[PATCH NEXT 0/12] netxen: driver update

[PATCH NEXT 0/12] netxen: driver update

[Dhananjay Phadke]

Dave,

Series of new features / enhancements and some clean-up
for net-next-2.6. These should merge cleanly with
c8a5a658b826508c7c61b57e ("netxen: reset ring consumer
during cleanup") when net-next tree is rebased to include
2.6.31 fixes.

Please apply.

Thanks,
	Dhananjay

[PATCH NEXT 01/12] netxen: refresh firmware info after reset

[Dhananjay Phadke]

o move dma mask update to netxen_start_firmware() so that
  if firmware changes across suspend (e.g. file -> flash)
  it reflects right dma mask.
o re-read firmware capabilities after firmware reset.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic_hw.c   |    3 +++
 drivers/net/netxen/netxen_nic_init.c |    4 ----
 drivers/net/netxen/netxen_nic_main.c |    9 ++++-----
 3 files changed, 7 insertions(+), 9 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic_hw.c b/drivers/net/netxen/netxen_nic_hw.c
index b9123d4..3c614c3 100644
--- a/drivers/net/netxen/netxen_nic_hw.c
+++ b/drivers/net/netxen/netxen_nic_hw.c
@@ -2047,6 +2047,9 @@ void netxen_nic_get_firmware_info(struct netxen_adapter *adapter)
 		dev_info(&pdev->dev, "firmware running in %s mode\n",
 		adapter->ahw.cut_through ? "cut-through" : "legacy");
 	}
+
+	if (adapter->fw_version >= NETXEN_VERSION_CODE(4, 0, 222))
+		adapter->capabilities = NXRD32(adapter, CRB_FW_CAPABILITIES_1);
 }
 
 int
diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c
index 5d3343e..847f78b 100644
--- a/drivers/net/netxen/netxen_nic_init.c
+++ b/drivers/net/netxen/netxen_nic_init.c
@@ -1083,10 +1083,6 @@ int netxen_init_firmware(struct netxen_adapter *adapter)
 	NXWR32(adapter, CRB_MPORT_MODE, MPORT_MULTI_FUNCTION_MODE);
 	NXWR32(adapter, CRB_CMDPEG_STATE, PHAN_INITIALIZE_ACK);
 
-	if (adapter->fw_version >= NETXEN_VERSION_CODE(4, 0, 222)) {
-		adapter->capabilities = NXRD32(adapter, CRB_FW_CAPABILITIES_1);
-	}
-
 	return err;
 }
 
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 637ac8b..4c7e4c2 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -764,6 +764,10 @@ wait_init:
 		return err;
 	}
 
+	nx_update_dma_mask(adapter);
+
+	netxen_nic_get_firmware_info(adapter);
+
 	return 0;
 }
 
@@ -1071,11 +1075,6 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	err = netxen_start_firmware(adapter, 1);
 	if (err)
 		goto err_out_iounmap;
-
-	nx_update_dma_mask(adapter);
-
-	netxen_nic_get_firmware_info(adapter);
-
 	/*
 	 * See if the firmware gave us a virtual-physical port mapping.
 	 */
-- 
1.6.0.2

Re: [PATCH NEXT 01/12] netxen: refresh firmware info after reset

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:36 -0700

> o move dma mask update to netxen_start_firmware() so that
>   if firmware changes across suspend (e.g. file -> flash)
>   it reflects right dma mask.
> o re-read firmware capabilities after firmware reset.
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 02/12] netxen: enable ip addr hashing

[Dhananjay Phadke]

NX3031 hardware requires local IP addresses for packet
accumulation (LRO). IP address hashing is required to
distinguish a local TCP flow from others (forwarded or
guest).

This patch adds listener for IP and netdev events and
configures IP address in the firmware.

Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |    5 ++
 drivers/net/netxen/netxen_nic_hw.c   |   24 +++++++
 drivers/net/netxen/netxen_nic_init.c |   52 +++++++-------
 drivers/net/netxen/netxen_nic_main.c |  126 ++++++++++++++++++++++++++++++++++
 4 files changed, 182 insertions(+), 25 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index f86e050..6b9e759 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -380,6 +380,7 @@ struct rcv_desc {
 };
 
 /* opcode field in status_desc */
+#define NETXEN_NIC_SYN_OFFLOAD  0x03
 #define NETXEN_NIC_RXPKT_DESC  0x04
 #define NETXEN_OLD_RXPKT_DESC  0x3f
 #define NETXEN_NIC_RESPONSE_DESC 0x05
@@ -1078,6 +1079,9 @@ typedef struct {
 
 #define NX_MAC_EVENT		0x1
 
+#define NX_IP_UP		2
+#define NX_IP_DOWN		3
+
 /*
  * Driver --> Firmware
  */
@@ -1443,6 +1447,7 @@ void netxen_p3_free_mac_list(struct netxen_adapter *adapter);
 int netxen_p3_nic_set_promisc(struct netxen_adapter *adapter, u32);
 int netxen_config_intr_coalesce(struct netxen_adapter *adapter);
 int netxen_config_rss(struct netxen_adapter *adapter, int enable);
+int netxen_config_ipaddr(struct netxen_adapter *adapter, u32 ip, int cmd);
 int netxen_linkevent_request(struct netxen_adapter *adapter, int enable);
 void netxen_advert_link_change(struct netxen_adapter *adapter, int linkup);
 
diff --git a/drivers/net/netxen/netxen_nic_hw.c b/drivers/net/netxen/netxen_nic_hw.c
index 3c614c3..088611b 100644
--- a/drivers/net/netxen/netxen_nic_hw.c
+++ b/drivers/net/netxen/netxen_nic_hw.c
@@ -706,6 +706,30 @@ int netxen_config_rss(struct netxen_adapter *adapter, int enable)
 	return rv;
 }
 
+int netxen_config_ipaddr(struct netxen_adapter *adapter, u32 ip, int cmd)
+{
+	nx_nic_req_t req;
+	u64 word;
+	int rv;
+
+	memset(&req, 0, sizeof(nx_nic_req_t));
+	req.qhdr = cpu_to_le64(NX_HOST_REQUEST << 23);
+
+	word = NX_NIC_H2C_OPCODE_CONFIG_IPADDR | ((u64)adapter->portnum << 16);
+	req.req_hdr = cpu_to_le64(word);
+
+	req.words[0] = cpu_to_le64(cmd);
+	req.words[1] = cpu_to_le64(ip);
+
+	rv = netxen_send_cmd_descs(adapter, (struct cmd_desc_type0 *)&req, 1);
+	if (rv != 0) {
+		printk(KERN_ERR "%s: could not notify %s IP 0x%x reuqest\n",
+				adapter->netdev->name,
+				(cmd == NX_IP_UP) ? "Add" : "Remove", ip);
+	}
+	return rv;
+}
+
 int netxen_linkevent_request(struct netxen_adapter *adapter, int enable)
 {
 	nx_nic_req_t req;
diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c
index 847f78b..5d7a2c4 100644
--- a/drivers/net/netxen/netxen_nic_init.c
+++ b/drivers/net/netxen/netxen_nic_init.c
@@ -880,22 +880,10 @@ netxen_validate_firmware(struct netxen_adapter *adapter, const char *fwname)
 	return 0;
 }
 
-void netxen_request_firmware(struct netxen_adapter *adapter)
+static int
+netxen_p3_has_mn(struct netxen_adapter *adapter)
 {
 	u32 capability, flashed_ver;
-	u8 fw_type;
-	struct pci_dev *pdev = adapter->pdev;
-	int rc = 0;
-
-	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
-		fw_type = NX_P2_MN_ROMIMAGE;
-		goto request_fw;
-	} else {
-		fw_type = NX_P3_CT_ROMIMAGE;
-		goto request_fw;
-	}
-
-request_mn:
 	capability = 0;
 
 	netxen_rom_fast_read(adapter,
@@ -903,23 +891,35 @@ request_mn:
 	flashed_ver = NETXEN_DECODE_VERSION(flashed_ver);
 
 	if (flashed_ver >= NETXEN_VERSION_CODE(4, 0, 220)) {
+
 		capability = NXRD32(adapter, NX_PEG_TUNE_CAPABILITY);
-		if (capability & NX_PEG_TUNE_MN_PRESENT) {
-			fw_type = NX_P3_MN_ROMIMAGE;
-			goto request_fw;
-		}
+		if (capability & NX_PEG_TUNE_MN_PRESENT)
+			return 1;
+	}
+	return 0;
+}
+
+void netxen_request_firmware(struct netxen_adapter *adapter)
+{
+	u8 fw_type;
+	struct pci_dev *pdev = adapter->pdev;
+	int rc = 0;
+
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
+		fw_type = NX_P2_MN_ROMIMAGE;
+		goto request_fw;
 	}
 
-	fw_type = NX_FLASH_ROMIMAGE;
-	adapter->fw = NULL;
-	goto done;
+	fw_type = netxen_p3_has_mn(adapter) ?
+		NX_P3_MN_ROMIMAGE : NX_P3_CT_ROMIMAGE;
 
 request_fw:
 	rc = request_firmware(&adapter->fw, fw_name[fw_type], &pdev->dev);
 	if (rc != 0) {
-		if (fw_type == NX_P3_CT_ROMIMAGE) {
+		if (fw_type == NX_P3_MN_ROMIMAGE) {
 			msleep(1);
-			goto request_mn;
+			fw_type = NX_P3_CT_ROMIMAGE;
+			goto request_fw;
 		}
 
 		fw_type = NX_FLASH_ROMIMAGE;
@@ -931,9 +931,10 @@ request_fw:
 	if (rc != 0) {
 		release_firmware(adapter->fw);
 
-		if (fw_type == NX_P3_CT_ROMIMAGE) {
+		if (fw_type == NX_P3_MN_ROMIMAGE) {
 			msleep(1);
-			goto request_mn;
+			fw_type = NX_P3_CT_ROMIMAGE;
+			goto request_fw;
 		}
 
 		fw_type = NX_FLASH_ROMIMAGE;
@@ -1292,6 +1293,7 @@ netxen_process_rcv_ring(struct nx_host_sds_ring *sds_ring, int max)
 		switch (opcode) {
 		case NETXEN_NIC_RXPKT_DESC:
 		case NETXEN_OLD_RXPKT_DESC:
+		case NETXEN_NIC_SYN_OFFLOAD:
 			break;
 		case NETXEN_NIC_RESPONSE_DESC:
 			netxen_handle_fw_message(desc_cnt, consumer, sds_ring);
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 4c7e4c2..b368326 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -39,6 +39,7 @@
 #include <linux/if_vlan.h>
 #include <net/ip.h>
 #include <linux/ipv6.h>
+#include <linux/inetdevice.h>
 
 MODULE_DESCRIPTION("NetXen Multi port (1/10) Gigabit Network Driver");
 MODULE_LICENSE("GPL");
@@ -1780,6 +1781,125 @@ static void netxen_nic_poll_controller(struct net_device *netdev)
 }
 #endif
 
+#define is_netxen_netdev(dev) (dev->netdev_ops == &netxen_netdev_ops)
+
+static int
+netxen_destip_supported(struct netxen_adapter *adapter)
+{
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id))
+		return 0;
+
+	if (adapter->ahw.cut_through)
+		return 0;
+
+	return 1;
+}
+
+static int netxen_netdev_event(struct notifier_block *this,
+				 unsigned long event, void *ptr)
+{
+	struct netxen_adapter *adapter;
+	struct net_device *dev = (struct net_device *)ptr;
+	struct in_device *indev;
+
+recheck:
+	if (dev == NULL)
+		goto done;
+
+	if (dev->priv_flags & IFF_802_1Q_VLAN) {
+		dev = vlan_dev_real_dev(dev);
+		goto recheck;
+	}
+
+	if (!is_netxen_netdev(dev))
+		goto done;
+
+	adapter = netdev_priv(dev);
+
+	if (!adapter || !netxen_destip_supported(adapter))
+		goto done;
+
+	if (adapter->is_up != NETXEN_ADAPTER_UP_MAGIC)
+		goto done;
+
+	indev = in_dev_get(dev);
+	if (!indev)
+		goto done;
+
+	for_ifa(indev) {
+		switch (event) {
+		case NETDEV_UP:
+			netxen_config_ipaddr(adapter,
+					ifa->ifa_address, NX_IP_UP);
+			break;
+		case NETDEV_DOWN:
+			netxen_config_ipaddr(adapter,
+					ifa->ifa_address, NX_IP_DOWN);
+			break;
+		default:
+			break;
+		}
+	} endfor_ifa(indev);
+
+	in_dev_put(indev);
+done:
+	return NOTIFY_DONE;
+}
+
+static int
+netxen_inetaddr_event(struct notifier_block *this,
+		unsigned long event, void *ptr)
+{
+	struct netxen_adapter *adapter;
+	struct net_device *dev;
+
+	struct in_ifaddr *ifa = (struct in_ifaddr *)ptr;
+
+	dev = ifa->ifa_dev ? ifa->ifa_dev->dev : NULL;
+
+recheck:
+	if (dev == NULL || !netif_running(dev))
+		goto done;
+
+	if (dev->priv_flags & IFF_802_1Q_VLAN) {
+		dev = vlan_dev_real_dev(dev);
+		goto recheck;
+	}
+
+	if (!is_netxen_netdev(dev))
+		goto done;
+
+	adapter = netdev_priv(dev);
+
+	if (!adapter || !netxen_destip_supported(adapter))
+		goto done;
+
+	if (adapter->is_up != NETXEN_ADAPTER_UP_MAGIC)
+		goto done;
+
+	switch (event) {
+	case NETDEV_UP:
+		netxen_config_ipaddr(adapter, ifa->ifa_address, NX_IP_UP);
+		break;
+	case NETDEV_DOWN:
+		netxen_config_ipaddr(adapter, ifa->ifa_address, NX_IP_DOWN);
+		break;
+	default:
+		break;
+	}
+
+done:
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block	netxen_netdev_cb = {
+	.notifier_call = netxen_netdev_event,
+};
+
+static struct notifier_block netxen_inetaddr_cb = {
+	.notifier_call = netxen_inetaddr_event,
+};
+
 static struct pci_driver netxen_driver = {
 	.name = netxen_nic_driver_name,
 	.id_table = netxen_pci_tbl,
@@ -1800,6 +1920,9 @@ static int __init netxen_init_module(void)
 	if ((netxen_workq = create_singlethread_workqueue("netxen")) == NULL)
 		return -ENOMEM;
 
+	register_netdevice_notifier(&netxen_netdev_cb);
+	register_inetaddr_notifier(&netxen_inetaddr_cb);
+
 	return pci_register_driver(&netxen_driver);
 }
 
@@ -1808,6 +1931,9 @@ module_init(netxen_init_module);
 static void __exit netxen_exit_module(void)
 {
 	pci_unregister_driver(&netxen_driver);
+
+	unregister_inetaddr_notifier(&netxen_inetaddr_cb);
+	unregister_netdevice_notifier(&netxen_netdev_cb);
 	destroy_workqueue(netxen_workq);
 }
 
-- 
1.6.0.2

Re: [PATCH NEXT 02/12] netxen: enable ip addr hashing

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:37 -0700

> NX3031 hardware requires local IP addresses for packet
> accumulation (LRO). IP address hashing is required to
> distinguish a local TCP flow from others (forwarded or
> guest).
> 
> This patch adds listener for IP and netdev events and
> configures IP address in the firmware.
> 
> Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 03/12] netxen: configure interrupt coalesce defaults

[Dhananjay Phadke]

Initialize and configure interrupt coalesing defaults
in the firmware, so that these also reflect in "ethool -c".

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic_main.c |   20 ++++++++++++++++++++
 1 files changed, 20 insertions(+), 0 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index b368326..76bdd5c 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -819,6 +819,20 @@ netxen_nic_free_irq(struct netxen_adapter *adapter)
 	}
 }
 
+static void
+netxen_nic_init_coalesce_defaults(struct netxen_adapter *adapter)
+{
+	adapter->coal.flags = NETXEN_NIC_INTR_DEFAULT;
+	adapter->coal.normal.data.rx_time_us =
+		NETXEN_DEFAULT_INTR_COALESCE_RX_TIME_US;
+	adapter->coal.normal.data.rx_packets =
+		NETXEN_DEFAULT_INTR_COALESCE_RX_PACKETS;
+	adapter->coal.normal.data.tx_time_us =
+		NETXEN_DEFAULT_INTR_COALESCE_TX_TIME_US;
+	adapter->coal.normal.data.tx_packets =
+		NETXEN_DEFAULT_INTR_COALESCE_TX_PACKETS;
+}
+
 static int
 netxen_nic_up(struct netxen_adapter *adapter, struct net_device *netdev)
 {
@@ -841,6 +855,9 @@ netxen_nic_up(struct netxen_adapter *adapter, struct net_device *netdev)
 	if (adapter->max_sds_rings > 1)
 		netxen_config_rss(adapter, 1);
 
+	if (NX_IS_REVISION_P3(adapter->ahw.revision_id))
+		netxen_config_intr_coalesce(adapter);
+
 	netxen_napi_enable(adapter);
 
 	if (adapter->capabilities & NX_FW_CAPABILITY_LINK_NOTIFICATION)
@@ -936,6 +953,9 @@ netxen_nic_attach(struct netxen_adapter *adapter)
 		goto err_out_free_rxbuf;
 	}
 
+	if (NX_IS_REVISION_P3(adapter->ahw.revision_id))
+		netxen_nic_init_coalesce_defaults(adapter);
+
 	adapter->is_up = NETXEN_ADAPTER_UP_MAGIC;
 	return 0;
 
-- 
1.6.0.2

Re: [PATCH NEXT 03/12] netxen: configure interrupt coalesce defaults

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:38 -0700

> Initialize and configure interrupt coalesing defaults
> in the firmware, so that these also reflect in "ethool -c".
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 04/12] netxen: annotate dma watchdog setup

[Dhananjay Phadke]

o remove superfluous code to setup PCI dma watchdog for NX2031.
o disable dma watchdog completely for NX3031 (not required).

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |   73 +--------------------------------
 drivers/net/netxen/netxen_nic_hdr.h  |    2 +-
 drivers/net/netxen/netxen_nic_init.c |   59 ++++++++++++++++-----------
 drivers/net/netxen/netxen_nic_main.c |    8 ++--
 4 files changed, 43 insertions(+), 99 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index 6b9e759..3368af0 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -1324,24 +1324,6 @@ struct netxen_adapter {
 	const struct firmware *fw;
 };
 
-/*
- * NetXen dma watchdog control structure
- *
- *	Bit 0		: enabled => R/O: 1 watchdog active, 0 inactive
- *	Bit 1		: disable_request => 1 req disable dma watchdog
- *	Bit 2		: enable_request =>  1 req enable dma watchdog
- *	Bit 3-31	: unused
- */
-
-#define netxen_set_dma_watchdog_disable_req(config_word) \
-	_netxen_set_bits(config_word, 1, 1, 1)
-#define netxen_set_dma_watchdog_enable_req(config_word) \
-	_netxen_set_bits(config_word, 2, 1, 1)
-#define netxen_get_dma_watchdog_enabled(config_word) \
-	((config_word) & 0x1)
-#define netxen_get_dma_watchdog_disabled(config_word) \
-	(((config_word) >> 1) & 0x1)
-
 int netxen_niu_xgbe_enable_phy_interrupts(struct netxen_adapter *adapter);
 int netxen_niu_gbe_enable_phy_interrupts(struct netxen_adapter *adapter);
 int netxen_niu_xgbe_disable_phy_interrupts(struct netxen_adapter *adapter);
@@ -1402,8 +1384,9 @@ unsigned long netxen_nic_pci_set_window_2M(struct netxen_adapter *adapter,
 		unsigned long long addr);
 
 /* Functions from netxen_nic_init.c */
-void netxen_free_adapter_offload(struct netxen_adapter *adapter);
-int netxen_initialize_adapter_offload(struct netxen_adapter *adapter);
+int netxen_init_dummy_dma(struct netxen_adapter *adapter);
+void netxen_free_dummy_dma(struct netxen_adapter *adapter);
+
 int netxen_phantom_init(struct netxen_adapter *adapter, int pegtune_val);
 int netxen_load_firmware(struct netxen_adapter *adapter);
 int netxen_need_fw_reset(struct netxen_adapter *adapter);
@@ -1510,56 +1493,6 @@ static inline void get_brd_name_by_type(u32 type, char *name)
 		name = "Unknown";
 }
 
-static inline int
-dma_watchdog_shutdown_request(struct netxen_adapter *adapter)
-{
-	u32 ctrl;
-
-	/* check if already inactive */
-	ctrl = adapter->hw_read_wx(adapter,
-			NETXEN_CAM_RAM(NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL));
-
-	if (netxen_get_dma_watchdog_enabled(ctrl) == 0)
-		return 1;
-
-	/* Send the disable request */
-	netxen_set_dma_watchdog_disable_req(ctrl);
-	NXWR32(adapter, NETXEN_CAM_RAM(NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL), ctrl);
-
-	return 0;
-}
-
-static inline int
-dma_watchdog_shutdown_poll_result(struct netxen_adapter *adapter)
-{
-	u32 ctrl;
-
-	ctrl = adapter->hw_read_wx(adapter,
-			NETXEN_CAM_RAM(NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL));
-
-	return (netxen_get_dma_watchdog_enabled(ctrl) == 0);
-}
-
-static inline int
-dma_watchdog_wakeup(struct netxen_adapter *adapter)
-{
-	u32 ctrl;
-
-	ctrl = adapter->hw_read_wx(adapter,
-			NETXEN_CAM_RAM(NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL));
-
-	if (netxen_get_dma_watchdog_enabled(ctrl))
-		return 1;
-
-	/* send the wakeup request */
-	netxen_set_dma_watchdog_enable_req(ctrl);
-
-	NXWR32(adapter, NETXEN_CAM_RAM(NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL), ctrl);
-
-	return 0;
-}
-
-
 static inline u32 netxen_tx_avail(struct nx_host_tx_ring *tx_ring)
 {
 	smp_mb();
diff --git a/drivers/net/netxen/netxen_nic_hdr.h b/drivers/net/netxen/netxen_nic_hdr.h
index 8241036..a732858 100644
--- a/drivers/net/netxen/netxen_nic_hdr.h
+++ b/drivers/net/netxen/netxen_nic_hdr.h
@@ -852,7 +852,7 @@ enum {
 #define NX_PEG_TUNE_MN_PRESENT		0x1
 #define NX_PEG_TUNE_CAPABILITY		(NETXEN_CAM_RAM(0x02c))
 
-#define NETXEN_CAM_RAM_DMA_WATCHDOG_CTRL		(0x14)
+#define NETXEN_DMA_WATCHDOG_CTRL	(NETXEN_CAM_RAM(0x14))
 #define NETXEN_PEG_ALIVE_COUNTER	(NETXEN_CAM_RAM(0xb0))
 
 #define	ISR_MSI_INT_TRIGGER(FUNC) (NETXEN_PCIX_PS_REG(PCIX_MSI_F(FUNC)))
diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c
index 5d7a2c4..9135845 100644
--- a/drivers/net/netxen/netxen_nic_init.c
+++ b/drivers/net/netxen/netxen_nic_init.c
@@ -954,19 +954,20 @@ netxen_release_firmware(struct netxen_adapter *adapter)
 		release_firmware(adapter->fw);
 }
 
-int netxen_initialize_adapter_offload(struct netxen_adapter *adapter)
+int netxen_init_dummy_dma(struct netxen_adapter *adapter)
 {
-	uint64_t addr;
-	uint32_t hi;
-	uint32_t lo;
+	u64 addr;
+	u32 hi, lo;
 
-	adapter->dummy_dma.addr =
-	    pci_alloc_consistent(adapter->pdev,
+	if (!NX_IS_REVISION_P2(adapter->ahw.revision_id))
+		return 0;
+
+	adapter->dummy_dma.addr = pci_alloc_consistent(adapter->pdev,
 				 NETXEN_HOST_DUMMY_DMA_SIZE,
 				 &adapter->dummy_dma.phys_addr);
 	if (adapter->dummy_dma.addr == NULL) {
-		printk("%s: ERROR: Could not allocate dummy DMA memory\n",
-		       __func__);
+		dev_err(&adapter->pdev->dev,
+			"ERROR: Could not allocate dummy DMA memory\n");
 		return -ENOMEM;
 	}
 
@@ -977,29 +978,41 @@ int netxen_initialize_adapter_offload(struct netxen_adapter *adapter)
 	NXWR32(adapter, CRB_HOST_DUMMY_BUF_ADDR_HI, hi);
 	NXWR32(adapter, CRB_HOST_DUMMY_BUF_ADDR_LO, lo);
 
-	if (NX_IS_REVISION_P3(adapter->ahw.revision_id)) {
-		uint32_t temp = 0;
-		NXWR32(adapter, CRB_HOST_DUMMY_BUF, temp);
-	}
-
 	return 0;
 }
 
-void netxen_free_adapter_offload(struct netxen_adapter *adapter)
+/*
+ * NetXen DMA watchdog control:
+ *
+ *	Bit 0		: enabled => R/O: 1 watchdog active, 0 inactive
+ *	Bit 1		: disable_request => 1 req disable dma watchdog
+ *	Bit 2		: enable_request =>  1 req enable dma watchdog
+ *	Bit 3-31	: unused
+ */
+void netxen_free_dummy_dma(struct netxen_adapter *adapter)
 {
 	int i = 100;
+	u32 ctrl;
+
+	if (!NX_IS_REVISION_P2(adapter->ahw.revision_id))
+		return;
 
 	if (!adapter->dummy_dma.addr)
 		return;
 
-	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
-		do {
-			if (dma_watchdog_shutdown_request(adapter) == 1)
-				break;
+	ctrl = NXRD32(adapter, NETXEN_DMA_WATCHDOG_CTRL);
+	if ((ctrl & 0x1) != 0) {
+		NXWR32(adapter, NETXEN_DMA_WATCHDOG_CTRL, (ctrl | 0x2));
+
+		while ((ctrl & 0x1) != 0) {
+
 			msleep(50);
-			if (dma_watchdog_shutdown_poll_result(adapter) == 1)
+
+			ctrl = NXRD32(adapter, NETXEN_DMA_WATCHDOG_CTRL);
+
+			if (--i == 0)
 				break;
-		} while (--i);
+		};
 	}
 
 	if (i) {
@@ -1008,10 +1021,8 @@ void netxen_free_adapter_offload(struct netxen_adapter *adapter)
 			    adapter->dummy_dma.addr,
 			    adapter->dummy_dma.phys_addr);
 		adapter->dummy_dma.addr = NULL;
-	} else {
-		printk(KERN_ERR "%s: dma_watchdog_shutdown failed\n",
-				adapter->netdev->name);
-	}
+	} else
+		dev_err(&adapter->pdev->dev, "dma_watchdog_shutdown failed\n");
 }
 
 int netxen_phantom_init(struct netxen_adapter *adapter, int pegtune_val)
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 76bdd5c..fb976cb 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -745,7 +745,7 @@ netxen_start_firmware(struct netxen_adapter *adapter, int request_fw)
 
 	}
 
-	err = netxen_initialize_adapter_offload(adapter);
+	err = netxen_init_dummy_dma(adapter);
 	if (err)
 		return err;
 
@@ -761,7 +761,7 @@ wait_init:
 	/* Handshake with the card before we register the devices. */
 	err = netxen_phantom_init(adapter, NETXEN_NIC_PEG_TUNE);
 	if (err) {
-		netxen_free_adapter_offload(adapter);
+		netxen_free_dummy_dma(adapter);
 		return err;
 	}
 
@@ -1154,7 +1154,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 err_out_disable_msi:
 	netxen_teardown_intr(adapter);
 
-	netxen_free_adapter_offload(adapter);
+	netxen_free_dummy_dma(adapter);
 
 err_out_iounmap:
 	netxen_cleanup_pci_map(adapter);
@@ -1189,7 +1189,7 @@ static void __devexit netxen_nic_remove(struct pci_dev *pdev)
 	}
 
 	if (adapter->portnum == 0)
-		netxen_free_adapter_offload(adapter);
+		netxen_free_dummy_dma(adapter);
 
 	netxen_teardown_intr(adapter);
 	netxen_free_sds_rings(&adapter->recv_ctx);
-- 
1.6.0.2

Re: [PATCH NEXT 04/12] netxen: annotate dma watchdog setup

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:39 -0700

> o remove superfluous code to setup PCI dma watchdog for NX2031.
> o disable dma watchdog completely for NX3031 (not required).
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 05/12] netxen: refactor tso code

[Dhananjay Phadke]

o move all tso / checksum offload code into netxen_tso_check().
o optimize the tso header copy into simple loop.
o clean up unnecessary unions from cmd_desc_type0 struct.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |   51 +++---------
 drivers/net/netxen/netxen_nic_main.c |  152 ++++++++++++++++------------------
 2 files changed, 84 insertions(+), 119 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index 3368af0..3a4cbd5 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -316,56 +316,29 @@ struct netxen_ring_ctx {
 	cpu_to_le16(((_flags) & 0x7f) | (((_opcode) & 0x3f) << 7))
 
 #define netxen_set_tx_frags_len(_desc, _frags, _len) \
-	(_desc)->num_of_buffers_total_length = \
+	(_desc)->nfrags__length = \
 	cpu_to_le32(((_frags) & 0xff) | (((_len) & 0xffffff) << 8))
 
 struct cmd_desc_type0 {
 	u8 tcp_hdr_offset;	/* For LSO only */
 	u8 ip_hdr_offset;	/* For LSO only */
-	/* Bit pattern: 0-6 flags, 7-12 opcode, 13-15 unused */
-	__le16 flags_opcode;
-	/* Bit pattern: 0-7 total number of segments,
-	   8-31 Total size of the packet */
-	__le32 num_of_buffers_total_length;
-	union {
-		struct {
-			__le32 addr_low_part2;
-			__le32 addr_high_part2;
-		};
-		__le64 addr_buffer2;
-	};
+	__le16 flags_opcode;	/* 15:13 unused, 12:7 opcode, 6:0 flags */
+	__le32 nfrags__length;	/* 31:8 total len, 7:0 frag count */
+
+	__le64 addr_buffer2;
 
-	__le16 reference_handle;	/* changed to u16 to add mss */
-	__le16 mss;		/* passed by NDIS_PACKET for LSO */
-	/* Bit pattern 0-3 port, 0-3 ctx id */
-	u8 port_ctxid;
+	__le16 reference_handle;
+	__le16 mss;
+	u8 port_ctxid;		/* 7:4 ctxid 3:0 port */
 	u8 total_hdr_length;	/* LSO only : MAC+IP+TCP Hdr size */
 	__le16 conn_id;		/* IPSec offoad only */
 
-	union {
-		struct {
-			__le32 addr_low_part3;
-			__le32 addr_high_part3;
-		};
-		__le64 addr_buffer3;
-	};
-	union {
-		struct {
-			__le32 addr_low_part1;
-			__le32 addr_high_part1;
-		};
-		__le64 addr_buffer1;
-	};
+	__le64 addr_buffer3;
+	__le64 addr_buffer1;
 
 	__le16 buffer_length[4];
 
-	union {
-		struct {
-			__le32 addr_low_part4;
-			__le32 addr_high_part4;
-		};
-		__le64 addr_buffer4;
-	};
+	__le64 addr_buffer4;
 
 	__le64 unused;
 
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index fb976cb..381bd8f 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -1310,13 +1310,18 @@ static int netxen_nic_close(struct net_device *netdev)
 	return 0;
 }
 
-static bool netxen_tso_check(struct net_device *netdev,
-		      struct cmd_desc_type0 *desc, struct sk_buff *skb)
+static void
+netxen_tso_check(struct net_device *netdev,
+		struct nx_host_tx_ring *tx_ring,
+		struct cmd_desc_type0 *first_desc,
+		struct sk_buff *skb)
 {
-	bool tso = false;
 	u8 opcode = TX_ETHER_PKT;
 	__be16 protocol = skb->protocol;
 	u16 flags = 0;
+	u32 producer;
+	int copied, offset, copy_len, hdr_len = 0, tso = 0;
+	struct cmd_desc_type0 *hwdesc;
 
 	if (protocol == cpu_to_be16(ETH_P_8021Q)) {
 		struct vlan_ethhdr *vh = (struct vlan_ethhdr *)skb->data;
@@ -1327,13 +1332,14 @@ static bool netxen_tso_check(struct net_device *netdev,
 	if ((netdev->features & (NETIF_F_TSO | NETIF_F_TSO6)) &&
 			skb_shinfo(skb)->gso_size > 0) {
 
-		desc->mss = cpu_to_le16(skb_shinfo(skb)->gso_size);
-		desc->total_hdr_length =
-			skb_transport_offset(skb) + tcp_hdrlen(skb);
+		hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+
+		first_desc->mss = cpu_to_le16(skb_shinfo(skb)->gso_size);
+		first_desc->total_hdr_length = hdr_len;
 
 		opcode = (protocol == cpu_to_be16(ETH_P_IPV6)) ?
 				TX_TCP_LSO6 : TX_TCP_LSO;
-		tso = true;
+		tso = 1;
 
 	} else if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u8 l4proto;
@@ -1354,10 +1360,39 @@ static bool netxen_tso_check(struct net_device *netdev,
 				opcode = TX_UDPV6_PKT;
 		}
 	}
-	desc->tcp_hdr_offset = skb_transport_offset(skb);
-	desc->ip_hdr_offset = skb_network_offset(skb);
-	netxen_set_tx_flags_opcode(desc, flags, opcode);
-	return tso;
+	first_desc->tcp_hdr_offset = skb_transport_offset(skb);
+	first_desc->ip_hdr_offset = skb_network_offset(skb);
+	netxen_set_tx_flags_opcode(first_desc, flags, opcode);
+
+	if (!tso)
+		return;
+
+	/* For LSO, we need to copy the MAC/IP/TCP headers into
+	 * the descriptor ring
+	 */
+	producer = tx_ring->producer;
+	copied = 0;
+	offset = 2;
+
+	while (copied < hdr_len) {
+
+		copy_len = min((int)sizeof(struct cmd_desc_type0) - offset,
+				(hdr_len - copied));
+
+		hwdesc = &tx_ring->desc_head[producer];
+		tx_ring->cmd_buf_arr[producer].skb = NULL;
+
+		skb_copy_from_linear_data_offset(skb, copied,
+				 (char *)hwdesc + offset, copy_len);
+
+		copied += copy_len;
+		offset = 0;
+
+		producer = get_next_index(producer, tx_ring->num_desc);
+	}
+
+	tx_ring->producer = producer;
+	barrier();
 }
 
 static void
@@ -1381,9 +1416,8 @@ netxen_clean_tx_dma_mapping(struct pci_dev *pdev,
 static inline void
 netxen_clear_cmddesc(u64 *desc)
 {
-	int i;
-	for (i = 0; i < 8; i++)
-		desc[i] = 0ULL;
+	desc[0] = 0ULL;
+	desc[2] = 0ULL;
 }
 
 static int
@@ -1391,18 +1425,18 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 {
 	struct netxen_adapter *adapter = netdev_priv(netdev);
 	struct nx_host_tx_ring *tx_ring = adapter->tx_ring;
-	unsigned int first_seg_len = skb->len - skb->data_len;
+	struct skb_frag_struct *frag;
 	struct netxen_cmd_buffer *pbuf;
 	struct netxen_skb_frag *buffrag;
-	struct cmd_desc_type0 *hwdesc;
-	struct pci_dev *pdev = adapter->pdev;
+	struct cmd_desc_type0 *hwdesc, *first_desc;
+	struct pci_dev *pdev;
 	dma_addr_t temp_dma;
 	int i, k;
+	unsigned long offset;
 
 	u32 producer;
-	int frag_count, no_of_desc;
+	int len, frag_count, no_of_desc;
 	u32 num_txd = tx_ring->num_desc;
-	bool is_tso = false;
 
 	frag_count = skb_shinfo(skb)->nr_frags + 1;
 
@@ -1416,32 +1450,30 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 
 	producer = tx_ring->producer;
 
-	hwdesc = &tx_ring->desc_head[producer];
-	netxen_clear_cmddesc((u64 *)hwdesc);
-	pbuf = &tx_ring->cmd_buf_arr[producer];
+	pdev = adapter->pdev;
+	len = skb->len - skb->data_len;
 
-	is_tso = netxen_tso_check(netdev, hwdesc, skb);
+	temp_dma = pci_map_single(pdev, skb->data, len, PCI_DMA_TODEVICE);
+	if (pci_dma_mapping_error(pdev, temp_dma))
+		goto drop_packet;
 
+	pbuf = &tx_ring->cmd_buf_arr[producer];
 	pbuf->skb = skb;
 	pbuf->frag_count = frag_count;
-	buffrag = &pbuf->frag_array[0];
-	temp_dma = pci_map_single(pdev, skb->data, first_seg_len,
-				      PCI_DMA_TODEVICE);
-	if (pci_dma_mapping_error(pdev, temp_dma))
-		goto drop_packet;
 
+	buffrag = &pbuf->frag_array[0];
 	buffrag->dma = temp_dma;
-	buffrag->length = first_seg_len;
+	buffrag->length = len;
+
+	first_desc = hwdesc = &tx_ring->desc_head[producer];
+	netxen_clear_cmddesc((u64 *)hwdesc);
 	netxen_set_tx_frags_len(hwdesc, frag_count, skb->len);
 	netxen_set_tx_port(hwdesc, adapter->portnum);
 
-	hwdesc->buffer_length[0] = cpu_to_le16(first_seg_len);
-	hwdesc->addr_buffer1 = cpu_to_le64(buffrag->dma);
+	hwdesc->buffer_length[0] = cpu_to_le16(len);
+	hwdesc->addr_buffer1 = cpu_to_le64(temp_dma);
 
 	for (i = 1, k = 1; i < frag_count; i++, k++) {
-		struct skb_frag_struct *frag;
-		int len, temp_len;
-		unsigned long offset;
 
 		/* move to next desc. if there is a need */
 		if ((i & 0x3) == 0) {
@@ -1452,11 +1484,11 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 			pbuf = &tx_ring->cmd_buf_arr[producer];
 			pbuf->skb = NULL;
 		}
+		buffrag = &pbuf->frag_array[i];
 		frag = &skb_shinfo(skb)->frags[i - 1];
 		len = frag->size;
 		offset = frag->page_offset;
 
-		temp_len = len;
 		temp_dma = pci_map_page(pdev, frag->page, offset,
 					len, PCI_DMA_TODEVICE);
 		if (pci_dma_mapping_error(pdev, temp_dma)) {
@@ -1464,11 +1496,10 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 			goto drop_packet;
 		}
 
-		buffrag++;
 		buffrag->dma = temp_dma;
-		buffrag->length = temp_len;
+		buffrag->length = len;
 
-		hwdesc->buffer_length[k] = cpu_to_le16(temp_len);
+		hwdesc->buffer_length[k] = cpu_to_le16(len);
 		switch (k) {
 		case 0:
 			hwdesc->addr_buffer1 = cpu_to_le64(temp_dma);
@@ -1483,53 +1514,14 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 			hwdesc->addr_buffer4 = cpu_to_le64(temp_dma);
 			break;
 		}
-		frag++;
 	}
-	producer = get_next_index(producer, num_txd);
+	tx_ring->producer = get_next_index(producer, num_txd);
 
-	/* For LSO, we need to copy the MAC/IP/TCP headers into
-	 * the descriptor ring
-	 */
-	if (is_tso) {
-		int hdr_len, first_hdr_len, more_hdr;
-		hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
-		if (hdr_len > (sizeof(struct cmd_desc_type0) - 2)) {
-			first_hdr_len = sizeof(struct cmd_desc_type0) - 2;
-			more_hdr = 1;
-		} else {
-			first_hdr_len = hdr_len;
-			more_hdr = 0;
-		}
-		/* copy the MAC/IP/TCP headers to the cmd descriptor list */
-		hwdesc = &tx_ring->desc_head[producer];
-		pbuf = &tx_ring->cmd_buf_arr[producer];
-		pbuf->skb = NULL;
-
-		/* copy the first 64 bytes */
-		memcpy(((void *)hwdesc) + 2,
-		       (void *)(skb->data), first_hdr_len);
-		producer = get_next_index(producer, num_txd);
-
-		if (more_hdr) {
-			hwdesc = &tx_ring->desc_head[producer];
-			pbuf = &tx_ring->cmd_buf_arr[producer];
-			pbuf->skb = NULL;
-			/* copy the next 64 bytes - should be enough except
-			 * for pathological case
-			 */
-			skb_copy_from_linear_data_offset(skb, first_hdr_len,
-							 hwdesc,
-							 (hdr_len -
-							  first_hdr_len));
-			producer = get_next_index(producer, num_txd);
-		}
-	}
-
-	tx_ring->producer = producer;
-	adapter->stats.txbytes += skb->len;
+	netxen_tso_check(netdev, tx_ring, first_desc, skb);
 
 	netxen_nic_update_cmd_producer(adapter, tx_ring);
 
+	adapter->stats.txbytes += skb->len;
 	adapter->stats.xmitcalled++;
 
 	return NETDEV_TX_OK;
-- 
1.6.0.2

Re: [PATCH NEXT 05/12] netxen: refactor tso code

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:40 -0700

> o move all tso / checksum offload code into netxen_tso_check().
> o optimize the tso header copy into simple loop.
> o clean up unnecessary unions from cmd_desc_type0 struct.
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 06/12] netxen: module params for ring sizes

[Dhananjay Phadke]

Add module parameters to set rx, jumbo and tx ring sizes
globally. 1GbE and 10GbE ports can have different ring
sizes.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h         |   19 +++++--
 drivers/net/netxen/netxen_nic_ethtool.c |   21 ++++----
 drivers/net/netxen/netxen_nic_main.c    |   89 +++++++++++++++++++++++++-----
 3 files changed, 98 insertions(+), 31 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index 3a4cbd5..e017827 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -200,13 +200,20 @@
 #define RCV_RING_JUMBO	1
 #define RCV_RING_LRO	2
 
-#define MAX_CMD_DESCRIPTORS		4096
-#define MAX_RCV_DESCRIPTORS		16384
-#define MAX_CMD_DESCRIPTORS_HOST	1024
-#define MAX_RCV_DESCRIPTORS_1G		2048
-#define MAX_RCV_DESCRIPTORS_10G		4096
-#define MAX_JUMBO_RCV_DESCRIPTORS	1024
+#define MIN_CMD_DESCRIPTORS		64
+#define MIN_RCV_DESCRIPTORS		64
+#define MIN_JUMBO_DESCRIPTORS		32
+
+#define MAX_CMD_DESCRIPTORS		1024
+#define MAX_RCV_DESCRIPTORS_1G		4096
+#define MAX_RCV_DESCRIPTORS_10G		8192
+#define MAX_JUMBO_RCV_DESCRIPTORS_1G	512
+#define MAX_JUMBO_RCV_DESCRIPTORS_10G	1024
 #define MAX_LRO_RCV_DESCRIPTORS		8
+
+#define DEFAULT_RCV_DESCRIPTORS_1G	2048
+#define DEFAULT_RCV_DESCRIPTORS_10G	4096
+
 #define NETXEN_CTX_SIGNATURE	0xdee0
 #define NETXEN_CTX_SIGNATURE_V2	0x0002dee0
 #define NETXEN_CTX_RESET	0xbad0
diff --git a/drivers/net/netxen/netxen_nic_ethtool.c b/drivers/net/netxen/netxen_nic_ethtool.c
index e16ea46..80d286d 100644
--- a/drivers/net/netxen/netxen_nic_ethtool.c
+++ b/drivers/net/netxen/netxen_nic_ethtool.c
@@ -494,20 +494,21 @@ netxen_nic_get_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
 {
 	struct netxen_adapter *adapter = netdev_priv(dev);
 
-	ring->rx_pending = 0;
-	ring->rx_jumbo_pending = 0;
-	ring->rx_pending += adapter->recv_ctx.
-		rds_rings[RCV_RING_NORMAL].num_desc;
-	ring->rx_jumbo_pending += adapter->recv_ctx.
-		rds_rings[RCV_RING_JUMBO].num_desc;
+	ring->rx_pending = adapter->num_rxd;
+	ring->rx_jumbo_pending = adapter->num_jumbo_rxd;
+	ring->rx_jumbo_pending += adapter->num_lro_rxd;
 	ring->tx_pending = adapter->num_txd;
 
-	if (adapter->ahw.port_type == NETXEN_NIC_GBE)
+	if (adapter->ahw.port_type == NETXEN_NIC_GBE) {
 		ring->rx_max_pending = MAX_RCV_DESCRIPTORS_1G;
-	else
+		ring->rx_jumbo_max_pending = MAX_JUMBO_RCV_DESCRIPTORS_1G;
+	} else {
 		ring->rx_max_pending = MAX_RCV_DESCRIPTORS_10G;
-	ring->tx_max_pending = MAX_CMD_DESCRIPTORS_HOST;
-	ring->rx_jumbo_max_pending = MAX_JUMBO_RCV_DESCRIPTORS;
+		ring->rx_jumbo_max_pending = MAX_JUMBO_RCV_DESCRIPTORS_10G;
+	}
+
+	ring->tx_max_pending = MAX_CMD_DESCRIPTORS;
+
 	ring->rx_mini_max_pending = 0;
 	ring->rx_mini_pending = 0;
 }
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 381bd8f..3f64a10 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -58,6 +58,26 @@ static int use_msi = 1;
 
 static int use_msi_x = 1;
 
+static u32 tx_desc = MAX_CMD_DESCRIPTORS;
+module_param(tx_desc, uint, S_IRUGO);
+MODULE_PARM_DESC(tx_desc, "Transmit descriptors count");
+
+static u32 rdesc_1g = DEFAULT_RCV_DESCRIPTORS_1G;
+module_param(rdesc_1g, uint, S_IRUGO);
+MODULE_PARM_DESC(rdesc_1g, "Receive descriptors count for 1GbE");
+
+static u32 rdesc_10g = DEFAULT_RCV_DESCRIPTORS_10G;
+module_param(rdesc_10g, uint, S_IRUGO);
+MODULE_PARM_DESC(rdesc_10g, "Receive descriptors count for 10GbE");
+
+static u32 jumbo_desc_1g = MAX_JUMBO_RCV_DESCRIPTORS_1G;
+module_param(jumbo_desc_1g, uint, S_IRUGO);
+MODULE_PARM_DESC(jumbo_desc_1g, "Jumbo receive descriptors count for 1GbE");
+
+static u32 jumbo_desc_10g = MAX_JUMBO_RCV_DESCRIPTORS_10G;
+module_param(jumbo_desc_10g, uint, S_IRUGO);
+MODULE_PARM_DESC(jumbo_desc_10g, "Jumbo receive descriptors count for 10GbE");
+
 /* Local functions to NetXen NIC driver */
 static int __devinit netxen_nic_probe(struct pci_dev *pdev,
 		const struct pci_device_id *ent);
@@ -283,12 +303,16 @@ nx_update_dma_mask(struct netxen_adapter *adapter)
 	return 0;
 }
 
-static void netxen_check_options(struct netxen_adapter *adapter)
+static void
+netxen_check_options(struct netxen_adapter *adapter)
 {
-	if (adapter->ahw.port_type == NETXEN_NIC_XGBE)
-		adapter->num_rxd = MAX_RCV_DESCRIPTORS_10G;
-	else if (adapter->ahw.port_type == NETXEN_NIC_GBE)
-		adapter->num_rxd = MAX_RCV_DESCRIPTORS_1G;
+	if (adapter->ahw.port_type == NETXEN_NIC_XGBE) {
+		adapter->num_rxd = rdesc_10g;
+		adapter->num_jumbo_rxd = jumbo_desc_10g;
+	} else if (adapter->ahw.port_type == NETXEN_NIC_GBE) {
+		adapter->num_rxd = rdesc_1g;
+		adapter->num_jumbo_rxd = jumbo_desc_1g;
+	}
 
 	adapter->msix_supported = 0;
 	if (NX_IS_REVISION_P3(adapter->ahw.revision_id)) {
@@ -306,11 +330,15 @@ static void netxen_check_options(struct netxen_adapter *adapter)
 		}
 	}
 
-	adapter->num_txd = MAX_CMD_DESCRIPTORS_HOST;
-	adapter->num_jumbo_rxd = MAX_JUMBO_RCV_DESCRIPTORS;
-	adapter->num_lro_rxd = MAX_LRO_RCV_DESCRIPTORS;
+	adapter->num_txd = tx_desc;
 
-	return;
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
+		adapter->num_lro_rxd = MAX_LRO_RCV_DESCRIPTORS;
+		adapter->max_rds_rings = 3;
+	} else {
+		adapter->num_lro_rxd = 0;
+		adapter->max_rds_rings = 2;
+	}
 }
 
 static int
@@ -908,11 +936,6 @@ netxen_nic_attach(struct netxen_adapter *adapter)
 		return -EIO;
 	}
 
-	if (adapter->fw_major < 4)
-		adapter->max_rds_rings = 3;
-	else
-		adapter->max_rds_rings = 2;
-
 	err = netxen_alloc_sw_resources(adapter);
 	if (err) {
 		printk(KERN_ERR "%s: Error in setting sw resources\n",
@@ -1923,11 +1946,47 @@ static struct pci_driver netxen_driver = {
 #endif
 };
 
-/* Driver Registration on NetXen card    */
+u32
+netxen_validate_ringparam(u32 val, u32 min, u32 max, char *r_name)
+{
+	u32 num_desc;
+	num_desc = max(val, min);
+	num_desc = min(num_desc, max);
+	num_desc = roundup_pow_of_two(num_desc);
+
+	if (val != num_desc) {
+		printk(KERN_INFO "%s: setting %s ring size %d instead of %d\n",
+		       netxen_nic_driver_name, r_name, num_desc, val);
+	}
+
+	return num_desc;
+}
+
+static void
+netxen_check_module_params(void)
+{
+	rdesc_1g = netxen_validate_ringparam(rdesc_1g,
+			MIN_RCV_DESCRIPTORS, MAX_RCV_DESCRIPTORS_1G,
+			"1GbE rx");
+	rdesc_10g = netxen_validate_ringparam(rdesc_10g,
+			MIN_RCV_DESCRIPTORS, MAX_RCV_DESCRIPTORS_10G,
+			"10GbE rx");
+	jumbo_desc_1g = netxen_validate_ringparam(jumbo_desc_1g,
+			MIN_JUMBO_DESCRIPTORS, MAX_JUMBO_RCV_DESCRIPTORS_1G,
+			"1GbE jumbo rx");
+	jumbo_desc_10g = netxen_validate_ringparam(jumbo_desc_10g,
+			MIN_JUMBO_DESCRIPTORS, MAX_JUMBO_RCV_DESCRIPTORS_10G,
+			"10GbE jumbo rx");
+	tx_desc = netxen_validate_ringparam(tx_desc,
+			MIN_CMD_DESCRIPTORS, MAX_CMD_DESCRIPTORS,
+			"10GbE jumbo rx");
+}
+
 
 static int __init netxen_init_module(void)
 {
 	printk(KERN_INFO "%s\n", netxen_nic_driver_string);
+	netxen_check_module_params();
 
 	if ((netxen_workq = create_singlethread_workqueue("netxen")) == NULL)
 		return -ENOMEM;
-- 
1.6.0.2

Re: [PATCH NEXT 06/12] netxen: module params for ring sizes

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:41 -0700

> Add module parameters to set rx, jumbo and tx ring sizes
> globally. 1GbE and 10GbE ports can have different ring
> sizes.
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

We do not add module parameters for stuff like this.

Invoke ethtool early enough, by whatever means, to set
such things as desired.

[PATCH NEXT 07/12] netxen: clean up frimware version checks

[Dhananjay Phadke]

NX2031 firmware version will never be > 4.0.0, so replace
(adapter->fw_major < 4) checks with pci revision ID check.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |    2 +-
 drivers/net/netxen/netxen_nic_ctx.c  |    6 +++---
 drivers/net/netxen/netxen_nic_hw.c   |    1 -
 drivers/net/netxen/netxen_nic_init.c |    2 +-
 drivers/net/netxen/netxen_nic_main.c |    6 +++---
 5 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index e017827..b58d0b2 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -1299,7 +1299,7 @@ struct netxen_adapter {
 
 	nx_nic_intr_coalesce_t coal;
 
-	u32 fw_major;
+	u32 resv5;
 	u32 fw_version;
 	const struct firmware *fw;
 };
diff --git a/drivers/net/netxen/netxen_nic_ctx.c b/drivers/net/netxen/netxen_nic_ctx.c
index 9f8ae47..9e04696 100644
--- a/drivers/net/netxen/netxen_nic_ctx.c
+++ b/drivers/net/netxen/netxen_nic_ctx.c
@@ -647,7 +647,7 @@ int netxen_alloc_hw_resources(struct netxen_adapter *adapter)
 		}
 		rds_ring->desc_head = (struct rcv_desc *)addr;
 
-		if (adapter->fw_major < 4)
+		if (NX_IS_REVISION_P2(adapter->ahw.revision_id))
 			rds_ring->crb_rcv_producer =
 				recv_crb_registers[port].crb_rcv_producer[ring];
 	}
@@ -675,7 +675,7 @@ int netxen_alloc_hw_resources(struct netxen_adapter *adapter)
 	}
 
 
-	if (adapter->fw_major >= 4) {
+	if (!NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 		err = nx_fw_cmd_create_rx_ctx(adapter);
 		if (err)
 			goto err_out_free;
@@ -705,7 +705,7 @@ void netxen_free_hw_resources(struct netxen_adapter *adapter)
 
 	int port = adapter->portnum;
 
-	if (adapter->fw_major >= 4) {
+	if (!NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 		nx_fw_cmd_destroy_rx_ctx(adapter);
 		nx_fw_cmd_destroy_tx_ctx(adapter);
 	} else {
diff --git a/drivers/net/netxen/netxen_nic_hw.c b/drivers/net/netxen/netxen_nic_hw.c
index 088611b..ddb9deb 100644
--- a/drivers/net/netxen/netxen_nic_hw.c
+++ b/drivers/net/netxen/netxen_nic_hw.c
@@ -2045,7 +2045,6 @@ void netxen_nic_get_firmware_info(struct netxen_adapter *adapter)
 	fw_minor = NXRD32(adapter, NETXEN_FW_VERSION_MINOR);
 	fw_build = NXRD32(adapter, NETXEN_FW_VERSION_SUB);
 
-	adapter->fw_major = fw_major;
 	adapter->fw_version = NETXEN_VERSION_CODE(fw_major, fw_minor, fw_build);
 
 	if (adapter->portnum == 0) {
diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c
index 9135845..ef7f1b1 100644
--- a/drivers/net/netxen/netxen_nic_init.c
+++ b/drivers/net/netxen/netxen_nic_init.c
@@ -1477,7 +1477,7 @@ netxen_post_rx_buffers(struct netxen_adapter *adapter, u32 ringid,
 		NXWR32(adapter, rds_ring->crb_rcv_producer,
 				(producer-1) & (rds_ring->num_desc-1));
 
-		if (adapter->fw_major < 4) {
+		if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 			/*
 			 * Write a doorbell msg to tell phanmon of change in
 			 * receive ring producer
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 3f64a10..8d4c30a 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -952,7 +952,7 @@ netxen_nic_attach(struct netxen_adapter *adapter)
 		goto err_out_free_sw;
 	}
 
-	if (adapter->fw_major < 4) {
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 		tx_ring = adapter->tx_ring;
 		tx_ring->crb_cmd_producer = crb_cmd_producer[adapter->portnum];
 		tx_ring->crb_cmd_consumer = crb_cmd_consumer[adapter->portnum];
@@ -1123,7 +1123,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	 * See if the firmware gave us a virtual-physical port mapping.
 	 */
 	adapter->physical_port = adapter->portnum;
-	if (adapter->fw_major < 4) {
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 		i = NXRD32(adapter, CRB_V2P(adapter->portnum));
 		if (i != 0x55555555)
 			adapter->physical_port = i;
@@ -1747,7 +1747,7 @@ static irqreturn_t netxen_intr(int irq, void *data)
 	}
 
 	/* clear interrupt */
-	if (adapter->fw_major < 4)
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id))
 		netxen_nic_disable_int(sds_ring);
 
 	adapter->pci_write_immediate(adapter,
-- 
1.6.0.2

Re: [PATCH NEXT 07/12] netxen: clean up frimware version checks

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:42 -0700

> NX2031 firmware version will never be > 4.0.0, so replace
> (adapter->fw_major < 4) checks with pci revision ID check.
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 08/12] netxen: refactor net_device setup code

[Dhananjay Phadke]

Move all net_device initialization into one function
netxen_setup_netdev().

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic_main.c |  138 ++++++++++++++++++---------------
 1 files changed, 75 insertions(+), 63 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 8d4c30a..8a4e6ce 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -86,7 +86,7 @@ static int netxen_nic_open(struct net_device *netdev);
 static int netxen_nic_close(struct net_device *netdev);
 static int netxen_nic_xmit_frame(struct sk_buff *, struct net_device *);
 static void netxen_tx_timeout(struct net_device *netdev);
-static void netxen_tx_timeout_task(struct work_struct *work);
+static void netxen_reset_task(struct work_struct *work);
 static void netxen_watchdog(unsigned long);
 static int netxen_nic_poll(struct napi_struct *napi, int budget);
 #ifdef CONFIG_NET_POLL_CONTROLLER
@@ -202,7 +202,7 @@ netxen_napi_add(struct netxen_adapter *adapter, struct net_device *netdev)
 	struct netxen_recv_context *recv_ctx = &adapter->recv_ctx;
 
 	if (netxen_alloc_sds_rings(recv_ctx, adapter->max_sds_rings))
-		return 1;
+		return -ENOMEM;
 
 	for (ring = 0; ring < adapter->max_sds_rings; ring++) {
 		sds_ring = &recv_ctx->sds_rings[ring];
@@ -1001,6 +1001,68 @@ netxen_nic_detach(struct netxen_adapter *adapter)
 	adapter->is_up = 0;
 }
 
+static int
+netxen_setup_netdev(struct netxen_adapter *adapter,
+		struct net_device *netdev)
+{
+	int err = 0;
+	struct pci_dev *pdev = adapter->pdev;
+
+	adapter->rx_csum = 1;
+	adapter->mc_enabled = 0;
+	if (NX_IS_REVISION_P3(adapter->ahw.revision_id))
+		adapter->max_mc_count = 38;
+	else
+		adapter->max_mc_count = 16;
+
+	netdev->netdev_ops	   = &netxen_netdev_ops;
+	netdev->watchdog_timeo     = 2*HZ;
+
+	netxen_nic_change_mtu(netdev, netdev->mtu);
+
+	SET_ETHTOOL_OPS(netdev, &netxen_nic_ethtool_ops);
+
+	netdev->features |= (NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO);
+	netdev->features |= (NETIF_F_GRO);
+	netdev->vlan_features |= (NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO);
+
+	if (NX_IS_REVISION_P3(adapter->ahw.revision_id)) {
+		netdev->features |= (NETIF_F_IPV6_CSUM | NETIF_F_TSO6);
+		netdev->vlan_features |= (NETIF_F_IPV6_CSUM | NETIF_F_TSO6);
+	}
+
+	if (adapter->pci_using_dac) {
+		netdev->features |= NETIF_F_HIGHDMA;
+		netdev->vlan_features |= NETIF_F_HIGHDMA;
+	}
+
+	netdev->irq = adapter->msix_entries[0].vector;
+
+	err = netxen_napi_add(adapter, netdev);
+	if (err)
+		return err;
+
+	init_timer(&adapter->watchdog_timer);
+	adapter->watchdog_timer.function = &netxen_watchdog;
+	adapter->watchdog_timer.data = (unsigned long)adapter;
+	INIT_WORK(&adapter->watchdog_task, netxen_watchdog_task);
+	INIT_WORK(&adapter->tx_timeout_task, netxen_reset_task);
+
+	if (netxen_read_mac_addr(adapter))
+		dev_warn(&pdev->dev, "failed to read mac addr\n");
+
+	netif_carrier_off(netdev);
+	netif_stop_queue(netdev);
+
+	err = register_netdev(netdev);
+	if (err) {
+		dev_err(&pdev->dev, "failed to register net device\n");
+		return err;
+	}
+
+	return 0;
+}
+
 static int __devinit
 netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
@@ -1038,9 +1100,8 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	netdev = alloc_etherdev(sizeof(struct netxen_adapter));
 	if(!netdev) {
-		printk(KERN_ERR"%s: Failed to allocate memory for the "
-				"device block.Check system memory resource"
-				" usage.\n", netxen_nic_driver_name);
+		dev_err(&pdev->dev, "failed to allocate net_device\n");
+		err = -ENOMEM;
 		goto err_out_free_res;
 	}
 
@@ -1068,38 +1129,10 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	/* This will be reset for mezz cards  */
 	adapter->portnum = pci_func_id;
-	adapter->rx_csum = 1;
-	adapter->mc_enabled = 0;
-	if (NX_IS_REVISION_P3(revision_id))
-		adapter->max_mc_count = 38;
-	else
-		adapter->max_mc_count = 16;
-
-	netdev->netdev_ops	   = &netxen_netdev_ops;
-	netdev->watchdog_timeo     = 2*HZ;
-
-	netxen_nic_change_mtu(netdev, netdev->mtu);
-
-	SET_ETHTOOL_OPS(netdev, &netxen_nic_ethtool_ops);
-
-	netdev->features |= (NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO);
-	netdev->features |= (NETIF_F_GRO);
-	netdev->vlan_features |= (NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO);
-
-	if (NX_IS_REVISION_P3(revision_id)) {
-		netdev->features |= (NETIF_F_IPV6_CSUM | NETIF_F_TSO6);
-		netdev->vlan_features |= (NETIF_F_IPV6_CSUM | NETIF_F_TSO6);
-	}
-
-	if (adapter->pci_using_dac) {
-		netdev->features |= NETIF_F_HIGHDMA;
-		netdev->vlan_features |= NETIF_F_HIGHDMA;
-	}
 
-	if (netxen_nic_get_board_info(adapter) != 0) {
-		printk("%s: Error getting board config info.\n",
-				netxen_nic_driver_name);
-		err = -EIO;
+	err = netxen_nic_get_board_info(adapter);
+	if (err) {
+		dev_err(&pdev->dev, "Error getting board config info.\n");
 		goto err_out_iounmap;
 	}
 
@@ -1119,6 +1152,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	err = netxen_start_firmware(adapter, 1);
 	if (err)
 		goto err_out_iounmap;
+
 	/*
 	 * See if the firmware gave us a virtual-physical port mapping.
 	 */
@@ -1133,31 +1167,9 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	netxen_setup_intr(adapter);
 
-	netdev->irq = adapter->msix_entries[0].vector;
-
-	if (netxen_napi_add(adapter, netdev))
-		goto err_out_disable_msi;
-
-	init_timer(&adapter->watchdog_timer);
-	adapter->watchdog_timer.function = &netxen_watchdog;
-	adapter->watchdog_timer.data = (unsigned long)adapter;
-	INIT_WORK(&adapter->watchdog_task, netxen_watchdog_task);
-	INIT_WORK(&adapter->tx_timeout_task, netxen_tx_timeout_task);
-
-	err = netxen_read_mac_addr(adapter);
+	err = netxen_setup_netdev(adapter, netdev);
 	if (err)
-		dev_warn(&pdev->dev, "failed to read mac addr\n");
-
-	netif_carrier_off(netdev);
-	netif_stop_queue(netdev);
-
-	if ((err = register_netdev(netdev))) {
-		printk(KERN_ERR "%s: register_netdev failed port #%d"
-			       " aborting\n", netxen_nic_driver_name,
-			       adapter->portnum);
-		err = -EIO;
 		goto err_out_disable_msi;
-	}
 
 	pci_set_drvdata(pdev, adapter);
 
@@ -1676,10 +1688,13 @@ static void netxen_tx_timeout(struct net_device *netdev)
 {
 	struct netxen_adapter *adapter = (struct netxen_adapter *)
 						netdev_priv(netdev);
+
+	dev_err(&netdev->dev, "transmit timeout, resetting.\n");
+
 	SCHEDULE_WORK(&adapter->tx_timeout_task);
 }
 
-static void netxen_tx_timeout_task(struct work_struct *work)
+static void netxen_reset_task(struct work_struct *work)
 {
 	struct netxen_adapter *adapter =
 		container_of(work, struct netxen_adapter, tx_timeout_task);
@@ -1687,9 +1702,6 @@ static void netxen_tx_timeout_task(struct work_struct *work)
 	if (!netif_running(adapter->netdev))
 		return;
 
-	printk(KERN_ERR "%s %s: transmit timeout, resetting.\n",
-	       netxen_nic_driver_name, adapter->netdev->name);
-
 	netxen_napi_disable(adapter);
 
 	adapter->netdev->trans_start = jiffies;
-- 
1.6.0.2

Re: [PATCH NEXT 08/12] netxen: refactor net_device setup code

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:43 -0700

> Move all net_device initialization into one function
> netxen_setup_netdev().
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 09/12] netxen: fix skb alloc size for legacy firmware

[Dhananjay Phadke]

Request 1532 bytes skb data size for NX3031. NX2031 firmware
needs 1760 sized buffers.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |   11 +++--------
 drivers/net/netxen/netxen_nic_init.c |   13 +++++++++----
 2 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index b58d0b2..e7702f6 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -143,18 +143,13 @@
 #define NX_ETHERMTU                    1500
 #define NX_MAX_ETHERHDR                32 /* This contains some padding */
 
-#define NX_RX_NORMAL_BUF_MAX_LEN       (NX_MAX_ETHERHDR + NX_ETHERMTU)
+#define NX_P2_RX_BUF_MAX_LEN           1760
+#define NX_P3_RX_BUF_MAX_LEN           (NX_MAX_ETHERHDR + NX_ETHERMTU)
 #define NX_P2_RX_JUMBO_BUF_MAX_LEN     (NX_MAX_ETHERHDR + P2_MAX_MTU)
 #define NX_P3_RX_JUMBO_BUF_MAX_LEN     (NX_MAX_ETHERHDR + P3_MAX_MTU)
 #define NX_CT_DEFAULT_RX_BUF_LEN	2048
 
-#define MAX_RX_BUFFER_LENGTH		1760
-#define MAX_RX_JUMBO_BUFFER_LENGTH 	8062
-#define MAX_RX_LRO_BUFFER_LENGTH	(8062)
-#define RX_DMA_MAP_LEN			(MAX_RX_BUFFER_LENGTH - 2)
-#define RX_JUMBO_DMA_MAP_LEN	\
-	(MAX_RX_JUMBO_BUFFER_LENGTH - 2)
-#define RX_LRO_DMA_MAP_LEN		(MAX_RX_LRO_BUFFER_LENGTH - 2)
+#define NX_RX_LRO_BUFFER_LENGTH		(8060)
 
 /*
  * Maximum number of ring contexts
diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c
index ef7f1b1..a6b244a 100644
--- a/drivers/net/netxen/netxen_nic_init.c
+++ b/drivers/net/netxen/netxen_nic_init.c
@@ -247,9 +247,14 @@ int netxen_alloc_sw_resources(struct netxen_adapter *adapter)
 				rds_ring->skb_size =
 					NX_CT_DEFAULT_RX_BUF_LEN;
 			} else {
-				rds_ring->dma_size = RX_DMA_MAP_LEN;
+				if (NX_IS_REVISION_P3(adapter->ahw.revision_id))
+					rds_ring->dma_size =
+						NX_P3_RX_BUF_MAX_LEN;
+				else
+					rds_ring->dma_size =
+						NX_P2_RX_BUF_MAX_LEN;
 				rds_ring->skb_size =
-					MAX_RX_BUFFER_LENGTH;
+					rds_ring->dma_size + NET_IP_ALIGN;
 			}
 			break;
 
@@ -267,8 +272,8 @@ int netxen_alloc_sw_resources(struct netxen_adapter *adapter)
 
 		case RCV_RING_LRO:
 			rds_ring->num_desc = adapter->num_lro_rxd;
-			rds_ring->dma_size = RX_LRO_DMA_MAP_LEN;
-			rds_ring->skb_size = MAX_RX_LRO_BUFFER_LENGTH;
+			rds_ring->dma_size = NX_RX_LRO_BUFFER_LENGTH;
+			rds_ring->skb_size = rds_ring->dma_size + NET_IP_ALIGN;
 			break;
 
 		}
-- 
1.6.0.2

Re: [PATCH NEXT 09/12] netxen: fix skb alloc size for legacy firmware

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:44 -0700

> Request 1532 bytes skb data size for NX3031. NX2031 firmware
> needs 1760 sized buffers.
> 
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 10/12] netxen: add vlan tx acceleration support

[Dhananjay Phadke]

Enable vlan tx acceleration for NX3031 if firmware advertises
capability.

Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h      |   11 ++++++-
 drivers/net/netxen/netxen_nic_main.c |   52 ++++++++++++++++++++++++++++++---
 2 files changed, 57 insertions(+), 6 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index e7702f6..d01cc09 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -304,6 +304,10 @@ struct netxen_ring_ctx {
 #define FLAGS_IPSEC_SA_ADD	0x04
 #define FLAGS_IPSEC_SA_DELETE	0x08
 #define FLAGS_VLAN_TAGGED	0x10
+#define FLAGS_VLAN_OOB		0x40
+
+#define netxen_set_tx_vlan_tci(cmd_desc, v)	\
+	(cmd_desc)->vlan_TCI = cpu_to_le16(v);
 
 #define netxen_set_cmd_desc_port(cmd_desc, var)	\
 	((cmd_desc)->port_ctxid |= ((var) & 0x0F))
@@ -342,7 +346,9 @@ struct cmd_desc_type0 {
 
 	__le64 addr_buffer4;
 
-	__le64 unused;
+	__le16 vlan_TCI;
+	__le16 reserved;
+	__le32 reserved2;
 
 } __attribute__ ((aligned(64)));
 
@@ -1111,6 +1117,9 @@ typedef struct {
 
 #define NX_FW_CAPABILITY_LINK_NOTIFICATION	(1 << 5)
 #define NX_FW_CAPABILITY_SWITCHING		(1 << 6)
+#define NX_FW_CAPABILITY_PEXQ			(1 << 7)
+#define NX_FW_CAPABILITY_BDG			(1 << 8)
+#define NX_FW_CAPABILITY_FVLANTX		(1 << 9)
 
 /* module types */
 #define LINKEVENT_MODULE_NOT_PRESENT			1
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 8a4e6ce..41ab99d 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -1036,6 +1036,9 @@ netxen_setup_netdev(struct netxen_adapter *adapter,
 		netdev->vlan_features |= NETIF_F_HIGHDMA;
 	}
 
+	if (adapter->capabilities & NX_FW_CAPABILITY_FVLANTX)
+		netdev->features |= (NETIF_F_HW_VLAN_TX);
+
 	netdev->irq = adapter->msix_entries[0].vector;
 
 	err = netxen_napi_add(adapter, netdev);
@@ -1353,15 +1356,24 @@ netxen_tso_check(struct net_device *netdev,
 {
 	u8 opcode = TX_ETHER_PKT;
 	__be16 protocol = skb->protocol;
-	u16 flags = 0;
+	u16 flags = 0, vid = 0;
 	u32 producer;
-	int copied, offset, copy_len, hdr_len = 0, tso = 0;
+	int copied, offset, copy_len, hdr_len = 0, tso = 0, vlan_oob = 0;
 	struct cmd_desc_type0 *hwdesc;
+	struct vlan_ethhdr *vh;
 
 	if (protocol == cpu_to_be16(ETH_P_8021Q)) {
-		struct vlan_ethhdr *vh = (struct vlan_ethhdr *)skb->data;
+
+		vh = (struct vlan_ethhdr *)skb->data;
 		protocol = vh->h_vlan_encapsulated_proto;
 		flags = FLAGS_VLAN_TAGGED;
+
+	} else if (vlan_tx_tag_present(skb)) {
+
+		flags = FLAGS_VLAN_OOB;
+		vid = vlan_tx_tag_get(skb);
+		netxen_set_tx_vlan_tci(first_desc, vid);
+		vlan_oob = 1;
 	}
 
 	if ((netdev->features & (NETIF_F_TSO | NETIF_F_TSO6)) &&
@@ -1371,6 +1383,13 @@ netxen_tso_check(struct net_device *netdev,
 
 		first_desc->mss = cpu_to_le16(skb_shinfo(skb)->gso_size);
 		first_desc->total_hdr_length = hdr_len;
+		if (vlan_oob) {
+			first_desc->total_hdr_length += VLAN_HLEN;
+			first_desc->tcp_hdr_offset = VLAN_HLEN;
+			first_desc->ip_hdr_offset = VLAN_HLEN;
+			/* Only in case of TSO on vlan device */
+			flags |= FLAGS_VLAN_TAGGED;
+		}
 
 		opcode = (protocol == cpu_to_be16(ETH_P_IPV6)) ?
 				TX_TCP_LSO6 : TX_TCP_LSO;
@@ -1395,8 +1414,9 @@ netxen_tso_check(struct net_device *netdev,
 				opcode = TX_UDPV6_PKT;
 		}
 	}
-	first_desc->tcp_hdr_offset = skb_transport_offset(skb);
-	first_desc->ip_hdr_offset = skb_network_offset(skb);
+
+	first_desc->tcp_hdr_offset += skb_transport_offset(skb);
+	first_desc->ip_hdr_offset += skb_network_offset(skb);
 	netxen_set_tx_flags_opcode(first_desc, flags, opcode);
 
 	if (!tso)
@@ -1409,6 +1429,28 @@ netxen_tso_check(struct net_device *netdev,
 	copied = 0;
 	offset = 2;
 
+	if (vlan_oob) {
+		/* Create a TSO vlan header template for firmware */
+
+		hwdesc = &tx_ring->desc_head[producer];
+		tx_ring->cmd_buf_arr[producer].skb = NULL;
+
+		copy_len = min((int)sizeof(struct cmd_desc_type0) - offset,
+				hdr_len + VLAN_HLEN);
+
+		vh = (struct vlan_ethhdr *)((char *)hwdesc + 2);
+		skb_copy_from_linear_data(skb, vh, 12);
+		vh->h_vlan_proto = htons(ETH_P_8021Q);
+		vh->h_vlan_TCI = htons(vid);
+		skb_copy_from_linear_data_offset(skb, 12,
+				(char *)vh + 16, copy_len - 16);
+
+		copied = copy_len;
+		offset = 0;
+
+		producer = get_next_index(producer, tx_ring->num_desc);
+	}
+
 	while (copied < hdr_len) {
 
 		copy_len = min((int)sizeof(struct cmd_desc_type0) - offset,
-- 
1.6.0.2

Re: [PATCH NEXT 10/12] netxen: add vlan tx acceleration support

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:45 -0700

> Enable vlan tx acceleration for NX3031 if firmware advertises
> capability.
> 
> Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 11/12] netxen: support for ethtool set ringparam

[Dhananjay Phadke]

Add support for ethtool -G to tune rx and tx ring sizes
per interface basis.

This is only supported for NX3031 based cards.

Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h         |    4 +++
 drivers/net/netxen/netxen_nic_ethtool.c |   44 +++++++++++++++++++++++++++++-
 drivers/net/netxen/netxen_nic_main.c    |   45 ++++++++++++++++++++++++-------
 3 files changed, 82 insertions(+), 11 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index d01cc09..9527f31 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -1427,6 +1427,10 @@ struct net_device_stats *netxen_nic_get_stats(struct net_device *netdev);
 void netxen_nic_update_cmd_producer(struct netxen_adapter *adapter,
 		struct nx_host_tx_ring *tx_ring);
 
+/* Functions from netxen_nic_main.c */
+u32 netxen_validate_ringparam(u32 val, u32 min, u32 max, char *);
+int netxen_nic_reset_context(struct netxen_adapter *);
+
 /*
  * NetXen Board information
  */
diff --git a/drivers/net/netxen/netxen_nic_ethtool.c b/drivers/net/netxen/netxen_nic_ethtool.c
index 80d286d..ec93aa7 100644
--- a/drivers/net/netxen/netxen_nic_ethtool.c
+++ b/drivers/net/netxen/netxen_nic_ethtool.c
@@ -490,7 +490,8 @@ netxen_nic_get_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom,
 }
 
 static void
-netxen_nic_get_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
+netxen_nic_get_ringparam(struct net_device *dev,
+		struct ethtool_ringparam *ring)
 {
 	struct netxen_adapter *adapter = netdev_priv(dev);
 
@@ -513,6 +514,46 @@ netxen_nic_get_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
 	ring->rx_mini_pending = 0;
 }
 
+static int
+netxen_nic_set_ringparam(struct net_device *dev,
+		struct ethtool_ringparam *ring)
+{
+	struct netxen_adapter *adapter = netdev_priv(dev);
+	u16 max_rcv_desc = MAX_RCV_DESCRIPTORS_10G;
+	u16 max_jumbo_desc = MAX_JUMBO_RCV_DESCRIPTORS_10G;
+	u16 num_rxd, num_jumbo_rxd, num_txd;
+
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id))
+		return -EOPNOTSUPP;
+
+	if (ring->rx_mini_pending)
+		return -EOPNOTSUPP;
+
+	if (adapter->ahw.port_type == NETXEN_NIC_GBE) {
+		max_rcv_desc = MAX_RCV_DESCRIPTORS_1G;
+		max_jumbo_desc = MAX_JUMBO_RCV_DESCRIPTORS_10G;
+	}
+
+	num_rxd = netxen_validate_ringparam(ring->rx_pending,
+			MIN_RCV_DESCRIPTORS, max_rcv_desc, "rx");
+
+	num_jumbo_rxd = netxen_validate_ringparam(ring->rx_jumbo_pending,
+			MIN_JUMBO_DESCRIPTORS, max_jumbo_desc, "rx jumbo");
+
+	num_txd = netxen_validate_ringparam(ring->tx_pending,
+			MIN_CMD_DESCRIPTORS, MAX_CMD_DESCRIPTORS, "tx");
+
+	if (num_rxd == adapter->num_rxd && num_txd == adapter->num_txd &&
+			num_jumbo_rxd == adapter->num_jumbo_rxd)
+		return 0;
+
+	adapter->num_rxd = num_rxd;
+	adapter->num_jumbo_rxd = num_jumbo_rxd;
+	adapter->num_txd = num_txd;
+
+	return netxen_nic_reset_context(adapter);
+}
+
 static void
 netxen_nic_get_pauseparam(struct net_device *dev,
 			  struct ethtool_pauseparam *pause)
@@ -894,6 +935,7 @@ struct ethtool_ops netxen_nic_ethtool_ops = {
 	.get_eeprom_len = netxen_nic_get_eeprom_len,
 	.get_eeprom = netxen_nic_get_eeprom,
 	.get_ringparam = netxen_nic_get_ringparam,
+	.set_ringparam = netxen_nic_set_ringparam,
 	.get_pauseparam = netxen_nic_get_pauseparam,
 	.set_pauseparam = netxen_nic_set_pauseparam,
 	.set_tx_csum = ethtool_op_set_tx_csum,
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 41ab99d..894b1fd 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -930,6 +930,9 @@ netxen_nic_attach(struct netxen_adapter *adapter)
 	struct nx_host_rds_ring *rds_ring;
 	struct nx_host_tx_ring *tx_ring;
 
+	if (adapter->is_up == NETXEN_ADAPTER_UP_MAGIC)
+		return 0;
+
 	err = netxen_init_firmware(adapter);
 	if (err != 0) {
 		printk(KERN_ERR "Failed to init firmware\n");
@@ -993,6 +996,9 @@ err_out_free_sw:
 static void
 netxen_nic_detach(struct netxen_adapter *adapter)
 {
+	if (adapter->is_up != NETXEN_ADAPTER_UP_MAGIC)
+		return;
+
 	netxen_free_hw_resources(adapter);
 	netxen_release_rx_buffers(adapter);
 	netxen_nic_free_irq(adapter);
@@ -1001,6 +1007,30 @@ netxen_nic_detach(struct netxen_adapter *adapter)
 	adapter->is_up = 0;
 }
 
+int
+netxen_nic_reset_context(struct netxen_adapter *adapter)
+{
+	int err = 0;
+	struct net_device *netdev = adapter->netdev;
+
+	if (adapter->is_up == NETXEN_ADAPTER_UP_MAGIC) {
+
+		if (netif_running(netdev))
+			netxen_nic_down(adapter, netdev);
+
+		netxen_nic_detach(adapter);
+
+		err = netxen_nic_attach(adapter);
+		if (err)
+			goto done;
+
+		if (netif_running(netdev))
+			err = netxen_nic_up(adapter, netdev);
+	}
+done:
+	return err;
+}
+
 static int
 netxen_setup_netdev(struct netxen_adapter *adapter,
 		struct net_device *netdev)
@@ -1222,9 +1252,7 @@ static void __devexit netxen_nic_remove(struct pci_dev *pdev)
 
 	unregister_netdev(netdev);
 
-	if (adapter->is_up == NETXEN_ADAPTER_UP_MAGIC) {
-		netxen_nic_detach(adapter);
-	}
+	netxen_nic_detach(adapter);
 
 	if (adapter->portnum == 0)
 		netxen_free_dummy_dma(adapter);
@@ -1256,8 +1284,7 @@ netxen_nic_suspend(struct pci_dev *pdev, pm_message_t state)
 	if (netif_running(netdev))
 		netxen_nic_down(adapter, netdev);
 
-	if (adapter->is_up == NETXEN_ADAPTER_UP_MAGIC)
-		netxen_nic_detach(adapter);
+	netxen_nic_detach(adapter);
 
 	pci_save_state(pdev);
 
@@ -1318,11 +1345,9 @@ static int netxen_nic_open(struct net_device *netdev)
 	if (adapter->driver_mismatch)
 		return -EIO;
 
-	if (adapter->is_up != NETXEN_ADAPTER_UP_MAGIC) {
-		err = netxen_nic_attach(adapter);
-		if (err)
-			return err;
-	}
+	err = netxen_nic_attach(adapter);
+	if (err)
+		return err;
 
 	err = netxen_nic_up(adapter, netdev);
 	if (err)
-- 
1.6.0.2

Re: [PATCH NEXT 11/12] netxen: support for ethtool set ringparam

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:46 -0700

> Add support for ethtool -G to tune rx and tx ring sizes
> per interface basis.
> 
> This is only supported for NX3031 based cards.
> 
> Signed-off-by: Amit Kumar Salecha <amit@netxen.com>
> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

[PATCH NEXT 12/12] netxen: update version to 4.0.41

[Dhananjay Phadke]

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
---
 drivers/net/netxen/netxen_nic.h |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h
index 9527f31..d4ac151 100644
--- a/drivers/net/netxen/netxen_nic.h
+++ b/drivers/net/netxen/netxen_nic.h
@@ -57,8 +57,8 @@
 
 #define _NETXEN_NIC_LINUX_MAJOR 4
 #define _NETXEN_NIC_LINUX_MINOR 0
-#define _NETXEN_NIC_LINUX_SUBVERSION 30
-#define NETXEN_NIC_LINUX_VERSIONID  "4.0.30"
+#define _NETXEN_NIC_LINUX_SUBVERSION 41
+#define NETXEN_NIC_LINUX_VERSIONID  "4.0.41"
 
 #define NETXEN_VERSION_CODE(a, b, c)	(((a) << 24) + ((b) << 16) + (c))
 #define _major(v)	(((v) >> 24) & 0xff)
-- 
1.6.0.2

Re: [PATCH NEXT 12/12] netxen: update version to 4.0.41

[David Miller]

From: Dhananjay Phadke <dhananjay@netxen.com>
Date: Sun, 26 Jul 2009 23:07:47 -0700

> Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>

Applied.

Re: [PATCH NEXT 12/12] netxen: update version to 4.0.41

[David Miller]

Ok, in order to make this build I'm going to have to insert
some parts of patch #6 which I rejected because that adds
ring size defaults/max/etc. defines which get used by later
patches.

RE: [PATCH NEXT 12/12] netxen: update version to 4.0.41

[Dhananjay Phadke]

Ok, the idea for module params was to set global defaults
(configured in modprobe.conf) vs. per interface ethtool calls.
The ethtool support is anyway added by # 12 patch.

Is it not good to have both options available? 

I will sync up my tree and see if the merge is functionally
correct.

Thanks,
Dhananjay

> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net] 
> Sent: Monday, July 27, 2009 10:54
> To: Dhananjay Phadke
> Cc: netdev@vger.kernel.org
> Subject: Re: [PATCH NEXT 12/12] netxen: update version to 4.0.41
> 
> 
> Ok, in order to make this build I'm going to have to insert
> some parts of patch #6 which I rejected because that adds
> ring size defaults/max/etc. defines which get used by later
> patches.
> 
> No virus found in this incoming message.
> Checked by AVG - www.avg.com 
> Version: 8.5.392 / Virus Database: 270.13.25/2256 - Release 
> Date: 07/27/09 05:58:00
> 

Re: [PATCH NEXT 12/12] netxen: update version to 4.0.41

[David Miller]

From: Dhananjay Phadke <dhananjay.phadke@qlogic.com>
Date: Mon, 27 Jul 2009 12:26:58 -0700

> Ok, the idea for module params was to set global defaults
> (configured in modprobe.conf) vs. per interface ethtool calls.
> The ethtool support is anyway added by # 12 patch.
> 
> Is it not good to have both options available? 

No, it is not good.

With module parameters we:

1) duplicate functionality

2) every device driver will name the module parms differently,
   therefore the interface for the user is different for every device
   type

We discuss this all the time, and this is a pretty deep rooted
belief and requirement for all network drivers.

Re: [PATCH NEXT 12/12] netxen: update version to 4.0.41

[Dhananjay Phadke]

Agreed that it duplicates work in individual drivers,
and the naming style followed by different vendors.

But the problem is ethtool can work only after driver
has successfully initialized interfaces. Otherwise,
ethtool gets no chance to tune up/down.

The driver defaults are for tuned for typical use cases.
On some frugal systems with lower amount of RAM, the
driver bails out failing to allocate enough ring space
or rx buffers [especially for 32-bit kernels]. This is
also a problem for multi-function (more than one PCI
function per physical port) mode, where number of
interfaces and hence the memory consumption gets
multiplied. For these cases something that can set
the bar before device initialization, is quite helpful.

Thanks,
Dhananjay



David Miller wrote:
> No, it is not good.
> 
> With module parameters we:
> 
> 1) duplicate functionality
> 
> 2) every device driver will name the module parms differently,
>    therefore the interface for the user is different for every device
>    type
> 
> We discuss this all the time, and this is a pretty deep rooted
> belief and requirement for all network drivers.