iscsi_tcp.c

		/* bind iSCSI connection and socket */
		tcp_conn->sock = sock;

		/* setup Socket parameters */
		sk = sock->sk;
		sk->sk_reuse = 1;
		sk->sk_sndtimeo = 15 * HZ; /* FIXME: make it configurable */
		sk->sk_allocation = GFP_ATOMIC;

		/* FIXME: disable Nagle's algorithm */

		/*
		 * Intercept TCP callbacks for sendfile like receive
		 * processing.
		 */
		conn->recv_lock = &sk->sk_callback_lock;
		iscsi_conn_set_callbacks(conn);
		tcp_conn->sendpage = tcp_conn->sock->ops->sendpage;
		/*
		 * set receive state machine into initial state
		 */
		tcp_conn->in_progress = IN_PROGRESS_WAIT_HEADER;
	}

	return 0;
}

static void
iscsi_tcp_cleanup_ctask(struct iscsi_conn *conn, struct iscsi_cmd_task *ctask)
{
	struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;
	struct iscsi_r2t_info *r2t;

	/* flush ctask's r2t queues */
	while (__kfifo_get(tcp_ctask->r2tqueue, (void*)&r2t, sizeof(void*)))
		__kfifo_put(tcp_ctask->r2tpool.queue, (void*)&r2t,
			    sizeof(void*));

	__iscsi_ctask_cleanup(conn, ctask);
}

static void
iscsi_tcp_suspend_conn_rx(struct iscsi_conn *conn)
{
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
	struct sock *sk;

	if (!tcp_conn->sock)
		return;

	sk = tcp_conn->sock->sk;
	write_lock_bh(&sk->sk_callback_lock);
	set_bit(ISCSI_SUSPEND_BIT, &conn->suspend_rx);
	write_unlock_bh(&sk->sk_callback_lock);
}

static void
iscsi_tcp_terminate_conn(struct iscsi_conn *conn)
{
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;

	if (!tcp_conn->sock)
		return;

	sock_hold(tcp_conn->sock->sk);
	iscsi_conn_restore_callbacks(conn);
	sock_put(tcp_conn->sock->sk);

	sock_release(tcp_conn->sock);
	tcp_conn->sock = NULL;
	conn->recv_lock = NULL;
}

/* called with host lock */
static void
iscsi_tcp_mgmt_init(struct iscsi_conn *conn, struct iscsi_mgmt_task *mtask,
		    char *data, uint32_t data_size)
{
	struct iscsi_tcp_mgmt_task *tcp_mtask = mtask->dd_data;

	iscsi_buf_init_virt(&tcp_mtask->headbuf, (char*)mtask->hdr,
				    sizeof(struct iscsi_hdr));
	tcp_mtask->xmstate = XMSTATE_IMM_HDR;

	if (mtask->data_count)
		iscsi_buf_init_iov(&tcp_mtask->sendbuf, (char*)mtask->data,
				    mtask->data_count);
}

static int
iscsi_r2tpool_alloc(struct iscsi_session *session)
{
	int i;
	int cmd_i;

	/*
	 * initialize per-task: R2T pool and xmit queue
	 */
	for (cmd_i = 0; cmd_i < session->cmds_max; cmd_i++) {
	        struct iscsi_cmd_task *ctask = session->cmds[cmd_i];
		struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;

		/*
		 * pre-allocated x4 as much r2ts to handle race when
		 * target acks DataOut faster than we data_xmit() queues
		 * could replenish r2tqueue.
		 */

		/* R2T pool */
		if (iscsi_pool_init(&tcp_ctask->r2tpool, session->max_r2t * 4,
				    (void***)&tcp_ctask->r2ts,
				    sizeof(struct iscsi_r2t_info))) {
			goto r2t_alloc_fail;
		}

		/* R2T xmit queue */
		tcp_ctask->r2tqueue = kfifo_alloc(
		      session->max_r2t * 4 * sizeof(void*), GFP_KERNEL, NULL);
		if (tcp_ctask->r2tqueue == ERR_PTR(-ENOMEM)) {
			iscsi_pool_free(&tcp_ctask->r2tpool,
					(void**)tcp_ctask->r2ts);
			goto r2t_alloc_fail;
		}

		/*
		 * number of
		 * Data-Out PDU's within R2T-sequence can be quite big;
		 * using mempool
		 */
		tcp_ctask->datapool = mempool_create_slab_pool(ISCSI_DTASK_DEFAULT_MAX,
							       taskcache);
		if (tcp_ctask->datapool == NULL) {
			kfifo_free(tcp_ctask->r2tqueue);
			iscsi_pool_free(&tcp_ctask->r2tpool,
					(void**)tcp_ctask->r2ts);
			goto r2t_alloc_fail;
		}
		INIT_LIST_HEAD(&tcp_ctask->dataqueue);
	}

	return 0;

r2t_alloc_fail:
	for (i = 0; i < cmd_i; i++) {
		struct iscsi_cmd_task *ctask = session->cmds[i];
		struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;

		mempool_destroy(tcp_ctask->datapool);
		kfifo_free(tcp_ctask->r2tqueue);
		iscsi_pool_free(&tcp_ctask->r2tpool,
				(void**)tcp_ctask->r2ts);
	}
	return -ENOMEM;
}

static void
iscsi_r2tpool_free(struct iscsi_session *session)
{
	int i;

	for (i = 0; i < session->cmds_max; i++) {
		struct iscsi_cmd_task *ctask = session->cmds[i];
		struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;

		mempool_destroy(tcp_ctask->datapool);
		kfifo_free(tcp_ctask->r2tqueue);
		iscsi_pool_free(&tcp_ctask->r2tpool,
				(void**)tcp_ctask->r2ts);
	}
}

static int
iscsi_conn_set_param(struct iscsi_cls_conn *cls_conn, enum iscsi_param param,
		     uint32_t value)
{
	struct iscsi_conn *conn = cls_conn->dd_data;
	struct iscsi_session *session = conn->session;
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;

	switch(param) {
	case ISCSI_PARAM_MAX_RECV_DLENGTH: {
		char *saveptr = tcp_conn->data;
		gfp_t flags = GFP_KERNEL;

		if (tcp_conn->data_size >= value) {
			conn->max_recv_dlength = value;
			break;
		}

		spin_lock_bh(&session->lock);
		if (conn->stop_stage == STOP_CONN_RECOVER)
			flags = GFP_ATOMIC;
		spin_unlock_bh(&session->lock);

		if (value <= PAGE_SIZE)
			tcp_conn->data = kmalloc(value, flags);
		else
			tcp_conn->data = (void*)__get_free_pages(flags,
							     get_order(value));
		if (tcp_conn->data == NULL) {
			tcp_conn->data = saveptr;
			return -ENOMEM;
		}
		if (tcp_conn->data_size <= PAGE_SIZE)
			kfree(saveptr);
		else
			free_pages((unsigned long)saveptr,
				   get_order(tcp_conn->data_size));
		conn->max_recv_dlength = value;
		tcp_conn->data_size = value;
		}
		break;
	case ISCSI_PARAM_MAX_XMIT_DLENGTH:
		conn->max_xmit_dlength =  value;
		break;
	case ISCSI_PARAM_HDRDGST_EN:
		conn->hdrdgst_en = value;
		tcp_conn->hdr_size = sizeof(struct iscsi_hdr);
		if (conn->hdrdgst_en) {
			tcp_conn->hdr_size += sizeof(__u32);
			if (!tcp_conn->tx_tfm)
				tcp_conn->tx_tfm = crypto_alloc_tfm("crc32c",
								    0);
			if (!tcp_conn->tx_tfm)
				return -ENOMEM;
			if (!tcp_conn->rx_tfm)
				tcp_conn->rx_tfm = crypto_alloc_tfm("crc32c",
								    0);
			if (!tcp_conn->rx_tfm) {
				crypto_free_tfm(tcp_conn->tx_tfm);
				return -ENOMEM;
			}
		} else {
			if (tcp_conn->tx_tfm)
				crypto_free_tfm(tcp_conn->tx_tfm);
			if (tcp_conn->rx_tfm)
				crypto_free_tfm(tcp_conn->rx_tfm);
		}
		break;
	case ISCSI_PARAM_DATADGST_EN:
		conn->datadgst_en = value;
		if (conn->datadgst_en) {
			if (!tcp_conn->data_tx_tfm)
				tcp_conn->data_tx_tfm =
				    crypto_alloc_tfm("crc32c", 0);
			if (!tcp_conn->data_tx_tfm)
				return -ENOMEM;
			if (!tcp_conn->data_rx_tfm)
				tcp_conn->data_rx_tfm =
				    crypto_alloc_tfm("crc32c", 0);
			if (!tcp_conn->data_rx_tfm) {
				crypto_free_tfm(tcp_conn->data_tx_tfm);
				return -ENOMEM;
			}
		} else {
			if (tcp_conn->data_tx_tfm)
				crypto_free_tfm(tcp_conn->data_tx_tfm);
			if (tcp_conn->data_rx_tfm)
				crypto_free_tfm(tcp_conn->data_rx_tfm);
		}
		tcp_conn->sendpage = conn->datadgst_en ?
			sock_no_sendpage : tcp_conn->sock->ops->sendpage;
		break;
	case ISCSI_PARAM_INITIAL_R2T_EN:
		session->initial_r2t_en = value;
		break;
	case ISCSI_PARAM_MAX_R2T:
		if (session->max_r2t == roundup_pow_of_two(value))
			break;
		iscsi_r2tpool_free(session);
		session->max_r2t = value;
		if (session->max_r2t & (session->max_r2t - 1))
			session->max_r2t = roundup_pow_of_two(session->max_r2t);
		if (iscsi_r2tpool_alloc(session))
			return -ENOMEM;
		break;
	case ISCSI_PARAM_IMM_DATA_EN:
		session->imm_data_en = value;
		break;
	case ISCSI_PARAM_FIRST_BURST:
		session->first_burst = value;
		break;
	case ISCSI_PARAM_MAX_BURST:
		session->max_burst = value;
		break;
	case ISCSI_PARAM_PDU_INORDER_EN:
		session->pdu_inorder_en = value;
		break;
	case ISCSI_PARAM_DATASEQ_INORDER_EN:
		session->dataseq_inorder_en = value;
		break;
	case ISCSI_PARAM_ERL:
		session->erl = value;
		break;
	case ISCSI_PARAM_IFMARKER_EN:
		BUG_ON(value);
		session->ifmarker_en = value;
		break;
	case ISCSI_PARAM_OFMARKER_EN:
		BUG_ON(value);
		session->ofmarker_en = value;
		break;
	case ISCSI_PARAM_EXP_STATSN:
		conn->exp_statsn = value;
		break;
	default:
		break;
	}

	return 0;
}

static int
iscsi_session_get_param(struct iscsi_cls_session *cls_session,
			enum iscsi_param param, uint32_t *value)
{
	struct Scsi_Host *shost = iscsi_session_to_shost(cls_session);
	struct iscsi_session *session = iscsi_hostdata(shost->hostdata);

	switch(param) {
	case ISCSI_PARAM_INITIAL_R2T_EN:
		*value = session->initial_r2t_en;
		break;
	case ISCSI_PARAM_MAX_R2T:
		*value = session->max_r2t;
		break;
	case ISCSI_PARAM_IMM_DATA_EN:
		*value = session->imm_data_en;
		break;
	case ISCSI_PARAM_FIRST_BURST:
		*value = session->first_burst;
		break;
	case ISCSI_PARAM_MAX_BURST:
		*value = session->max_burst;
		break;
	case ISCSI_PARAM_PDU_INORDER_EN:
		*value = session->pdu_inorder_en;
		break;
	case ISCSI_PARAM_DATASEQ_INORDER_EN:
		*value = session->dataseq_inorder_en;
		break;
	case ISCSI_PARAM_ERL:
		*value = session->erl;
		break;
	case ISCSI_PARAM_IFMARKER_EN:
		*value = session->ifmarker_en;
		break;
	case ISCSI_PARAM_OFMARKER_EN:
		*value = session->ofmarker_en;
		break;
	default:
		return -EINVAL;
	}

	return 0;
}

static int
iscsi_conn_get_param(struct iscsi_cls_conn *cls_conn,
		     enum iscsi_param param, uint32_t *value)
{
	struct iscsi_conn *conn = cls_conn->dd_data;
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
	struct inet_sock *inet;

	switch(param) {
	case ISCSI_PARAM_MAX_RECV_DLENGTH:
		*value = conn->max_recv_dlength;
		break;
	case ISCSI_PARAM_MAX_XMIT_DLENGTH:
		*value = conn->max_xmit_dlength;
		break;
	case ISCSI_PARAM_HDRDGST_EN:
		*value = conn->hdrdgst_en;
		break;
	case ISCSI_PARAM_DATADGST_EN:
		*value = conn->datadgst_en;
		break;
	case ISCSI_PARAM_CONN_PORT:
		mutex_lock(&conn->xmitmutex);
		if (!tcp_conn->sock) {
			mutex_unlock(&conn->xmitmutex);
			return -EINVAL;
		}

		inet = inet_sk(tcp_conn->sock->sk);
		*value = be16_to_cpu(inet->dport);
		mutex_unlock(&conn->xmitmutex);
	case ISCSI_PARAM_EXP_STATSN:
		*value = conn->exp_statsn;
		break;
	default:
		return -EINVAL;
	}

	return 0;
}

static int
iscsi_conn_get_str_param(struct iscsi_cls_conn *cls_conn,
			 enum iscsi_param param, char *buf)
{
	struct iscsi_conn *conn = cls_conn->dd_data;
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
	struct sock *sk;
	struct inet_sock *inet;
	struct ipv6_pinfo *np;
	int len = 0;

	switch (param) {
	case ISCSI_PARAM_CONN_ADDRESS:
		mutex_lock(&conn->xmitmutex);
		if (!tcp_conn->sock) {
			mutex_unlock(&conn->xmitmutex);
			return -EINVAL;
		}

		sk = tcp_conn->sock->sk;
		if (sk->sk_family == PF_INET) {
			inet = inet_sk(sk);
			len = sprintf(buf, "%u.%u.%u.%u\n",
				      NIPQUAD(inet->daddr));
		} else {
			np = inet6_sk(sk);
			len = sprintf(buf,
				"%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x\n",
				NIP6(np->daddr));
		}
		mutex_unlock(&conn->xmitmutex);
		break;
	default:
		return -EINVAL;
	}

	return len;
}

static void
iscsi_conn_get_stats(struct iscsi_cls_conn *cls_conn, struct iscsi_stats *stats)
{
	struct iscsi_conn *conn = cls_conn->dd_data;
	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;

	stats->txdata_octets = conn->txdata_octets;
	stats->rxdata_octets = conn->rxdata_octets;
	stats->scsicmd_pdus = conn->scsicmd_pdus_cnt;
	stats->dataout_pdus = conn->dataout_pdus_cnt;
	stats->scsirsp_pdus = conn->scsirsp_pdus_cnt;
	stats->datain_pdus = conn->datain_pdus_cnt;
	stats->r2t_pdus = conn->r2t_pdus_cnt;
	stats->tmfcmd_pdus = conn->tmfcmd_pdus_cnt;
	stats->tmfrsp_pdus = conn->tmfrsp_pdus_cnt;
	stats->custom_length = 3;
	strcpy(stats->custom[0].desc, "tx_sendpage_failures");
	stats->custom[0].value = tcp_conn->sendpage_failures_cnt;
	strcpy(stats->custom[1].desc, "rx_discontiguous_hdr");
	stats->custom[1].value = tcp_conn->discontiguous_hdr_cnt;
	strcpy(stats->custom[2].desc, "eh_abort_cnt");
	stats->custom[2].value = conn->eh_abort_cnt;
}

static struct iscsi_cls_session *
iscsi_tcp_session_create(struct iscsi_transport *iscsit,
			 struct scsi_transport_template *scsit,
			 uint32_t initial_cmdsn, uint32_t *hostno)
{
	struct iscsi_cls_session *cls_session;
	struct iscsi_session *session;
	uint32_t hn;
	int cmd_i;

	cls_session = iscsi_session_setup(iscsit, scsit,
					 sizeof(struct iscsi_tcp_cmd_task),
					 sizeof(struct iscsi_tcp_mgmt_task),
					 initial_cmdsn, &hn);
	if (!cls_session)
		return NULL;
	*hostno = hn;

	session = class_to_transport_session(cls_session);
	for (cmd_i = 0; cmd_i < session->cmds_max; cmd_i++) {
		struct iscsi_cmd_task *ctask = session->cmds[cmd_i];
		struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;

		ctask->hdr = &tcp_ctask->hdr;
	}

	for (cmd_i = 0; cmd_i < session->mgmtpool_max; cmd_i++) {
		struct iscsi_mgmt_task *mtask = session->mgmt_cmds[cmd_i];
		struct iscsi_tcp_mgmt_task *tcp_mtask = mtask->dd_data;

		mtask->hdr = &tcp_mtask->hdr;
	}

	if (iscsi_r2tpool_alloc(class_to_transport_session(cls_session)))
		goto r2tpool_alloc_fail;

	return cls_session;

r2tpool_alloc_fail:
	iscsi_session_teardown(cls_session);
	return NULL;
}

static void iscsi_tcp_session_destroy(struct iscsi_cls_session *cls_session)
{
	struct iscsi_session *session = class_to_transport_session(cls_session);
	struct iscsi_data_task *dtask, *n;
	int cmd_i;

	for (cmd_i = 0; cmd_i < session->cmds_max; cmd_i++) {
		struct iscsi_cmd_task *ctask = session->cmds[cmd_i];
		struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;

		list_for_each_entry_safe(dtask, n, &tcp_ctask->dataqueue,
					 item) {
			list_del(&dtask->item);
			mempool_free(dtask, tcp_ctask->datapool);
		}
	}

	iscsi_r2tpool_free(class_to_transport_session(cls_session));
	iscsi_session_teardown(cls_session);
}

static struct scsi_host_template iscsi_sht = {
	.name			= "iSCSI Initiator over TCP/IP, v."
				  ISCSI_VERSION_STR,
	.queuecommand           = iscsi_queuecommand,
	.change_queue_depth	= iscsi_change_queue_depth,
	.can_queue		= ISCSI_XMIT_CMDS_MAX - 1,
	.sg_tablesize		= ISCSI_SG_TABLESIZE,
	.cmd_per_lun		= ISCSI_DEF_CMD_PER_LUN,
	.eh_abort_handler       = iscsi_eh_abort,
	.eh_host_reset_handler	= iscsi_eh_host_reset,
	.use_clustering         = DISABLE_CLUSTERING,
	.proc_name		= "iscsi_tcp",
	.this_id		= -1,
};

static struct iscsi_transport iscsi_tcp_transport = {
	.owner			= THIS_MODULE,
	.name			= "tcp",
	.caps			= CAP_RECOVERY_L0 | CAP_MULTI_R2T | CAP_HDRDGST
				  | CAP_DATADGST,
	.param_mask		= ISCSI_MAX_RECV_DLENGTH |
				  ISCSI_MAX_XMIT_DLENGTH |
				  ISCSI_HDRDGST_EN |
				  ISCSI_DATADGST_EN |
				  ISCSI_INITIAL_R2T_EN |
				  ISCSI_MAX_R2T |
				  ISCSI_IMM_DATA_EN |
				  ISCSI_FIRST_BURST |
				  ISCSI_MAX_BURST |
				  ISCSI_PDU_INORDER_EN |
				  ISCSI_DATASEQ_INORDER_EN |
				  ISCSI_ERL |
				  ISCSI_CONN_PORT |
				  ISCSI_CONN_ADDRESS |
				  ISCSI_EXP_STATSN,
	.host_template		= &iscsi_sht,
	.conndata_size		= sizeof(struct iscsi_conn),
	.max_conn		= 1,
	.max_cmd_len		= ISCSI_TCP_MAX_CMD_LEN,
	/* session management */
	.create_session		= iscsi_tcp_session_create,
	.destroy_session	= iscsi_tcp_session_destroy,
	/* connection management */
	.create_conn		= iscsi_tcp_conn_create,
	.bind_conn		= iscsi_tcp_conn_bind,
	.destroy_conn		= iscsi_tcp_conn_destroy,
	.set_param		= iscsi_conn_set_param,
	.get_conn_param		= iscsi_conn_get_param,
	.get_conn_str_param	= iscsi_conn_get_str_param,
	.get_session_param	= iscsi_session_get_param,
	.start_conn		= iscsi_conn_start,
	.stop_conn		= iscsi_conn_stop,
	/* these are called as part of conn recovery */
	.suspend_conn_recv	= iscsi_tcp_suspend_conn_rx,
	.terminate_conn		= iscsi_tcp_terminate_conn,
	/* IO */
	.send_pdu		= iscsi_conn_send_pdu,
	.get_stats		= iscsi_conn_get_stats,
	.init_cmd_task		= iscsi_tcp_cmd_init,
	.init_mgmt_task		= iscsi_tcp_mgmt_init,
	.xmit_cmd_task		= iscsi_tcp_ctask_xmit,
	.xmit_mgmt_task		= iscsi_tcp_mtask_xmit,
	.cleanup_cmd_task	= iscsi_tcp_cleanup_ctask,
	/* recovery */
	.session_recovery_timedout = iscsi_session_recovery_timedout,
};

static int __init
iscsi_tcp_init(void)
{
	if (iscsi_max_lun < 1) {
		printk(KERN_ERR "iscsi_tcp: Invalid max_lun value of %u\n",
		       iscsi_max_lun);
		return -EINVAL;
	}
	iscsi_tcp_transport.max_lun = iscsi_max_lun;

	taskcache = kmem_cache_create("iscsi_taskcache",
			sizeof(struct iscsi_data_task), 0,
			SLAB_HWCACHE_ALIGN, NULL, NULL);
	if (!taskcache)
		return -ENOMEM;

	if (!iscsi_register_transport(&iscsi_tcp_transport))
		kmem_cache_destroy(taskcache);

	return 0;
}

static void __exit
iscsi_tcp_exit(void)
{
	iscsi_unregister_transport(&iscsi_tcp_transport);
	kmem_cache_destroy(taskcache);
}

module_init(iscsi_tcp_init);
module_exit(iscsi_tcp_exit);