lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date:	Mon, 14 Jul 2008 17:41:02 +0300
From:	Yevgeny Petrilin <yevgenyp@...lanox.co.il>
To:	jeff@...zik.org
CC:	netdev@...r.kernel.org, Liran Liss <liranl@...lanox.co.il>,
	tziporet@...lanox.co.il, Roland Dreier <rdreier@...co.com>
Subject: [PATCH RFC 03/10] mlx4_en: net_device implementation

[PATCH] mlx4_en : en_netdev.c

This file implements the network device interface.
Such device is created for every Ethernet port of the device.

Signed-off-by: Liran Liss <liranl@...lanox.co.il>
Signed-off-by: Yevgeny Petrilin <yevgenyp@...lanox.co.il>
---
 drivers/net/mlx4/en_netdev.c | 1076 ++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 1076 insertions(+), 0 deletions(-)
 create mode 100644 drivers/net/mlx4/en_netdev.c

diff --git a/drivers/net/mlx4/en_netdev.c b/drivers/net/mlx4/en_netdev.c
new file mode 100644
index 0000000..5c1f654
--- /dev/null
+++ b/drivers/net/mlx4/en_netdev.c
@@ -0,0 +1,1076 @@
+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/etherdevice.h>
+#include <linux/tcp.h>
+#include <linux/if_vlan.h>
+#include <linux/delay.h>
+
+#include <linux/mlx4/driver.h>
+#include <linux/mlx4/device.h>
+#include <linux/mlx4/cmd.h>
+#include <linux/mlx4/cq.h>
+
+#include "mlx4_en.h"
+#include "en_port.h"
+
+
+static void mlx4_en_vlan_rx_register(struct net_device *dev,
+				     struct vlan_group *grp)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	mlx4_dbg(mdev, "Regsitering VLAN group:%p\n", grp);
+	priv->vlgrp = grp;
+
+	down(&mdev->state_sem);
+	if (mdev->state == DEVICE_UP && priv->port_state == PORT_UP) {
+		err = mlx4_SET_VLAN_FLTR(mdev->dev, priv->port, grp);
+		if (err)
+			mlx4_err(mdev, "Failed configuring VLAN filter\n");
+	}
+	up(&mdev->state_sem);
+}
+
+static void mlx4_en_vlan_rx_add_vid(struct net_device *dev, unsigned short vid)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	if (!priv->vlgrp)
+		return;
+
+	mlx4_dbg(mdev, "adding VLAN:%d (vlgrp entry:%p)\n",
+		 vid, vlan_group_get_device(priv->vlgrp, vid));
+
+	/* Add VID to port VLAN filter */
+	down(&mdev->state_sem);
+	if (mdev->state == DEVICE_UP && priv->port_state == PORT_UP) {
+		err = mlx4_SET_VLAN_FLTR(mdev->dev, priv->port, priv->vlgrp);
+		if (err)
+			mlx4_err(mdev, "Failed configuring VLAN filter\n");
+	}
+	up(&mdev->state_sem);
+}
+
+static void mlx4_en_vlan_rx_kill_vid(struct net_device *dev, unsigned short vid)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	if (!priv->vlgrp)
+		return;
+
+	mlx4_dbg(mdev, "Killing VID:%d (vlgrp:%p vlgrp "
+		 "entry:%p)\n", vid, priv->vlgrp,
+		 vlan_group_get_device(priv->vlgrp, vid));
+	vlan_group_set_device(priv->vlgrp, vid, NULL);
+
+	/* Remove VID from port VLAN filter */
+	down(&mdev->state_sem);
+	if (mdev->state == DEVICE_UP && priv->port_state == PORT_UP) {
+		err = mlx4_SET_VLAN_FLTR(mdev->dev, priv->port, priv->vlgrp);
+		if (err)
+			mlx4_err(mdev, "Failed configuring VLAN filter\n");
+	}
+	up(&mdev->state_sem);
+}
+
+static u64 mlx4_en_mac_to_u64(u8 *addr)
+{
+	u64 mac = 0;
+	int i;
+
+	for (i = 0; i < ETH_ALEN; i++) {
+		mac <<= 8;
+		mac |= addr[i];
+	}
+	return mac;
+}
+
+static int mlx4_en_set_mac(struct net_device *dev, void *addr)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct sockaddr *saddr = addr;
+
+	if (!is_valid_ether_addr(saddr->sa_data)) {
+		mlx4_dbg(mdev, "Invalid mac address!\n");
+		return -EADDRNOTAVAIL;
+	}
+	memcpy(dev->dev_addr, saddr->sa_data, ETH_ALEN);
+	priv->mac = mlx4_en_mac_to_u64(dev->dev_addr);
+	queue_work(mdev->workqueue, &priv->mac_task);
+	return 0;
+}
+
+static void mlx4_en_do_set_mac(struct work_struct *work)
+{
+	struct mlx4_en_priv *priv = container_of(work, struct mlx4_en_priv,
+						 mac_task);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err = 0;
+
+	down(&mdev->state_sem);
+	if (priv->port_state == PORT_UP) {
+		/* Remove old MAC and insert the new one */
+		mlx4_unregister_mac(mdev->dev, priv->port, priv->mac_index);
+		err = mlx4_register_mac(mdev->dev, priv->port,
+					priv->mac, &priv->mac_index);
+		if (err)
+			mlx4_err(mdev, "Failed changing HW MAC address\n");
+	} else
+		mlx4_dbg(mdev, "Port is down, exiting...\n");
+
+	up(&mdev->state_sem);
+}
+
+static void mlx4_en_clear_list(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct dev_mc_list *plist = priv->mc_list;
+	struct dev_mc_list *next;
+
+	while (plist) {
+		next = plist->next;
+		kfree(plist);
+		plist = next;
+	}
+	priv->mc_list = NULL;
+}
+
+static void mlx4_en_cache_mclist(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct dev_mc_list *mclist;
+	struct dev_mc_list *tmp;
+	struct dev_mc_list *plist = NULL;
+
+	for (mclist = dev->mc_list; mclist; mclist = mclist->next) {
+		tmp = kmalloc(sizeof(struct dev_mc_list), GFP_ATOMIC);
+		if (!tmp) {
+			mlx4_err(mdev, "failed to allocate multicast list\n");
+			mlx4_en_clear_list(dev);
+			return;
+		}
+		memcpy(tmp, mclist, sizeof(struct dev_mc_list));
+		tmp->next = NULL;
+		if (plist)
+			plist->next = tmp;
+		else
+			priv->mc_list = tmp;
+		plist = tmp;
+	}
+}
+
+
+static void mlx4_en_set_multicast(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+
+	if (priv->port_state != PORT_UP)
+		return;
+
+	queue_work(priv->mdev->workqueue, &priv->mcast_task);
+}
+
+static void mlx4_en_do_set_multicast(struct work_struct *work)
+{
+	struct mlx4_en_priv *priv = container_of(work, struct mlx4_en_priv,
+						 mcast_task);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct net_device *dev = priv->dev;
+	struct dev_mc_list *mclist;
+	u64 mcast_addr = 0;
+	int err;
+
+	mlx4_dbg(mdev, "multicast task called for port:%d\n", priv->port);
+
+	down(&mdev->state_sem);
+	if (mdev->state != DEVICE_UP) {
+		mlx4_dbg(mdev, "Card is not up, ignoring multicast change.\n");
+		goto out;
+	}
+	if (priv->port_state == PORT_DOWN) {
+		mlx4_dbg(mdev, "Port is down, ignoring multicast change.\n");
+		goto out;
+	}
+
+	/*
+	 * Promsicuous mode: disable all filters
+	 */
+
+	if (dev->flags & IFF_PROMISC) {
+		if (!(priv->flags & MLX4_EN_FLAG_PROMISC)) {
+			mlx4_warn(mdev, "Port:%d entering promiscuous mode\n",
+				  priv->port);
+			priv->flags |= MLX4_EN_FLAG_PROMISC;
+
+			/* Enable promiscouos mode */
+			err = mlx4_SET_PORT_qpn_calc(mdev->dev, priv->port,
+						     priv->base_qpn, 1);
+			if (err)
+				mlx4_err(mdev, "Failed enabling "
+					 "promiscous mode\n");
+
+			/* Disable port multicast filter (unconditionally) */
+			err = mlx4_SET_MCAST_FLTR(mdev->dev, priv->port, 0,
+						  0, MLX4_MCAST_DISABLE);
+			if (err)
+				mlx4_err(mdev, "Failed disabling "
+					 "multicast filter\n");
+
+			/* Disable port VLAN filter */
+			err = mlx4_SET_VLAN_FLTR(mdev->dev, priv->port, NULL);
+			if (err)
+				mlx4_err(mdev, "Failed disabling "
+					 "VLAN filter\n");
+		}
+		goto out;
+	}
+
+	/*
+	 * Not in promiscous mode
+	 */
+
+	if (priv->flags & MLX4_EN_FLAG_PROMISC) {
+		mlx4_warn(mdev, "Port:%d leaving promiscuous mode\n",
+			  priv->port);
+		priv->flags &= ~MLX4_EN_FLAG_PROMISC;
+
+		/* Disable promiscouos mode */
+		err = mlx4_SET_PORT_qpn_calc(mdev->dev, priv->port,
+					     priv->base_qpn, 0);
+		if (err)
+			mlx4_err(mdev, "Failed disabling promiscous mode\n");
+
+		/* Enable port VLAN filter */
+		err = mlx4_SET_VLAN_FLTR(mdev->dev, priv->port, priv->vlgrp);
+		if (err)
+			mlx4_err(mdev, "Failed enabling VLAN filter\n");
+	}
+
+	/* Enable/disable the multicast filter according to IFF_ALLMULTI */
+	if (dev->flags & IFF_ALLMULTI) {
+		err = mlx4_SET_MCAST_FLTR(mdev->dev, priv->port, 0,
+					  0, MLX4_MCAST_DISABLE);
+		if (err)
+			mlx4_err(mdev, "Failed disabling multicast filter\n");
+	} else {
+		err = mlx4_SET_MCAST_FLTR(mdev->dev, priv->port, 0,
+					  0, MLX4_MCAST_DISABLE);
+		if (err)
+			mlx4_err(mdev, "Failed disabling multicast filter\n");
+
+		/* Flush mcast filter and init it with broadcast address */
+		mlx4_SET_MCAST_FLTR(mdev->dev, priv->port, ETH_BCAST,
+				    1, MLX4_MCAST_CONFIG);
+
+		/* Update multicast list - we cache all addresses so they won't
+		 * change while HW is updated holding the command semaphor */
+		netif_tx_lock_bh(dev);
+		mlx4_en_cache_mclist(dev);
+		netif_tx_unlock_bh(dev);
+		for (mclist = priv->mc_list; mclist; mclist = mclist->next) {
+			mcast_addr = mlx4_en_mac_to_u64(mclist->dmi_addr);
+			mlx4_SET_MCAST_FLTR(mdev->dev, priv->port,
+					    mcast_addr, 0, MLX4_MCAST_CONFIG);
+		}
+		err = mlx4_SET_MCAST_FLTR(mdev->dev, priv->port, 0,
+					  0, MLX4_MCAST_ENABLE);
+		if (err)
+			mlx4_err(mdev, "Failed enabling multicast filter\n");
+
+		mlx4_en_clear_list(dev);
+	}
+out:
+	up(&mdev->state_sem);
+}
+
+void mlx4_en_netpoll(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_cq *cq;
+	unsigned long flags;
+	int i;
+
+
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		cq = &priv->rx_cq[i];
+		spin_lock_irqsave(&cq->lock, flags);
+		napi_synchronize(&cq->napi);
+		mlx4_en_process_rx_cq(dev, cq, 0);
+		spin_unlock_irqrestore(&cq->lock, flags);
+	}
+}
+
+
+static void mlx4_en_tx_timeout(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	mlx4_warn(mdev, "Tx timeout called on port:%d\n", priv->port);
+
+	if (netif_carrier_ok(dev)) {
+		priv->port_stats.rst_cnt++;
+		mlx4_dbg(mdev, "Scheduling watchdog\n");
+		queue_work(mdev->workqueue, &priv->watchdog_task);
+	}
+}
+
+
+static struct net_device_stats *mlx4_en_get_stats(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+
+	spin_lock_bh(&priv->stats_lock);
+	memcpy(&dev->stats, &priv->stats, sizeof(priv->stats));
+	spin_unlock_bh(&priv->stats_lock);
+
+	return &dev->stats;
+}
+
+void mlx4_en_set_default_moderation(struct mlx4_en_priv *priv)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct mlx4_en_cq *cq;
+	int i;
+
+	/* If we haven't received a specific coalescing setting
+	 * (module param), we set the moderation paramters as follows:
+	 * - moder_cnt is set to the number of mtu sized packets to
+	 *   satisfy our coelsing target.
+	 * - moder_time is set to a fixed value.
+	 */
+	priv->tx_frames = (mdev->profile.tx_moder_cnt ==
+			   MLX4_EN_MODER_AUTO_CONF) ?
+				MLX4_EN_TX_COAL_TARGET /
+				priv->dev->mtu + 1 :
+				mdev->profile.tx_moder_cnt;
+	priv->tx_usecs = (mdev->profile.tx_moder_time ==
+			  MLX4_EN_MODER_AUTO_CONF) ?
+				MLX4_EN_TX_COAL_TIME :
+				mdev->profile.tx_moder_time;
+	priv->rx_frames = (mdev->profile.rx_moder_cnt ==
+			   MLX4_EN_MODER_AUTO_CONF) ?
+				MLX4_EN_RX_COAL_TARGET /
+				priv->dev->mtu + 1 :
+				mdev->profile.rx_moder_cnt;
+	priv->rx_usecs = (mdev->profile.rx_moder_time ==
+			  MLX4_EN_MODER_AUTO_CONF) ?
+				MLX4_EN_RX_COAL_TIME :
+				mdev->profile.rx_moder_time;
+	mlx4_dbg(mdev, "Default coalesing params for mtu:%d - tx_frames:%d "
+			"tx_usecs:%d rx_frames:%d rx_usecs:%d\n",
+			priv->dev->mtu, priv->tx_frames, priv->tx_usecs,
+			priv->rx_frames, priv->rx_usecs);
+
+	/* Setup cq moderation params */
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		cq = &priv->rx_cq[i];
+		cq->moder_cnt = priv->rx_frames;
+		cq->moder_time = priv->rx_usecs;
+	}
+
+	/* Reset auto-moderation params */
+	priv->pkt_rate_low = MLX4_EN_RX_RATE_LOW;
+	priv->rx_usecs_low = MLX4_EN_RX_COAL_TIME_LOW;
+	priv->pkt_rate_high = MLX4_EN_RX_RATE_HIGH;
+	priv->rx_usecs_high = MLX4_EN_RX_COAL_TIME_HIGH;
+	priv->sample_interval = MLX4_EN_SAMPLE_INTERVAL;
+	priv->adaptive_rx_coal = mdev->profile.auto_moder;
+	priv->last_moder_time = MLX4_EN_MODER_AUTO_CONF;
+	priv->last_moder_jiffies = 0;
+	priv->last_moder_packets = 0;
+	priv->last_moder_tx_packets = 0;
+	priv->last_moder_bytes = 0;
+}
+
+static void mlx4_en_auto_moderation(struct mlx4_en_priv *priv)
+{
+	unsigned long period = (unsigned long) (jiffies - priv->last_moder_jiffies);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct mlx4_en_cq *cq;
+	unsigned long packets;
+	unsigned long rate;
+	unsigned long avg_pkt_size;
+	unsigned long rx_packets;
+	unsigned long rx_bytes;
+	unsigned long tx_packets;
+	unsigned long tx_pkt_diff;
+	unsigned long rx_pkt_diff;
+	int moder_time;
+	int i, err;
+
+	if (!priv->adaptive_rx_coal || period < priv->sample_interval * HZ)
+		return;
+
+	if (DO_SW_COUNTERS) {
+		struct mlx4_en_rx_ring *rx_ring;
+		struct mlx4_en_tx_ring *tx_ring;
+
+		rx_packets = 0;
+		rx_bytes = 0;
+		tx_packets = 0;
+		for (i = 0; i < priv->rx_ring_num; i++) {
+			rx_ring = &priv->rx_ring[i];
+			rx_bytes += rx_ring->bytes;
+			rx_packets += rx_ring->packets;
+		}
+		for (i = 0; i < priv->tx_ring_num; i++) {
+			tx_ring = &priv->tx_ring[i];
+			tx_packets += tx_ring->packets;
+		}
+	} else {
+		spin_lock_bh(&priv->stats_lock);
+		rx_packets = priv->stats.rx_packets;
+		rx_bytes = priv->stats.rx_bytes;
+		tx_packets = priv->stats.tx_packets;
+		spin_unlock_bh(&priv->stats_lock);
+	}
+	if (!priv->last_moder_jiffies || !period)
+		goto out;
+
+	tx_pkt_diff = ((unsigned long) (tx_packets -
+					priv->last_moder_tx_packets));
+	rx_pkt_diff = ((unsigned long) (rx_packets -
+					priv->last_moder_packets));
+	packets = MAX(tx_pkt_diff, rx_pkt_diff);
+	rate = packets * HZ / period;
+	avg_pkt_size = packets ? ((unsigned long) (rx_bytes -
+				 priv->last_moder_bytes)) / packets : 0;
+
+	/* Apply auto-moderation only when packet rate exceeds a rate that
+	 * it matters */
+	if (rate > MLX4_EN_RX_RATE_THRESH) {
+		/* If tx and rx packet rates are not balanced, assume that
+		 * traffic is mainly BW bound and apply maximum moderation.
+		 * Otherwise, moderate according to packet rate */
+		if (2 * tx_pkt_diff > 3 * rx_pkt_diff ||
+		    2 * rx_pkt_diff > 3 * tx_pkt_diff) {
+			moder_time = priv->rx_usecs_high;
+		} else {
+			if (rate < priv->pkt_rate_low)
+				moder_time = priv->rx_usecs_low;
+			else if (rate > priv->pkt_rate_high)
+				moder_time = priv->rx_usecs_high;
+			else
+				moder_time = (rate - priv->pkt_rate_low) *
+					(priv->rx_usecs_high - priv->rx_usecs_low) /
+					(priv->pkt_rate_high - priv->pkt_rate_low) +
+					priv->rx_usecs_low;
+		}
+	} else {
+		/* When packet rate is low, use default moderation rather than
+		 * 0 to prevent interrupt storms if traffic suddenly increases */
+		moder_time = priv->rx_usecs;
+	}
+
+	mlx4_dbg(mdev, "tx rate:%lu rx_rate:%lu\n",
+		 tx_pkt_diff * HZ / period, rx_pkt_diff * HZ / period);
+
+	mlx4_dbg(mdev, "Rx moder_time changed from:%d to %d period:%lu [jiff] "
+			"packets:%lu avg_pkt_size:%lu rate:%lu [p/s])\n",
+			priv->last_moder_time, moder_time, period, packets,
+			avg_pkt_size, rate);
+
+	if (moder_time != priv->last_moder_time) {
+		priv->last_moder_time = moder_time;
+		for (i = 0; i < priv->rx_ring_num; i++) {
+			cq = &priv->rx_cq[i];
+			cq->moder_time = moder_time;
+			err = mlx4_en_set_cq_moder(priv, cq);
+			if (err) {
+				mlx4_err(mdev, "Failed modifying moderation "
+					       "for cq:%d on port:%d\n",
+						i, priv->port);
+				break;
+			}
+		}
+	}
+
+out:
+	priv->last_moder_packets = rx_packets;
+	priv->last_moder_tx_packets = tx_packets;
+	priv->last_moder_bytes = rx_bytes;
+	priv->last_moder_jiffies = jiffies;
+}
+
+static void mlx4_en_do_get_stats(struct work_struct *work)
+{
+	struct delayed_work *delay = container_of(work, struct delayed_work,
+						  work);
+	struct mlx4_en_priv *priv = container_of(delay, struct mlx4_en_priv,
+						 stats_task);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+	int i;
+
+	mlx4_dbg(mdev, "Updating statistics for port:%d\n", priv->port);
+
+	err = mlx4_en_DUMP_ETH_STATS(mdev, priv->port, 0);
+	if (err)
+		mlx4_dbg(mdev, "Could not update stats for "
+				"port:%d\n", priv->port);
+
+	if (DO_SW_COUNTERS) {
+		spin_lock_bh(&priv->stats_lock);
+		priv->stats.tx_bytes = 0;
+		priv->stats.tx_packets = 0;
+		for (i = 0; i < priv->tx_ring_num; i++) {
+			priv->stats.tx_bytes += priv->tx_ring[i].bytes;
+			priv->stats.tx_packets += priv->tx_ring[i].packets;
+		}
+		priv->stats.rx_bytes = 0;
+		priv->stats.rx_packets = 0;
+		for (i = 0; i < priv->rx_ring_num; i++) {
+			priv->stats.rx_bytes += priv->rx_ring[i].bytes;
+			priv->stats.rx_packets += priv->rx_ring[i].packets;
+		}
+		spin_unlock_bh(&priv->stats_lock);
+	}
+
+	down(&mdev->state_sem);
+	if (mdev->state == DEVICE_UP) {
+		if (priv->port_state == PORT_UP)
+			mlx4_en_auto_moderation(priv);
+
+		queue_delayed_work(mdev->workqueue, &priv->stats_task, STATS_DELAY);
+	}
+	up(&mdev->state_sem);
+}
+
+static void mlx4_en_linkstate(struct work_struct *work)
+{
+	struct mlx4_en_priv *priv = container_of(work, struct mlx4_en_priv,
+						 linkstate_task);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int linkstate = priv->link_state;
+
+	down(&mdev->state_sem);
+	/* If observable port state changed set carrier state and
+	 * report to system log */
+	if (priv->last_link_state != linkstate) {
+		if (linkstate == MLX4_DEV_EVENT_PORT_DOWN) {
+			mlx4_info(mdev, "Port %d - link down\n", priv->port);
+			netif_carrier_off(priv->dev);
+		} else {
+			mlx4_info(mdev, "Port %d - link up\n", priv->port);
+			netif_carrier_on(priv->dev);
+		}
+	}
+	priv->last_link_state = linkstate;
+	up(&mdev->state_sem);
+}
+
+
+int mlx4_en_start_port(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct mlx4_en_cq *cq;
+	struct mlx4_en_tx_ring *tx_ring;
+	struct mlx4_en_rx_ring *rx_ring;
+	int rx_index = 0;
+	int tx_index = 0;
+	u16 stride;
+	int err = 0;
+	int i;
+	int j;
+
+	mlx4_dbg(mdev, "starting port:%d\n", priv->port);
+	if (priv->port_state == PORT_UP) {
+		mlx4_dbg(mdev, "start port called while port already up\n");
+		return 0;
+	}
+
+	/* Calculate Rx buf size */
+	dev->mtu = MIN(dev->mtu, priv->max_mtu);
+	mlx4_en_calc_rx_buf(dev);
+	mlx4_dbg(mdev, "Rx buf size:%d\n", priv->rx_skb_size);
+	stride = roundup_pow_of_two(sizeof(struct mlx4_en_rx_desc) +
+				    DS_SIZE * priv->num_frags);
+	/* Configure rx cq's and rings */
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		cq = &priv->rx_cq[i];
+		rx_ring = &priv->rx_ring[i];
+
+		err = mlx4_en_create_cq(priv, cq, priv->prof->rx_ring_size, i, RX);
+		if (err) {
+			mlx4_err(mdev, "Failed allocating Rx CQ\n");
+			goto rx_err;
+		}
+		for (j = 0; j < cq->size; j++)
+			cq->buf[j].op_tr_own = MLX4_CQE_OWNER_MASK;
+		err = mlx4_en_set_cq_moder(priv, cq);
+		if (err) {
+			mlx4_err(mdev, "Failed setting cq moderation parameters");
+			mlx4_en_destroy_cq(priv, cq);
+			goto rx_err;
+		}
+		mlx4_cq_arm(&cq->mcq, MLX4_CQ_DB_REQ_NOT, mdev->uar_map,
+			    &mdev->uar_lock);
+
+		err = mlx4_en_create_rx_ring(priv, rx_ring,
+					     priv->prof->rx_ring_size,
+					     stride, cq->mcq.cqn);
+		if (err) {
+			mlx4_err(mdev, "Failed allocating Rx ring\n");
+			mlx4_en_destroy_cq(priv, cq);
+			goto rx_err;
+		}
+		++rx_index;
+	}
+
+	err = mlx4_en_config_rss_steer(priv);
+	if (err) {
+		mlx4_err(mdev, "Failed configuring rss steering\n");
+		goto rx_err;
+	}
+
+	/* Configure tx cq's and rings */
+	for (i = 0; i < priv->tx_ring_num; i++) {
+		/* Configure cq */
+		cq = &priv->tx_cq[i];
+		err = mlx4_en_create_cq(priv, cq, priv->prof->tx_ring_size, i, TX);
+		if (err) {
+			mlx4_err(mdev, "Failed allocating Tx CQ\n");
+			goto tx_err;
+		}
+		mlx4_dbg(mdev, "Resetting index of collapsed CQ:%d "
+				"to -1\n", i);
+			cq->buf->index = 0xffff;
+
+		/* Configure ring */
+		tx_ring = &priv->tx_ring[i];
+		err = mlx4_en_activate_tx_ring(priv, tx_ring, cq->mcq.cqn,
+					       priv->rx_ring[0].srq.srqn);
+		if (err) {
+			mlx4_err(mdev, "Failed allocating Tx ring\n");
+			mlx4_en_destroy_cq(priv, cq);
+			goto tx_err;
+		}
+		/* Set initial ownership of all Tx TXBBs to SW (1) */
+		for (j = 0; j < tx_ring->buf_size; j += STAMP_STRIDE)
+			*((u32 *) (tx_ring->buf + j)) = 0xffffffff;
+		++tx_index;
+	}
+
+	/* Configure port */
+	err = mlx4_SET_PORT_general(mdev->dev, priv->port,
+				    priv->rx_skb_size + ETH_FCS_LEN,
+				    mdev->profile.tx_pause,
+				    mdev->profile.tx_ppp,
+				    mdev->profile.rx_pause,
+				    mdev->profile.rx_ppp);
+	if (err) {
+		mlx4_err(mdev, "Failed setting port general configurations"
+			       " for port %d, with error %d\n", priv->port, err);
+		goto tx_err;
+	}
+	/* Set default qp number */
+	err = mlx4_SET_PORT_qpn_calc(mdev->dev, priv->port, priv->base_qpn, 0);
+	if (err) {
+		mlx4_err(mdev, "Failed setting default qp numbers\n");
+		goto tx_err;
+	}
+	/* Set port mac number */
+	mlx4_dbg(mdev, "Setting mac for port %d\n", priv->port);
+	err = mlx4_register_mac(mdev->dev, priv->port,
+				priv->mac, &priv->mac_index);
+	if (err) {
+		mlx4_err(mdev, "Failed setting port mac\n");
+		goto tx_err;
+	}
+
+	/* Init port */
+	mlx4_dbg(mdev, "Initializing port\n");
+	err = mlx4_INIT_PORT(mdev->dev, priv->port);
+	if (err) {
+		mlx4_err(mdev, "Failed Initializing port\n");
+		goto mac_err;
+	}
+
+	/* Schedule multicast task to populate multicast list */
+	queue_work(mdev->workqueue, &priv->mcast_task);
+
+	priv->port_state = PORT_UP;
+	netif_start_queue(dev);
+	return 0;
+
+mac_err:
+	mlx4_unregister_mac(mdev->dev, priv->port, priv->mac_index);
+tx_err:
+	while (tx_index--) {
+		mlx4_en_deactivate_tx_ring(priv, &priv->tx_ring[tx_index]);
+		mlx4_en_destroy_cq(priv, &priv->tx_cq[tx_index]);
+	}
+
+	mlx4_en_release_rss_steer(priv);
+rx_err:
+	while (rx_index--) {
+		mlx4_en_destroy_rx_ring(priv, &priv->rx_ring[rx_index]);
+		mlx4_en_destroy_cq(priv, &priv->rx_cq[rx_index]);
+	}
+
+	return err; /* need to close devices */
+}
+
+
+void mlx4_en_stop_port(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int i;
+
+	mlx4_dbg(mdev, "Stop_port called for port:%d\n", priv->port);
+
+	if (priv->port_state == PORT_DOWN) {
+		mlx4_dbg(mdev, "stop port (%d) called while port already down\n",
+			 priv->port);
+		return;
+	}
+	netif_stop_queue(dev);
+
+	/* Synchronize with tx routine */
+	netif_tx_lock_bh(dev);
+	priv->port_state = PORT_DOWN;
+	netif_tx_unlock_bh(dev);
+
+	/* close port*/
+	mlx4_CLOSE_PORT(mdev->dev, priv->port);
+
+	/* Unregister Mac address for the port */
+	mlx4_unregister_mac(mdev->dev, priv->port, priv->mac_index);
+
+	/* Free TX Rings */
+	for (i = 0; i < priv->tx_ring_num; i++) {
+		mlx4_en_deactivate_tx_ring(priv, &priv->tx_ring[i]);
+		mlx4_en_destroy_cq(priv, &priv->tx_cq[i]);
+	}
+	msleep(10);
+
+	for (i = 0; i < priv->tx_ring_num; i++)
+		mlx4_en_free_tx_buf(dev, &priv->tx_ring[i]);
+
+	/* Free RSS qps */
+	mlx4_en_release_rss_steer(priv);
+
+	/* Free RX Rings */
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		mlx4_en_destroy_rx_ring(priv, &priv->rx_ring[i]);
+		while (test_bit(NAPI_STATE_SCHED, &priv->rx_cq[i].napi.state))
+			msleep(1);
+		mlx4_en_destroy_cq(priv, &priv->rx_cq[i]);
+	}
+}
+
+static void mlx4_en_restart(struct work_struct *work)
+{
+	struct mlx4_en_priv *priv = container_of(work, struct mlx4_en_priv,
+						 watchdog_task);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	struct net_device *dev = priv->dev;
+
+	mlx4_dbg(mdev, "Watchdog task called for port %d\n", priv->port);
+	mlx4_en_stop_port(dev);
+	if (mlx4_en_start_port(dev))
+	    mlx4_err(mdev, "Failed restarting port %d\n", priv->port);
+}
+
+
+int mlx4_en_open(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int i;
+	int err = 0;
+
+	mlx4_dbg(mdev, "Open called for port:%d\n", priv->port);
+
+	down(&mdev->state_sem);
+
+	if (mdev->state != DEVICE_UP) {
+		mlx4_err(mdev, "Cannot open - device down/disabled\n");
+		err = -EBUSY;
+		goto out;
+	}
+
+	/* Reset HW statistics and performance counters */
+	if (mlx4_en_DUMP_ETH_STATS(mdev, priv->port, 1))
+		mlx4_dbg(mdev, "Failed dumping statistics\n");
+
+	memset(&priv->stats, 0, sizeof(priv->stats));
+	memset(&priv->pstats, 0, sizeof(priv->pstats));
+	if (DO_SW_COUNTERS) {
+		for (i = 0; i < priv->tx_ring_num; i++) {
+			priv->tx_ring[i].bytes = 0;
+			priv->tx_ring[i].packets = 0;
+		}
+		for (i = 0; i < priv->rx_ring_num; i++) {
+			priv->rx_ring[i].bytes = 0;
+			priv->rx_ring[i].packets = 0;
+		}
+	}
+	mlx4_en_set_default_moderation(priv);
+	err = mlx4_en_start_port(dev);
+	if (err)
+		mlx4_err(mdev, "Failed starting port:%d\n", priv->port);
+
+out:
+	up(&mdev->state_sem);
+	return err;
+}
+
+
+int mlx4_en_close(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	mlx4_info(mdev, "Close called for port:%d\n", priv->port);
+
+	down(&mdev->state_sem);
+
+	mlx4_en_stop_port(dev);
+	netif_carrier_off(dev);
+
+	up(&mdev->state_sem);
+	return 0;
+}
+
+
+void mlx4_en_destroy_netdev(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int i;
+
+	mlx4_dbg(mdev, "Destroying netdev on port:%d\n", priv->port);
+
+	/* Unregsiter device - this will close the port if it was up */
+	if (priv->registered)
+		unregister_netdev(dev);
+
+	mlx4_free_hwq_res(mdev->dev, &priv->res, MLX4_EN_PAGE_SIZE);
+
+	cancel_delayed_work(&priv->stats_task);
+	cancel_delayed_work(&priv->refill_task);
+	/* flush any pending task for this netdev */
+	flush_workqueue(mdev->workqueue);
+
+	/* Detach the netdev so tasks would not attempt to access it */
+	down(&mdev->state_sem);
+	mdev->pndev[priv->port] = NULL;
+	up(&mdev->state_sem);
+
+	for (i = 0; i < priv->tx_ring_num; i++)
+		mlx4_en_destroy_tx_ring(priv, &priv->tx_ring[i]);
+
+	free_netdev(dev);
+}
+
+static int mlx4_en_change_mtu(struct net_device *dev, int new_mtu)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err = 0;
+
+	mlx4_dbg(mdev, "Change MTU called - current:%d new:%d\n",
+		 dev->mtu, new_mtu);
+
+	if ((new_mtu < MLX4_EN_MIN_MTU) || (new_mtu > priv->max_mtu)) {
+		mlx4_err(mdev, "Bad MTU size:%d.\n", new_mtu);
+		return -EPERM;
+	}
+	dev->mtu = new_mtu;
+
+	if (netif_running(dev)) {
+		down(&mdev->state_sem);
+		if (mdev->state != DEVICE_UP) {
+			/* NIC is probably restarting - let watchdog task reset
+			 * the port */
+			mlx4_dbg(mdev, "Change MTU called with card down!?\n");
+		} else {
+			mlx4_en_stop_port(dev);
+			mlx4_en_set_default_moderation(priv);
+			err = mlx4_en_start_port(dev);
+			if (err) {
+				mlx4_err(mdev, "Failed restarting port:%d\n",
+					 priv->port);
+				queue_work(mdev->workqueue, &priv->watchdog_task);
+			}
+		}
+		up(&mdev->state_sem);
+	}
+	return 0;
+}
+
+int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
+			struct mlx4_en_port_profile *prof)
+{
+	struct net_device *dev;
+	struct mlx4_en_priv *priv;
+	int i;
+	int err;
+
+	dev = alloc_etherdev(sizeof(struct mlx4_en_priv));
+	if (dev == NULL) {
+		mlx4_err(mdev, "Net device allocation failed\n");
+		return -ENOMEM;
+	}
+
+	SET_NETDEV_DEV(dev, &mdev->dev->pdev->dev);
+
+	/*
+	 * Initialize driver private data
+	 */
+
+	priv = netdev_priv(dev);
+	memset(priv, 0, sizeof(struct mlx4_en_priv));
+	priv->dev = dev;
+	priv->mdev = mdev;
+	priv->prof = prof;
+	priv->port = port;
+	priv->port_state = PORT_DOWN;
+	priv->rx_csum = 1;
+	priv->rx_buf = prof->rx_buf;
+	priv->flags = prof->flags;
+	priv->tx_ring_num = prof->tx_ring_num;
+	priv->rx_ring_num = prof->rx_ring_num;
+	priv->mc_list = NULL;
+	priv->mac_index = -1;
+	spin_lock_init(&priv->stats_lock);
+	INIT_WORK(&priv->mcast_task, mlx4_en_do_set_multicast);
+	INIT_WORK(&priv->mac_task, mlx4_en_do_set_mac);
+	INIT_DELAYED_WORK(&priv->refill_task, mlx4_en_rx_refill);
+	INIT_WORK(&priv->watchdog_task, mlx4_en_restart);
+	INIT_WORK(&priv->linkstate_task, mlx4_en_linkstate);
+	INIT_DELAYED_WORK(&priv->stats_task, mlx4_en_do_get_stats);
+
+	/* Query for default mac and max mtu */
+	priv->max_mtu = mdev->dev->caps.eth_mtu_cap[priv->port];
+	priv->mac = mdev->dev->caps.def_mac[priv->port];
+	mlx4_dbg(mdev, " Max mtu is: %d, Mac is: 0x%llx\n", priv->max_mtu, priv->mac);
+
+	/* Create tx Rings */
+	for (i = 0; i < prof->tx_ring_num; i++) {
+		err = mlx4_en_create_tx_ring(priv, &priv->tx_ring[i],
+					     prof->tx_ring_size, TXBB_SIZE);
+		if (err) {
+			mlx4_err(mdev, "Failed to create tx ring\n");
+			goto out;
+		}
+	}
+
+	/* Populate Rx default RSS mappings */
+	mlx4_en_set_default_rss_map(mdev, &priv->rss_map, prof->rx_ring_num *
+						RSS_FACTOR, prof->rx_ring_num);
+	/* Allocate page for receive rings */
+	err = mlx4_alloc_hwq_res(mdev->dev, &priv->res,
+				MLX4_EN_PAGE_SIZE, MLX4_EN_PAGE_SIZE);
+	if (err) {
+		mlx4_err(mdev, "Failed to allocate page for rx qps\n");
+		goto out;
+	}
+
+	/* Populate Tx priority mappings */
+	mlx4_en_set_prio_map(mdev, priv->tx_prio_map, prof->tx_ring_num);
+
+	/*
+	 * Initialize netdev entry points
+	 */
+
+	dev->open = &mlx4_en_open;
+	dev->stop = &mlx4_en_close;
+	dev->hard_start_xmit = &mlx4_en_xmit;
+	dev->get_stats = &mlx4_en_get_stats;
+	dev->set_multicast_list = &mlx4_en_set_multicast;
+	dev->set_mac_address = &mlx4_en_set_mac;
+	dev->change_mtu = &mlx4_en_change_mtu;
+	dev->tx_timeout = &mlx4_en_tx_timeout;
+	dev->watchdog_timeo = MLX4_EN_WATCHDOG_TIMEOUT;
+	dev->vlan_rx_register = mlx4_en_vlan_rx_register;
+	dev->vlan_rx_add_vid = mlx4_en_vlan_rx_add_vid;
+	dev->vlan_rx_kill_vid = mlx4_en_vlan_rx_kill_vid;
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	dev->poll_controller = mlx4_en_netpoll;
+#endif
+	SET_ETHTOOL_OPS(dev, &mlx4_en_ethtool_ops);
+
+	/* Set defualt MAC */
+	dev->addr_len = ETH_ALEN;
+	for (i = 0; i < ETH_ALEN; i++)
+		dev->dev_addr[ETH_ALEN - 1 - i] =
+		(u8) (priv->mac >> (8 * i));
+
+	/*
+	 * Set driver featurs
+	 */
+	dev->features |= NETIF_F_SG;
+	dev->features |= NETIF_F_HW_CSUM;
+	if (mdev->mlx4_en_flags & MLX4_EN_FLAG_HIGH_DMA)
+		dev->features |= NETIF_F_HIGHDMA;
+	dev->features |= NETIF_F_HW_VLAN_TX |
+			 NETIF_F_HW_VLAN_RX |
+			 NETIF_F_HW_VLAN_FILTER;
+#ifdef NETIF_F_TSO
+	dev->features |= NETIF_F_TSO;
+#endif
+#ifdef NETIF_F_TSO6
+	dev->features |= NETIF_F_TSO6;
+#endif
+
+	mdev->pndev[port] = dev;
+
+	netif_carrier_off(dev);
+	err = register_netdev(dev);
+	if (err) {
+		mlx4_err(mdev, "Netdev registration failed\n");
+		goto out;
+	}
+	priv->registered = 1;
+	queue_delayed_work(mdev->workqueue, &priv->stats_task, STATS_DELAY);
+	return 0;
+
+out:
+	mlx4_en_destroy_netdev(dev);
+	return err;
+}
+
-- 
1.5.3.7




--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists