);

free_dev:
	kfree(srp_dev);

free_attr:
	kfree(dev_attr);
}

static void srp_remove_one(struct ib_device *device)
{
	struct srp_device *srp_dev;
	struct srp_host *host, *tmp_host;
	LIST_HEAD(target_list);
	struct srp_target_port *target, *tmp_target;

	srp_dev = ib_get_client_data(device, &srp_client);

	list_for_each_entry_safe(host, tmp_host, &srp_dev->dev_list, list) {
		class_device_unregister(&host->class_dev);
		/*
		 * Wait for the sysfs entry to go away, so that no new
		 * target ports can be created.
		 */
		wait_for_completion(&host->released);

		/*
		 * Mark all target ports as removed, so we stop queueing
		 * commands and don't try to reconnect.
		 */
		spin_lock(&host->target_lock);
		list_for_each_entry(target, &host->target_list, list) {
			spin_lock_irq(target->scsi_host->host_lock);
			target->state = SRP_TARGET_REMOVED;
			spin_unlock_irq(target->scsi_host->host_lock);
		}
		spin_unlock(&host->target_lock);

		/*
		 * Wait for any reconnection tasks that may have
		 * started before we marked our target ports as
		 * removed, and any target port removal tasks.
		 */
		flush_scheduled_work();

		list_for_each_entry_safe(target, tmp_target,
					 &host->target_list, list) {
			srp_remove_host(target->scsi_host);
			scsi_remove_host(target->scsi_host);
			srp_disconnect_target(target);
			ib_destroy_cm_id(target->cm_id);
			srp_free_target_ib(target);
			scsi_host_put(target->scsi_host);
		}

		kfree(host);
	}

	if (srp_dev->fmr_pool)
		ib_destroy_fmr_pool(srp_dev->fmr_pool);
	ib_dereg_mr(srp_dev->mr);
	ib_dealloc_pd(srp_dev->pd);

	kfree(srp_dev);
}

static struct srp_function_template ib_srp_transport_functions = {
};

static int __init srp_init_module(void)
{
	int ret;

	ib_srp_transport_template =
		srp_attach_transport(&ib_srp_transport_functions);
	if (!ib_srp_transport_template)
		return -ENOMEM;

	srp_template.sg_tablesize = srp_sg_tablesize;
	srp_max_iu_len = (sizeof (struct srp_cmd) +
			  sizeof (struct srp_indirect_buf) +
			  srp_sg_tablesize * 16);

	ret = class_register(&srp_class);
	if (ret) {
		printk(KERN_ERR PFX "couldn't register class infiniband_srp\n");
		srp_release_transport(ib_srp_transport_template);
		return ret;
	}

	ib_sa_register_client(&srp_sa_client);

	ret = ib_register_client(&srp_client);
	if (ret) {
		printk(KERN_ERR PFX "couldn't register IB client\n");
		srp_release_transport(ib_srp_transport_template);
		ib_sa_unregister_client(&srp_sa_client);
		class_unregister(&srp_class);
		return ret;
	}

	return 0;
}

static void __exit srp_cleanup_module(void)
{
	ib_unregister_client(&srp_client);
	ib_sa_unregister_client(&srp_sa_client);
	class_unregister(&srp_class);
	srp_release_transport(ib_srp_transport_template);
}

module_init(srp_init_module);
module_exit(srp_cleanup_module);