* [net-next-2.6 5/7] igb: add support for VF Transmit rate limit using iproute2
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Lior Levy, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Lior Levy <lior.levy@intel.com>
Implemented igb_ndo_set_vf_bw function which is being used
by iproute2 tool. In addition, updated igb_ndo_get_vf_config function
to show the actual rate limit to the user.
The rate limitation can be configured only when the link is up.
The rate limit value can be ranged between 0 and actual
link speed measured in Mbps. A value of '0' disables the rate limit for
this specific VF.
iproute2 usage will be 'ip link set ethX vf Y rate Z'.
After the command is made, the rate will be changed instantly.
To view the current rate limit, use 'ip link show ethX'.
The rates will be zeroed only upon driver reload or a link speed change.
This feature is being supported only by 82576 device.
Signed-off-by: Lior Levy <lior.levy@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_defines.h | 7 +++
drivers/net/igb/e1000_regs.h | 4 ++
drivers/net/igb/igb.h | 2 +
drivers/net/igb/igb_main.c | 88 ++++++++++++++++++++++++++++++++++++++-
4 files changed, 99 insertions(+), 2 deletions(-)
diff --git a/drivers/net/igb/e1000_defines.h b/drivers/net/igb/e1000_defines.h
index 6319ed9..ff46c91 100644
--- a/drivers/net/igb/e1000_defines.h
+++ b/drivers/net/igb/e1000_defines.h
@@ -770,4 +770,11 @@
#define E1000_PCIEMISC_LX_DECISION 0x00000080 /* Lx power decision based
on DMA coal */
+/* Tx Rate-Scheduler Config fields */
+#define E1000_RTTBCNRC_RS_ENA 0x80000000
+#define E1000_RTTBCNRC_RF_DEC_MASK 0x00003FFF
+#define E1000_RTTBCNRC_RF_INT_SHIFT 14
+#define E1000_RTTBCNRC_RF_INT_MASK \
+ (E1000_RTTBCNRC_RF_DEC_MASK << E1000_RTTBCNRC_RF_INT_SHIFT)
+
#endif
diff --git a/drivers/net/igb/e1000_regs.h b/drivers/net/igb/e1000_regs.h
index 8ac83c5..3a6f847 100644
--- a/drivers/net/igb/e1000_regs.h
+++ b/drivers/net/igb/e1000_regs.h
@@ -106,6 +106,10 @@
#define E1000_RQDPC(_n) (0x0C030 + ((_n) * 0x40))
+/* TX Rate Limit Registers */
+#define E1000_RTTDQSEL 0x3604 /* Tx Desc Plane Queue Select - WO */
+#define E1000_RTTBCNRC 0x36B0 /* Tx BCN Rate-Scheduler Config - WO */
+
/* Split and Replication RX Control - RW */
#define E1000_RXPBS 0x02404 /* Rx Packet Buffer Size - RW */
/*
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 92a4ef0..bbc5ebf 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -77,6 +77,7 @@ struct vf_data_storage {
unsigned long last_nack;
u16 pf_vlan; /* When set, guest VLAN config not allowed. */
u16 pf_qos;
+ u16 tx_rate;
};
#define IGB_VF_FLAG_CTS 0x00000001 /* VF is clear to send data */
@@ -323,6 +324,7 @@ struct igb_adapter {
u16 rx_ring_count;
unsigned int vfs_allocated_count;
struct vf_data_storage *vf_data;
+ int vf_rate_link_speed;
u32 rss_queues;
u32 wvbr;
};
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index cb6bf7b..88f8925 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -150,6 +150,7 @@ static int igb_ndo_set_vf_vlan(struct net_device *netdev,
static int igb_ndo_set_vf_bw(struct net_device *netdev, int vf, int tx_rate);
static int igb_ndo_get_vf_config(struct net_device *netdev, int vf,
struct ifla_vf_info *ivi);
+static void igb_check_vf_rate_limit(struct igb_adapter *);
#ifdef CONFIG_PM
static int igb_suspend(struct pci_dev *, pm_message_t);
@@ -3511,6 +3512,7 @@ static void igb_watchdog_task(struct work_struct *work)
netif_carrier_on(netdev);
igb_ping_all_vfs(adapter);
+ igb_check_vf_rate_limit(adapter);
/* link state has changed, schedule phy info update */
if (!test_bit(__IGB_DOWN, &adapter->state))
@@ -6599,9 +6601,91 @@ static int igb_ndo_set_vf_mac(struct net_device *netdev, int vf, u8 *mac)
return igb_set_vf_mac(adapter, vf, mac);
}
+static int igb_link_mbps(int internal_link_speed)
+{
+ switch (internal_link_speed) {
+ case SPEED_100:
+ return 100;
+ case SPEED_1000:
+ return 1000;
+ default:
+ return 0;
+ }
+}
+
+static void igb_set_vf_rate_limit(struct e1000_hw *hw, int vf, int tx_rate,
+ int link_speed)
+{
+ int rf_dec, rf_int;
+ u32 bcnrc_val;
+
+ if (tx_rate != 0) {
+ /* Calculate the rate factor values to set */
+ rf_int = link_speed / tx_rate;
+ rf_dec = (link_speed - (rf_int * tx_rate));
+ rf_dec = (rf_dec * (1<<E1000_RTTBCNRC_RF_INT_SHIFT)) / tx_rate;
+
+ bcnrc_val = E1000_RTTBCNRC_RS_ENA;
+ bcnrc_val |= ((rf_int<<E1000_RTTBCNRC_RF_INT_SHIFT) &
+ E1000_RTTBCNRC_RF_INT_MASK);
+ bcnrc_val |= (rf_dec & E1000_RTTBCNRC_RF_DEC_MASK);
+ } else {
+ bcnrc_val = 0;
+ }
+
+ wr32(E1000_RTTDQSEL, vf); /* vf X uses queue X */
+ wr32(E1000_RTTBCNRC, bcnrc_val);
+}
+
+static void igb_check_vf_rate_limit(struct igb_adapter *adapter)
+{
+ int actual_link_speed, i;
+ bool reset_rate = false;
+
+ /* VF TX rate limit was not set or not supported */
+ if ((adapter->vf_rate_link_speed == 0) ||
+ (adapter->hw.mac.type != e1000_82576))
+ return;
+
+ actual_link_speed = igb_link_mbps(adapter->link_speed);
+ if (actual_link_speed != adapter->vf_rate_link_speed) {
+ reset_rate = true;
+ adapter->vf_rate_link_speed = 0;
+ dev_info(&adapter->pdev->dev,
+ "Link speed has been changed. VF Transmit "
+ "rate is disabled\n");
+ }
+
+ for (i = 0; i < adapter->vfs_allocated_count; i++) {
+ if (reset_rate)
+ adapter->vf_data[i].tx_rate = 0;
+
+ igb_set_vf_rate_limit(&adapter->hw, i,
+ adapter->vf_data[i].tx_rate,
+ actual_link_speed);
+ }
+}
+
static int igb_ndo_set_vf_bw(struct net_device *netdev, int vf, int tx_rate)
{
- return -EOPNOTSUPP;
+ struct igb_adapter *adapter = netdev_priv(netdev);
+ struct e1000_hw *hw = &adapter->hw;
+ int actual_link_speed;
+
+ if (hw->mac.type != e1000_82576)
+ return -EOPNOTSUPP;
+
+ actual_link_speed = igb_link_mbps(adapter->link_speed);
+ if ((vf >= adapter->vfs_allocated_count) ||
+ (!(rd32(E1000_STATUS) & E1000_STATUS_LU)) ||
+ (tx_rate < 0) || (tx_rate > actual_link_speed))
+ return -EINVAL;
+
+ adapter->vf_rate_link_speed = actual_link_speed;
+ adapter->vf_data[vf].tx_rate = (u16)tx_rate;
+ igb_set_vf_rate_limit(hw, vf, tx_rate, actual_link_speed);
+
+ return 0;
}
static int igb_ndo_get_vf_config(struct net_device *netdev,
@@ -6612,7 +6696,7 @@ static int igb_ndo_get_vf_config(struct net_device *netdev,
return -EINVAL;
ivi->vf = vf;
memcpy(&ivi->mac, adapter->vf_data[vf].vf_mac_addresses, ETH_ALEN);
- ivi->tx_rate = 0;
+ ivi->tx_rate = adapter->vf_data[vf].tx_rate;
ivi->vlan = adapter->vf_data[vf].pf_vlan;
ivi->qos = adapter->vf_data[vf].pf_qos;
return 0;
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 7/7] igb: update version string
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Carolyn Wyborny, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Carolyn Wyborny <carolyn.wyborny@intel.com>
This will synchronize the version with the out of tree driver which
shares its functionality.
Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index a55fa17..579dbba 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -50,7 +50,7 @@
#endif
#include "igb.h"
-#define DRV_VERSION "2.1.0-k2"
+#define DRV_VERSION "2.4.13-k2"
char igb_driver_name[] = "igb";
char igb_driver_version[] = DRV_VERSION;
static const char igb_driver_string[] =
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 6/7] igb: Update Intel copyright notice for driver source.
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Carolyn Wyborny, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Carolyn Wyborny <carolyn.wyborny@intel.com>
This fix updates copyright information to include current year 2011.
Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 88f8925..a55fa17 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -55,7 +55,7 @@ char igb_driver_name[] = "igb";
char igb_driver_version[] = DRV_VERSION;
static const char igb_driver_string[] =
"Intel(R) Gigabit Ethernet Network Driver";
-static const char igb_copyright[] = "Copyright (c) 2007-2009 Intel Corporation.";
+static const char igb_copyright[] = "Copyright (c) 2007-2011 Intel Corporation.";
static const struct e1000_info *igb_info_tbl[] = {
[board_82575] = &e1000_82575_info,
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 4/7] igbvf: remove Tx hang detection
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Lior Levy, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Lior Levy <lior.levy@intel.com>
Removed Tx hang detection mechanism from igbvf.
This mechanism has no affect and can cause false alarm message in some cases.
Signed-off-by: Lior Levy <lior.levy@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igbvf/igbvf.h | 3 --
drivers/net/igbvf/netdev.c | 60 --------------------------------------------
2 files changed, 0 insertions(+), 63 deletions(-)
diff --git a/drivers/net/igbvf/igbvf.h b/drivers/net/igbvf/igbvf.h
index 990c329..d5dad5d 100644
--- a/drivers/net/igbvf/igbvf.h
+++ b/drivers/net/igbvf/igbvf.h
@@ -201,9 +201,6 @@ struct igbvf_adapter {
unsigned int restart_queue;
u32 txd_cmd;
- bool detect_tx_hung;
- u8 tx_timeout_factor;
-
u32 tx_int_delay;
u32 tx_abs_int_delay;
diff --git a/drivers/net/igbvf/netdev.c b/drivers/net/igbvf/netdev.c
index 6352c81..42fdf59 100644
--- a/drivers/net/igbvf/netdev.c
+++ b/drivers/net/igbvf/netdev.c
@@ -396,35 +396,6 @@ static void igbvf_put_txbuf(struct igbvf_adapter *adapter,
buffer_info->time_stamp = 0;
}
-static void igbvf_print_tx_hang(struct igbvf_adapter *adapter)
-{
- struct igbvf_ring *tx_ring = adapter->tx_ring;
- unsigned int i = tx_ring->next_to_clean;
- unsigned int eop = tx_ring->buffer_info[i].next_to_watch;
- union e1000_adv_tx_desc *eop_desc = IGBVF_TX_DESC_ADV(*tx_ring, eop);
-
- /* detected Tx unit hang */
- dev_err(&adapter->pdev->dev,
- "Detected Tx Unit Hang:\n"
- " TDH <%x>\n"
- " TDT <%x>\n"
- " next_to_use <%x>\n"
- " next_to_clean <%x>\n"
- "buffer_info[next_to_clean]:\n"
- " time_stamp <%lx>\n"
- " next_to_watch <%x>\n"
- " jiffies <%lx>\n"
- " next_to_watch.status <%x>\n",
- readl(adapter->hw.hw_addr + tx_ring->head),
- readl(adapter->hw.hw_addr + tx_ring->tail),
- tx_ring->next_to_use,
- tx_ring->next_to_clean,
- tx_ring->buffer_info[eop].time_stamp,
- eop,
- jiffies,
- eop_desc->wb.status);
-}
-
/**
* igbvf_setup_tx_resources - allocate Tx resources (Descriptors)
* @adapter: board private structure
@@ -771,7 +742,6 @@ static void igbvf_set_itr(struct igbvf_adapter *adapter)
static bool igbvf_clean_tx_irq(struct igbvf_ring *tx_ring)
{
struct igbvf_adapter *adapter = tx_ring->adapter;
- struct e1000_hw *hw = &adapter->hw;
struct net_device *netdev = adapter->netdev;
struct igbvf_buffer *buffer_info;
struct sk_buff *skb;
@@ -832,22 +802,6 @@ static bool igbvf_clean_tx_irq(struct igbvf_ring *tx_ring)
}
}
- if (adapter->detect_tx_hung) {
- /* Detect a transmit hang in hardware, this serializes the
- * check with the clearing of time_stamp and movement of i */
- adapter->detect_tx_hung = false;
- if (tx_ring->buffer_info[i].time_stamp &&
- time_after(jiffies, tx_ring->buffer_info[i].time_stamp +
- (adapter->tx_timeout_factor * HZ)) &&
- !(er32(STATUS) & E1000_STATUS_TXOFF)) {
-
- tx_desc = IGBVF_TX_DESC_ADV(*tx_ring, i);
- /* detected Tx unit hang */
- igbvf_print_tx_hang(adapter);
-
- netif_stop_queue(netdev);
- }
- }
adapter->net_stats.tx_bytes += total_bytes;
adapter->net_stats.tx_packets += total_packets;
return count < tx_ring->count;
@@ -1863,17 +1817,6 @@ static void igbvf_watchdog_task(struct work_struct *work)
&adapter->link_duplex);
igbvf_print_link_info(adapter);
- /* adjust timeout factor according to speed/duplex */
- adapter->tx_timeout_factor = 1;
- switch (adapter->link_speed) {
- case SPEED_10:
- adapter->tx_timeout_factor = 16;
- break;
- case SPEED_100:
- /* maybe add some timeout factor ? */
- break;
- }
-
netif_carrier_on(netdev);
netif_wake_queue(netdev);
}
@@ -1907,9 +1850,6 @@ static void igbvf_watchdog_task(struct work_struct *work)
/* Cause software interrupt to ensure Rx ring is cleaned */
ew32(EICS, adapter->rx_ring->eims_value);
- /* Force detection of hung controller every watchdog period */
- adapter->detect_tx_hung = 1;
-
/* Reset the timer */
if (!test_bit(__IGBVF_DOWN, &adapter->state))
mod_timer(&adapter->watchdog_timer,
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 3/7] ixgbevf: Fix name of function in function header comment
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Greg Rose, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Greg Rose <gregory.v.rose@intel.com>
Some of the function names in function header comments did not match
actual name of the function.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbevf/ixgbevf_main.c | 16 ++++++++--------
1 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ixgbevf/ixgbevf_main.c b/drivers/net/ixgbevf/ixgbevf_main.c
index 1f36f8f..43af761 100644
--- a/drivers/net/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ixgbevf/ixgbevf_main.c
@@ -107,7 +107,7 @@ static inline void ixgbevf_release_rx_desc(struct ixgbe_hw *hw,
}
/*
- * ixgbe_set_ivar - set the IVAR registers, mapping interrupt causes to vectors
+ * ixgbevf_set_ivar - set IVAR registers - maps interrupt causes to vectors
* @adapter: pointer to adapter struct
* @direction: 0 for Rx, 1 for Tx, -1 for other causes
* @queue: queue to map the corresponding interrupt to
@@ -1017,7 +1017,7 @@ static irqreturn_t ixgbevf_msix_clean_tx(int irq, void *data)
}
/**
- * ixgbe_msix_clean_rx - single unshared vector rx clean (all queues)
+ * ixgbevf_msix_clean_rx - single unshared vector rx clean (all queues)
* @irq: unused
* @data: pointer to our q_vector struct for this interrupt vector
**/
@@ -1972,7 +1972,7 @@ static void ixgbevf_acquire_msix_vectors(struct ixgbevf_adapter *adapter,
}
/*
- * ixgbe_set_num_queues: Allocate queues for device, feature dependant
+ * ixgbevf_set_num_queues: Allocate queues for device, feature dependant
* @adapter: board private structure to initialize
*
* This is the top level queue allocation routine. The order here is very
@@ -3534,9 +3534,9 @@ static struct pci_driver ixgbevf_driver = {
};
/**
- * ixgbe_init_module - Driver Registration Routine
+ * ixgbevf_init_module - Driver Registration Routine
*
- * ixgbe_init_module is the first routine called when the driver is
+ * ixgbevf_init_module is the first routine called when the driver is
* loaded. All it does is register with the PCI subsystem.
**/
static int __init ixgbevf_init_module(void)
@@ -3554,9 +3554,9 @@ static int __init ixgbevf_init_module(void)
module_init(ixgbevf_init_module);
/**
- * ixgbe_exit_module - Driver Exit Cleanup Routine
+ * ixgbevf_exit_module - Driver Exit Cleanup Routine
*
- * ixgbe_exit_module is called just before the driver is removed
+ * ixgbevf_exit_module is called just before the driver is removed
* from memory.
**/
static void __exit ixgbevf_exit_module(void)
@@ -3566,7 +3566,7 @@ static void __exit ixgbevf_exit_module(void)
#ifdef DEBUG
/**
- * ixgbe_get_hw_dev_name - return device name string
+ * ixgbevf_get_hw_dev_name - return device name string
* used by hardware layer to print debugging information
**/
char *ixgbevf_get_hw_dev_name(struct ixgbe_hw *hw)
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 2/7] ixgbevf: Enable jumbo frame support for X540 VF
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Greg Rose, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Greg Rose <gregory.v.rose@intel.com>
The X540 controller allows jumbo frame setup on a per VF basis. Enable
use of jumbo frames when the VF device belongs to the X540 controller.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbevf/defines.h | 2 ++
drivers/net/ixgbevf/ixgbevf_main.c | 19 +++++++++++++++++--
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ixgbevf/defines.h b/drivers/net/ixgbevf/defines.h
index de643eb..78abb6f 100644
--- a/drivers/net/ixgbevf/defines.h
+++ b/drivers/net/ixgbevf/defines.h
@@ -65,6 +65,8 @@ typedef u32 ixgbe_link_speed;
#define IXGBE_RXCTRL_DMBYPS 0x00000002 /* Descriptor Monitor Bypass */
#define IXGBE_RXDCTL_ENABLE 0x02000000 /* Enable specific Rx Queue */
#define IXGBE_RXDCTL_VME 0x40000000 /* VLAN mode enable */
+#define IXGBE_RXDCTL_RLPMLMASK 0x00003FFF /* Only supported on the X540 */
+#define IXGBE_RXDCTL_RLPML_EN 0x00008000
/* DCA Control */
#define IXGBE_DCA_TXCTRL_TX_WB_RO_EN (1 << 11) /* Tx Desc writeback RO bit */
diff --git a/drivers/net/ixgbevf/ixgbevf_main.c b/drivers/net/ixgbevf/ixgbevf_main.c
index 464e6c9..1f36f8f 100644
--- a/drivers/net/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ixgbevf/ixgbevf_main.c
@@ -51,7 +51,7 @@ char ixgbevf_driver_name[] = "ixgbevf";
static const char ixgbevf_driver_string[] =
"Intel(R) 82599 Virtual Function";
-#define DRV_VERSION "1.0.19-k0"
+#define DRV_VERSION "1.1.0-k0"
const char ixgbevf_driver_version[] = DRV_VERSION;
static char ixgbevf_copyright[] =
"Copyright (c) 2009 - 2010 Intel Corporation.";
@@ -1665,6 +1665,11 @@ static int ixgbevf_up_complete(struct ixgbevf_adapter *adapter)
j = adapter->rx_ring[i].reg_idx;
rxdctl = IXGBE_READ_REG(hw, IXGBE_VFRXDCTL(j));
rxdctl |= IXGBE_RXDCTL_ENABLE;
+ if (hw->mac.type == ixgbe_mac_X540_vf) {
+ rxdctl &= ~IXGBE_RXDCTL_RLPMLMASK;
+ rxdctl |= ((netdev->mtu + ETH_HLEN + ETH_FCS_LEN) |
+ IXGBE_RXDCTL_RLPML_EN);
+ }
IXGBE_WRITE_REG(hw, IXGBE_VFRXDCTL(j), rxdctl);
ixgbevf_rx_desc_queue_enable(adapter, i);
}
@@ -3217,10 +3222,16 @@ static int ixgbevf_set_mac(struct net_device *netdev, void *p)
static int ixgbevf_change_mtu(struct net_device *netdev, int new_mtu)
{
struct ixgbevf_adapter *adapter = netdev_priv(netdev);
+ struct ixgbe_hw *hw = &adapter->hw;
int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
+ int max_possible_frame = MAXIMUM_ETHERNET_VLAN_SIZE;
+ u32 msg[2];
+
+ if (adapter->hw.mac.type == ixgbe_mac_X540_vf)
+ max_possible_frame = IXGBE_MAX_JUMBO_FRAME_SIZE;
/* MTU < 68 is an error and causes problems on some kernels */
- if ((new_mtu < 68) || (max_frame > MAXIMUM_ETHERNET_VLAN_SIZE))
+ if ((new_mtu < 68) || (max_frame > max_possible_frame))
return -EINVAL;
hw_dbg(&adapter->hw, "changing MTU from %d to %d\n",
@@ -3228,6 +3239,10 @@ static int ixgbevf_change_mtu(struct net_device *netdev, int new_mtu)
/* must set new MTU before calling down or up */
netdev->mtu = new_mtu;
+ msg[0] = IXGBE_VF_SET_LPE;
+ msg[1] = max_frame;
+ hw->mbx.ops.write_posted(hw, msg, 2);
+
if (netif_running(netdev))
ixgbevf_reinit_locked(adapter);
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 1/7] ixgbe: Enable Jumbo Frames on the X540 10Gigabit Controller
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Greg Rose, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1298545109-8990-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Greg Rose <gregory.v.rose@intel.com>
The X540 controller supports jumbo frames in SR-IOV mode. Allow
configuration of jumbo frames either in the PF driver or on behalf of
a VF.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 18 ++++++++++++++++--
drivers/net/ixgbe/ixgbe_sriov.c | 29 ++++++++++++++++++++++++++++-
drivers/net/ixgbe/ixgbe_type.h | 2 ++
3 files changed, 46 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index eca762d..f0d0c5a 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -3077,6 +3077,14 @@ void ixgbe_configure_rx_ring(struct ixgbe_adapter *adapter,
ixgbe_configure_srrctl(adapter, ring);
ixgbe_configure_rscctl(adapter, ring);
+ /* If operating in IOV mode set RLPML for X540 */
+ if ((adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) &&
+ hw->mac.type == ixgbe_mac_X540) {
+ rxdctl &= ~IXGBE_RXDCTL_RLPMLMASK;
+ rxdctl |= ((ring->netdev->mtu + ETH_HLEN +
+ ETH_FCS_LEN + VLAN_HLEN) | IXGBE_RXDCTL_RLPML_EN);
+ }
+
if (hw->mac.type == ixgbe_mac_82598EB) {
/*
* enable cache line friendly hardware writes:
@@ -5441,8 +5449,14 @@ static int ixgbe_change_mtu(struct net_device *netdev, int new_mtu)
int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
/* MTU < 68 is an error and causes problems on some kernels */
- if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE))
- return -EINVAL;
+ if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED &&
+ hw->mac.type != ixgbe_mac_X540) {
+ if ((new_mtu < 68) || (max_frame > MAXIMUM_ETHERNET_VLAN_SIZE))
+ return -EINVAL;
+ } else {
+ if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE))
+ return -EINVAL;
+ }
e_info(probe, "changing MTU from %d to %d\n", netdev->mtu, new_mtu);
/* must set new MTU before calling down or up */
diff --git a/drivers/net/ixgbe/ixgbe_sriov.c b/drivers/net/ixgbe/ixgbe_sriov.c
index 187b3a1..fb4868d 100644
--- a/drivers/net/ixgbe/ixgbe_sriov.c
+++ b/drivers/net/ixgbe/ixgbe_sriov.c
@@ -110,6 +110,33 @@ static int ixgbe_set_vf_vlan(struct ixgbe_adapter *adapter, int add, int vid,
return adapter->hw.mac.ops.set_vfta(&adapter->hw, vid, vf, (bool)add);
}
+void ixgbe_set_vf_lpe(struct ixgbe_adapter *adapter, u32 *msgbuf)
+{
+ struct ixgbe_hw *hw = &adapter->hw;
+ int new_mtu = msgbuf[1];
+ u32 max_frs;
+ int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
+
+ /* Only X540 supports jumbo frames in IOV mode */
+ if (adapter->hw.mac.type != ixgbe_mac_X540)
+ return;
+
+ /* MTU < 68 is an error and causes problems on some kernels */
+ if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE)) {
+ e_err(drv, "VF mtu %d out of range\n", new_mtu);
+ return;
+ }
+
+ max_frs = (IXGBE_READ_REG(hw, IXGBE_MAXFRS) &
+ IXGBE_MHADD_MFS_MASK) >> IXGBE_MHADD_MFS_SHIFT;
+ if (max_frs < new_mtu) {
+ max_frs = new_mtu << IXGBE_MHADD_MFS_SHIFT;
+ IXGBE_WRITE_REG(hw, IXGBE_MAXFRS, max_frs);
+ }
+
+ e_info(hw, "VF requests change max MTU to %d\n", new_mtu);
+}
+
static void ixgbe_set_vmolr(struct ixgbe_hw *hw, u32 vf, bool aupe)
{
u32 vmolr = IXGBE_READ_REG(hw, IXGBE_VMOLR(vf));
@@ -302,7 +329,7 @@ static int ixgbe_rcv_msg_from_vf(struct ixgbe_adapter *adapter, u32 vf)
hash_list, vf);
break;
case IXGBE_VF_SET_LPE:
- WARN_ON((msgbuf[0] & 0xFFFF) == IXGBE_VF_SET_LPE);
+ ixgbe_set_vf_lpe(adapter, msgbuf);
break;
case IXGBE_VF_SET_VLAN:
add = (msgbuf[0] & IXGBE_VT_MSGINFO_MASK)
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index fd3358f..ab65d13 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -1680,6 +1680,8 @@
#define IXGBE_RXCTRL_DMBYPS 0x00000002 /* Descriptor Monitor Bypass */
#define IXGBE_RXDCTL_ENABLE 0x02000000 /* Enable specific Rx Queue */
#define IXGBE_RXDCTL_VME 0x40000000 /* VLAN mode enable */
+#define IXGBE_RXDCTL_RLPMLMASK 0x00003FFF /* Only supported on the X540 */
+#define IXGBE_RXDCTL_RLPML_EN 0x00008000
#define IXGBE_FCTRL_SBP 0x00000002 /* Store Bad Packet */
#define IXGBE_FCTRL_MPE 0x00000100 /* Multicast Promiscuous Ena*/
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 0/7][pull request] Intel Wired LAN Driver Updates
From: Jeff Kirsher @ 2011-02-24 10:58 UTC (permalink / raw)
To: davem; +Cc: Jeff Kirsher, netdev, gospo, bphilips
The following series contains jumbo frame support for X540 devices,
comment cleanup/fixes for ixgbevf & igb, and the addition of Tx rate
limiting feature for igb.
The following are changes since commit 55ae22d08fc9b531bc8a88b7306004e7053bb425:
Merge branch 'tipc-Feb23-2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/net-next-2.6
and are available in the git repository at:
master.kernel.org:/pub/scm/linux/kernel/git/jkirsher/net-next-2.6 master
Carolyn Wyborny (2):
igb: Update Intel copyright notice for driver source.
igb: update version string
Greg Rose (3):
ixgbe: Enable Jumbo Frames on the X540 10Gigabit Controller
ixgbevf: Enable jumbo frame support for X540 VF
ixgbevf: Fix name of function in function header comment
Lior Levy (2):
igbvf: remove Tx hang detection
igb: add support for VF Transmit rate limit using iproute2
drivers/net/igb/e1000_defines.h | 7 +++
drivers/net/igb/e1000_regs.h | 4 ++
drivers/net/igb/igb.h | 2 +
drivers/net/igb/igb_main.c | 92 ++++++++++++++++++++++++++++++++++--
drivers/net/igbvf/igbvf.h | 3 -
drivers/net/igbvf/netdev.c | 60 -----------------------
drivers/net/ixgbe/ixgbe_main.c | 18 ++++++-
drivers/net/ixgbe/ixgbe_sriov.c | 29 +++++++++++-
drivers/net/ixgbe/ixgbe_type.h | 2 +
drivers/net/ixgbevf/defines.h | 2 +
drivers/net/ixgbevf/ixgbevf_main.c | 35 ++++++++++----
11 files changed, 174 insertions(+), 80 deletions(-)
--
1.7.4
^ permalink raw reply
* Re: [Lxc-users] Bad checksums and lost packets with macvlan on dummy
From: Daniel Lezcano @ 2011-02-24 10:20 UTC (permalink / raw)
To: Andrian Nord; +Cc: lxc-users, Patrick McHardy, Linux Netdev List
In-Reply-To: <20110223170512.GA10277@nord.niifaq.ru>
On 02/23/2011 06:13 PM, Andrian Nord wrote:
> On Mon, Feb 21, 2011 at 05:07:31PM +0100, Daniel Lezcano wrote:
>> I Cc'ed the netdev mailing list and Patrick in case my analysis is wrong
>> or incomplete.
> I'm confirming, that this happens only when macvlan's are onto dummy net
> device. In case of some physical interface under macvlan there is no lost
> packages and no broken checksums.
Hmm, that's interesting ...
I saw you were using the command 'nc6', do you use netcat with ipv6 ?
^ permalink raw reply
* [PATCH ref0] net: add Faraday FTMAC100 10/100 Ethernet driver
From: Po-Yu Chuang @ 2011-02-24 9:29 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, bhutchings, eric.dumazet, joe, dilinger, mirqus,
davem, Po-Yu Chuang
In-Reply-To: <1298535761.2814.1.camel@edumazet-laptop>
From: Po-Yu Chuang <ratbert@faraday-tech.com>
*** NOTE ***
This patch is only for reference. Its performance is worse than v5.
Please do not apply this.
FTMAC100 Ethernet Media Access Controller supports 10/100 Mbps and
MII. This driver has been working on some ARM/NDS32 SoC's including
Faraday A320 and Andes AG101.
Signed-off-by: Po-Yu Chuang <ratbert@faraday-tech.com>
---
v2:
always use NAPI
do not use our own net_device_stats structure
don't set trans_start and last_rx
stats.rx_packets and stats.rx_bytes include dropped packets
add missed netif_napi_del()
initialize spinlocks in probe function
remove rx_lock and hw_lock
use netdev_[err/info/dbg] instead of dev_* ones
use netdev_alloc_skb_ip_align()
remove ftmac100_get_stats()
use is_valid_ether_addr() instead of is_zero_ether_addr()
add const to ftmac100_ethtool_ops and ftmac100_netdev_ops
use net_ratelimit() instead of printk_ratelimit()
no explicit inline
use %pM to print MAC address
add comment before wmb
use napi poll() to handle all interrupts
v3:
undo "stats.rx_packets and stats.rx_bytes include dropped packets"
ftmac100_mdio_read() returns 0 if error
fix comment typos
use pr_fmt and pr_info
define INT_MASK_ALL_ENABLED
define MACCR_ENABLE_ALL
do not count length error many times
use bool/true/false
use cpu_to_le32/le32_to_cpu to access descriptors
indent style fix
v4:
should not access skb after netif_receive_skb()
use resource_size()
better way to use cpu_to_le32/le32_to_cpu
use spin_lock() for tx_lock
combine all netdev_info() together in ftmac100_poll()
v5:
use dev_kfree_skb() in ftmac100_tx_complete_packet()
cpu_to_le32/le32_to_cpu usage fix
drop GFP_DMA
ref0:
cpu_to_le32/le32_to_cpu usage fix
remove "tx queue full" message
reduce critical section protected by tx_lock
add check of MAX_PKT_SIZE and RX_BUF_SIZE
add __exit to ftmac100_remove()
zero copy - use skb_fill_page_desc() and __pskb_pull_tail().
drivers/net/Kconfig | 9 +
drivers/net/Makefile | 1 +
drivers/net/ftmac100.c | 1196 ++++++++++++++++++++++++++++++++++++++++++++++++
drivers/net/ftmac100.h | 180 ++++++++
4 files changed, 1386 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/ftmac100.c
create mode 100644 drivers/net/ftmac100.h
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 4f1755b..6b12274 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2014,6 +2014,15 @@ config BCM63XX_ENET
This driver supports the ethernet MACs in the Broadcom 63xx
MIPS chipset family (BCM63XX).
+config FTMAC100
+ tristate "Faraday FTMAC100 10/100 Ethernet support"
+ depends on ARM
+ select MII
+ help
+ This driver supports the FTMAC100 10/100 Ethernet controller
+ from Faraday. It is used on Faraday A320, Andes AG101 and some
+ other ARM/NDS32 SoC's.
+
source "drivers/net/fs_enet/Kconfig"
source "drivers/net/octeon/Kconfig"
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index b90738d..7c21711 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -147,6 +147,7 @@ obj-$(CONFIG_FORCEDETH) += forcedeth.o
obj-$(CONFIG_NE_H8300) += ne-h8300.o 8390.o
obj-$(CONFIG_AX88796) += ax88796.o
obj-$(CONFIG_BCM63XX_ENET) += bcm63xx_enet.o
+obj-$(CONFIG_FTMAC100) += ftmac100.o
obj-$(CONFIG_TSI108_ETH) += tsi108_eth.o
obj-$(CONFIG_MV643XX_ETH) += mv643xx_eth.o
diff --git a/drivers/net/ftmac100.c b/drivers/net/ftmac100.c
new file mode 100644
index 0000000..eaa6544
--- /dev/null
+++ b/drivers/net/ftmac100.c
@@ -0,0 +1,1196 @@
+/*
+ * Faraday FTMAC100 10/100 Ethernet
+ *
+ * (C) Copyright 2009-2011 Faraday Technology
+ * Po-Yu Chuang <ratbert@faraday-tech.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/dma-mapping.h>
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/mii.h>
+#include <linux/module.h>
+#include <linux/netdevice.h>
+#include <linux/platform_device.h>
+
+#include "ftmac100.h"
+
+#define DRV_NAME "ftmac100"
+#define DRV_VERSION "0.2"
+
+#define RX_QUEUE_ENTRIES 128 /* must be power of 2 */
+#define TX_QUEUE_ENTRIES 16 /* must be power of 2 */
+
+#define MAX_PKT_SIZE 1518
+#define RX_BUF_SIZE 2044 /* must be smaller than 0x7ff */
+
+#if MAX_PKT_SIZE > 0x7ff
+#error invalid MAX_PKT_SIZE
+#endif
+
+#if RX_BUF_SIZE > 0x7ff || RX_BUF_SIZE > PAGE_SIZE
+#error invalid RX_BUF_SIZE
+#endif
+
+/******************************************************************************
+ * private data
+ *****************************************************************************/
+struct ftmac100_descs {
+ struct ftmac100_rxdes rxdes[RX_QUEUE_ENTRIES];
+ struct ftmac100_txdes txdes[TX_QUEUE_ENTRIES];
+};
+
+struct ftmac100 {
+ struct resource *res;
+ void __iomem *base;
+ int irq;
+
+ struct ftmac100_descs *descs;
+ dma_addr_t descs_dma_addr;
+
+ unsigned int rx_pointer;
+ unsigned int tx_clean_pointer;
+ unsigned int tx_pointer;
+ unsigned int tx_pending;
+
+ spinlock_t tx_lock;
+
+ struct net_device *netdev;
+ struct device *dev;
+ struct napi_struct napi;
+
+ struct mii_if_info mii;
+};
+
+static int ftmac100_alloc_rx_page(struct ftmac100 *priv, struct ftmac100_rxdes *rxdes);
+
+/******************************************************************************
+ * internal functions (hardware register access)
+ *****************************************************************************/
+#define INT_MASK_ALL_ENABLED (FTMAC100_INT_RPKT_FINISH | \
+ FTMAC100_INT_NORXBUF | \
+ FTMAC100_INT_XPKT_OK | \
+ FTMAC100_INT_XPKT_LOST | \
+ FTMAC100_INT_RPKT_LOST | \
+ FTMAC100_INT_AHB_ERR | \
+ FTMAC100_INT_PHYSTS_CHG)
+
+#define INT_MASK_ALL_DISABLED 0
+
+static void ftmac100_enable_all_int(struct ftmac100 *priv)
+{
+ iowrite32(INT_MASK_ALL_ENABLED, priv->base + FTMAC100_OFFSET_IMR);
+}
+
+static void ftmac100_disable_all_int(struct ftmac100 *priv)
+{
+ iowrite32(INT_MASK_ALL_DISABLED, priv->base + FTMAC100_OFFSET_IMR);
+}
+
+static void ftmac100_set_rx_ring_base(struct ftmac100 *priv, dma_addr_t addr)
+{
+ iowrite32(addr, priv->base + FTMAC100_OFFSET_RXR_BADR);
+}
+
+static void ftmac100_set_tx_ring_base(struct ftmac100 *priv, dma_addr_t addr)
+{
+ iowrite32(addr, priv->base + FTMAC100_OFFSET_TXR_BADR);
+}
+
+static void ftmac100_txdma_start_polling(struct ftmac100 *priv)
+{
+ iowrite32(1, priv->base + FTMAC100_OFFSET_TXPD);
+}
+
+static int ftmac100_reset(struct ftmac100 *priv)
+{
+ struct net_device *netdev = priv->netdev;
+ int i;
+
+ /* NOTE: reset clears all registers */
+ iowrite32(FTMAC100_MACCR_SW_RST, priv->base + FTMAC100_OFFSET_MACCR);
+
+ for (i = 0; i < 5; i++) {
+ unsigned int maccr;
+
+ maccr = ioread32(priv->base + FTMAC100_OFFSET_MACCR);
+ if (!(maccr & FTMAC100_MACCR_SW_RST)) {
+ /*
+ * FTMAC100_MACCR_SW_RST cleared does not indicate
+ * that hardware reset completed (what the f*ck).
+ * We still need to wait for a while.
+ */
+ usleep_range(500, 1000);
+ return 0;
+ }
+
+ usleep_range(1000, 10000);
+ }
+
+ netdev_err(netdev, "software reset failed\n");
+ return -EIO;
+}
+
+static void ftmac100_set_mac(struct ftmac100 *priv, const unsigned char *mac)
+{
+ unsigned int maddr = mac[0] << 8 | mac[1];
+ unsigned int laddr = mac[2] << 24 | mac[3] << 16 | mac[4] << 8 | mac[5];
+
+ iowrite32(maddr, priv->base + FTMAC100_OFFSET_MAC_MADR);
+ iowrite32(laddr, priv->base + FTMAC100_OFFSET_MAC_LADR);
+}
+
+#define MACCR_ENABLE_ALL (FTMAC100_MACCR_XMT_EN | \
+ FTMAC100_MACCR_RCV_EN | \
+ FTMAC100_MACCR_XDMA_EN | \
+ FTMAC100_MACCR_RDMA_EN | \
+ FTMAC100_MACCR_CRC_APD | \
+ FTMAC100_MACCR_FULLDUP | \
+ FTMAC100_MACCR_RX_RUNT | \
+ FTMAC100_MACCR_RX_BROADPKT)
+
+static int ftmac100_start_hw(struct ftmac100 *priv)
+{
+ struct net_device *netdev = priv->netdev;
+
+ if (ftmac100_reset(priv))
+ return -EIO;
+
+ /* setup ring buffer base registers */
+ ftmac100_set_rx_ring_base(priv,
+ priv->descs_dma_addr +
+ offsetof(struct ftmac100_descs, rxdes));
+ ftmac100_set_tx_ring_base(priv,
+ priv->descs_dma_addr +
+ offsetof(struct ftmac100_descs, txdes));
+
+ iowrite32(FTMAC100_APTC_RXPOLL_CNT(1), priv->base + FTMAC100_OFFSET_APTC);
+
+ ftmac100_set_mac(priv, netdev->dev_addr);
+
+ iowrite32(MACCR_ENABLE_ALL, priv->base + FTMAC100_OFFSET_MACCR);
+ return 0;
+}
+
+static void ftmac100_stop_hw(struct ftmac100 *priv)
+{
+ iowrite32(0, priv->base + FTMAC100_OFFSET_MACCR);
+}
+
+/******************************************************************************
+ * internal functions (receive descriptor)
+ *****************************************************************************/
+static bool ftmac100_rxdes_first_segment(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_FRS);
+}
+
+static bool ftmac100_rxdes_last_segment(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_LRS);
+}
+
+static bool ftmac100_rxdes_owned_by_dma(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_RXDMA_OWN);
+}
+
+static void ftmac100_rxdes_set_dma_own(struct ftmac100_rxdes *rxdes)
+{
+ /* clear status bits */
+ rxdes->rxdes0 = cpu_to_le32(FTMAC100_RXDES0_RXDMA_OWN);
+}
+
+static bool ftmac100_rxdes_rx_error(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_RX_ERR);
+}
+
+static bool ftmac100_rxdes_crc_error(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_CRC_ERR);
+}
+
+static bool ftmac100_rxdes_frame_too_long(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_FTL);
+}
+
+static bool ftmac100_rxdes_runt(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_RUNT);
+}
+
+static bool ftmac100_rxdes_odd_nibble(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_RX_ODD_NB);
+}
+
+static unsigned int ftmac100_rxdes_frame_length(struct ftmac100_rxdes *rxdes)
+{
+ return le32_to_cpu(rxdes->rxdes0) & FTMAC100_RXDES0_RFL;
+}
+
+static bool ftmac100_rxdes_multicast(struct ftmac100_rxdes *rxdes)
+{
+ return rxdes->rxdes0 & cpu_to_le32(FTMAC100_RXDES0_MULTICAST);
+}
+
+static void ftmac100_rxdes_set_buffer_size(struct ftmac100_rxdes *rxdes,
+ unsigned int size)
+{
+ rxdes->rxdes1 &= cpu_to_le32(FTMAC100_RXDES1_EDORR);
+ rxdes->rxdes1 |= cpu_to_le32(FTMAC100_RXDES1_RXBUF_SIZE(size));
+}
+
+static void ftmac100_rxdes_set_end_of_ring(struct ftmac100_rxdes *rxdes)
+{
+ rxdes->rxdes1 |= cpu_to_le32(FTMAC100_RXDES1_EDORR);
+}
+
+static void ftmac100_rxdes_set_dma_addr(struct ftmac100_rxdes *rxdes,
+ dma_addr_t addr)
+{
+ rxdes->rxdes2 = cpu_to_le32(addr);
+}
+
+static dma_addr_t ftmac100_rxdes_get_dma_addr(struct ftmac100_rxdes *rxdes)
+{
+ return le32_to_cpu(rxdes->rxdes2);
+}
+
+/*
+ * rxdes3 is not used by hardware. We use it to keep track of page.
+ * Since hardware does not touch it, we can skip cpu_to_le32()/le32_to_cpu().
+ */
+static void ftmac100_rxdes_set_page(struct ftmac100_rxdes *rxdes, struct page *page)
+{
+ rxdes->rxdes3 = (unsigned int)page;
+}
+
+static struct page *ftmac100_rxdes_get_page(struct ftmac100_rxdes *rxdes)
+{
+ return (struct page *)rxdes->rxdes3;
+}
+
+/******************************************************************************
+ * internal functions (receive)
+ *****************************************************************************/
+static int ftmac100_next_rx_pointer(int pointer)
+{
+ return (pointer + 1) & (RX_QUEUE_ENTRIES - 1);
+}
+
+static void ftmac100_rx_pointer_advance(struct ftmac100 *priv)
+{
+ priv->rx_pointer = ftmac100_next_rx_pointer(priv->rx_pointer);
+}
+
+static struct ftmac100_rxdes *ftmac100_current_rxdes(struct ftmac100 *priv)
+{
+ return &priv->descs->rxdes[priv->rx_pointer];
+}
+
+static struct ftmac100_rxdes *
+ftmac100_rx_locate_first_segment(struct ftmac100 *priv)
+{
+ struct ftmac100_rxdes *rxdes = ftmac100_current_rxdes(priv);
+
+ while (!ftmac100_rxdes_owned_by_dma(rxdes)) {
+ if (ftmac100_rxdes_first_segment(rxdes))
+ return rxdes;
+
+ ftmac100_rxdes_set_dma_own(rxdes);
+ ftmac100_rx_pointer_advance(priv);
+ rxdes = ftmac100_current_rxdes(priv);
+ }
+
+ return NULL;
+}
+
+static bool ftmac100_rx_packet_error(struct ftmac100 *priv,
+ struct ftmac100_rxdes *rxdes)
+{
+ struct net_device *netdev = priv->netdev;
+ bool error = false;
+
+ if (unlikely(ftmac100_rxdes_rx_error(rxdes))) {
+ if (net_ratelimit())
+ netdev_info(netdev, "rx err\n");
+
+ netdev->stats.rx_errors++;
+ error = true;
+ }
+
+ if (unlikely(ftmac100_rxdes_crc_error(rxdes))) {
+ if (net_ratelimit())
+ netdev_info(netdev, "rx crc err\n");
+
+ netdev->stats.rx_crc_errors++;
+ error = true;
+ }
+
+ if (unlikely(ftmac100_rxdes_frame_too_long(rxdes))) {
+ if (net_ratelimit())
+ netdev_info(netdev, "rx frame too long\n");
+
+ netdev->stats.rx_length_errors++;
+ error = true;
+ } else if (unlikely(ftmac100_rxdes_runt(rxdes))) {
+ if (net_ratelimit())
+ netdev_info(netdev, "rx runt\n");
+
+ netdev->stats.rx_length_errors++;
+ error = true;
+ } else if (unlikely(ftmac100_rxdes_odd_nibble(rxdes))) {
+ if (net_ratelimit())
+ netdev_info(netdev, "rx odd nibble\n");
+
+ netdev->stats.rx_length_errors++;
+ error = true;
+ }
+
+ return error;
+}
+
+static void ftmac100_rx_drop_packet(struct ftmac100 *priv)
+{
+ struct net_device *netdev = priv->netdev;
+ struct ftmac100_rxdes *rxdes = ftmac100_current_rxdes(priv);
+ bool done = false;
+
+ if (net_ratelimit())
+ netdev_dbg(netdev, "drop packet %p\n", rxdes);
+
+ do {
+ if (ftmac100_rxdes_last_segment(rxdes))
+ done = true;
+
+ ftmac100_rxdes_set_dma_own(rxdes);
+ ftmac100_rx_pointer_advance(priv);
+ rxdes = ftmac100_current_rxdes(priv);
+ } while (!done && !ftmac100_rxdes_owned_by_dma(rxdes));
+
+ netdev->stats.rx_dropped++;
+}
+
+static bool ftmac100_rx_packet(struct ftmac100 *priv, int *processed)
+{
+ struct net_device *netdev = priv->netdev;
+ struct ftmac100_rxdes *rxdes;
+ struct sk_buff *skb;
+ struct page *page;
+ dma_addr_t map;
+ int length;
+
+ rxdes = ftmac100_rx_locate_first_segment(priv);
+ if (!rxdes)
+ return false;
+
+ if (unlikely(ftmac100_rx_packet_error(priv, rxdes))) {
+ ftmac100_rx_drop_packet(priv);
+ return true;
+ }
+
+ /*
+ * It is impossible to get multi-segment packets
+ * because we always provide big enough receive buffers.
+ */
+ if (unlikely(!ftmac100_rxdes_last_segment(rxdes)))
+ BUG();
+
+ /* start processing */
+ skb = netdev_alloc_skb_ip_align(netdev, ETH_HLEN);
+ if (unlikely(!skb)) {
+ if (net_ratelimit())
+ netdev_err(netdev, "rx skb alloc failed\n");
+
+ ftmac100_rx_drop_packet(priv);
+ return true;
+ }
+
+ if (unlikely(ftmac100_rxdes_multicast(rxdes)))
+ netdev->stats.multicast++;
+
+ map = ftmac100_rxdes_get_dma_addr(rxdes);
+ dma_unmap_page(priv->dev, map, RX_BUF_SIZE, DMA_FROM_DEVICE);
+
+ length = ftmac100_rxdes_frame_length(rxdes);
+ page = ftmac100_rxdes_get_page(rxdes);
+ skb_fill_page_desc(skb, 0, page, 0, length);
+ skb->len += length;
+ skb->data_len += length;
+ skb->truesize += length;
+ __pskb_pull_tail(skb, ETH_HLEN);
+
+ ftmac100_alloc_rx_page(priv, rxdes);
+
+ ftmac100_rx_pointer_advance(priv);
+
+ skb->protocol = eth_type_trans(skb, netdev);
+
+ netdev->stats.rx_packets++;
+ netdev->stats.rx_bytes += skb->len;
+
+ /* push packet to protocol stack */
+ netif_receive_skb(skb);
+
+ (*processed)++;
+ return true;
+}
+
+/******************************************************************************
+ * internal functions (transmit descriptor)
+ *****************************************************************************/
+static void ftmac100_txdes_reset(struct ftmac100_txdes *txdes)
+{
+ /* clear all except end of ring bit */
+ txdes->txdes0 = 0;
+ txdes->txdes1 &= cpu_to_le32(FTMAC100_TXDES1_EDOTR);
+ txdes->txdes2 = 0;
+ txdes->txdes3 = 0;
+}
+
+static bool ftmac100_txdes_owned_by_dma(struct ftmac100_txdes *txdes)
+{
+ return txdes->txdes0 & cpu_to_le32(FTMAC100_TXDES0_TXDMA_OWN);
+}
+
+static void ftmac100_txdes_set_dma_own(struct ftmac100_txdes *txdes)
+{
+ /*
+ * Make sure dma own bit will not be set before any other
+ * descriptor fields.
+ */
+ wmb();
+ txdes->txdes0 |= cpu_to_le32(FTMAC100_TXDES0_TXDMA_OWN);
+}
+
+static bool ftmac100_txdes_excessive_collision(struct ftmac100_txdes *txdes)
+{
+ return txdes->txdes0 & cpu_to_le32(FTMAC100_TXDES0_TXPKT_EXSCOL);
+}
+
+static bool ftmac100_txdes_late_collision(struct ftmac100_txdes *txdes)
+{
+ return txdes->txdes0 & cpu_to_le32(FTMAC100_TXDES0_TXPKT_LATECOL);
+}
+
+static void ftmac100_txdes_set_end_of_ring(struct ftmac100_txdes *txdes)
+{
+ txdes->txdes1 |= cpu_to_le32(FTMAC100_TXDES1_EDOTR);
+}
+
+static void ftmac100_txdes_set_first_segment(struct ftmac100_txdes *txdes)
+{
+ txdes->txdes1 |= cpu_to_le32(FTMAC100_TXDES1_FTS);
+}
+
+static void ftmac100_txdes_set_last_segment(struct ftmac100_txdes *txdes)
+{
+ txdes->txdes1 |= cpu_to_le32(FTMAC100_TXDES1_LTS);
+}
+
+static void ftmac100_txdes_set_txint(struct ftmac100_txdes *txdes)
+{
+ txdes->txdes1 |= cpu_to_le32(FTMAC100_TXDES1_TXIC);
+}
+
+static void ftmac100_txdes_set_buffer_size(struct ftmac100_txdes *txdes,
+ unsigned int len)
+{
+ txdes->txdes1 |= cpu_to_le32(FTMAC100_TXDES1_TXBUF_SIZE(len));
+}
+
+static void ftmac100_txdes_set_dma_addr(struct ftmac100_txdes *txdes,
+ dma_addr_t addr)
+{
+ txdes->txdes2 = cpu_to_le32(addr);
+}
+
+static dma_addr_t ftmac100_txdes_get_dma_addr(struct ftmac100_txdes *txdes)
+{
+ return le32_to_cpu(txdes->txdes2);
+}
+
+/*
+ * txdes3 is not used by hardware. We use it to keep track of socket buffer.
+ * Since hardware does not touch it, we can skip cpu_to_le32()/le32_to_cpu().
+ */
+static void ftmac100_txdes_set_skb(struct ftmac100_txdes *txdes, struct sk_buff *skb)
+{
+ txdes->txdes3 = (unsigned int)skb;
+}
+
+static struct sk_buff *ftmac100_txdes_get_skb(struct ftmac100_txdes *txdes)
+{
+ return (struct sk_buff *)txdes->txdes3;
+}
+
+/******************************************************************************
+ * internal functions (transmit)
+ *****************************************************************************/
+static int ftmac100_next_tx_pointer(int pointer)
+{
+ return (pointer + 1) & (TX_QUEUE_ENTRIES - 1);
+}
+
+static void ftmac100_tx_pointer_advance(struct ftmac100 *priv)
+{
+ priv->tx_pointer = ftmac100_next_tx_pointer(priv->tx_pointer);
+}
+
+static void ftmac100_tx_clean_pointer_advance(struct ftmac100 *priv)
+{
+ priv->tx_clean_pointer = ftmac100_next_tx_pointer(priv->tx_clean_pointer);
+}
+
+static struct ftmac100_txdes *ftmac100_current_txdes(struct ftmac100 *priv)
+{
+ return &priv->descs->txdes[priv->tx_pointer];
+}
+
+static struct ftmac100_txdes *ftmac100_current_clean_txdes(struct ftmac100 *priv)
+{
+ return &priv->descs->txdes[priv->tx_clean_pointer];
+}
+
+static bool ftmac100_tx_complete_packet(struct ftmac100 *priv)
+{
+ struct net_device *netdev = priv->netdev;
+ struct ftmac100_txdes *txdes;
+ struct sk_buff *skb;
+ dma_addr_t map;
+
+ if (priv->tx_pending == 0)
+ return false;
+
+ txdes = ftmac100_current_clean_txdes(priv);
+
+ if (ftmac100_txdes_owned_by_dma(txdes))
+ return false;
+
+ skb = ftmac100_txdes_get_skb(txdes);
+ map = ftmac100_txdes_get_dma_addr(txdes);
+
+ if (unlikely(ftmac100_txdes_excessive_collision(txdes) ||
+ ftmac100_txdes_late_collision(txdes))) {
+ /*
+ * packet transmitted to ethernet lost due to late collision
+ * or excessive collision
+ */
+ netdev->stats.tx_aborted_errors++;
+ } else {
+ netdev->stats.tx_packets++;
+ netdev->stats.tx_bytes += skb->len;
+ }
+
+ dma_unmap_single(priv->dev, map, skb_headlen(skb), DMA_TO_DEVICE);
+ dev_kfree_skb(skb);
+
+ ftmac100_txdes_reset(txdes);
+
+ ftmac100_tx_clean_pointer_advance(priv);
+
+ spin_lock(&priv->tx_lock);
+ priv->tx_pending--;
+ spin_unlock(&priv->tx_lock);
+ netif_wake_queue(netdev);
+
+ return true;
+}
+
+static void ftmac100_tx_complete(struct ftmac100 *priv)
+{
+ while (ftmac100_tx_complete_packet(priv))
+ ;
+}
+
+static int ftmac100_xmit(struct ftmac100 *priv, struct sk_buff *skb,
+ dma_addr_t map)
+{
+ struct net_device *netdev = priv->netdev;
+ struct ftmac100_txdes *txdes;
+ unsigned int len = (skb->len < ETH_ZLEN) ? ETH_ZLEN : skb->len;
+
+ txdes = ftmac100_current_txdes(priv);
+ ftmac100_tx_pointer_advance(priv);
+
+ /* setup TX descriptor */
+ ftmac100_txdes_set_skb(txdes, skb);
+ ftmac100_txdes_set_dma_addr(txdes, map);
+
+ ftmac100_txdes_set_first_segment(txdes);
+ ftmac100_txdes_set_last_segment(txdes);
+ ftmac100_txdes_set_txint(txdes);
+ ftmac100_txdes_set_buffer_size(txdes, len);
+
+ spin_lock(&priv->tx_lock);
+ priv->tx_pending++;
+ if (priv->tx_pending == TX_QUEUE_ENTRIES)
+ netif_stop_queue(netdev);
+
+ /* start transmit */
+ ftmac100_txdes_set_dma_own(txdes);
+ spin_unlock(&priv->tx_lock);
+
+ ftmac100_txdma_start_polling(priv);
+ return NETDEV_TX_OK;
+}
+
+/******************************************************************************
+ * internal functions (buffer)
+ *****************************************************************************/
+static int ftmac100_alloc_rx_page(struct ftmac100 *priv, struct ftmac100_rxdes *rxdes)
+{
+ struct net_device *netdev = priv->netdev;
+ struct page *page;
+ dma_addr_t map;
+
+ page = alloc_page(GFP_KERNEL);
+ if (!page) {
+ if (net_ratelimit())
+ netdev_err(netdev, "failed to allocate rx page\n");
+ return -ENOMEM;
+ }
+
+ map = dma_map_page(priv->dev, page, 0, RX_BUF_SIZE, DMA_FROM_DEVICE);
+ if (unlikely(dma_mapping_error(priv->dev, map))) {
+ if (net_ratelimit())
+ netdev_err(netdev, "failed to map rx page\n");
+ __free_page(page);
+ return -ENOMEM;
+ }
+
+ ftmac100_rxdes_set_page(rxdes, page);
+ ftmac100_rxdes_set_dma_addr(rxdes, map);
+ ftmac100_rxdes_set_buffer_size(rxdes, RX_BUF_SIZE);
+ ftmac100_rxdes_set_dma_own(rxdes);
+ return 0;
+}
+
+static void ftmac100_free_buffers(struct ftmac100 *priv)
+{
+ int i;
+
+ for (i = 0; i < RX_QUEUE_ENTRIES; i++) {
+ struct ftmac100_rxdes *rxdes = &priv->descs->rxdes[i];
+ struct page *page = ftmac100_rxdes_get_page(rxdes);
+ dma_addr_t map = ftmac100_rxdes_get_dma_addr(rxdes);
+
+ if (!page)
+ continue;
+
+ dma_unmap_page(priv->dev, map, RX_BUF_SIZE, DMA_FROM_DEVICE);
+ __free_page(page);
+ }
+
+ for (i = 0; i < TX_QUEUE_ENTRIES; i++) {
+ struct ftmac100_txdes *txdes = &priv->descs->txdes[i];
+ struct sk_buff *skb = ftmac100_txdes_get_skb(txdes);
+ dma_addr_t map = ftmac100_txdes_get_dma_addr(txdes);
+
+ if (!skb)
+ continue;
+
+ dma_unmap_single(priv->dev, map, skb_headlen(skb), DMA_TO_DEVICE);
+ dev_kfree_skb(skb);
+ }
+
+ dma_free_coherent(priv->dev, sizeof(struct ftmac100_descs),
+ priv->descs, priv->descs_dma_addr);
+}
+
+static int ftmac100_alloc_buffers(struct ftmac100 *priv)
+{
+ int i;
+
+ priv->descs = dma_alloc_coherent(priv->dev, sizeof(struct ftmac100_descs),
+ &priv->descs_dma_addr, GFP_KERNEL);
+ if (!priv->descs)
+ return -ENOMEM;
+
+ memset(priv->descs, 0, sizeof(struct ftmac100_descs));
+
+ /* initialize RX ring */
+ ftmac100_rxdes_set_end_of_ring(&priv->descs->rxdes[RX_QUEUE_ENTRIES - 1]);
+
+ for (i = 0; i < RX_QUEUE_ENTRIES; i++) {
+ struct ftmac100_rxdes *rxdes = &priv->descs->rxdes[i];
+
+ if (ftmac100_alloc_rx_page(priv, rxdes))
+ goto err;
+ }
+
+ /* initialize TX ring */
+ ftmac100_txdes_set_end_of_ring(&priv->descs->txdes[TX_QUEUE_ENTRIES - 1]);
+ return 0;
+
+err:
+ ftmac100_free_buffers(priv);
+ return -ENOMEM;
+}
+
+/******************************************************************************
+ * struct mii_if_info functions
+ *****************************************************************************/
+static int ftmac100_mdio_read(struct net_device *netdev, int phy_id, int reg)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ unsigned int phycr;
+ int i;
+
+ phycr = FTMAC100_PHYCR_PHYAD(phy_id) |
+ FTMAC100_PHYCR_REGAD(reg) |
+ FTMAC100_PHYCR_MIIRD;
+
+ iowrite32(phycr, priv->base + FTMAC100_OFFSET_PHYCR);
+
+ for (i = 0; i < 10; i++) {
+ phycr = ioread32(priv->base + FTMAC100_OFFSET_PHYCR);
+
+ if ((phycr & FTMAC100_PHYCR_MIIRD) == 0)
+ return phycr & FTMAC100_PHYCR_MIIRDATA;
+
+ usleep_range(100, 1000);
+ }
+
+ netdev_err(netdev, "mdio read timed out\n");
+ return 0;
+}
+
+static void ftmac100_mdio_write(struct net_device *netdev, int phy_id, int reg,
+ int data)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ unsigned int phycr;
+ int i;
+
+ phycr = FTMAC100_PHYCR_PHYAD(phy_id) |
+ FTMAC100_PHYCR_REGAD(reg) |
+ FTMAC100_PHYCR_MIIWR;
+
+ data = FTMAC100_PHYWDATA_MIIWDATA(data);
+
+ iowrite32(data, priv->base + FTMAC100_OFFSET_PHYWDATA);
+ iowrite32(phycr, priv->base + FTMAC100_OFFSET_PHYCR);
+
+ for (i = 0; i < 10; i++) {
+ phycr = ioread32(priv->base + FTMAC100_OFFSET_PHYCR);
+
+ if ((phycr & FTMAC100_PHYCR_MIIWR) == 0)
+ return;
+
+ usleep_range(100, 1000);
+ }
+
+ netdev_err(netdev, "mdio write timed out\n");
+}
+
+/******************************************************************************
+ * struct ethtool_ops functions
+ *****************************************************************************/
+static void ftmac100_get_drvinfo(struct net_device *netdev,
+ struct ethtool_drvinfo *info)
+{
+ strcpy(info->driver, DRV_NAME);
+ strcpy(info->version, DRV_VERSION);
+ strcpy(info->bus_info, dev_name(&netdev->dev));
+}
+
+static int ftmac100_get_settings(struct net_device *netdev, struct ethtool_cmd *cmd)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ return mii_ethtool_gset(&priv->mii, cmd);
+}
+
+static int ftmac100_set_settings(struct net_device *netdev, struct ethtool_cmd *cmd)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ return mii_ethtool_sset(&priv->mii, cmd);
+}
+
+static int ftmac100_nway_reset(struct net_device *netdev)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ return mii_nway_restart(&priv->mii);
+}
+
+static u32 ftmac100_get_link(struct net_device *netdev)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ return mii_link_ok(&priv->mii);
+}
+
+static const struct ethtool_ops ftmac100_ethtool_ops = {
+ .set_settings = ftmac100_set_settings,
+ .get_settings = ftmac100_get_settings,
+ .get_drvinfo = ftmac100_get_drvinfo,
+ .nway_reset = ftmac100_nway_reset,
+ .get_link = ftmac100_get_link,
+};
+
+/******************************************************************************
+ * interrupt handler
+ *****************************************************************************/
+static irqreturn_t ftmac100_interrupt(int irq, void *dev_id)
+{
+ struct net_device *netdev = dev_id;
+ struct ftmac100 *priv = netdev_priv(netdev);
+
+ if (likely(netif_running(netdev))) {
+ /* Disable interrupts for polling */
+ ftmac100_disable_all_int(priv);
+ napi_schedule(&priv->napi);
+ }
+
+ return IRQ_HANDLED;
+}
+
+/******************************************************************************
+ * struct napi_struct functions
+ *****************************************************************************/
+static int ftmac100_poll(struct napi_struct *napi, int budget)
+{
+ struct ftmac100 *priv = container_of(napi, struct ftmac100, napi);
+ struct net_device *netdev = priv->netdev;
+ unsigned int status;
+ bool completed = true;
+ int rx = 0;
+
+ status = ioread32(priv->base + FTMAC100_OFFSET_ISR);
+
+ if (status & (FTMAC100_INT_RPKT_FINISH | FTMAC100_INT_NORXBUF)) {
+ /*
+ * FTMAC100_INT_RPKT_FINISH:
+ * RX DMA has received packets into RX buffer successfully
+ *
+ * FTMAC100_INT_NORXBUF:
+ * RX buffer unavailable
+ */
+ bool retry;
+
+ do {
+ retry = ftmac100_rx_packet(priv, &rx);
+ } while (retry && rx < budget);
+
+ if (retry && rx == budget)
+ completed = false;
+ }
+
+ if (status & (FTMAC100_INT_XPKT_OK | FTMAC100_INT_XPKT_LOST)) {
+ /*
+ * FTMAC100_INT_XPKT_OK:
+ * packet transmitted to ethernet successfully
+ *
+ * FTMAC100_INT_XPKT_LOST:
+ * packet transmitted to ethernet lost due to late
+ * collision or excessive collision
+ */
+ ftmac100_tx_complete(priv);
+ }
+
+ if (status & (FTMAC100_INT_NORXBUF | FTMAC100_INT_RPKT_LOST |
+ FTMAC100_INT_AHB_ERR | FTMAC100_INT_PHYSTS_CHG)) {
+ if (net_ratelimit())
+ netdev_info(netdev, "[ISR] = 0x%x: %s%s%s%s\n", status,
+ status & FTMAC100_INT_NORXBUF ? "NORXBUF " : "",
+ status & FTMAC100_INT_RPKT_LOST ? "RPKT_LOST " : "",
+ status & FTMAC100_INT_AHB_ERR ? "AHB_ERR " : "",
+ status & FTMAC100_INT_PHYSTS_CHG ? "PHYSTS_CHG" : "");
+
+ if (status & FTMAC100_INT_NORXBUF) {
+ /* RX buffer unavailable */
+ netdev->stats.rx_over_errors++;
+ }
+
+ if (status & FTMAC100_INT_RPKT_LOST) {
+ /* received packet lost due to RX FIFO full */
+ netdev->stats.rx_fifo_errors++;
+ }
+
+ if (status & FTMAC100_INT_PHYSTS_CHG) {
+ /* PHY link status change */
+ mii_check_link(&priv->mii);
+ }
+ }
+
+ if (completed) {
+ /* stop polling */
+ napi_complete(napi);
+ ftmac100_enable_all_int(priv);
+ }
+
+ return rx;
+}
+
+/******************************************************************************
+ * struct net_device_ops functions
+ *****************************************************************************/
+static int ftmac100_open(struct net_device *netdev)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ int err;
+
+ err = ftmac100_alloc_buffers(priv);
+ if (err) {
+ netdev_err(netdev, "failed to allocate buffers\n");
+ goto err_alloc;
+ }
+
+ err = request_irq(priv->irq, ftmac100_interrupt, 0, netdev->name, netdev);
+ if (err) {
+ netdev_err(netdev, "failed to request irq %d\n", priv->irq);
+ goto err_irq;
+ }
+
+ priv->rx_pointer = 0;
+ priv->tx_clean_pointer = 0;
+ priv->tx_pointer = 0;
+ priv->tx_pending = 0;
+
+ err = ftmac100_start_hw(priv);
+ if (err)
+ goto err_hw;
+
+ napi_enable(&priv->napi);
+ netif_start_queue(netdev);
+
+ ftmac100_enable_all_int(priv);
+
+ return 0;
+
+err_hw:
+ free_irq(priv->irq, netdev);
+err_irq:
+ ftmac100_free_buffers(priv);
+err_alloc:
+ return err;
+}
+
+static int ftmac100_stop(struct net_device *netdev)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+
+ ftmac100_disable_all_int(priv);
+ netif_stop_queue(netdev);
+ napi_disable(&priv->napi);
+ ftmac100_stop_hw(priv);
+ free_irq(priv->irq, netdev);
+ ftmac100_free_buffers(priv);
+
+ return 0;
+}
+
+static int ftmac100_hard_start_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ dma_addr_t map;
+
+ if (unlikely(skb->len > MAX_PKT_SIZE)) {
+ if (net_ratelimit())
+ netdev_dbg(netdev, "tx packet too big\n");
+
+ netdev->stats.tx_dropped++;
+ dev_kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
+
+ map = dma_map_single(priv->dev, skb->data, skb_headlen(skb), DMA_TO_DEVICE);
+ if (unlikely(dma_mapping_error(priv->dev, map))) {
+ /* drop packet */
+ if (net_ratelimit())
+ netdev_err(netdev, "map socket buffer failed\n");
+
+ netdev->stats.tx_dropped++;
+ dev_kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
+
+ return ftmac100_xmit(priv, skb, map);
+}
+
+/* optional */
+static int ftmac100_do_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
+{
+ struct ftmac100 *priv = netdev_priv(netdev);
+ struct mii_ioctl_data *data = if_mii(ifr);
+
+ return generic_mii_ioctl(&priv->mii, data, cmd, NULL);
+}
+
+static const struct net_device_ops ftmac100_netdev_ops = {
+ .ndo_open = ftmac100_open,
+ .ndo_stop = ftmac100_stop,
+ .ndo_start_xmit = ftmac100_hard_start_xmit,
+ .ndo_set_mac_address = eth_mac_addr,
+ .ndo_validate_addr = eth_validate_addr,
+ .ndo_do_ioctl = ftmac100_do_ioctl,
+};
+
+/******************************************************************************
+ * struct platform_driver functions
+ *****************************************************************************/
+static int ftmac100_probe(struct platform_device *pdev)
+{
+ struct resource *res;
+ int irq;
+ struct net_device *netdev;
+ struct ftmac100 *priv;
+ int err;
+
+ if (!pdev)
+ return -ENODEV;
+
+ res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!res)
+ return -ENXIO;
+
+ irq = platform_get_irq(pdev, 0);
+ if (irq < 0)
+ return irq;
+
+ /* setup net_device */
+ netdev = alloc_etherdev(sizeof(*priv));
+ if (!netdev) {
+ err = -ENOMEM;
+ goto err_alloc_etherdev;
+ }
+
+ SET_NETDEV_DEV(netdev, &pdev->dev);
+ SET_ETHTOOL_OPS(netdev, &ftmac100_ethtool_ops);
+ netdev->netdev_ops = &ftmac100_netdev_ops;
+
+ platform_set_drvdata(pdev, netdev);
+
+ /* setup private data */
+ priv = netdev_priv(netdev);
+ priv->netdev = netdev;
+ priv->dev = &pdev->dev;
+
+ spin_lock_init(&priv->tx_lock);
+
+ /* initialize NAPI */
+ netif_napi_add(netdev, &priv->napi, ftmac100_poll, 64);
+
+ /* map io memory */
+ priv->res = request_mem_region(res->start, resource_size(res),
+ dev_name(&pdev->dev));
+ if (!priv->res) {
+ dev_err(&pdev->dev, "Could not reserve memory region\n");
+ err = -ENOMEM;
+ goto err_req_mem;
+ }
+
+ priv->base = ioremap(res->start, res->end - res->start);
+ if (!priv->base) {
+ dev_err(&pdev->dev, "Failed to ioremap ethernet registers\n");
+ err = -EIO;
+ goto err_ioremap;
+ }
+
+ priv->irq = irq;
+
+ /* initialize struct mii_if_info */
+ priv->mii.phy_id = 0;
+ priv->mii.phy_id_mask = 0x1f;
+ priv->mii.reg_num_mask = 0x1f;
+ priv->mii.dev = netdev;
+ priv->mii.mdio_read = ftmac100_mdio_read;
+ priv->mii.mdio_write = ftmac100_mdio_write;
+
+ /* register network device */
+ err = register_netdev(netdev);
+ if (err) {
+ dev_err(&pdev->dev, "Failed to register netdev\n");
+ goto err_register_netdev;
+ }
+
+ netdev_info(netdev, "irq %d, mapped at %p\n", priv->irq, priv->base);
+
+ if (!is_valid_ether_addr(netdev->dev_addr)) {
+ random_ether_addr(netdev->dev_addr);
+ netdev_info(netdev, "generated random MAC address %pM\n",
+ netdev->dev_addr);
+ }
+
+ return 0;
+
+err_register_netdev:
+ iounmap(priv->base);
+err_ioremap:
+ release_resource(priv->res);
+err_req_mem:
+ netif_napi_del(&priv->napi);
+ platform_set_drvdata(pdev, NULL);
+ free_netdev(netdev);
+err_alloc_etherdev:
+ return err;
+}
+
+static int __exit ftmac100_remove(struct platform_device *pdev)
+{
+ struct net_device *netdev;
+ struct ftmac100 *priv;
+
+ netdev = platform_get_drvdata(pdev);
+ priv = netdev_priv(netdev);
+
+ unregister_netdev(netdev);
+
+ iounmap(priv->base);
+ release_resource(priv->res);
+
+ netif_napi_del(&priv->napi);
+ platform_set_drvdata(pdev, NULL);
+ free_netdev(netdev);
+ return 0;
+}
+
+static struct platform_driver ftmac100_driver = {
+ .probe = ftmac100_probe,
+ .remove = __exit_p(ftmac100_remove),
+ .driver = {
+ .name = DRV_NAME,
+ .owner = THIS_MODULE,
+ },
+};
+
+/******************************************************************************
+ * initialization / finalization
+ *****************************************************************************/
+static int __init ftmac100_init(void)
+{
+ pr_info("Loading version " DRV_VERSION " ...\n");
+ return platform_driver_register(&ftmac100_driver);
+}
+
+static void __exit ftmac100_exit(void)
+{
+ platform_driver_unregister(&ftmac100_driver);
+}
+
+module_init(ftmac100_init);
+module_exit(ftmac100_exit);
+
+MODULE_AUTHOR("Po-Yu Chuang <ratbert@faraday-tech.com>");
+MODULE_DESCRIPTION("FTMAC100 driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/net/ftmac100.h b/drivers/net/ftmac100.h
new file mode 100644
index 0000000..46a0c47
--- /dev/null
+++ b/drivers/net/ftmac100.h
@@ -0,0 +1,180 @@
+/*
+ * Faraday FTMAC100 10/100 Ethernet
+ *
+ * (C) Copyright 2009-2011 Faraday Technology
+ * Po-Yu Chuang <ratbert@faraday-tech.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef __FTMAC100_H
+#define __FTMAC100_H
+
+#define FTMAC100_OFFSET_ISR 0x00
+#define FTMAC100_OFFSET_IMR 0x04
+#define FTMAC100_OFFSET_MAC_MADR 0x08
+#define FTMAC100_OFFSET_MAC_LADR 0x0c
+#define FTMAC100_OFFSET_MAHT0 0x10
+#define FTMAC100_OFFSET_MAHT1 0x14
+#define FTMAC100_OFFSET_TXPD 0x18
+#define FTMAC100_OFFSET_RXPD 0x1c
+#define FTMAC100_OFFSET_TXR_BADR 0x20
+#define FTMAC100_OFFSET_RXR_BADR 0x24
+#define FTMAC100_OFFSET_ITC 0x28
+#define FTMAC100_OFFSET_APTC 0x2c
+#define FTMAC100_OFFSET_DBLAC 0x30
+#define FTMAC100_OFFSET_MACCR 0x88
+#define FTMAC100_OFFSET_MACSR 0x8c
+#define FTMAC100_OFFSET_PHYCR 0x90
+#define FTMAC100_OFFSET_PHYWDATA 0x94
+#define FTMAC100_OFFSET_FCR 0x98
+#define FTMAC100_OFFSET_BPR 0x9c
+#define FTMAC100_OFFSET_TS 0xc4
+#define FTMAC100_OFFSET_DMAFIFOS 0xc8
+#define FTMAC100_OFFSET_TM 0xcc
+#define FTMAC100_OFFSET_TX_MCOL_SCOL 0xd4
+#define FTMAC100_OFFSET_RPF_AEP 0xd8
+#define FTMAC100_OFFSET_XM_PG 0xdc
+#define FTMAC100_OFFSET_RUNT_TLCC 0xe0
+#define FTMAC100_OFFSET_CRCER_FTL 0xe4
+#define FTMAC100_OFFSET_RLC_RCC 0xe8
+#define FTMAC100_OFFSET_BROC 0xec
+#define FTMAC100_OFFSET_MULCA 0xf0
+#define FTMAC100_OFFSET_RP 0xf4
+#define FTMAC100_OFFSET_XP 0xf8
+
+/*
+ * Interrupt status register & interrupt mask register
+ */
+#define FTMAC100_INT_RPKT_FINISH (1 << 0)
+#define FTMAC100_INT_NORXBUF (1 << 1)
+#define FTMAC100_INT_XPKT_FINISH (1 << 2)
+#define FTMAC100_INT_NOTXBUF (1 << 3)
+#define FTMAC100_INT_XPKT_OK (1 << 4)
+#define FTMAC100_INT_XPKT_LOST (1 << 5)
+#define FTMAC100_INT_RPKT_SAV (1 << 6)
+#define FTMAC100_INT_RPKT_LOST (1 << 7)
+#define FTMAC100_INT_AHB_ERR (1 << 8)
+#define FTMAC100_INT_PHYSTS_CHG (1 << 9)
+
+/*
+ * Interrupt timer control register
+ */
+#define FTMAC100_ITC_RXINT_CNT(x) (((x) & 0xf) << 0)
+#define FTMAC100_ITC_RXINT_THR(x) (((x) & 0x7) << 4)
+#define FTMAC100_ITC_RXINT_TIME_SEL (1 << 7)
+#define FTMAC100_ITC_TXINT_CNT(x) (((x) & 0xf) << 8)
+#define FTMAC100_ITC_TXINT_THR(x) (((x) & 0x7) << 12)
+#define FTMAC100_ITC_TXINT_TIME_SEL (1 << 15)
+
+/*
+ * Automatic polling timer control register
+ */
+#define FTMAC100_APTC_RXPOLL_CNT(x) (((x) & 0xf) << 0)
+#define FTMAC100_APTC_RXPOLL_TIME_SEL (1 << 4)
+#define FTMAC100_APTC_TXPOLL_CNT(x) (((x) & 0xf) << 8)
+#define FTMAC100_APTC_TXPOLL_TIME_SEL (1 << 12)
+
+/*
+ * DMA burst length and arbitration control register
+ */
+#define FTMAC100_DBLAC_INCR4_EN (1 << 0)
+#define FTMAC100_DBLAC_INCR8_EN (1 << 1)
+#define FTMAC100_DBLAC_INCR16_EN (1 << 2)
+#define FTMAC100_DBLAC_RXFIFO_LTHR(x) (((x) & 0x7) << 3)
+#define FTMAC100_DBLAC_RXFIFO_HTHR(x) (((x) & 0x7) << 6)
+#define FTMAC100_DBLAC_RX_THR_EN (1 << 9)
+
+/*
+ * MAC control register
+ */
+#define FTMAC100_MACCR_XDMA_EN (1 << 0)
+#define FTMAC100_MACCR_RDMA_EN (1 << 1)
+#define FTMAC100_MACCR_SW_RST (1 << 2)
+#define FTMAC100_MACCR_LOOP_EN (1 << 3)
+#define FTMAC100_MACCR_CRC_DIS (1 << 4)
+#define FTMAC100_MACCR_XMT_EN (1 << 5)
+#define FTMAC100_MACCR_ENRX_IN_HALFTX (1 << 6)
+#define FTMAC100_MACCR_RCV_EN (1 << 8)
+#define FTMAC100_MACCR_HT_MULTI_EN (1 << 9)
+#define FTMAC100_MACCR_RX_RUNT (1 << 10)
+#define FTMAC100_MACCR_RX_FTL (1 << 11)
+#define FTMAC100_MACCR_RCV_ALL (1 << 12)
+#define FTMAC100_MACCR_CRC_APD (1 << 14)
+#define FTMAC100_MACCR_FULLDUP (1 << 15)
+#define FTMAC100_MACCR_RX_MULTIPKT (1 << 16)
+#define FTMAC100_MACCR_RX_BROADPKT (1 << 17)
+
+/*
+ * PHY control register
+ */
+#define FTMAC100_PHYCR_MIIRDATA 0xffff
+#define FTMAC100_PHYCR_PHYAD(x) (((x) & 0x1f) << 16)
+#define FTMAC100_PHYCR_REGAD(x) (((x) & 0x1f) << 21)
+#define FTMAC100_PHYCR_MIIRD (1 << 26)
+#define FTMAC100_PHYCR_MIIWR (1 << 27)
+
+/*
+ * PHY write data register
+ */
+#define FTMAC100_PHYWDATA_MIIWDATA(x) ((x) & 0xffff)
+
+/*
+ * Transmit descriptor, aligned to 16 bytes
+ */
+struct ftmac100_txdes {
+ unsigned int txdes0;
+ unsigned int txdes1;
+ unsigned int txdes2; /* TXBUF_BADR */
+ unsigned int txdes3; /* not used by HW */
+} __attribute__ ((aligned(16)));
+
+#define FTMAC100_TXDES0_TXPKT_LATECOL (1 << 0)
+#define FTMAC100_TXDES0_TXPKT_EXSCOL (1 << 1)
+#define FTMAC100_TXDES0_TXDMA_OWN (1 << 31)
+
+#define FTMAC100_TXDES1_TXBUF_SIZE(x) ((x) & 0x7ff)
+#define FTMAC100_TXDES1_LTS (1 << 27)
+#define FTMAC100_TXDES1_FTS (1 << 28)
+#define FTMAC100_TXDES1_TX2FIC (1 << 29)
+#define FTMAC100_TXDES1_TXIC (1 << 30)
+#define FTMAC100_TXDES1_EDOTR (1 << 31)
+
+/*
+ * Receive descriptor, aligned to 16 bytes
+ */
+struct ftmac100_rxdes {
+ unsigned int rxdes0;
+ unsigned int rxdes1;
+ unsigned int rxdes2; /* RXBUF_BADR */
+ unsigned int rxdes3; /* not used by HW */
+} __attribute__ ((aligned(16)));
+
+#define FTMAC100_RXDES0_RFL 0x7ff
+#define FTMAC100_RXDES0_MULTICAST (1 << 16)
+#define FTMAC100_RXDES0_BROADCAST (1 << 17)
+#define FTMAC100_RXDES0_RX_ERR (1 << 18)
+#define FTMAC100_RXDES0_CRC_ERR (1 << 19)
+#define FTMAC100_RXDES0_FTL (1 << 20)
+#define FTMAC100_RXDES0_RUNT (1 << 21)
+#define FTMAC100_RXDES0_RX_ODD_NB (1 << 22)
+#define FTMAC100_RXDES0_LRS (1 << 28)
+#define FTMAC100_RXDES0_FRS (1 << 29)
+#define FTMAC100_RXDES0_RXDMA_OWN (1 << 31)
+
+#define FTMAC100_RXDES1_RXBUF_SIZE(x) ((x) & 0x7ff)
+#define FTMAC100_RXDES1_EDORR (1 << 31)
+
+#endif /* __FTMAC100_H */
--
1.6.3.3
^ permalink raw reply related
* Re: [PATCH v4] net: add Faraday FTMAC100 10/100 Ethernet driver
From: Eric Dumazet @ 2011-02-24 8:22 UTC (permalink / raw)
To: Po-Yu Chuang
Cc: David Miller, mirqus, netdev, linux-kernel, bhutchings, joe,
dilinger, ratbert
In-Reply-To: <AANLkTikE_OU_qiBUdiDP39js8jQzGqqo1FYW2uHmw3He@mail.gmail.com>
Le jeudi 24 février 2011 à 16:07 +0800, Po-Yu Chuang a écrit :
> Hi David,
>
> On Thu, Feb 24, 2011 at 3:51 PM, David Miller <davem@davemloft.net> wrote:
> > From: Po-Yu Chuang <ratbert.chuang@gmail.com>
> > Date: Thu, 24 Feb 2011 15:27:55 +0800
> >
> >> I guess the problem is because a HW restriction that the rx buffer must be
> >> 64 bits aligned. Since I cannot make rx buffer starts at offset 2 bytes, the
> >> IP header, TCP header and data are not 4 bytes aligned. The performance
> >> drops drastically.
> >
> > I cannot believe that after 20 years of commodity ethernet networking
> > chips were first designed, people are still designing hardware that
> > doesn't do this right.
>
> Ha ha...
> Well, this restriction was removed in the later IPs of our company. :-p
>
> >
> > Just emit garbage bytes into the sub-word alignment padding if the chip
> > wants to word align it's DMA writes.
>
> Not sure what do you mean. The problem is that HW does not accept a
> base address of RX buffer which is not 8 bytes aligned.
I still dont understand the problem, maybe you should post your work as
RFC
Why not using for the first part the skb buffer (eventually copy the
first 128 bytes to get aligned IP/TCP header), and add frags for
following parts ?
^ permalink raw reply
* Re: [PATCH v4] net: add Faraday FTMAC100 10/100 Ethernet driver
From: Po-Yu Chuang @ 2011-02-24 8:07 UTC (permalink / raw)
To: David Miller
Cc: mirqus, netdev, linux-kernel, bhutchings, eric.dumazet, joe,
dilinger, ratbert
In-Reply-To: <20110223.235129.02262510.davem@davemloft.net>
Hi David,
On Thu, Feb 24, 2011 at 3:51 PM, David Miller <davem@davemloft.net> wrote:
> From: Po-Yu Chuang <ratbert.chuang@gmail.com>
> Date: Thu, 24 Feb 2011 15:27:55 +0800
>
>> I guess the problem is because a HW restriction that the rx buffer must be
>> 64 bits aligned. Since I cannot make rx buffer starts at offset 2 bytes, the
>> IP header, TCP header and data are not 4 bytes aligned. The performance
>> drops drastically.
>
> I cannot believe that after 20 years of commodity ethernet networking
> chips were first designed, people are still designing hardware that
> doesn't do this right.
Ha ha...
Well, this restriction was removed in the later IPs of our company. :-p
>
> Just emit garbage bytes into the sub-word alignment padding if the chip
> wants to word align it's DMA writes.
Not sure what do you mean. The problem is that HW does not accept a
base address of RX buffer which is not 8 bytes aligned.
> Even the 15 year old Dec Tulip chips do this properly.
best regards,
Po-Yu Chuang
^ permalink raw reply
* Re: [PATCH] cxgb{3,4}: streamline Kconfig options
From: Jan Beulich @ 2011-02-24 8:00 UTC (permalink / raw)
To: David Miller; +Cc: divy, dm, linux-kbuild, netdev
In-Reply-To: <20110223.122709.193705956.davem@davemloft.net>
>>> On 23.02.11 at 21:27, David Miller <davem@davemloft.net> wrote:
> From: "Jan Beulich" <JBeulich@novell.com>
> Date: Wed, 23 Feb 2011 09:46:10 +0000
>
>>>>> On 22.02.11 at 19:14, David Miller <davem@davemloft.net> wrote:
>>> From: "Jan Beulich" <JBeulich@novell.com>
>>> Date: Thu, 17 Feb 2011 13:29:30 +0000
>>>
>>>> The CHELSIO_T{3,4}_DEPENDS options are really awkward, and can be
>>>> easily dropped if the reverse dependencies of SCSI_CXGB{3,4}_ISCSI on
>>>> the former get converted to normal (forward) ones referring to
>>>> CHELSIO_T{3,4}.
>>>>
>>>> Signed-off-by: Jan Beulich <jbeulich@novell.com>
>>>
>>> I think the goal of these strange rules is not to be complicated
>>> on purpose, but rather to cause the iSCSI drivers to appear without
>>> the user having to know that he needs to enable the networking
>>> driver in order for that to happen.
>>
>> While I realize that this might have been the reason, it's completely
>> contrary to how everyone else writes dependencies, and hence I
>> think these should be removed.
>
> If you knew you were changing the behavior of the config option in
> this way, you sure didn't think it was worth mentioning in your commit
> message.
I stated in the comment what I think this is - awkward.
> I definitely would never expect to have to enable a scsi option to get
> some network driver visible to enable in the config, and therefore I
> could see the opposite being insanely frustrating too.
The resulting dependency seems quite logical to me: Some higher
level networking functionality (iSCSI) depends on some lower level
networking functionality (an actual driver).
> You can't ignore these issues and just say "that's not the normal way
> so I'm going to change it anyways."
Admittedly I considered only my personal perspective.
Now, to get the whole discussion productive again - where do we
go from here? I don't think these drivers are so special that they
really need to behave backwards to how (almost?) everything else
is done... If changing it the way I did in the first try isn't deemed
acceptable, would it be at least acceptable to remove those
helper options (or, not as welcome from my perspective not the
least because of the odd dependency on INET instead of NET,
fold them into a single more generic one that others could also
benefit from)?
As to that INET vs NET dependency - is it possible that the
network drivers really just need NET, but the iSCSI ones need
INET? In which case the only common dependency would be
PCI - certainly not worth a custom helper option.
Jan
^ permalink raw reply
* Re: [PATCH v4] net: add Faraday FTMAC100 10/100 Ethernet driver
From: David Miller @ 2011-02-24 7:51 UTC (permalink / raw)
To: ratbert.chuang
Cc: mirqus, netdev, linux-kernel, bhutchings, eric.dumazet, joe,
dilinger, ratbert
In-Reply-To: <AANLkTimorcNAEbSpuE_dUXKuOBZDcDcGSJ2cGefniThc@mail.gmail.com>
From: Po-Yu Chuang <ratbert.chuang@gmail.com>
Date: Thu, 24 Feb 2011 15:27:55 +0800
> I guess the problem is because a HW restriction that the rx buffer must be
> 64 bits aligned. Since I cannot make rx buffer starts at offset 2 bytes, the
> IP header, TCP header and data are not 4 bytes aligned. The performance
> drops drastically.
I cannot believe that after 20 years of commodity ethernet networking
chips were first designed, people are still designing hardware that
doesn't do this right.
Just emit garbage bytes into the sub-word alignment padding if the chip
wants to word align it's DMA writes.
Even the 15 year old Dec Tulip chips do this properly.
^ permalink raw reply
* Re: [PATCH v4] net: add Faraday FTMAC100 10/100 Ethernet driver
From: Po-Yu Chuang @ 2011-02-24 7:27 UTC (permalink / raw)
To: David Miller
Cc: mirqus, netdev, linux-kernel, bhutchings, eric.dumazet, joe,
dilinger, ratbert
In-Reply-To: <20110131.203556.193730771.davem@davemloft.net>
Hi David,
On Tue, Feb 1, 2011 at 12:35 PM, David Miller <davem@davemloft.net> wrote:
> From: Po-Yu Chuang <ratbert.chuang@gmail.com>
> Date: Tue, 1 Feb 2011 11:56:16 +0800
>
>> If I simply allocate a page for each rx ring entry, I still need to allocate
>> an skb and copy at least packet header in first page to skb->data. Then
>> add the page of rest of payload to skb by skb_fill_page_desc().
>
> You should attach the pages, the use __pskb_pull_tail() to bring in the
> headers to the linear skb->data area.
>
> See drivers/net/niu.c:niu_process_rx_pkt().
I tried two ways to implement zero-copy.
One is to preallocate skb big enough for any rx packet and use the skb
as rx buffer.
The other is use page as rx buffer, use skb_fill_page_desc() to add a
data page to
skb and then pull only header to skb by __pskb_pull_tail() as you suggested.
Two implementations are slower than the original memcpy version.
(benchmarked with iperf)
I guess the problem is because a HW restriction that the rx buffer must be
64 bits aligned. Since I cannot make rx buffer starts at offset 2 bytes, the
IP header, TCP header and data are not 4 bytes aligned. The performance
drops drastically.
Therefore, I will submit later a v6 which is still using memcpy().
best regards,
Po-Yu Chuang
^ permalink raw reply
* Re: [PATCH net-next-2.6 v4] net_sched: SFB flow scheduler
From: Eric Dumazet @ 2011-02-24 7:18 UTC (permalink / raw)
To: Stephen Hemminger
Cc: David Miller, Juliusz.Chroboczek, linville, kaber, netdev, andi
In-Reply-To: <1298531670.2898.22.camel@edumazet-laptop>
Le jeudi 24 février 2011 à 08:14 +0100, Eric Dumazet a écrit :
> Le mercredi 23 février 2011 à 22:51 -0800, Stephen Hemminger a écrit :
> > Where is the iproute (q_sfb) piece?
> >
> >
>
> I'll send it this morning after polishing.
>
> As a matter of fact, I might add Patrick idea to use nested attributes,
> before official SFB release.
>
BTW, one can use old iproute (tc command) for SFB with default params.
^ permalink raw reply
* [PATCH 26/26] xfrm: Const'ify xfrm_address_t args to xfrm_state_find.
From: David Miller @ 2011-02-24 7:17 UTC (permalink / raw)
To: netdev
This required a const'ification in xfrm_init_tempstate() too.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
include/net/xfrm.h | 4 ++--
net/xfrm/xfrm_state.c | 4 ++--
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 3205e5e..44dccfc 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -1334,8 +1334,8 @@ extern int xfrm_state_walk(struct net *net, struct xfrm_state_walk *walk,
int (*func)(struct xfrm_state *, int, void*), void *);
extern void xfrm_state_walk_done(struct xfrm_state_walk *walk);
extern struct xfrm_state *xfrm_state_alloc(struct net *net);
-extern struct xfrm_state *xfrm_state_find(xfrm_address_t *daddr,
- xfrm_address_t *saddr,
+extern struct xfrm_state *xfrm_state_find(const xfrm_address_t *daddr,
+ const xfrm_address_t *saddr,
const struct flowi *fl,
struct xfrm_tmpl *tmpl,
struct xfrm_policy *pol, int *err,
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 9d9ac7c..8496b3d 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -659,7 +659,7 @@ EXPORT_SYMBOL(xfrm_sad_getinfo);
static int
xfrm_init_tempstate(struct xfrm_state *x, const struct flowi *fl,
const struct xfrm_tmpl *tmpl,
- xfrm_address_t *daddr, xfrm_address_t *saddr,
+ const xfrm_address_t *daddr, const xfrm_address_t *saddr,
unsigned short family)
{
struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
@@ -790,7 +790,7 @@ static void xfrm_state_look_at(struct xfrm_policy *pol, struct xfrm_state *x,
}
struct xfrm_state *
-xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
+xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
const struct flowi *fl, struct xfrm_tmpl *tmpl,
struct xfrm_policy *pol, int *err,
unsigned short family)
--
1.7.4.1
^ permalink raw reply related
* [PATCH 25/26] xfrm: Remove unused 'saddr' and 'daddr' args to xfrm_state_look_at.
From: David Miller @ 2011-02-24 7:17 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_state.c | 5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 8a57a1e..9d9ac7c 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -754,7 +754,6 @@ static void xfrm_hash_grow_check(struct net *net, int have_hash_collision)
static void xfrm_state_look_at(struct xfrm_policy *pol, struct xfrm_state *x,
const struct flowi *fl, unsigned short family,
- xfrm_address_t *daddr, xfrm_address_t *saddr,
struct xfrm_state **best, int *acq_in_progress,
int *error)
{
@@ -820,7 +819,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
tmpl->mode == x->props.mode &&
tmpl->id.proto == x->id.proto &&
(tmpl->id.spi == x->id.spi || !tmpl->id.spi))
- xfrm_state_look_at(pol, x, fl, encap_family, daddr, saddr,
+ xfrm_state_look_at(pol, x, fl, encap_family,
&best, &acquire_in_progress, &error);
}
if (best)
@@ -836,7 +835,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
tmpl->mode == x->props.mode &&
tmpl->id.proto == x->id.proto &&
(tmpl->id.spi == x->id.spi || !tmpl->id.spi))
- xfrm_state_look_at(pol, x, fl, encap_family, daddr, saddr,
+ xfrm_state_look_at(pol, x, fl, encap_family,
&best, &acquire_in_progress, &error);
}
--
1.7.4.1
^ permalink raw reply related
* [PATCH 24/26] xfrm: Const'ify xfrm_address_t args to __xfrm_state_lookup{,_byaddr}.
From: David Miller @ 2011-02-24 7:17 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_state.c | 10 ++++++++--
1 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index ac6c48a..8a57a1e 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -678,7 +678,10 @@ xfrm_init_tempstate(struct xfrm_state *x, const struct flowi *fl,
return 0;
}
-static struct xfrm_state *__xfrm_state_lookup(struct net *net, u32 mark, xfrm_address_t *daddr, __be32 spi, u8 proto, unsigned short family)
+static struct xfrm_state *__xfrm_state_lookup(struct net *net, u32 mark,
+ const xfrm_address_t *daddr,
+ __be32 spi, u8 proto,
+ unsigned short family)
{
unsigned int h = xfrm_spi_hash(net, daddr, spi, proto, family);
struct xfrm_state *x;
@@ -700,7 +703,10 @@ static struct xfrm_state *__xfrm_state_lookup(struct net *net, u32 mark, xfrm_ad
return NULL;
}
-static struct xfrm_state *__xfrm_state_lookup_byaddr(struct net *net, u32 mark, xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto, unsigned short family)
+static struct xfrm_state *__xfrm_state_lookup_byaddr(struct net *net, u32 mark,
+ const xfrm_address_t *daddr,
+ const xfrm_address_t *saddr,
+ u8 proto, unsigned short family)
{
unsigned int h = xfrm_src_hash(net, daddr, saddr, family);
struct xfrm_state *x;
--
1.7.4.1
^ permalink raw reply related
* [PATCH 23/26] xfrm: Const'ify xfrm_tmpl arg to xfrm_init_tempstate.
From: David Miller @ 2011-02-24 7:17 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_state.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 0383d83..ac6c48a 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -658,7 +658,7 @@ EXPORT_SYMBOL(xfrm_sad_getinfo);
static int
xfrm_init_tempstate(struct xfrm_state *x, const struct flowi *fl,
- struct xfrm_tmpl *tmpl,
+ const struct xfrm_tmpl *tmpl,
xfrm_address_t *daddr, xfrm_address_t *saddr,
unsigned short family)
{
--
1.7.4.1
^ permalink raw reply related
* [PATCH 22/26] xfrm: Const'ify xfrm_address_t args to xfrm_*_hash.
From: David Miller @ 2011-02-24 7:17 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_state.c | 11 ++++++-----
1 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 555bedd..0383d83 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -50,8 +50,8 @@ static void xfrm_audit_state_replay(struct xfrm_state *x,
#endif /* CONFIG_AUDITSYSCALL */
static inline unsigned int xfrm_dst_hash(struct net *net,
- xfrm_address_t *daddr,
- xfrm_address_t *saddr,
+ const xfrm_address_t *daddr,
+ const xfrm_address_t *saddr,
u32 reqid,
unsigned short family)
{
@@ -59,15 +59,16 @@ static inline unsigned int xfrm_dst_hash(struct net *net,
}
static inline unsigned int xfrm_src_hash(struct net *net,
- xfrm_address_t *daddr,
- xfrm_address_t *saddr,
+ const xfrm_address_t *daddr,
+ const xfrm_address_t *saddr,
unsigned short family)
{
return __xfrm_src_hash(daddr, saddr, family, net->xfrm.state_hmask);
}
static inline unsigned int
-xfrm_spi_hash(struct net *net, xfrm_address_t *daddr, __be32 spi, u8 proto, unsigned short family)
+xfrm_spi_hash(struct net *net, const xfrm_address_t *daddr,
+ __be32 spi, u8 proto, unsigned short family)
{
return __xfrm_spi_hash(daddr, spi, proto, family, net->xfrm.state_hmask);
}
--
1.7.4.1
^ permalink raw reply related
* [PATCH 21/26] xfrm: Const'ify sec_path arg to secpath_has_nontransport.
From: David Miller @ 2011-02-24 7:16 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_policy.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 2de0bc2..41a91d2 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1994,7 +1994,7 @@ int __xfrm_decode_session(struct sk_buff *skb, struct flowi *fl,
}
EXPORT_SYMBOL(__xfrm_decode_session);
-static inline int secpath_has_nontransport(struct sec_path *sp, int k, int *idxp)
+static inline int secpath_has_nontransport(const struct sec_path *sp, int k, int *idxp)
{
for (; k < sp->len; k++) {
if (sp->xvec[k]->props.mode != XFRM_MODE_TRANSPORT) {
--
1.7.4.1
^ permalink raw reply related
* [PATCH 20/26] xfrm: Const'ify ptr args to xfrm_policy_ok.
From: David Miller @ 2011-02-24 7:16 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_policy.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index f766e5f..2de0bc2 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1956,7 +1956,7 @@ xfrm_state_ok(const struct xfrm_tmpl *tmpl, const struct xfrm_state *x,
* Otherwise "-2 - errored_index" is returned.
*/
static inline int
-xfrm_policy_ok(struct xfrm_tmpl *tmpl, struct sec_path *sp, int start,
+xfrm_policy_ok(const struct xfrm_tmpl *tmpl, const struct sec_path *sp, int start,
unsigned short family)
{
int idx = start;
--
1.7.4.1
^ permalink raw reply related
* [PATCH 19/26] xfrm: Const'ify ptr args to xfrm_state_ok.
From: David Miller @ 2011-02-24 7:16 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/xfrm/xfrm_policy.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 3a4221a..f766e5f 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1933,7 +1933,7 @@ xfrm_secpath_reject(int idx, struct sk_buff *skb, const struct flowi *fl)
*/
static inline int
-xfrm_state_ok(struct xfrm_tmpl *tmpl, struct xfrm_state *x,
+xfrm_state_ok(const struct xfrm_tmpl *tmpl, const struct xfrm_state *x,
unsigned short family)
{
if (xfrm_state_kern(x))
--
1.7.4.1
^ permalink raw reply related
* [PATCH 18/26] xfrm: Const'ify ptr args to xfrm_state_*_check and xfrm_state_kern.
From: David Miller @ 2011-02-24 7:16 UTC (permalink / raw)
To: netdev
Signed-off-by: David S. Miller <davem@davemloft.net>
---
include/net/xfrm.h | 24 ++++++++++++------------
1 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index f6d2f63..3205e5e 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -1178,8 +1178,8 @@ void xfrm_flowi_addr_get(const struct flowi *fl,
}
static __inline__ int
-__xfrm4_state_addr_check(struct xfrm_state *x,
- xfrm_address_t *daddr, xfrm_address_t *saddr)
+__xfrm4_state_addr_check(const struct xfrm_state *x,
+ const xfrm_address_t *daddr, const xfrm_address_t *saddr)
{
if (daddr->a4 == x->id.daddr.a4 &&
(saddr->a4 == x->props.saddr.a4 || !saddr->a4 || !x->props.saddr.a4))
@@ -1188,8 +1188,8 @@ __xfrm4_state_addr_check(struct xfrm_state *x,
}
static __inline__ int
-__xfrm6_state_addr_check(struct xfrm_state *x,
- xfrm_address_t *daddr, xfrm_address_t *saddr)
+__xfrm6_state_addr_check(const struct xfrm_state *x,
+ const xfrm_address_t *daddr, const xfrm_address_t *saddr)
{
if (!ipv6_addr_cmp((struct in6_addr *)daddr, (struct in6_addr *)&x->id.daddr) &&
(!ipv6_addr_cmp((struct in6_addr *)saddr, (struct in6_addr *)&x->props.saddr)||
@@ -1200,8 +1200,8 @@ __xfrm6_state_addr_check(struct xfrm_state *x,
}
static __inline__ int
-xfrm_state_addr_check(struct xfrm_state *x,
- xfrm_address_t *daddr, xfrm_address_t *saddr,
+xfrm_state_addr_check(const struct xfrm_state *x,
+ const xfrm_address_t *daddr, const xfrm_address_t *saddr,
unsigned short family)
{
switch (family) {
@@ -1214,23 +1214,23 @@ xfrm_state_addr_check(struct xfrm_state *x,
}
static __inline__ int
-xfrm_state_addr_flow_check(struct xfrm_state *x, const struct flowi *fl,
+xfrm_state_addr_flow_check(const struct xfrm_state *x, const struct flowi *fl,
unsigned short family)
{
switch (family) {
case AF_INET:
return __xfrm4_state_addr_check(x,
- (xfrm_address_t *)&fl->fl4_dst,
- (xfrm_address_t *)&fl->fl4_src);
+ (const xfrm_address_t *)&fl->fl4_dst,
+ (const xfrm_address_t *)&fl->fl4_src);
case AF_INET6:
return __xfrm6_state_addr_check(x,
- (xfrm_address_t *)&fl->fl6_dst,
- (xfrm_address_t *)&fl->fl6_src);
+ (const xfrm_address_t *)&fl->fl6_dst,
+ (const xfrm_address_t *)&fl->fl6_src);
}
return 0;
}
-static inline int xfrm_state_kern(struct xfrm_state *x)
+static inline int xfrm_state_kern(const struct xfrm_state *x)
{
return atomic_read(&x->tunnel_users);
}
--
1.7.4.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox