Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking changes from David S Miller:

 1) Remove the ipv4 routing cache.  Now lookups go directly into the FIB
    trie and use prebuilt routes cached there.

    No more garbage collection, no more rDOS attacks on the routing
    cache.  Instead we now get predictable and consistent performance,
    no matter what the pattern of traffic we service.

    This has been almost 2 years in the making.  Special thanks to
    Julian Anastasov, Eric Dumazet, Steffen Klassert, and others who
    have helped along the way.

    I'm sure that with a change of this magnitude there will be some
    kind of fallout, but such things ought the be simple to fix at this
    point.  Luckily I'm not European so I'll be around all of August to
    fix things :-)

    The major stages of this work here are each fronted by a forced
    merge commit whose commit message contains a top-level description
    of the motivations and implementation issues.

 2) Pre-demux of established ipv4 TCP sockets, saves a route demux on
    input.

 3) TCP SYN/ACK performance tweaks from Eric Dumazet.

 4) Add namespace support for netfilter L4 conntrack helpers, from Gao
    Feng.

 5) Add config mechanism for Energy Efficient Ethernet to ethtool, from
    Yuval Mintz.

 6) Remove quadratic behavior from /proc/net/unix, from Eric Dumazet.

 7) Support for connection tracker helpers in userspace, from Pablo
    Neira Ayuso.

 8) Allow userspace driven TX load balancing functions in TEAM driver,
    from Jiri Pirko.

 9) Kill off NLMSG_PUT and RTA_PUT macros, more gross stuff with
    embedded gotos.

10) TCP Small Queues, essentially minimize the amount of TCP data queued
    up in the packet scheduler layer.  Whereas the existing BQL (Byte
    Queue Limits) limits the pkt_sched --> netdevice queuing levels,
    this controls the TCP --> pkt_sched queueing levels.

    From Eric Dumazet.

11) Reduce the number of get_page/put_page ops done on SKB fragments,
    from Alexander Duyck.

12) Implement protection against blind resets in TCP (RFC 5961), from
    Eric Dumazet.

13) Support the client side of TCP Fast Open, basically the ability to
    send data in the SYN exchange, from Yuchung Cheng.

    Basically, the sender queues up data with a sendmsg() call using
    MSG_FASTOPEN, then they do the connect() which emits the queued up
    fastopen data.

14) Avoid all the problems we get into in TCP when timers or PMTU events
    hit a locked socket.  The TCP Small Queues changes added a
    tcp_release_cb() that allows us to queue work up to the
    release_sock() caller, and that's what we use here too.  From Eric
    Dumazet.

15) Zero copy on TX support for TUN driver, from Michael S. Tsirkin.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1870 commits)
  genetlink: define lockdep_genl_is_held() when CONFIG_LOCKDEP
  r8169: revert "add byte queue limit support".
  ipv4: Change rt->rt_iif encoding.
  net: Make skb->skb_iif always track skb->dev
  ipv4: Prepare for change of rt->rt_iif encoding.
  ipv4: Remove all RTCF_DIRECTSRC handliing.
  ipv4: Really ignore ICMP address requests/replies.
  decnet: Don't set RTCF_DIRECTSRC.
  net/ipv4/ip_vti.c: Fix __rcu warnings detected by sparse.
  ipv4: Remove redundant assignment
  rds: set correct msg_namelen
  openvswitch: potential NULL deref in sample()
  tcp: dont drop MTU reduction indications
  bnx2x: Add new 57840 device IDs
  tcp: avoid oops in tcp_metrics and reset tcpm_stamp
  niu: Change niu_rbr_fill() to use unlikely() to check niu_rbr_add_page() return value
  niu: Fix to check for dma mapping errors.
  net: Fix references to out-of-scope variables in put_cmsg_compat()
  net: ethernet: davinci_emac: add pm_runtime support
  net: ethernet: davinci_emac: Remove unnecessary #include
  ...

5 files changed, 146 insertions, 14 deletions

diff --git a/drivers/scsi/bnx2i/57xx_iscsi_hsi.h b/drivers/scsi/bnx2i/57xx_iscsi_hsi.h
index dc0a08e..f2db5fe 100644
--- a/drivers/scsi/bnx2i/57xx_iscsi_hsi.h
+++ b/drivers/scsi/bnx2i/57xx_iscsi_hsi.h
@@ -267,7 +267,13 @@ struct bnx2i_cmd_request {
  * task statistics for write response
  */
 struct bnx2i_write_resp_task_stat {
-	u32 num_data_ins;
+#if defined(__BIG_ENDIAN)
+	u16 num_r2ts;
+	u16 num_data_outs;
+#elif defined(__LITTLE_ENDIAN)
+	u16 num_data_outs;
+	u16 num_r2ts;
+#endif
 };
 
 /*
@@ -275,11 +281,11 @@ struct bnx2i_write_resp_task_stat {
  */
 struct bnx2i_read_resp_task_stat {
 #if defined(__BIG_ENDIAN)
-	u16 num_data_outs;
-	u16 num_r2ts;
+	u16 reserved;
+	u16 num_data_ins;
 #elif defined(__LITTLE_ENDIAN)
-	u16 num_r2ts;
-	u16 num_data_outs;
+	u16 num_data_ins;
+	u16 reserved;
 #endif
 };
 
diff --git a/drivers/scsi/bnx2i/bnx2i.h b/drivers/scsi/bnx2i/bnx2i.h
index 7e77cf6..3f9e706 100644
--- a/drivers/scsi/bnx2i/bnx2i.h
+++ b/drivers/scsi/bnx2i/bnx2i.h
@@ -44,6 +44,8 @@
 #include "57xx_iscsi_hsi.h"
 #include "57xx_iscsi_constants.h"
 
+#include "../../net/ethernet/broadcom/bnx2x/bnx2x_mfw_req.h"
+
 #define BNX2_ISCSI_DRIVER_NAME		"bnx2i"
 
 #define BNX2I_MAX_ADAPTERS		8
@@ -126,6 +128,43 @@
 #define REG_WR(__hba, offset, val)			\
 		writel(val, __hba->regview + offset)
 
+#ifdef CONFIG_32BIT
+#define GET_STATS_64(__hba, dst, field)				\
+	do {							\
+		spin_lock_bh(&__hba->stat_lock);		\
+		dst->field##_lo = __hba->stats.field##_lo;	\
+		dst->field##_hi = __hba->stats.field##_hi;	\
+		spin_unlock_bh(&__hba->stat_lock);		\
+	} while (0)
+
+#define ADD_STATS_64(__hba, field, len)				\
+	do {							\
+		if (spin_trylock(&__hba->stat_lock)) {		\
+			if (__hba->stats.field##_lo + len <	\
+			    __hba->stats.field##_lo)		\
+				__hba->stats.field##_hi++;	\
+			__hba->stats.field##_lo += len;		\
+			spin_unlock(&__hba->stat_lock);		\
+		}						\
+	} while (0)
+
+#else
+#define GET_STATS_64(__hba, dst, field)				\
+	do {							\
+		u64 val, *out;					\
+								\
+		val = __hba->bnx2i_stats.field;			\
+		out = (u64 *)&__hba->stats.field##_lo;		\
+		*out = cpu_to_le64(val);			\
+		out = (u64 *)&dst->field##_lo;			\
+		*out = cpu_to_le64(val);			\
+	} while (0)
+
+#define ADD_STATS_64(__hba, field, len)				\
+	do {							\
+		__hba->bnx2i_stats.field += len;		\
+	} while (0)
+#endif
 
 /**
  * struct generic_pdu_resc - login pdu resource structure
@@ -288,6 +327,15 @@ struct iscsi_cid_queue {
 	struct bnx2i_conn **conn_cid_tbl;
 };
 
+
+struct bnx2i_stats_info {
+	u64 rx_pdus;
+	u64 rx_bytes;
+	u64 tx_pdus;
+	u64 tx_bytes;
+};
+
+
 /**
  * struct bnx2i_hba - bnx2i adapter structure
  *
@@ -341,6 +389,8 @@ struct iscsi_cid_queue {
  * @ctx_ccell_tasks:       captures number of ccells and tasks supported by
  *                         currently offloaded connection, used to decode
  *                         context memory
+ * @stat_lock:		   spin lock used by the statistic collector (32 bit)
+ * @stats:		   local iSCSI statistic collection place holder
  *
  * Adapter Data Structure
  */
@@ -427,6 +477,12 @@ struct bnx2i_hba {
 	u32 num_sess_opened;
 	u32 num_conn_opened;
 	unsigned int ctx_ccell_tasks;
+
+#ifdef CONFIG_32BIT
+	spinlock_t stat_lock;
+#endif
+	struct bnx2i_stats_info bnx2i_stats;
+	struct iscsi_stats_info stats;
 };
 
 
@@ -750,6 +806,8 @@ extern void bnx2i_ulp_init(struct cnic_dev *dev);
 extern void bnx2i_ulp_exit(struct cnic_dev *dev);
 extern void bnx2i_start(void *handle);
 extern void bnx2i_stop(void *handle);
+extern int bnx2i_get_stats(void *handle);
+
 extern struct bnx2i_hba *get_adapter_list_head(void);
 
 struct bnx2i_conn *bnx2i_get_conn_from_id(struct bnx2i_hba *hba,
diff --git a/drivers/scsi/bnx2i/bnx2i_hwi.c b/drivers/scsi/bnx2i/bnx2i_hwi.c
index 86a12b4..33d6630 100644
--- a/drivers/scsi/bnx2i/bnx2i_hwi.c
+++ b/drivers/scsi/bnx2i/bnx2i_hwi.c
@@ -1350,6 +1350,7 @@ int bnx2i_process_scsi_cmd_resp(struct iscsi_session *session,
 				struct cqe *cqe)
 {
 	struct iscsi_conn *conn = bnx2i_conn->cls_conn->dd_data;
+	struct bnx2i_hba *hba = bnx2i_conn->hba;
 	struct bnx2i_cmd_response *resp_cqe;
 	struct bnx2i_cmd *bnx2i_cmd;
 	struct iscsi_task *task;
@@ -1367,16 +1368,26 @@ int bnx2i_process_scsi_cmd_resp(struct iscsi_session *session,
 
 	if (bnx2i_cmd->req.op_attr & ISCSI_CMD_REQUEST_READ) {
 		conn->datain_pdus_cnt +=
-			resp_cqe->task_stat.read_stat.num_data_outs;
+			resp_cqe->task_stat.read_stat.num_data_ins;
 		conn->rxdata_octets +=
 			bnx2i_cmd->req.total_data_transfer_length;
+		ADD_STATS_64(hba, rx_pdus,
+			     resp_cqe->task_stat.read_stat.num_data_ins);
+		ADD_STATS_64(hba, rx_bytes,
+			     bnx2i_cmd->req.total_data_transfer_length);
 	} else {
 		conn->dataout_pdus_cnt +=
-			resp_cqe->task_stat.read_stat.num_data_outs;
+			resp_cqe->task_stat.write_stat.num_data_outs;
 		conn->r2t_pdus_cnt +=
-			resp_cqe->task_stat.read_stat.num_r2ts;
+			resp_cqe->task_stat.write_stat.num_r2ts;
 		conn->txdata_octets +=
 			bnx2i_cmd->req.total_data_transfer_length;
+		ADD_STATS_64(hba, tx_pdus,
+			     resp_cqe->task_stat.write_stat.num_data_outs);
+		ADD_STATS_64(hba, tx_bytes,
+			     bnx2i_cmd->req.total_data_transfer_length);
+		ADD_STATS_64(hba, rx_pdus,
+			     resp_cqe->task_stat.write_stat.num_r2ts);
 	}
 	bnx2i_iscsi_unmap_sg_list(bnx2i_cmd);
 
@@ -1961,6 +1972,7 @@ static int bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 {
 	struct iscsi_conn *conn = bnx2i_conn->cls_conn->dd_data;
 	struct iscsi_session *session = conn->session;
+	struct bnx2i_hba *hba = bnx2i_conn->hba;
 	struct qp_info *qp;
 	struct bnx2i_nop_in_msg *nopin;
 	int tgt_async_msg;
@@ -1973,7 +1985,7 @@ static int bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 
 	if (!qp->cq_virt) {
 		printk(KERN_ALERT "bnx2i (%s): cq resr freed in bh execution!",
-			bnx2i_conn->hba->netdev->name);
+		       hba->netdev->name);
 		goto out;
 	}
 	while (1) {
@@ -1985,9 +1997,9 @@ static int bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 			if (nopin->op_code == ISCSI_OP_NOOP_IN &&
 			    nopin->itt == (u16) RESERVED_ITT) {
 				printk(KERN_ALERT "bnx2i: Unsolicited "
-					"NOP-In detected for suspended "
-					"connection dev=%s!\n",
-					bnx2i_conn->hba->netdev->name);
+				       "NOP-In detected for suspended "
+				       "connection dev=%s!\n",
+				       hba->netdev->name);
 				bnx2i_unsol_pdu_adjust_rq(bnx2i_conn);
 				goto cqe_out;
 			}
@@ -2001,7 +2013,7 @@ static int bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 			/* Run the kthread engine only for data cmds
 			   All other cmds will be completed in this bh! */
 			bnx2i_queue_scsi_cmd_resp(session, bnx2i_conn, nopin);
-			break;
+			goto done;
 		case ISCSI_OP_LOGIN_RSP:
 			bnx2i_process_login_resp(session, bnx2i_conn,
 						 qp->cq_cons_qe);
@@ -2044,11 +2056,15 @@ static int bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 			printk(KERN_ALERT "bnx2i: unknown opcode 0x%x\n",
 					  nopin->op_code);
 		}
+
+		ADD_STATS_64(hba, rx_pdus, 1);
+		ADD_STATS_64(hba, rx_bytes, nopin->data_length);
+done:
 		if (!tgt_async_msg) {
 			if (!atomic_read(&bnx2i_conn->ep->num_active_cmds))
 				printk(KERN_ALERT "bnx2i (%s): no active cmd! "
 				       "op 0x%x\n",
-				       bnx2i_conn->hba->netdev->name,
+				       hba->netdev->name,
 				       nopin->op_code);
 			else
 				atomic_dec(&bnx2i_conn->ep->num_active_cmds);
@@ -2692,6 +2708,7 @@ struct cnic_ulp_ops bnx2i_cnic_cb = {
 	.cm_remote_close = bnx2i_cm_remote_close,
 	.cm_remote_abort = bnx2i_cm_remote_abort,
 	.iscsi_nl_send_msg = bnx2i_send_nl_mesg,
+	.cnic_get_stats = bnx2i_get_stats,
 	.owner = THIS_MODULE
 };
 
diff --git a/drivers/scsi/bnx2i/bnx2i_init.c b/drivers/scsi/bnx2i/bnx2i_init.c
index 8b68167..b17637a 100644
--- a/drivers/scsi/bnx2i/bnx2i_init.c
+++ b/drivers/scsi/bnx2i/bnx2i_init.c
@@ -381,6 +381,46 @@ void bnx2i_ulp_exit(struct cnic_dev *dev)
 
 
 /**
+ * bnx2i_get_stats - Retrieve various statistic from iSCSI offload
+ * @handle:	bnx2i_hba
+ *
+ * function callback exported via bnx2i - cnic driver interface to
+ *      retrieve various iSCSI offload related statistics.
+ */
+int bnx2i_get_stats(void *handle)
+{
+	struct bnx2i_hba *hba = handle;
+	struct iscsi_stats_info *stats;
+
+	if (!hba)
+		return -EINVAL;
+
+	stats = (struct iscsi_stats_info *)hba->cnic->stats_addr;
+
+	if (!stats)
+		return -ENOMEM;
+
+	strlcpy(stats->version, DRV_MODULE_VERSION, sizeof(stats->version));
+	memcpy(stats->mac_add1 + 2, hba->cnic->mac_addr, ETH_ALEN);
+
+	stats->max_frame_size = hba->netdev->mtu;
+	stats->txq_size = hba->max_sqes;
+	stats->rxq_size = hba->max_cqes;
+
+	stats->txq_avg_depth = 0;
+	stats->rxq_avg_depth = 0;
+
+	GET_STATS_64(hba, stats, rx_pdus);
+	GET_STATS_64(hba, stats, rx_bytes);
+
+	GET_STATS_64(hba, stats, tx_pdus);
+	GET_STATS_64(hba, stats, tx_bytes);
+
+	return 0;
+}
+
+
+/**
  * bnx2i_percpu_thread_create - Create a receive thread for an
  *				online CPU
  *
diff --git a/drivers/scsi/bnx2i/bnx2i_iscsi.c b/drivers/scsi/bnx2i/bnx2i_iscsi.c
index 621538b..3b34c13 100644
--- a/drivers/scsi/bnx2i/bnx2i_iscsi.c
+++ b/drivers/scsi/bnx2i/bnx2i_iscsi.c
@@ -874,6 +874,11 @@ struct bnx2i_hba *bnx2i_alloc_hba(struct cnic_dev *cnic)
 		hba->conn_ctx_destroy_tmo = 2 * HZ;
 	}
 
+#ifdef CONFIG_32BIT
+	spin_lock_init(&hba->stat_lock);
+#endif
+	memset(&hba->stats, 0, sizeof(struct iscsi_stats_info));
+
 	if (iscsi_host_add(shost, &hba->pcidev->dev))
 		goto free_dump_mem;
 	return hba;
@@ -1181,12 +1186,18 @@ static int
 bnx2i_mtask_xmit(struct iscsi_conn *conn, struct iscsi_task *task)
 {
 	struct bnx2i_conn *bnx2i_conn = conn->dd_data;
+	struct bnx2i_hba *hba = bnx2i_conn->hba;
 	struct bnx2i_cmd *cmd = task->dd_data;
 
 	memset(bnx2i_conn->gen_pdu.req_buf, 0, ISCSI_DEF_MAX_RECV_SEG_LEN);
 
 	bnx2i_setup_cmd_wqe_template(cmd);
 	bnx2i_conn->gen_pdu.req_buf_size = task->data_count;
+
+	/* Tx PDU/data length count */
+	ADD_STATS_64(hba, tx_pdus, 1);
+	ADD_STATS_64(hba, tx_bytes, task->data_count);
+
 	if (task->data_count) {
 		memcpy(bnx2i_conn->gen_pdu.req_buf, task->data,
 		       task->data_count);