* [net-next PATCH 02/26] igb: move setting of buffsz out of repeated path in alloc_rx_buffers
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
@ 2009-02-07 9:15 ` Jeff Kirsher
2009-02-07 9:15 ` [net-next PATCH 03/26] igb: move initialization of number of queues into set_interrupt_capability Jeff Kirsher
` (24 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:15 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
buffsz is being repeatedly set when allocaing buffers. Since this value
should only need to be set once in the function I am moving it out of the
looped portion of the path.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 19 ++++++++-----------
1 files changed, 8 insertions(+), 11 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 21f0c22..ec5855a 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -3943,10 +3943,17 @@ static void igb_alloc_rx_buffers_adv(struct igb_ring *rx_ring,
struct igb_buffer *buffer_info;
struct sk_buff *skb;
unsigned int i;
+ int bufsz;
i = rx_ring->next_to_use;
buffer_info = &rx_ring->buffer_info[i];
+ if (adapter->rx_ps_hdr_size)
+ bufsz = adapter->rx_ps_hdr_size;
+ else
+ bufsz = adapter->rx_buffer_len;
+ bufsz += NET_IP_ALIGN;
+
while (cleaned_count--) {
rx_desc = E1000_RX_DESC_ADV(*rx_ring, i);
@@ -3962,23 +3969,14 @@ static void igb_alloc_rx_buffers_adv(struct igb_ring *rx_ring,
buffer_info->page_offset ^= PAGE_SIZE / 2;
}
buffer_info->page_dma =
- pci_map_page(pdev,
- buffer_info->page,
+ pci_map_page(pdev, buffer_info->page,
buffer_info->page_offset,
PAGE_SIZE / 2,
PCI_DMA_FROMDEVICE);
}
if (!buffer_info->skb) {
- int bufsz;
-
- if (adapter->rx_ps_hdr_size)
- bufsz = adapter->rx_ps_hdr_size;
- else
- bufsz = adapter->rx_buffer_len;
- bufsz += NET_IP_ALIGN;
skb = netdev_alloc_skb(netdev, bufsz);
-
if (!skb) {
adapter->alloc_rx_buff_failed++;
goto no_buffers;
@@ -3994,7 +3992,6 @@ static void igb_alloc_rx_buffers_adv(struct igb_ring *rx_ring,
buffer_info->dma = pci_map_single(pdev, skb->data,
bufsz,
PCI_DMA_FROMDEVICE);
-
}
/* Refresh the desc even if buffer_addrs didn't change because
* each write-back erases this info. */
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 03/26] igb: move initialization of number of queues into set_interrupt_capability
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
2009-02-07 9:15 ` [net-next PATCH 02/26] igb: move setting of buffsz out of repeated path in alloc_rx_buffers Jeff Kirsher
@ 2009-02-07 9:15 ` Jeff Kirsher
2009-02-07 9:16 ` [net-next PATCH 04/26] igb: remove check for needing an io port Jeff Kirsher
` (23 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:15 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch moves the initialization of the number of queues into
set_interrupt_capability. This allows the number of queues to increase in
the unlikely event that the system initially fails to allocate enough msi-x
interrupts, does a suspend/resume, and then can allocate enough interrupts
on resume.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 10 +++++-----
1 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index ec5855a..6ef0adb 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -543,6 +543,11 @@ static void igb_set_interrupt_capability(struct igb_adapter *adapter)
int err;
int numvecs, i;
+ /* Number of supported queues. */
+ /* Having more queues than CPUs doesn't make sense. */
+ adapter->num_rx_queues = min_t(u32, IGB_MAX_RX_QUEUES, num_online_cpus());
+ adapter->num_tx_queues = min_t(u32, IGB_MAX_TX_QUEUES, num_online_cpus());
+
numvecs = adapter->num_tx_queues + adapter->num_rx_queues + 1;
adapter->msix_entries = kcalloc(numvecs, sizeof(struct msix_entry),
GFP_KERNEL);
@@ -1450,11 +1455,6 @@ static int __devinit igb_sw_init(struct igb_adapter *adapter)
adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
- /* Number of supported queues. */
- /* Having more queues than CPUs doesn't make sense. */
- adapter->num_rx_queues = min_t(u32, IGB_MAX_RX_QUEUES, num_online_cpus());
- adapter->num_tx_queues = min_t(u32, IGB_MAX_TX_QUEUES, num_online_cpus());
-
/* This call may decrease the number of queues depending on
* interrupt mode. */
igb_set_interrupt_capability(adapter);
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 04/26] igb: remove check for needing an io port
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
2009-02-07 9:15 ` [net-next PATCH 02/26] igb: move setting of buffsz out of repeated path in alloc_rx_buffers Jeff Kirsher
2009-02-07 9:15 ` [net-next PATCH 03/26] igb: move initialization of number of queues into set_interrupt_capability Jeff Kirsher
@ 2009-02-07 9:16 ` Jeff Kirsher
2009-02-07 9:16 ` [net-next PATCH 05/26] igb: add link check function Jeff Kirsher
` (22 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:16 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Since igb supports only pci-e nics and there is no plan to support any
legacy pci parts in the driver there isn't really much need for checking to
see if an io port is needed.
In the unlikely event that we do begin supporting legacy pci parts then we
can see about adding this code back to the driver.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb.h | 4 ----
drivers/net/igb/igb_main.c | 50 +++++++++-----------------------------------
2 files changed, 10 insertions(+), 44 deletions(-)
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 30657dd..530d7aa 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -273,10 +273,6 @@ struct igb_adapter {
unsigned int flags;
u32 eeprom_wol;
- /* for ioport free */
- int bars;
- int need_ioport;
-
struct igb_ring *multi_tx_table[IGB_MAX_TX_QUEUES];
unsigned int tx_ring_count;
unsigned int rx_ring_count;
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 6ef0adb..13b10ba 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -974,21 +974,6 @@ void igb_reset(struct igb_adapter *adapter)
igb_get_phy_info(&adapter->hw);
}
-/**
- * igb_is_need_ioport - determine if an adapter needs ioport resources or not
- * @pdev: PCI device information struct
- *
- * Returns true if an adapter needs ioport resources
- **/
-static int igb_is_need_ioport(struct pci_dev *pdev)
-{
- switch (pdev->device) {
- /* Currently there are no adapters that need ioport resources */
- default:
- return false;
- }
-}
-
static const struct net_device_ops igb_netdev_ops = {
.ndo_open = igb_open,
.ndo_stop = igb_close,
@@ -1032,17 +1017,8 @@ static int __devinit igb_probe(struct pci_dev *pdev,
u16 eeprom_data = 0, state = 0;
u16 eeprom_apme_mask = IGB_EEPROM_APME;
u32 part_num;
- int bars, need_ioport;
- /* do not allocate ioport bars when not needed */
- need_ioport = igb_is_need_ioport(pdev);
- if (need_ioport) {
- bars = pci_select_bars(pdev, IORESOURCE_MEM | IORESOURCE_IO);
- err = pci_enable_device(pdev);
- } else {
- bars = pci_select_bars(pdev, IORESOURCE_MEM);
- err = pci_enable_device_mem(pdev);
- }
+ err = pci_enable_device_mem(pdev);
if (err)
return err;
@@ -1085,7 +1061,9 @@ static int __devinit igb_probe(struct pci_dev *pdev,
break;
}
- err = pci_request_selected_regions(pdev, bars, igb_driver_name);
+ err = pci_request_selected_regions(pdev, pci_select_bars(pdev,
+ IORESOURCE_MEM),
+ igb_driver_name);
if (err)
goto err_pci_reg;
@@ -1113,8 +1091,6 @@ static int __devinit igb_probe(struct pci_dev *pdev,
hw = &adapter->hw;
hw->back = adapter;
adapter->msg_enable = NETIF_MSG_DRV | NETIF_MSG_PROBE;
- adapter->bars = bars;
- adapter->need_ioport = need_ioport;
mmio_start = pci_resource_start(pdev, 0);
mmio_len = pci_resource_len(pdev, 0);
@@ -1361,7 +1337,8 @@ err_hw_init:
err_ioremap:
free_netdev(netdev);
err_alloc_etherdev:
- pci_release_selected_regions(pdev, bars);
+ pci_release_selected_regions(pdev, pci_select_bars(pdev,
+ IORESOURCE_MEM));
err_pci_reg:
err_dma:
pci_disable_device(pdev);
@@ -1420,7 +1397,8 @@ static void __devexit igb_remove(struct pci_dev *pdev)
iounmap(adapter->hw.hw_addr);
if (adapter->hw.flash_address)
iounmap(adapter->hw.flash_address);
- pci_release_selected_regions(pdev, adapter->bars);
+ pci_release_selected_regions(pdev, pci_select_bars(pdev,
+ IORESOURCE_MEM));
free_netdev(netdev);
@@ -4309,10 +4287,7 @@ static int igb_resume(struct pci_dev *pdev)
pci_set_power_state(pdev, PCI_D0);
pci_restore_state(pdev);
- if (adapter->need_ioport)
- err = pci_enable_device(pdev);
- else
- err = pci_enable_device_mem(pdev);
+ err = pci_enable_device_mem(pdev);
if (err) {
dev_err(&pdev->dev,
"igb: Cannot enable PCI device from suspend\n");
@@ -4423,12 +4398,7 @@ static pci_ers_result_t igb_io_slot_reset(struct pci_dev *pdev)
pci_ers_result_t result;
int err;
- if (adapter->need_ioport)
- err = pci_enable_device(pdev);
- else
- err = pci_enable_device_mem(pdev);
-
- if (err) {
+ if (pci_enable_device_mem(pdev)) {
dev_err(&pdev->dev,
"Cannot re-enable PCI device after reset.\n");
result = PCI_ERS_RESULT_DISCONNECT;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 05/26] igb: add link check function
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (2 preceding siblings ...)
2009-02-07 9:16 ` [net-next PATCH 04/26] igb: remove check for needing an io port Jeff Kirsher
@ 2009-02-07 9:16 ` Jeff Kirsher
2009-02-07 9:16 ` [net-next PATCH 06/26] igb: make dev_spec a union and remove dynamic allocation Jeff Kirsher
` (21 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:16 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Add a link check function to contain all activities related to verifying
that the link is present. The current approach is a bit cludgy and needs
to be cleaned up.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 59 +++++++++++++++++++++++++++++++-------------
1 files changed, 42 insertions(+), 17 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 13b10ba..b59088e 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2261,6 +2261,46 @@ static void igb_update_phy_info(unsigned long data)
}
/**
+ * igb_has_link - check shared code for link and determine up/down
+ * @adapter: pointer to driver private info
+ **/
+static bool igb_has_link(struct igb_adapter *adapter)
+{
+ struct e1000_hw *hw = &adapter->hw;
+ bool link_active = false;
+ s32 ret_val = 0;
+
+ /* get_link_status is set on LSC (link status) interrupt or
+ * rx sequence error interrupt. get_link_status will stay
+ * false until the e1000_check_for_link establishes link
+ * for copper adapters ONLY
+ */
+ switch (hw->phy.media_type) {
+ case e1000_media_type_copper:
+ if (hw->mac.get_link_status) {
+ ret_val = hw->mac.ops.check_for_link(hw);
+ link_active = !hw->mac.get_link_status;
+ } else {
+ link_active = true;
+ }
+ break;
+ case e1000_media_type_fiber:
+ ret_val = hw->mac.ops.check_for_link(hw);
+ link_active = !!(rd32(E1000_STATUS) & E1000_STATUS_LU);
+ break;
+ case e1000_media_type_internal_serdes:
+ ret_val = hw->mac.ops.check_for_link(hw);
+ link_active = hw->mac.serdes_has_link;
+ break;
+ default:
+ case e1000_media_type_unknown:
+ break;
+ }
+
+ return link_active;
+}
+
+/**
* igb_watchdog - Timer Call-back
* @data: pointer to adapter cast into an unsigned long
**/
@@ -2285,25 +2325,10 @@ static void igb_watchdog_task(struct work_struct *work)
s32 ret_val;
int i;
- if ((netif_carrier_ok(netdev)) &&
- (rd32(E1000_STATUS) & E1000_STATUS_LU))
+ link = igb_has_link(adapter);
+ if ((netif_carrier_ok(netdev)) && link)
goto link_up;
- ret_val = hw->mac.ops.check_for_link(&adapter->hw);
- if ((ret_val == E1000_ERR_PHY) &&
- (hw->phy.type == e1000_phy_igp_3) &&
- (rd32(E1000_CTRL) &
- E1000_PHY_CTRL_GBE_DISABLE))
- dev_info(&adapter->pdev->dev,
- "Gigabit has been disabled, downgrading speed\n");
-
- if ((hw->phy.media_type == e1000_media_type_internal_serdes) &&
- !(rd32(E1000_TXCW) & E1000_TXCW_ANE))
- link = mac->serdes_has_link;
- else
- link = rd32(E1000_STATUS) &
- E1000_STATUS_LU;
-
if (link) {
if (!netif_carrier_ok(netdev)) {
u32 ctrl;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 06/26] igb: make dev_spec a union and remove dynamic allocation
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (3 preceding siblings ...)
2009-02-07 9:16 ` [net-next PATCH 05/26] igb: add link check function Jeff Kirsher
@ 2009-02-07 9:16 ` Jeff Kirsher
2009-02-07 9:17 ` [net-next PATCH 07/26] igb: read address from RAH/RAL instead of from EEPROM Jeff Kirsher
` (20 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:16 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch makes dev_spec a union and simplifies it so that it does not
require dynamic allocation and freeing in the driver.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_82575.c | 34 +++++-----------------------------
drivers/net/igb/e1000_hw.h | 9 +++++++--
drivers/net/igb/e1000_mac.c | 13 -------------
drivers/net/igb/e1000_mac.h | 1 -
drivers/net/igb/igb_main.c | 2 --
5 files changed, 12 insertions(+), 47 deletions(-)
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index f5e4cad..ed9e8c0 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -62,17 +62,12 @@ static bool igb_sgmii_active_82575(struct e1000_hw *);
static s32 igb_reset_init_script_82575(struct e1000_hw *);
static s32 igb_read_mac_addr_82575(struct e1000_hw *);
-
-struct e1000_dev_spec_82575 {
- bool sgmii_active;
-};
-
static s32 igb_get_invariants_82575(struct e1000_hw *hw)
{
struct e1000_phy_info *phy = &hw->phy;
struct e1000_nvm_info *nvm = &hw->nvm;
struct e1000_mac_info *mac = &hw->mac;
- struct e1000_dev_spec_82575 *dev_spec;
+ struct e1000_dev_spec_82575 * dev_spec = &hw->dev_spec._82575;
u32 eecd;
s32 ret_val;
u16 size;
@@ -94,17 +89,6 @@ static s32 igb_get_invariants_82575(struct e1000_hw *hw)
break;
}
- /* MAC initialization */
- hw->dev_spec_size = sizeof(struct e1000_dev_spec_82575);
-
- /* Device-specific structure allocation */
- hw->dev_spec = kzalloc(hw->dev_spec_size, GFP_KERNEL);
-
- if (!hw->dev_spec)
- return -ENOMEM;
-
- dev_spec = (struct e1000_dev_spec_82575 *)hw->dev_spec;
-
/* Set media type */
/*
* The 82575 uses bits 22:23 for link mode. The mode can be changed
@@ -1234,20 +1218,12 @@ out:
**/
static bool igb_sgmii_active_82575(struct e1000_hw *hw)
{
- struct e1000_dev_spec_82575 *dev_spec;
- bool ret_val;
+ struct e1000_dev_spec_82575 *dev_spec = &hw->dev_spec._82575;
- if (hw->mac.type != e1000_82575) {
- ret_val = false;
- goto out;
- }
-
- dev_spec = (struct e1000_dev_spec_82575 *)hw->dev_spec;
+ if (hw->mac.type != e1000_82575 && hw->mac.type != e1000_82576)
+ return false;
- ret_val = dev_spec->sgmii_active;
-
-out:
- return ret_val;
+ return dev_spec->sgmii_active;
}
/**
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index 99504a6..06f72ae 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -565,9 +565,12 @@ struct e1000_fc_info {
enum e1000_fc_type original_type;
};
+struct e1000_dev_spec_82575 {
+ bool sgmii_active;
+};
+
struct e1000_hw {
void *back;
- void *dev_spec;
u8 __iomem *hw_addr;
u8 __iomem *flash_address;
@@ -580,7 +583,9 @@ struct e1000_hw {
struct e1000_bus_info bus;
struct e1000_host_mng_dhcp_cookie mng_cookie;
- u32 dev_spec_size;
+ union {
+ struct e1000_dev_spec_82575 _82575;
+ } dev_spec;
u16 device_id;
u16 subsystem_vendor_id;
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index 97f0049..16fa083 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -37,19 +37,6 @@
static s32 igb_set_default_fc(struct e1000_hw *hw);
static s32 igb_set_fc_watermarks(struct e1000_hw *hw);
-/**
- * igb_remove_device - Free device specific structure
- * @hw: pointer to the HW structure
- *
- * If a device specific structure was allocated, this function will
- * free it.
- **/
-void igb_remove_device(struct e1000_hw *hw)
-{
- /* Freeing the dev_spec member of e1000_hw structure */
- kfree(hw->dev_spec);
-}
-
static s32 igb_read_pcie_cap_reg(struct e1000_hw *hw, u32 reg, u16 *value)
{
struct igb_adapter *adapter = hw->back;
diff --git a/drivers/net/igb/e1000_mac.h b/drivers/net/igb/e1000_mac.h
index cbee6af..4ef40d5 100644
--- a/drivers/net/igb/e1000_mac.h
+++ b/drivers/net/igb/e1000_mac.h
@@ -63,7 +63,6 @@ void igb_mta_set(struct e1000_hw *hw, u32 hash_value);
void igb_put_hw_semaphore(struct e1000_hw *hw);
void igb_rar_set(struct e1000_hw *hw, u8 *addr, u32 index);
s32 igb_check_alt_mac_addr(struct e1000_hw *hw);
-void igb_remove_device(struct e1000_hw *hw);
void igb_reset_adaptive(struct e1000_hw *hw);
void igb_update_adaptive(struct e1000_hw *hw);
void igb_write_vfta(struct e1000_hw *hw, u32 offset, u32 value);
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index b59088e..cb3ac34 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1329,7 +1329,6 @@ err_eeprom:
if (hw->flash_address)
iounmap(hw->flash_address);
- igb_remove_device(hw);
igb_free_queues(adapter);
err_sw_init:
err_hw_init:
@@ -1389,7 +1388,6 @@ static void __devexit igb_remove(struct pci_dev *pdev)
if (!igb_check_reset_block(&adapter->hw))
igb_reset_phy(&adapter->hw);
- igb_remove_device(&adapter->hw);
igb_reset_interrupt_capability(adapter);
igb_free_queues(adapter);
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 07/26] igb: read address from RAH/RAL instead of from EEPROM
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (4 preceding siblings ...)
2009-02-07 9:16 ` [net-next PATCH 06/26] igb: make dev_spec a union and remove dynamic allocation Jeff Kirsher
@ 2009-02-07 9:17 ` Jeff Kirsher
2009-02-07 9:17 ` [net-next PATCH 08/26] igb: rename phy ops Jeff Kirsher
` (19 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:17 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Instead of pulling the mac address from EEPROM it is easier to pull it from
the RAL/RAH registers and then just copy it into the address structures.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_defines.h | 2 ++
drivers/net/igb/e1000_nvm.c | 28 +++++++++++-----------------
2 files changed, 13 insertions(+), 17 deletions(-)
diff --git a/drivers/net/igb/e1000_defines.h b/drivers/net/igb/e1000_defines.h
index 40d0342..54a1489 100644
--- a/drivers/net/igb/e1000_defines.h
+++ b/drivers/net/igb/e1000_defines.h
@@ -481,6 +481,8 @@
* manageability enabled, allowing us room for 15 multicast addresses.
*/
#define E1000_RAH_AV 0x80000000 /* Receive descriptor valid */
+#define E1000_RAL_MAC_ADDR_LEN 4
+#define E1000_RAH_MAC_ADDR_LEN 2
/* Error Codes */
#define E1000_ERR_NVM 1
diff --git a/drivers/net/igb/e1000_nvm.c b/drivers/net/igb/e1000_nvm.c
index a84e4e4..5942da1 100644
--- a/drivers/net/igb/e1000_nvm.c
+++ b/drivers/net/igb/e1000_nvm.c
@@ -515,29 +515,23 @@ out:
**/
s32 igb_read_mac_addr(struct e1000_hw *hw)
{
- s32 ret_val = 0;
- u16 offset, nvm_data, i;
+ u32 rar_high;
+ u32 rar_low;
+ u16 i;
- for (i = 0; i < ETH_ALEN; i += 2) {
- offset = i >> 1;
- ret_val = hw->nvm.ops.read_nvm(hw, offset, 1, &nvm_data);
- if (ret_val) {
- hw_dbg("NVM Read Error\n");
- goto out;
- }
- hw->mac.perm_addr[i] = (u8)(nvm_data & 0xFF);
- hw->mac.perm_addr[i+1] = (u8)(nvm_data >> 8);
- }
+ rar_high = rd32(E1000_RAH(0));
+ rar_low = rd32(E1000_RAL(0));
+
+ for (i = 0; i < E1000_RAL_MAC_ADDR_LEN; i++)
+ hw->mac.perm_addr[i] = (u8)(rar_low >> (i*8));
- /* Flip last bit of mac address if we're on second port */
- if (hw->bus.func == E1000_FUNC_1)
- hw->mac.perm_addr[5] ^= 1;
+ for (i = 0; i < E1000_RAH_MAC_ADDR_LEN; i++)
+ hw->mac.perm_addr[i+4] = (u8)(rar_high >> (i*8));
for (i = 0; i < ETH_ALEN; i++)
hw->mac.addr[i] = hw->mac.perm_addr[i];
-out:
- return ret_val;
+ return 0;
}
/**
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 08/26] igb: rename phy ops
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (5 preceding siblings ...)
2009-02-07 9:17 ` [net-next PATCH 07/26] igb: read address from RAH/RAL instead of from EEPROM Jeff Kirsher
@ 2009-02-07 9:17 ` Jeff Kirsher
2009-02-07 9:17 ` [net-next PATCH 09/26] igb: rename nvm ops Jeff Kirsher
` (18 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:17 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch renames write_phy_reg to write_reg and read_phy_reg to read_reg.
It seems redundant to call out phy in an operation that is part of the
phy_ops struct.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_82575.c | 38 ++--
drivers/net/igb/e1000_hw.h | 10 +
drivers/net/igb/e1000_mac.c | 8 -
drivers/net/igb/e1000_phy.c | 347 +++++++++++++++--------------------------
drivers/net/igb/e1000_phy.h | 1
drivers/net/igb/igb.h | 12 +
6 files changed, 162 insertions(+), 254 deletions(-)
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index ed9e8c0..9a66e34 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -179,13 +179,13 @@ static s32 igb_get_invariants_82575(struct e1000_hw *hw)
/* PHY function pointers */
if (igb_sgmii_active_82575(hw)) {
- phy->ops.reset_phy = igb_phy_hw_reset_sgmii_82575;
- phy->ops.read_phy_reg = igb_read_phy_reg_sgmii_82575;
- phy->ops.write_phy_reg = igb_write_phy_reg_sgmii_82575;
+ phy->ops.reset = igb_phy_hw_reset_sgmii_82575;
+ phy->ops.read_reg = igb_read_phy_reg_sgmii_82575;
+ phy->ops.write_reg = igb_write_phy_reg_sgmii_82575;
} else {
- phy->ops.reset_phy = igb_phy_hw_reset;
- phy->ops.read_phy_reg = igb_read_phy_reg_igp;
- phy->ops.write_phy_reg = igb_write_phy_reg_igp;
+ phy->ops.reset = igb_phy_hw_reset;
+ phy->ops.read_reg = igb_read_phy_reg_igp;
+ phy->ops.write_reg = igb_write_phy_reg_igp;
}
/* Set phy->phy_addr and phy->id. */
@@ -435,7 +435,7 @@ static s32 igb_phy_hw_reset_sgmii_82575(struct e1000_hw *hw)
* SFP documentation requires the following to configure the SPF module
* to work on SGMII. No further documentation is given.
*/
- ret_val = hw->phy.ops.write_phy_reg(hw, 0x1B, 0x8084);
+ ret_val = hw->phy.ops.write_reg(hw, 0x1B, 0x8084);
if (ret_val)
goto out;
@@ -464,28 +464,28 @@ static s32 igb_set_d0_lplu_state_82575(struct e1000_hw *hw, bool active)
s32 ret_val;
u16 data;
- ret_val = phy->ops.read_phy_reg(hw, IGP02E1000_PHY_POWER_MGMT, &data);
+ ret_val = phy->ops.read_reg(hw, IGP02E1000_PHY_POWER_MGMT, &data);
if (ret_val)
goto out;
if (active) {
data |= IGP02E1000_PM_D0_LPLU;
- ret_val = phy->ops.write_phy_reg(hw, IGP02E1000_PHY_POWER_MGMT,
+ ret_val = phy->ops.write_reg(hw, IGP02E1000_PHY_POWER_MGMT,
data);
if (ret_val)
goto out;
/* When LPLU is enabled, we should disable SmartSpeed */
- ret_val = phy->ops.read_phy_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
&data);
data &= ~IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = phy->ops.write_phy_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
+ ret_val = phy->ops.write_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
data);
if (ret_val)
goto out;
} else {
data &= ~IGP02E1000_PM_D0_LPLU;
- ret_val = phy->ops.write_phy_reg(hw, IGP02E1000_PHY_POWER_MGMT,
+ ret_val = phy->ops.write_reg(hw, IGP02E1000_PHY_POWER_MGMT,
data);
/*
* LPLU and SmartSpeed are mutually exclusive. LPLU is used
@@ -494,24 +494,24 @@ static s32 igb_set_d0_lplu_state_82575(struct e1000_hw *hw, bool active)
* SmartSpeed, so performance is maintained.
*/
if (phy->smart_speed == e1000_smart_speed_on) {
- ret_val = phy->ops.read_phy_reg(hw,
+ ret_val = phy->ops.read_reg(hw,
IGP01E1000_PHY_PORT_CONFIG, &data);
if (ret_val)
goto out;
data |= IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = phy->ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
IGP01E1000_PHY_PORT_CONFIG, data);
if (ret_val)
goto out;
} else if (phy->smart_speed == e1000_smart_speed_off) {
- ret_val = phy->ops.read_phy_reg(hw,
+ ret_val = phy->ops.read_reg(hw,
IGP01E1000_PHY_PORT_CONFIG, &data);
if (ret_val)
goto out;
data &= ~IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = phy->ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
IGP01E1000_PHY_PORT_CONFIG, data);
if (ret_val)
goto out;
@@ -1035,7 +1035,7 @@ static s32 igb_setup_copper_link_82575(struct e1000_hw *hw)
* depending on user settings.
*/
hw_dbg("Forcing Speed and Duplex\n");
- ret_val = igb_phy_force_speed_duplex(hw);
+ ret_val = hw->phy.ops.force_speed_duplex(hw);
if (ret_val) {
hw_dbg("Error Forcing Speed and Duplex\n");
goto out;
@@ -1423,9 +1423,9 @@ static struct e1000_mac_operations e1000_mac_ops_82575 = {
};
static struct e1000_phy_operations e1000_phy_ops_82575 = {
- .acquire_phy = igb_acquire_phy_82575,
+ .acquire = igb_acquire_phy_82575,
.get_cfg_done = igb_get_cfg_done_82575,
- .release_phy = igb_release_phy_82575,
+ .release = igb_release_phy_82575,
};
static struct e1000_nvm_operations e1000_nvm_ops_82575 = {
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index 06f72ae..5acb849 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -422,18 +422,18 @@ struct e1000_mac_operations {
};
struct e1000_phy_operations {
- s32 (*acquire_phy)(struct e1000_hw *);
+ s32 (*acquire)(struct e1000_hw *);
s32 (*check_reset_block)(struct e1000_hw *);
s32 (*force_speed_duplex)(struct e1000_hw *);
s32 (*get_cfg_done)(struct e1000_hw *hw);
s32 (*get_cable_length)(struct e1000_hw *);
s32 (*get_phy_info)(struct e1000_hw *);
- s32 (*read_phy_reg)(struct e1000_hw *, u32, u16 *);
- void (*release_phy)(struct e1000_hw *);
- s32 (*reset_phy)(struct e1000_hw *);
+ s32 (*read_reg)(struct e1000_hw *, u32, u16 *);
+ void (*release)(struct e1000_hw *);
+ s32 (*reset)(struct e1000_hw *);
s32 (*set_d0_lplu_state)(struct e1000_hw *, bool);
s32 (*set_d3_lplu_state)(struct e1000_hw *, bool);
- s32 (*write_phy_reg)(struct e1000_hw *, u32, u16);
+ s32 (*write_reg)(struct e1000_hw *, u32, u16);
};
struct e1000_nvm_operations {
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index 16fa083..d0b695c 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -707,11 +707,11 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
* has completed. We read this twice because this reg has
* some "sticky" (latched) bits.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS,
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS,
&mii_status_reg);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS,
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS,
&mii_status_reg);
if (ret_val)
goto out;
@@ -729,11 +729,11 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
* Page Ability Register (Address 5) to determine how
* flow control was negotiated.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_AUTONEG_ADV,
+ ret_val = hw->phy.ops.read_reg(hw, PHY_AUTONEG_ADV,
&mii_nway_adv_reg);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_LP_ABILITY,
+ ret_val = hw->phy.ops.read_reg(hw, PHY_LP_ABILITY,
&mii_nway_lp_ability_reg);
if (ret_val)
goto out;
diff --git a/drivers/net/igb/e1000_phy.c b/drivers/net/igb/e1000_phy.c
index 17fddb9..d73ea71 100644
--- a/drivers/net/igb/e1000_phy.c
+++ b/drivers/net/igb/e1000_phy.c
@@ -31,10 +31,6 @@
#include "e1000_mac.h"
#include "e1000_phy.h"
-static s32 igb_get_phy_cfg_done(struct e1000_hw *hw);
-static void igb_release_phy(struct e1000_hw *hw);
-static s32 igb_acquire_phy(struct e1000_hw *hw);
-static s32 igb_phy_reset_dsp(struct e1000_hw *hw);
static s32 igb_phy_setup_autoneg(struct e1000_hw *hw);
static void igb_phy_force_speed_duplex_setup(struct e1000_hw *hw,
u16 *phy_ctrl);
@@ -91,13 +87,13 @@ s32 igb_get_phy_id(struct e1000_hw *hw)
s32 ret_val = 0;
u16 phy_id;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_ID1, &phy_id);
+ ret_val = phy->ops.read_reg(hw, PHY_ID1, &phy_id);
if (ret_val)
goto out;
phy->id = (u32)(phy_id << 16);
udelay(20);
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_ID2, &phy_id);
+ ret_val = phy->ops.read_reg(hw, PHY_ID2, &phy_id);
if (ret_val)
goto out;
@@ -118,11 +114,11 @@ static s32 igb_phy_reset_dsp(struct e1000_hw *hw)
{
s32 ret_val;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_PHY_GEN_CONTROL, 0xC1);
+ ret_val = hw->phy.ops.write_reg(hw, M88E1000_PHY_GEN_CONTROL, 0xC1);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_PHY_GEN_CONTROL, 0);
+ ret_val = hw->phy.ops.write_reg(hw, M88E1000_PHY_GEN_CONTROL, 0);
out:
return ret_val;
@@ -257,9 +253,12 @@ out:
**/
s32 igb_read_phy_reg_igp(struct e1000_hw *hw, u32 offset, u16 *data)
{
- s32 ret_val;
+ s32 ret_val = 0;
+
+ if (!(hw->phy.ops.acquire))
+ goto out;
- ret_val = igb_acquire_phy(hw);
+ ret_val = hw->phy.ops.acquire(hw);
if (ret_val)
goto out;
@@ -268,16 +267,15 @@ s32 igb_read_phy_reg_igp(struct e1000_hw *hw, u32 offset, u16 *data)
IGP01E1000_PHY_PAGE_SELECT,
(u16)offset);
if (ret_val) {
- igb_release_phy(hw);
+ hw->phy.ops.release(hw);
goto out;
}
}
- ret_val = igb_read_phy_reg_mdic(hw,
- MAX_PHY_REG_ADDRESS & offset,
- data);
+ ret_val = igb_read_phy_reg_mdic(hw, MAX_PHY_REG_ADDRESS & offset,
+ data);
- igb_release_phy(hw);
+ hw->phy.ops.release(hw);
out:
return ret_val;
@@ -294,9 +292,12 @@ out:
**/
s32 igb_write_phy_reg_igp(struct e1000_hw *hw, u32 offset, u16 data)
{
- s32 ret_val;
+ s32 ret_val = 0;
- ret_val = igb_acquire_phy(hw);
+ if (!(hw->phy.ops.acquire))
+ goto out;
+
+ ret_val = hw->phy.ops.acquire(hw);
if (ret_val)
goto out;
@@ -305,16 +306,15 @@ s32 igb_write_phy_reg_igp(struct e1000_hw *hw, u32 offset, u16 data)
IGP01E1000_PHY_PAGE_SELECT,
(u16)offset);
if (ret_val) {
- igb_release_phy(hw);
+ hw->phy.ops.release(hw);
goto out;
}
}
- ret_val = igb_write_phy_reg_mdic(hw,
- MAX_PHY_REG_ADDRESS & offset,
+ ret_val = igb_write_phy_reg_mdic(hw, MAX_PHY_REG_ADDRESS & offset,
data);
- igb_release_phy(hw);
+ hw->phy.ops.release(hw);
out:
return ret_val;
@@ -339,8 +339,7 @@ s32 igb_copper_link_setup_m88(struct e1000_hw *hw)
}
/* Enable CRS on TX. This must be set for half-duplex operation. */
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_CTRL, &phy_data);
if (ret_val)
goto out;
@@ -383,8 +382,7 @@ s32 igb_copper_link_setup_m88(struct e1000_hw *hw)
if (phy->disable_polarity_correction == 1)
phy_data |= M88E1000_PSCR_POLARITY_REVERSAL;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- phy_data);
+ ret_val = phy->ops.write_reg(hw, M88E1000_PHY_SPEC_CTRL, phy_data);
if (ret_val)
goto out;
@@ -393,8 +391,7 @@ s32 igb_copper_link_setup_m88(struct e1000_hw *hw)
* Force TX_CLK in the Extended PHY Specific Control Register
* to 25MHz clock.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw,
- M88E1000_EXT_PHY_SPEC_CTRL,
+ ret_val = phy->ops.read_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL,
&phy_data);
if (ret_val)
goto out;
@@ -413,8 +410,7 @@ s32 igb_copper_link_setup_m88(struct e1000_hw *hw)
phy_data |= (M88E1000_EPSCR_MASTER_DOWNSHIFT_1X |
M88E1000_EPSCR_SLAVE_DOWNSHIFT_1X);
}
- ret_val = hw->phy.ops.write_phy_reg(hw,
- M88E1000_EXT_PHY_SPEC_CTRL,
+ ret_val = phy->ops.write_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL,
phy_data);
if (ret_val)
goto out;
@@ -449,7 +445,7 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
goto out;
}
- ret_val = hw->phy.ops.reset_phy(hw);
+ ret_val = phy->ops.reset(hw);
if (ret_val) {
hw_dbg("Error resetting the PHY.\n");
goto out;
@@ -464,8 +460,8 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
*/
if (phy->type == e1000_phy_igp) {
/* disable lplu d3 during driver init */
- if (hw->phy.ops.set_d3_lplu_state)
- ret_val = hw->phy.ops.set_d3_lplu_state(hw, false);
+ if (phy->ops.set_d3_lplu_state)
+ ret_val = phy->ops.set_d3_lplu_state(hw, false);
if (ret_val) {
hw_dbg("Error Disabling LPLU D3\n");
goto out;
@@ -473,13 +469,13 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
}
/* disable lplu d0 during driver init */
- ret_val = hw->phy.ops.set_d0_lplu_state(hw, false);
+ ret_val = phy->ops.set_d0_lplu_state(hw, false);
if (ret_val) {
hw_dbg("Error Disabling LPLU D0\n");
goto out;
}
/* Configure mdi-mdix settings */
- ret_val = hw->phy.ops.read_phy_reg(hw, IGP01E1000_PHY_PORT_CTRL, &data);
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_CTRL, &data);
if (ret_val)
goto out;
@@ -497,7 +493,7 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
data |= IGP01E1000_PSCR_AUTO_MDIX;
break;
}
- ret_val = hw->phy.ops.write_phy_reg(hw, IGP01E1000_PHY_PORT_CTRL, data);
+ ret_val = phy->ops.write_reg(hw, IGP01E1000_PHY_PORT_CTRL, data);
if (ret_val)
goto out;
@@ -510,33 +506,31 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
*/
if (phy->autoneg_advertised == ADVERTISE_1000_FULL) {
/* Disable SmartSpeed */
- ret_val = hw->phy.ops.read_phy_reg(hw,
- IGP01E1000_PHY_PORT_CONFIG,
- &data);
+ ret_val = phy->ops.read_reg(hw,
+ IGP01E1000_PHY_PORT_CONFIG,
+ &data);
if (ret_val)
goto out;
data &= ~IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = hw->phy.ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
IGP01E1000_PHY_PORT_CONFIG,
data);
if (ret_val)
goto out;
/* Set auto Master/Slave resolution process */
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_1000T_CTRL,
- &data);
+ ret_val = phy->ops.read_reg(hw, PHY_1000T_CTRL, &data);
if (ret_val)
goto out;
data &= ~CR_1000T_MS_ENABLE;
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_1000T_CTRL,
- data);
+ ret_val = phy->ops.write_reg(hw, PHY_1000T_CTRL, data);
if (ret_val)
goto out;
}
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_1000T_CTRL, &data);
+ ret_val = phy->ops.read_reg(hw, PHY_1000T_CTRL, &data);
if (ret_val)
goto out;
@@ -560,7 +554,7 @@ s32 igb_copper_link_setup_igp(struct e1000_hw *hw)
default:
break;
}
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_1000T_CTRL, data);
+ ret_val = phy->ops.write_reg(hw, PHY_1000T_CTRL, data);
if (ret_val)
goto out;
}
@@ -609,12 +603,12 @@ s32 igb_copper_link_autoneg(struct e1000_hw *hw)
* Restart auto-negotiation by setting the Auto Neg Enable bit and
* the Auto Neg Restart bit in the PHY control register.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_CONTROL, &phy_ctrl);
+ ret_val = phy->ops.read_reg(hw, PHY_CONTROL, &phy_ctrl);
if (ret_val)
goto out;
phy_ctrl |= (MII_CR_AUTO_NEG_EN | MII_CR_RESTART_AUTO_NEG);
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_CONTROL, phy_ctrl);
+ ret_val = phy->ops.write_reg(hw, PHY_CONTROL, phy_ctrl);
if (ret_val)
goto out;
@@ -656,15 +650,13 @@ static s32 igb_phy_setup_autoneg(struct e1000_hw *hw)
phy->autoneg_advertised &= phy->autoneg_mask;
/* Read the MII Auto-Neg Advertisement Register (Address 4). */
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_AUTONEG_ADV,
- &mii_autoneg_adv_reg);
+ ret_val = phy->ops.read_reg(hw, PHY_AUTONEG_ADV, &mii_autoneg_adv_reg);
if (ret_val)
goto out;
if (phy->autoneg_mask & ADVERTISE_1000_FULL) {
/* Read the MII 1000Base-T Control Register (Address 9). */
- ret_val = hw->phy.ops.read_phy_reg(hw,
- PHY_1000T_CTRL,
+ ret_val = phy->ops.read_reg(hw, PHY_1000T_CTRL,
&mii_1000t_ctrl_reg);
if (ret_val)
goto out;
@@ -785,17 +777,16 @@ static s32 igb_phy_setup_autoneg(struct e1000_hw *hw)
goto out;
}
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_AUTONEG_ADV,
- mii_autoneg_adv_reg);
+ ret_val = phy->ops.write_reg(hw, PHY_AUTONEG_ADV, mii_autoneg_adv_reg);
if (ret_val)
goto out;
hw_dbg("Auto-Neg Advertising %x\n", mii_autoneg_adv_reg);
if (phy->autoneg_mask & ADVERTISE_1000_FULL) {
- ret_val = hw->phy.ops.write_phy_reg(hw,
- PHY_1000T_CTRL,
- mii_1000t_ctrl_reg);
+ ret_val = phy->ops.write_reg(hw,
+ PHY_1000T_CTRL,
+ mii_1000t_ctrl_reg);
if (ret_val)
goto out;
}
@@ -819,13 +810,13 @@ s32 igb_phy_force_speed_duplex_igp(struct e1000_hw *hw)
u16 phy_data;
bool link;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_CONTROL, &phy_data);
+ ret_val = phy->ops.read_reg(hw, PHY_CONTROL, &phy_data);
if (ret_val)
goto out;
igb_phy_force_speed_duplex_setup(hw, &phy_data);
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_CONTROL, phy_data);
+ ret_val = phy->ops.write_reg(hw, PHY_CONTROL, phy_data);
if (ret_val)
goto out;
@@ -833,16 +824,14 @@ s32 igb_phy_force_speed_duplex_igp(struct e1000_hw *hw)
* Clear Auto-Crossover to force MDI manually. IGP requires MDI
* forced whenever speed and duplex are forced.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, IGP01E1000_PHY_PORT_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_CTRL, &phy_data);
if (ret_val)
goto out;
phy_data &= ~IGP01E1000_PSCR_AUTO_MDIX;
phy_data &= ~IGP01E1000_PSCR_FORCE_MDI_MDIX;
- ret_val = hw->phy.ops.write_phy_reg(hw, IGP01E1000_PHY_PORT_CTRL,
- phy_data);
+ ret_val = phy->ops.write_reg(hw, IGP01E1000_PHY_PORT_CTRL, phy_data);
if (ret_val)
goto out;
@@ -897,20 +886,18 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
* Clear Auto-Crossover to force MDI manually. M88E1000 requires MDI
* forced whenever speed and duplex are forced.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_CTRL, &phy_data);
if (ret_val)
goto out;
phy_data &= ~M88E1000_PSCR_AUTO_X_MODE;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- phy_data);
+ ret_val = phy->ops.write_reg(hw, M88E1000_PHY_SPEC_CTRL, phy_data);
if (ret_val)
goto out;
hw_dbg("M88E1000 PSCR: %X\n", phy_data);
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_CONTROL, &phy_data);
+ ret_val = phy->ops.read_reg(hw, PHY_CONTROL, &phy_data);
if (ret_val)
goto out;
@@ -919,7 +906,7 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
/* Reset the phy to commit changes. */
phy_data |= MII_CR_RESET;
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_CONTROL, phy_data);
+ ret_val = phy->ops.write_reg(hw, PHY_CONTROL, phy_data);
if (ret_val)
goto out;
@@ -940,7 +927,7 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
* We didn't get link.
* Reset the DSP and cross our fingers.
*/
- ret_val = hw->phy.ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
M88E1000_PHY_PAGE_SELECT,
0x001d);
if (ret_val)
@@ -957,8 +944,7 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
goto out;
}
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL, &phy_data);
if (ret_val)
goto out;
@@ -968,8 +954,7 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
* the reset value of 2.5MHz.
*/
phy_data |= M88E1000_EPSCR_TX_CLK_25;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL,
- phy_data);
+ ret_val = phy->ops.write_reg(hw, M88E1000_EXT_PHY_SPEC_CTRL, phy_data);
if (ret_val)
goto out;
@@ -977,14 +962,12 @@ s32 igb_phy_force_speed_duplex_m88(struct e1000_hw *hw)
* In addition, we must re-enable CRS on Tx for both half and full
* duplex.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_CTRL, &phy_data);
if (ret_val)
goto out;
phy_data |= M88E1000_PSCR_ASSERT_CRS_ON_TX;
- ret_val = hw->phy.ops.write_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- phy_data);
+ ret_val = phy->ops.write_reg(hw, M88E1000_PHY_SPEC_CTRL, phy_data);
out:
return ret_val;
@@ -1071,15 +1054,13 @@ s32 igb_set_d3_lplu_state(struct e1000_hw *hw, bool active)
s32 ret_val;
u16 data;
- ret_val = hw->phy.ops.read_phy_reg(hw, IGP02E1000_PHY_POWER_MGMT,
- &data);
+ ret_val = phy->ops.read_reg(hw, IGP02E1000_PHY_POWER_MGMT, &data);
if (ret_val)
goto out;
if (!active) {
data &= ~IGP02E1000_PM_D3_LPLU;
- ret_val = hw->phy.ops.write_phy_reg(hw,
- IGP02E1000_PHY_POWER_MGMT,
+ ret_val = phy->ops.write_reg(hw, IGP02E1000_PHY_POWER_MGMT,
data);
if (ret_val)
goto out;
@@ -1090,27 +1071,27 @@ s32 igb_set_d3_lplu_state(struct e1000_hw *hw, bool active)
* SmartSpeed, so performance is maintained.
*/
if (phy->smart_speed == e1000_smart_speed_on) {
- ret_val = hw->phy.ops.read_phy_reg(hw,
+ ret_val = phy->ops.read_reg(hw,
IGP01E1000_PHY_PORT_CONFIG,
&data);
if (ret_val)
goto out;
data |= IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = hw->phy.ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
IGP01E1000_PHY_PORT_CONFIG,
data);
if (ret_val)
goto out;
} else if (phy->smart_speed == e1000_smart_speed_off) {
- ret_val = hw->phy.ops.read_phy_reg(hw,
+ ret_val = phy->ops.read_reg(hw,
IGP01E1000_PHY_PORT_CONFIG,
&data);
if (ret_val)
goto out;
data &= ~IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = hw->phy.ops.write_phy_reg(hw,
+ ret_val = phy->ops.write_reg(hw,
IGP01E1000_PHY_PORT_CONFIG,
data);
if (ret_val)
@@ -1120,22 +1101,19 @@ s32 igb_set_d3_lplu_state(struct e1000_hw *hw, bool active)
(phy->autoneg_advertised == E1000_ALL_NOT_GIG) ||
(phy->autoneg_advertised == E1000_ALL_10_SPEED)) {
data |= IGP02E1000_PM_D3_LPLU;
- ret_val = hw->phy.ops.write_phy_reg(hw,
- IGP02E1000_PHY_POWER_MGMT,
+ ret_val = phy->ops.write_reg(hw, IGP02E1000_PHY_POWER_MGMT,
data);
if (ret_val)
goto out;
/* When LPLU is enabled, we should disable SmartSpeed */
- ret_val = hw->phy.ops.read_phy_reg(hw,
- IGP01E1000_PHY_PORT_CONFIG,
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
&data);
if (ret_val)
goto out;
data &= ~IGP01E1000_PSCFR_SMART_SPEED;
- ret_val = hw->phy.ops.write_phy_reg(hw,
- IGP01E1000_PHY_PORT_CONFIG,
+ ret_val = phy->ops.write_reg(hw, IGP01E1000_PHY_PORT_CONFIG,
data);
}
@@ -1176,7 +1154,7 @@ s32 igb_check_downshift(struct e1000_hw *hw)
goto out;
}
- ret_val = hw->phy.ops.read_phy_reg(hw, offset, &phy_data);
+ ret_val = phy->ops.read_reg(hw, offset, &phy_data);
if (!ret_val)
phy->speed_downgraded = (phy_data & mask) ? true : false;
@@ -1199,7 +1177,7 @@ static s32 igb_check_polarity_m88(struct e1000_hw *hw)
s32 ret_val;
u16 data;
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_STATUS, &data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_STATUS, &data);
if (!ret_val)
phy->cable_polarity = (data & M88E1000_PSSR_REV_POLARITY)
@@ -1228,8 +1206,7 @@ static s32 igb_check_polarity_igp(struct e1000_hw *hw)
* Polarity is determined based on the speed of
* our connection.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, IGP01E1000_PHY_PORT_STATUS,
- &data);
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_STATUS, &data);
if (ret_val)
goto out;
@@ -1246,7 +1223,7 @@ static s32 igb_check_polarity_igp(struct e1000_hw *hw)
mask = IGP01E1000_PSSR_POLARITY_REVERSED;
}
- ret_val = hw->phy.ops.read_phy_reg(hw, offset, &data);
+ ret_val = phy->ops.read_reg(hw, offset, &data);
if (!ret_val)
phy->cable_polarity = (data & mask)
@@ -1271,10 +1248,10 @@ static s32 igb_wait_autoneg(struct e1000_hw *hw)
/* Break after autoneg completes or PHY_AUTO_NEG_LIMIT expires. */
for (i = PHY_AUTO_NEG_LIMIT; i > 0; i--) {
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS, &phy_status);
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS, &phy_status);
if (ret_val)
break;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS, &phy_status);
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS, &phy_status);
if (ret_val)
break;
if (phy_status & MII_SR_AUTONEG_COMPLETE)
@@ -1310,10 +1287,10 @@ s32 igb_phy_has_link(struct e1000_hw *hw, u32 iterations,
* twice due to the link bit being sticky. No harm doing
* it across the board.
*/
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS, &phy_status);
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS, &phy_status);
if (ret_val)
break;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_STATUS, &phy_status);
+ ret_val = hw->phy.ops.read_reg(hw, PHY_STATUS, &phy_status);
if (ret_val)
break;
if (phy_status & MII_SR_LINK_STATUS)
@@ -1350,8 +1327,7 @@ s32 igb_get_cable_length_m88(struct e1000_hw *hw)
s32 ret_val;
u16 phy_data, index;
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_STATUS,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_STATUS, &phy_data);
if (ret_val)
goto out;
@@ -1372,8 +1348,8 @@ out:
*
* The automatic gain control (agc) normalizes the amplitude of the
* received signal, adjusting for the attenuation produced by the
- * cable. By reading the AGC registers, which reperesent the
- * cobination of course and fine gain value, the value can be put
+ * cable. By reading the AGC registers, which represent the
+ * combination of coarse and fine gain value, the value can be put
* into a lookup table to obtain the approximate cable length
* for each channel.
**/
@@ -1392,14 +1368,13 @@ s32 igb_get_cable_length_igp_2(struct e1000_hw *hw)
/* Read the AGC registers for all channels */
for (i = 0; i < IGP02E1000_PHY_CHANNEL_NUM; i++) {
- ret_val = hw->phy.ops.read_phy_reg(hw, agc_reg_array[i],
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, agc_reg_array[i], &phy_data);
if (ret_val)
goto out;
/*
* Getting bits 15:9, which represent the combination of
- * course and fine gain values. The result is a number
+ * coarse and fine gain values. The result is a number
* that can be put into the lookup table to obtain the
* approximate cable length.
*/
@@ -1456,7 +1431,7 @@ s32 igb_get_phy_info_m88(struct e1000_hw *hw)
u16 phy_data;
bool link;
- if (hw->phy.media_type != e1000_media_type_copper) {
+ if (phy->media_type != e1000_media_type_copper) {
hw_dbg("Phy info is only valid for copper media\n");
ret_val = -E1000_ERR_CONFIG;
goto out;
@@ -1472,33 +1447,29 @@ s32 igb_get_phy_info_m88(struct e1000_hw *hw)
goto out;
}
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_CTRL,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_CTRL, &phy_data);
if (ret_val)
goto out;
phy->polarity_correction = (phy_data & M88E1000_PSCR_POLARITY_REVERSAL)
- ? true
- : false;
+ ? true : false;
ret_val = igb_check_polarity_m88(hw);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, M88E1000_PHY_SPEC_STATUS,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, M88E1000_PHY_SPEC_STATUS, &phy_data);
if (ret_val)
goto out;
phy->is_mdix = (phy_data & M88E1000_PSSR_MDIX) ? true : false;
if ((phy_data & M88E1000_PSSR_SPEED) == M88E1000_PSSR_1000MBS) {
- ret_val = hw->phy.ops.get_cable_length(hw);
+ ret_val = phy->ops.get_cable_length(hw);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_1000T_STATUS,
- &phy_data);
+ ret_val = phy->ops.read_reg(hw, PHY_1000T_STATUS, &phy_data);
if (ret_val)
goto out;
@@ -1552,8 +1523,7 @@ s32 igb_get_phy_info_igp(struct e1000_hw *hw)
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, IGP01E1000_PHY_PORT_STATUS,
- &data);
+ ret_val = phy->ops.read_reg(hw, IGP01E1000_PHY_PORT_STATUS, &data);
if (ret_val)
goto out;
@@ -1561,12 +1531,11 @@ s32 igb_get_phy_info_igp(struct e1000_hw *hw)
if ((data & IGP01E1000_PSSR_SPEED_MASK) ==
IGP01E1000_PSSR_SPEED_1000MBPS) {
- ret_val = hw->phy.ops.get_cable_length(hw);
+ ret_val = phy->ops.get_cable_length(hw);
if (ret_val)
goto out;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_1000T_STATUS,
- &data);
+ ret_val = phy->ops.read_reg(hw, PHY_1000T_STATUS, &data);
if (ret_val)
goto out;
@@ -1599,12 +1568,12 @@ s32 igb_phy_sw_reset(struct e1000_hw *hw)
s32 ret_val;
u16 phy_ctrl;
- ret_val = hw->phy.ops.read_phy_reg(hw, PHY_CONTROL, &phy_ctrl);
+ ret_val = hw->phy.ops.read_reg(hw, PHY_CONTROL, &phy_ctrl);
if (ret_val)
goto out;
phy_ctrl |= MII_CR_RESET;
- ret_val = hw->phy.ops.write_phy_reg(hw, PHY_CONTROL, phy_ctrl);
+ ret_val = hw->phy.ops.write_reg(hw, PHY_CONTROL, phy_ctrl);
if (ret_val)
goto out;
@@ -1635,7 +1604,7 @@ s32 igb_phy_hw_reset(struct e1000_hw *hw)
goto out;
}
- ret_val = igb_acquire_phy(hw);
+ ret_val = phy->ops.acquire(hw);
if (ret_val)
goto out;
@@ -1650,74 +1619,14 @@ s32 igb_phy_hw_reset(struct e1000_hw *hw)
udelay(150);
- igb_release_phy(hw);
+ phy->ops.release(hw);
- ret_val = igb_get_phy_cfg_done(hw);
+ ret_val = phy->ops.get_cfg_done(hw);
out:
return ret_val;
}
-/* Internal function pointers */
-
-/**
- * igb_get_phy_cfg_done - Generic PHY configuration done
- * @hw: pointer to the HW structure
- *
- * Return success if silicon family did not implement a family specific
- * get_cfg_done function.
- **/
-static s32 igb_get_phy_cfg_done(struct e1000_hw *hw)
-{
- if (hw->phy.ops.get_cfg_done)
- return hw->phy.ops.get_cfg_done(hw);
-
- return 0;
-}
-
-/**
- * igb_release_phy - Generic release PHY
- * @hw: pointer to the HW structure
- *
- * Return if silicon family does not require a semaphore when accessing the
- * PHY.
- **/
-static void igb_release_phy(struct e1000_hw *hw)
-{
- if (hw->phy.ops.release_phy)
- hw->phy.ops.release_phy(hw);
-}
-
-/**
- * igb_acquire_phy - Generic acquire PHY
- * @hw: pointer to the HW structure
- *
- * Return success if silicon family does not require a semaphore when
- * accessing the PHY.
- **/
-static s32 igb_acquire_phy(struct e1000_hw *hw)
-{
- if (hw->phy.ops.acquire_phy)
- return hw->phy.ops.acquire_phy(hw);
-
- return 0;
-}
-
-/**
- * igb_phy_force_speed_duplex - Generic force PHY speed/duplex
- * @hw: pointer to the HW structure
- *
- * When the silicon family has not implemented a forced speed/duplex
- * function for the PHY, simply return 0.
- **/
-s32 igb_phy_force_speed_duplex(struct e1000_hw *hw)
-{
- if (hw->phy.ops.force_speed_duplex)
- return hw->phy.ops.force_speed_duplex(hw);
-
- return 0;
-}
-
/**
* igb_phy_init_script_igp3 - Inits the IGP3 PHY
* @hw: pointer to the HW structure
@@ -1730,75 +1639,75 @@ s32 igb_phy_init_script_igp3(struct e1000_hw *hw)
/* PHY init IGP 3 */
/* Enable rise/fall, 10-mode work in class-A */
- hw->phy.ops.write_phy_reg(hw, 0x2F5B, 0x9018);
+ hw->phy.ops.write_reg(hw, 0x2F5B, 0x9018);
/* Remove all caps from Replica path filter */
- hw->phy.ops.write_phy_reg(hw, 0x2F52, 0x0000);
+ hw->phy.ops.write_reg(hw, 0x2F52, 0x0000);
/* Bias trimming for ADC, AFE and Driver (Default) */
- hw->phy.ops.write_phy_reg(hw, 0x2FB1, 0x8B24);
+ hw->phy.ops.write_reg(hw, 0x2FB1, 0x8B24);
/* Increase Hybrid poly bias */
- hw->phy.ops.write_phy_reg(hw, 0x2FB2, 0xF8F0);
+ hw->phy.ops.write_reg(hw, 0x2FB2, 0xF8F0);
/* Add 4% to TX amplitude in Giga mode */
- hw->phy.ops.write_phy_reg(hw, 0x2010, 0x10B0);
+ hw->phy.ops.write_reg(hw, 0x2010, 0x10B0);
/* Disable trimming (TTT) */
- hw->phy.ops.write_phy_reg(hw, 0x2011, 0x0000);
+ hw->phy.ops.write_reg(hw, 0x2011, 0x0000);
/* Poly DC correction to 94.6% + 2% for all channels */
- hw->phy.ops.write_phy_reg(hw, 0x20DD, 0x249A);
+ hw->phy.ops.write_reg(hw, 0x20DD, 0x249A);
/* ABS DC correction to 95.9% */
- hw->phy.ops.write_phy_reg(hw, 0x20DE, 0x00D3);
+ hw->phy.ops.write_reg(hw, 0x20DE, 0x00D3);
/* BG temp curve trim */
- hw->phy.ops.write_phy_reg(hw, 0x28B4, 0x04CE);
+ hw->phy.ops.write_reg(hw, 0x28B4, 0x04CE);
/* Increasing ADC OPAMP stage 1 currents to max */
- hw->phy.ops.write_phy_reg(hw, 0x2F70, 0x29E4);
+ hw->phy.ops.write_reg(hw, 0x2F70, 0x29E4);
/* Force 1000 ( required for enabling PHY regs configuration) */
- hw->phy.ops.write_phy_reg(hw, 0x0000, 0x0140);
+ hw->phy.ops.write_reg(hw, 0x0000, 0x0140);
/* Set upd_freq to 6 */
- hw->phy.ops.write_phy_reg(hw, 0x1F30, 0x1606);
+ hw->phy.ops.write_reg(hw, 0x1F30, 0x1606);
/* Disable NPDFE */
- hw->phy.ops.write_phy_reg(hw, 0x1F31, 0xB814);
+ hw->phy.ops.write_reg(hw, 0x1F31, 0xB814);
/* Disable adaptive fixed FFE (Default) */
- hw->phy.ops.write_phy_reg(hw, 0x1F35, 0x002A);
+ hw->phy.ops.write_reg(hw, 0x1F35, 0x002A);
/* Enable FFE hysteresis */
- hw->phy.ops.write_phy_reg(hw, 0x1F3E, 0x0067);
+ hw->phy.ops.write_reg(hw, 0x1F3E, 0x0067);
/* Fixed FFE for short cable lengths */
- hw->phy.ops.write_phy_reg(hw, 0x1F54, 0x0065);
+ hw->phy.ops.write_reg(hw, 0x1F54, 0x0065);
/* Fixed FFE for medium cable lengths */
- hw->phy.ops.write_phy_reg(hw, 0x1F55, 0x002A);
+ hw->phy.ops.write_reg(hw, 0x1F55, 0x002A);
/* Fixed FFE for long cable lengths */
- hw->phy.ops.write_phy_reg(hw, 0x1F56, 0x002A);
+ hw->phy.ops.write_reg(hw, 0x1F56, 0x002A);
/* Enable Adaptive Clip Threshold */
- hw->phy.ops.write_phy_reg(hw, 0x1F72, 0x3FB0);
+ hw->phy.ops.write_reg(hw, 0x1F72, 0x3FB0);
/* AHT reset limit to 1 */
- hw->phy.ops.write_phy_reg(hw, 0x1F76, 0xC0FF);
+ hw->phy.ops.write_reg(hw, 0x1F76, 0xC0FF);
/* Set AHT master delay to 127 msec */
- hw->phy.ops.write_phy_reg(hw, 0x1F77, 0x1DEC);
+ hw->phy.ops.write_reg(hw, 0x1F77, 0x1DEC);
/* Set scan bits for AHT */
- hw->phy.ops.write_phy_reg(hw, 0x1F78, 0xF9EF);
+ hw->phy.ops.write_reg(hw, 0x1F78, 0xF9EF);
/* Set AHT Preset bits */
- hw->phy.ops.write_phy_reg(hw, 0x1F79, 0x0210);
+ hw->phy.ops.write_reg(hw, 0x1F79, 0x0210);
/* Change integ_factor of channel A to 3 */
- hw->phy.ops.write_phy_reg(hw, 0x1895, 0x0003);
+ hw->phy.ops.write_reg(hw, 0x1895, 0x0003);
/* Change prop_factor of channels BCD to 8 */
- hw->phy.ops.write_phy_reg(hw, 0x1796, 0x0008);
+ hw->phy.ops.write_reg(hw, 0x1796, 0x0008);
/* Change cg_icount + enable integbp for channels BCD */
- hw->phy.ops.write_phy_reg(hw, 0x1798, 0xD008);
+ hw->phy.ops.write_reg(hw, 0x1798, 0xD008);
/*
* Change cg_icount + enable integbp + change prop_factor_master
* to 8 for channel A
*/
- hw->phy.ops.write_phy_reg(hw, 0x1898, 0xD918);
+ hw->phy.ops.write_reg(hw, 0x1898, 0xD918);
/* Disable AHT in Slave mode on channel A */
- hw->phy.ops.write_phy_reg(hw, 0x187A, 0x0800);
+ hw->phy.ops.write_reg(hw, 0x187A, 0x0800);
/*
* Enable LPLU and disable AN to 1000 in non-D0a states,
* Enable SPD+B2B
*/
- hw->phy.ops.write_phy_reg(hw, 0x0019, 0x008D);
+ hw->phy.ops.write_reg(hw, 0x0019, 0x008D);
/* Enable restart AN on an1000_dis change */
- hw->phy.ops.write_phy_reg(hw, 0x001B, 0x2080);
+ hw->phy.ops.write_reg(hw, 0x001B, 0x2080);
/* Enable wh_fifo read clock in 10/100 modes */
- hw->phy.ops.write_phy_reg(hw, 0x0014, 0x0045);
+ hw->phy.ops.write_reg(hw, 0x0014, 0x0045);
/* Restart AN, Speed selection is 1000 */
- hw->phy.ops.write_phy_reg(hw, 0x0000, 0x1340);
+ hw->phy.ops.write_reg(hw, 0x0000, 0x1340);
return 0;
}
diff --git a/drivers/net/igb/e1000_phy.h b/drivers/net/igb/e1000_phy.h
index 8f8fe0a..7499e5d 100644
--- a/drivers/net/igb/e1000_phy.h
+++ b/drivers/net/igb/e1000_phy.h
@@ -44,7 +44,6 @@ enum e1000_smart_speed {
s32 igb_check_downshift(struct e1000_hw *hw);
s32 igb_check_reset_block(struct e1000_hw *hw);
s32 igb_copper_link_autoneg(struct e1000_hw *hw);
-s32 igb_phy_force_speed_duplex(struct e1000_hw *hw);
s32 igb_copper_link_setup_igp(struct e1000_hw *hw);
s32 igb_copper_link_setup_m88(struct e1000_hw *hw);
s32 igb_phy_force_speed_duplex_igp(struct e1000_hw *hw);
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 530d7aa..a2a812d 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -313,24 +313,24 @@ extern void igb_set_ethtool_ops(struct net_device *);
static inline s32 igb_reset_phy(struct e1000_hw *hw)
{
- if (hw->phy.ops.reset_phy)
- return hw->phy.ops.reset_phy(hw);
+ if (hw->phy.ops.reset)
+ return hw->phy.ops.reset(hw);
return 0;
}
static inline s32 igb_read_phy_reg(struct e1000_hw *hw, u32 offset, u16 *data)
{
- if (hw->phy.ops.read_phy_reg)
- return hw->phy.ops.read_phy_reg(hw, offset, data);
+ if (hw->phy.ops.read_reg)
+ return hw->phy.ops.read_reg(hw, offset, data);
return 0;
}
static inline s32 igb_write_phy_reg(struct e1000_hw *hw, u32 offset, u16 data)
{
- if (hw->phy.ops.write_phy_reg)
- return hw->phy.ops.write_phy_reg(hw, offset, data);
+ if (hw->phy.ops.write_reg)
+ return hw->phy.ops.write_reg(hw, offset, data);
return 0;
}
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 09/26] igb: rename nvm ops
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (6 preceding siblings ...)
2009-02-07 9:17 ` [net-next PATCH 08/26] igb: rename phy ops Jeff Kirsher
@ 2009-02-07 9:17 ` Jeff Kirsher
2009-02-07 9:18 ` [net-next PATCH 10/26] igb: remove unused rx_hdr_split statistic Jeff Kirsher
` (17 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:17 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
All of the nvm ops have the tag _nvm added to the end which is redundant
since all of the calls to the ops have to go through the nvm ops struct
anyway.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_82575.c | 8 ++++----
drivers/net/igb/e1000_hw.h | 8 ++++----
drivers/net/igb/e1000_mac.c | 9 ++++-----
drivers/net/igb/e1000_nvm.c | 14 +++++++-------
drivers/net/igb/igb_ethtool.c | 14 +++++++-------
drivers/net/igb/igb_main.c | 3 +--
6 files changed, 27 insertions(+), 29 deletions(-)
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index 9a66e34..527d4c8 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -1429,10 +1429,10 @@ static struct e1000_phy_operations e1000_phy_ops_82575 = {
};
static struct e1000_nvm_operations e1000_nvm_ops_82575 = {
- .acquire_nvm = igb_acquire_nvm_82575,
- .read_nvm = igb_read_nvm_eerd,
- .release_nvm = igb_release_nvm_82575,
- .write_nvm = igb_write_nvm_spi,
+ .acquire = igb_acquire_nvm_82575,
+ .read = igb_read_nvm_eerd,
+ .release = igb_release_nvm_82575,
+ .write = igb_write_nvm_spi,
};
const struct e1000_info e1000_82575_info = {
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index 5acb849..acb42a2 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -437,10 +437,10 @@ struct e1000_phy_operations {
};
struct e1000_nvm_operations {
- s32 (*acquire_nvm)(struct e1000_hw *);
- s32 (*read_nvm)(struct e1000_hw *, u16, u16, u16 *);
- void (*release_nvm)(struct e1000_hw *);
- s32 (*write_nvm)(struct e1000_hw *, u16, u16, u16 *);
+ s32 (*acquire)(struct e1000_hw *);
+ s32 (*read)(struct e1000_hw *, u16, u16, u16 *);
+ void (*release)(struct e1000_hw *);
+ s32 (*write)(struct e1000_hw *, u16, u16, u16 *);
};
struct e1000_info {
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index d0b695c..6682206 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -135,7 +135,7 @@ s32 igb_check_alt_mac_addr(struct e1000_hw *hw)
u16 offset, nvm_alt_mac_addr_offset, nvm_data;
u8 alt_mac_addr[ETH_ALEN];
- ret_val = hw->nvm.ops.read_nvm(hw, NVM_ALT_MAC_ADDR_PTR, 1,
+ ret_val = hw->nvm.ops.read(hw, NVM_ALT_MAC_ADDR_PTR, 1,
&nvm_alt_mac_addr_offset);
if (ret_val) {
hw_dbg("NVM Read Error\n");
@@ -152,7 +152,7 @@ s32 igb_check_alt_mac_addr(struct e1000_hw *hw)
for (i = 0; i < ETH_ALEN; i += 2) {
offset = nvm_alt_mac_addr_offset + (i >> 1);
- ret_val = hw->nvm.ops.read_nvm(hw, offset, 1, &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, offset, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
goto out;
@@ -575,8 +575,7 @@ static s32 igb_set_default_fc(struct e1000_hw *hw)
* control setting, then the variable hw->fc will
* be initialized based on a value in the EEPROM.
*/
- ret_val = hw->nvm.ops.read_nvm(hw, NVM_INIT_CONTROL2_REG, 1,
- &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, NVM_INIT_CONTROL2_REG, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
@@ -1028,7 +1027,7 @@ static s32 igb_valid_led_default(struct e1000_hw *hw, u16 *data)
{
s32 ret_val;
- ret_val = hw->nvm.ops.read_nvm(hw, NVM_ID_LED_SETTINGS, 1, data);
+ ret_val = hw->nvm.ops.read(hw, NVM_ID_LED_SETTINGS, 1, data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
goto out;
diff --git a/drivers/net/igb/e1000_nvm.c b/drivers/net/igb/e1000_nvm.c
index 5942da1..3379864 100644
--- a/drivers/net/igb/e1000_nvm.c
+++ b/drivers/net/igb/e1000_nvm.c
@@ -419,7 +419,7 @@ s32 igb_write_nvm_spi(struct e1000_hw *hw, u16 offset, u16 words, u16 *data)
goto out;
}
- ret_val = hw->nvm.ops.acquire_nvm(hw);
+ ret_val = hw->nvm.ops.acquire(hw);
if (ret_val)
goto out;
@@ -468,7 +468,7 @@ s32 igb_write_nvm_spi(struct e1000_hw *hw, u16 offset, u16 words, u16 *data)
msleep(10);
release:
- hw->nvm.ops.release_nvm(hw);
+ hw->nvm.ops.release(hw);
out:
return ret_val;
@@ -487,14 +487,14 @@ s32 igb_read_part_num(struct e1000_hw *hw, u32 *part_num)
s32 ret_val;
u16 nvm_data;
- ret_val = hw->nvm.ops.read_nvm(hw, NVM_PBA_OFFSET_0, 1, &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, NVM_PBA_OFFSET_0, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
goto out;
}
*part_num = (u32)(nvm_data << 16);
- ret_val = hw->nvm.ops.read_nvm(hw, NVM_PBA_OFFSET_1, 1, &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, NVM_PBA_OFFSET_1, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
goto out;
@@ -548,7 +548,7 @@ s32 igb_validate_nvm_checksum(struct e1000_hw *hw)
u16 i, nvm_data;
for (i = 0; i < (NVM_CHECKSUM_REG + 1); i++) {
- ret_val = hw->nvm.ops.read_nvm(hw, i, 1, &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, i, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error\n");
goto out;
@@ -581,7 +581,7 @@ s32 igb_update_nvm_checksum(struct e1000_hw *hw)
u16 i, nvm_data;
for (i = 0; i < NVM_CHECKSUM_REG; i++) {
- ret_val = hw->nvm.ops.read_nvm(hw, i, 1, &nvm_data);
+ ret_val = hw->nvm.ops.read(hw, i, 1, &nvm_data);
if (ret_val) {
hw_dbg("NVM Read Error while updating checksum.\n");
goto out;
@@ -589,7 +589,7 @@ s32 igb_update_nvm_checksum(struct e1000_hw *hw)
checksum += nvm_data;
}
checksum = (u16) NVM_SUM - checksum;
- ret_val = hw->nvm.ops.write_nvm(hw, NVM_CHECKSUM_REG, 1, &checksum);
+ ret_val = hw->nvm.ops.write(hw, NVM_CHECKSUM_REG, 1, &checksum);
if (ret_val)
hw_dbg("NVM Write Error while updating checksum.\n");
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index 4606e63..a5bf8ef 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -593,12 +593,12 @@ static int igb_get_eeprom(struct net_device *netdev,
return -ENOMEM;
if (hw->nvm.type == e1000_nvm_eeprom_spi)
- ret_val = hw->nvm.ops.read_nvm(hw, first_word,
+ ret_val = hw->nvm.ops.read(hw, first_word,
last_word - first_word + 1,
eeprom_buff);
else {
for (i = 0; i < last_word - first_word + 1; i++) {
- ret_val = hw->nvm.ops.read_nvm(hw, first_word + i, 1,
+ ret_val = hw->nvm.ops.read(hw, first_word + i, 1,
&eeprom_buff[i]);
if (ret_val)
break;
@@ -645,14 +645,14 @@ static int igb_set_eeprom(struct net_device *netdev,
if (eeprom->offset & 1) {
/* need read/modify/write of first changed EEPROM word */
/* only the second byte of the word is being modified */
- ret_val = hw->nvm.ops.read_nvm(hw, first_word, 1,
+ ret_val = hw->nvm.ops.read(hw, first_word, 1,
&eeprom_buff[0]);
ptr++;
}
if (((eeprom->offset + eeprom->len) & 1) && (ret_val == 0)) {
/* need read/modify/write of last changed EEPROM word */
/* only the first byte of the word is being modified */
- ret_val = hw->nvm.ops.read_nvm(hw, last_word, 1,
+ ret_val = hw->nvm.ops.read(hw, last_word, 1,
&eeprom_buff[last_word - first_word]);
}
@@ -665,7 +665,7 @@ static int igb_set_eeprom(struct net_device *netdev,
for (i = 0; i < last_word - first_word + 1; i++)
eeprom_buff[i] = cpu_to_le16(eeprom_buff[i]);
- ret_val = hw->nvm.ops.write_nvm(hw, first_word,
+ ret_val = hw->nvm.ops.write(hw, first_word,
last_word - first_word + 1, eeprom_buff);
/* Update the checksum over the first part of the EEPROM if needed
@@ -689,7 +689,7 @@ static void igb_get_drvinfo(struct net_device *netdev,
/* EEPROM image version # is reported as firmware version # for
* 82575 controllers */
- adapter->hw.nvm.ops.read_nvm(&adapter->hw, 5, 1, &eeprom_data);
+ adapter->hw.nvm.ops.read(&adapter->hw, 5, 1, &eeprom_data);
sprintf(firmware_version, "%d.%d-%d",
(eeprom_data & 0xF000) >> 12,
(eeprom_data & 0x0FF0) >> 4,
@@ -1056,7 +1056,7 @@ static int igb_eeprom_test(struct igb_adapter *adapter, u64 *data)
*data = 0;
/* Read and add up the contents of the EEPROM */
for (i = 0; i < (NVM_CHECKSUM_REG + 1); i++) {
- if ((adapter->hw.nvm.ops.read_nvm(&adapter->hw, i, 1, &temp))
+ if ((adapter->hw.nvm.ops.read(&adapter->hw, i, 1, &temp))
< 0) {
*data = 1;
break;
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index cb3ac34..e3a3582 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1243,8 +1243,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
if (hw->bus.func == 0 ||
hw->device_id == E1000_DEV_ID_82575EB_COPPER)
- hw->nvm.ops.read_nvm(hw, NVM_INIT_CONTROL3_PORT_A, 1,
- &eeprom_data);
+ hw->nvm.ops.read(hw, NVM_INIT_CONTROL3_PORT_A, 1, &eeprom_data);
if (eeprom_data & eeprom_apme_mask)
adapter->eeprom_wol |= E1000_WUFC_MAG;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 10/26] igb: remove unused rx_hdr_split statistic
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (7 preceding siblings ...)
2009-02-07 9:17 ` [net-next PATCH 09/26] igb: rename nvm ops Jeff Kirsher
@ 2009-02-07 9:18 ` Jeff Kirsher
2009-02-07 9:18 ` [net-next PATCH 11/26] igb: update feature flags supported in ethtool Jeff Kirsher
` (16 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:18 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This statistic is not used and so it is safe to remove
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb.h | 1 -
drivers/net/igb/igb_ethtool.c | 1 -
2 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index a2a812d..88fdfe4 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -238,7 +238,6 @@ struct igb_adapter {
u64 hw_csum_err;
u64 hw_csum_good;
- u64 rx_hdr_split;
u32 alloc_rx_buff_failed;
bool rx_csum;
u32 gorc;
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index a5bf8ef..84be46c 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -88,7 +88,6 @@ static const struct igb_stats igb_gstrings_stats[] = {
{ "rx_long_byte_count", IGB_STAT(stats.gorc) },
{ "rx_csum_offload_good", IGB_STAT(hw_csum_good) },
{ "rx_csum_offload_errors", IGB_STAT(hw_csum_err) },
- { "rx_header_split", IGB_STAT(rx_hdr_split) },
{ "alloc_rx_buff_failed", IGB_STAT(alloc_rx_buff_failed) },
{ "tx_smbus", IGB_STAT(stats.mgptc) },
{ "rx_smbus", IGB_STAT(stats.mgprc) },
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 11/26] igb: update feature flags supported in ethtool
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (8 preceding siblings ...)
2009-02-07 9:18 ` [net-next PATCH 10/26] igb: remove unused rx_hdr_split statistic Jeff Kirsher
@ 2009-02-07 9:18 ` Jeff Kirsher
2009-02-07 9:18 ` [net-next PATCH 12/26] igb: update testing done by ethtool Jeff Kirsher
` (15 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:18 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This driver is currently using HW_CSUM which is not correct. Update this
to use the IP_CSUM and IPV6_CSUM flags. In addition consolidate the TSO
flag setting.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_ethtool.c | 16 +++++++---------
drivers/net/igb/igb_main.c | 5 +++--
2 files changed, 10 insertions(+), 11 deletions(-)
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index 84be46c..d7bdc6c 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -287,15 +287,15 @@ static int igb_set_rx_csum(struct net_device *netdev, u32 data)
static u32 igb_get_tx_csum(struct net_device *netdev)
{
- return (netdev->features & NETIF_F_HW_CSUM) != 0;
+ return (netdev->features & NETIF_F_IP_CSUM) != 0;
}
static int igb_set_tx_csum(struct net_device *netdev, u32 data)
{
if (data)
- netdev->features |= NETIF_F_HW_CSUM;
+ netdev->features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
else
- netdev->features &= ~NETIF_F_HW_CSUM;
+ netdev->features &= ~(NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
return 0;
}
@@ -304,15 +304,13 @@ static int igb_set_tso(struct net_device *netdev, u32 data)
{
struct igb_adapter *adapter = netdev_priv(netdev);
- if (data)
+ if (data) {
netdev->features |= NETIF_F_TSO;
- else
- netdev->features &= ~NETIF_F_TSO;
-
- if (data)
netdev->features |= NETIF_F_TSO6;
- else
+ } else {
+ netdev->features &= ~NETIF_F_TSO;
netdev->features &= ~NETIF_F_TSO6;
+ }
dev_info(&adapter->pdev->dev, "TSO is %s\n",
data ? "Enabled" : "Disabled");
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index e3a3582..8c27e0a 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1158,11 +1158,12 @@ static int __devinit igb_probe(struct pci_dev *pdev,
"PHY reset is blocked due to SOL/IDER session.\n");
netdev->features = NETIF_F_SG |
- NETIF_F_HW_CSUM |
+ NETIF_F_IP_CSUM |
NETIF_F_HW_VLAN_TX |
NETIF_F_HW_VLAN_RX |
NETIF_F_HW_VLAN_FILTER;
+ netdev->features |= NETIF_F_IPV6_CSUM;
netdev->features |= NETIF_F_TSO;
netdev->features |= NETIF_F_TSO6;
@@ -1172,7 +1173,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
netdev->vlan_features |= NETIF_F_TSO;
netdev->vlan_features |= NETIF_F_TSO6;
- netdev->vlan_features |= NETIF_F_HW_CSUM;
+ netdev->vlan_features |= NETIF_F_IP_CSUM;
netdev->vlan_features |= NETIF_F_SG;
if (pci_using_dac)
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 12/26] igb: update testing done by ethtool
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (9 preceding siblings ...)
2009-02-07 9:18 ` [net-next PATCH 11/26] igb: update feature flags supported in ethtool Jeff Kirsher
@ 2009-02-07 9:18 ` Jeff Kirsher
2009-02-07 9:19 ` [net-next PATCH 13/26] igb: add counter for dma out of sync errors Jeff Kirsher
` (14 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:18 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Most of the code for the testing has pretty much become stale at this point
and is need of update. This update just streamlines most of the code,
widens the range of interrupt testing.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_ethtool.c | 87 +++++++++++++++++++++++++++++------------
1 files changed, 61 insertions(+), 26 deletions(-)
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index d7bdc6c..33c23a1 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -855,23 +855,26 @@ static struct igb_reg_test reg_test_82576[] = {
{ E1000_RDBAL(0), 0x100, 4, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
{ E1000_RDBAH(0), 0x100, 4, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
{ E1000_RDLEN(0), 0x100, 4, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
- { E1000_RDBAL(4), 0x40, 8, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
- { E1000_RDBAH(4), 0x40, 8, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
- { E1000_RDLEN(4), 0x40, 8, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
- /* Enable all four RX queues before testing. */
- { E1000_RXDCTL(0), 0x100, 1, WRITE_NO_TEST, 0, E1000_RXDCTL_QUEUE_ENABLE },
+ { E1000_RDBAL(4), 0x40, 12, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
+ { E1000_RDBAH(4), 0x40, 12, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
+ { E1000_RDLEN(4), 0x40, 12, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
+ /* Enable all RX queues before testing. */
+ { E1000_RXDCTL(0), 0x100, 4, WRITE_NO_TEST, 0, E1000_RXDCTL_QUEUE_ENABLE },
+ { E1000_RXDCTL(4), 0x40, 12, WRITE_NO_TEST, 0, E1000_RXDCTL_QUEUE_ENABLE },
/* RDH is read-only for 82576, only test RDT. */
{ E1000_RDT(0), 0x100, 4, PATTERN_TEST, 0x0000FFFF, 0x0000FFFF },
+ { E1000_RDT(4), 0x40, 12, PATTERN_TEST, 0x0000FFFF, 0x0000FFFF },
{ E1000_RXDCTL(0), 0x100, 4, WRITE_NO_TEST, 0, 0 },
+ { E1000_RXDCTL(4), 0x40, 12, WRITE_NO_TEST, 0, 0 },
{ E1000_FCRTH, 0x100, 1, PATTERN_TEST, 0x0000FFF0, 0x0000FFF0 },
{ E1000_FCTTV, 0x100, 1, PATTERN_TEST, 0x0000FFFF, 0x0000FFFF },
{ E1000_TIPG, 0x100, 1, PATTERN_TEST, 0x3FFFFFFF, 0x3FFFFFFF },
{ E1000_TDBAL(0), 0x100, 4, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
{ E1000_TDBAH(0), 0x100, 4, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
{ E1000_TDLEN(0), 0x100, 4, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
- { E1000_TDBAL(4), 0x40, 8, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
- { E1000_TDBAH(4), 0x40, 8, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
- { E1000_TDLEN(4), 0x40, 8, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
+ { E1000_TDBAL(4), 0x40, 12, PATTERN_TEST, 0xFFFFFF80, 0xFFFFFFFF },
+ { E1000_TDBAH(4), 0x40, 12, PATTERN_TEST, 0xFFFFFFFF, 0xFFFFFFFF },
+ { E1000_TDLEN(4), 0x40, 12, PATTERN_TEST, 0x000FFFF0, 0x000FFFFF },
{ E1000_RCTL, 0x100, 1, SET_READ_TEST, 0xFFFFFFFF, 0x00000000 },
{ E1000_RCTL, 0x100, 1, SET_READ_TEST, 0x04CFB0FE, 0x003FFFFB },
{ E1000_RCTL, 0x100, 1, SET_READ_TEST, 0x04CFB0FE, 0xFFFFFFFF },
@@ -918,12 +921,13 @@ static struct igb_reg_test reg_test_82575[] = {
static bool reg_pattern_test(struct igb_adapter *adapter, u64 *data,
int reg, u32 mask, u32 write)
{
+ struct e1000_hw *hw = &adapter->hw;
u32 pat, val;
u32 _test[] =
{0x5A5A5A5A, 0xA5A5A5A5, 0x00000000, 0xFFFFFFFF};
for (pat = 0; pat < ARRAY_SIZE(_test); pat++) {
- writel((_test[pat] & write), (adapter->hw.hw_addr + reg));
- val = readl(adapter->hw.hw_addr + reg);
+ wr32(reg, (_test[pat] & write));
+ val = rd32(reg);
if (val != (_test[pat] & write & mask)) {
dev_err(&adapter->pdev->dev, "pattern test reg %04X "
"failed: got 0x%08X expected 0x%08X\n",
@@ -938,9 +942,10 @@ static bool reg_pattern_test(struct igb_adapter *adapter, u64 *data,
static bool reg_set_and_check(struct igb_adapter *adapter, u64 *data,
int reg, u32 mask, u32 write)
{
+ struct e1000_hw *hw = &adapter->hw;
u32 val;
- writel((write & mask), (adapter->hw.hw_addr + reg));
- val = readl(adapter->hw.hw_addr + reg);
+ wr32(reg, write & mask);
+ val = rd32(reg);
if ((write & mask) != (val & mask)) {
dev_err(&adapter->pdev->dev, "set/check reg %04X test failed:"
" got 0x%08X expected 0x%08X\n", reg,
@@ -1006,12 +1011,14 @@ static int igb_reg_test(struct igb_adapter *adapter, u64 *data)
for (i = 0; i < test->array_len; i++) {
switch (test->test_type) {
case PATTERN_TEST:
- REG_PATTERN_TEST(test->reg + (i * test->reg_offset),
+ REG_PATTERN_TEST(test->reg +
+ (i * test->reg_offset),
test->mask,
test->write);
break;
case SET_READ_TEST:
- REG_SET_AND_CHECK(test->reg + (i * test->reg_offset),
+ REG_SET_AND_CHECK(test->reg +
+ (i * test->reg_offset),
test->mask,
test->write);
break;
@@ -1083,16 +1090,17 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
{
struct e1000_hw *hw = &adapter->hw;
struct net_device *netdev = adapter->netdev;
- u32 mask, i = 0, shared_int = true;
+ u32 mask, ics_mask, i = 0, shared_int = true;
u32 irq = adapter->pdev->irq;
*data = 0;
/* Hook up test interrupt handler just for this test */
- if (adapter->msix_entries) {
+ if (adapter->msix_entries)
/* NOTE: we don't test MSI-X interrupts here, yet */
return 0;
- } else if (adapter->flags & IGB_FLAG_HAS_MSI) {
+
+ if (adapter->flags & IGB_FLAG_HAS_MSI) {
shared_int = false;
if (request_irq(irq, &igb_test_intr, 0, netdev->name, netdev)) {
*data = 1;
@@ -1108,16 +1116,31 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
}
dev_info(&adapter->pdev->dev, "testing %s interrupt\n",
(shared_int ? "shared" : "unshared"));
-
/* Disable all the interrupts */
wr32(E1000_IMC, 0xFFFFFFFF);
msleep(10);
+ /* Define all writable bits for ICS */
+ switch(hw->mac.type) {
+ case e1000_82575:
+ ics_mask = 0x37F47EDD;
+ break;
+ case e1000_82576:
+ ics_mask = 0x77D4FBFD;
+ break;
+ default:
+ ics_mask = 0x7FFFFFFF;
+ break;
+ }
+
/* Test each interrupt */
- for (; i < 10; i++) {
+ for (; i < 31; i++) {
/* Interrupt to test */
mask = 1 << i;
+ if (!(mask & ics_mask))
+ continue;
+
if (!shared_int) {
/* Disable the interrupt to be reported in
* the cause register and then force the same
@@ -1126,8 +1149,12 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
* test failed.
*/
adapter->test_icr = 0;
- wr32(E1000_IMC, ~mask & 0x00007FFF);
- wr32(E1000_ICS, ~mask & 0x00007FFF);
+
+ /* Flush any pending interrupts */
+ wr32(E1000_ICR, ~0);
+
+ wr32(E1000_IMC, mask);
+ wr32(E1000_ICS, mask);
msleep(10);
if (adapter->test_icr & mask) {
@@ -1143,6 +1170,10 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
* test failed.
*/
adapter->test_icr = 0;
+
+ /* Flush any pending interrupts */
+ wr32(E1000_ICR, ~0);
+
wr32(E1000_IMS, mask);
wr32(E1000_ICS, mask);
msleep(10);
@@ -1160,11 +1191,15 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
* test failed.
*/
adapter->test_icr = 0;
- wr32(E1000_IMC, ~mask & 0x00007FFF);
- wr32(E1000_ICS, ~mask & 0x00007FFF);
+
+ /* Flush any pending interrupts */
+ wr32(E1000_ICR, ~0);
+
+ wr32(E1000_IMC, ~mask);
+ wr32(E1000_ICS, ~mask);
msleep(10);
- if (adapter->test_icr) {
+ if (adapter->test_icr & mask) {
*data = 5;
break;
}
@@ -1172,7 +1207,7 @@ static int igb_intr_test(struct igb_adapter *adapter, u64 *data)
}
/* Disable all the interrupts */
- wr32(E1000_IMC, 0xFFFFFFFF);
+ wr32(E1000_IMC, ~0);
msleep(10);
/* Unhook test interrupt handler */
@@ -1450,7 +1485,7 @@ static int igb_setup_loopback_test(struct igb_adapter *adapter)
E1000_CTRL_TFCE |
E1000_CTRL_LRST);
reg |= E1000_CTRL_SLU |
- E1000_CTRL_FD;
+ E1000_CTRL_FD;
wr32(E1000_CTRL, reg);
/* Unset switch control to serdes energy detect */
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 13/26] igb: add counter for dma out of sync errors
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (10 preceding siblings ...)
2009-02-07 9:18 ` [net-next PATCH 12/26] igb: update testing done by ethtool Jeff Kirsher
@ 2009-02-07 9:19 ` Jeff Kirsher
2009-02-07 9:19 ` [net-next PATCH 14/26] igb: cleanup igb_netpoll to be more friendly with napi & GRO Jeff Kirsher
` (13 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Add a counter for dma out of sync errors reported via interrupt.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_defines.h | 6 +++++-
drivers/net/igb/e1000_hw.h | 1 +
drivers/net/igb/igb_ethtool.c | 1 +
drivers/net/igb/igb_main.c | 19 +++++++++++++++++--
4 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/drivers/net/igb/e1000_defines.h b/drivers/net/igb/e1000_defines.h
index 54a1489..bff62dd 100644
--- a/drivers/net/igb/e1000_defines.h
+++ b/drivers/net/igb/e1000_defines.h
@@ -413,6 +413,7 @@
/* LAN connected device generates an interrupt */
#define E1000_ICR_PHYINT 0x00001000
#define E1000_ICR_EPRST 0x00100000 /* ME handware reset occurs */
+#define E1000_ICR_DOUTSYNC 0x10000000 /* NIC DMA out of sync */
/* Extended Interrupt Cause Read */
#define E1000_EICR_RX_QUEUE0 0x00000001 /* Rx Queue 0 Interrupt */
@@ -441,7 +442,8 @@
E1000_IMS_TXDW | \
E1000_IMS_RXDMT0 | \
E1000_IMS_RXSEQ | \
- E1000_IMS_LSC)
+ E1000_IMS_LSC | \
+ E1000_IMS_DOUTSYNC)
/* Interrupt Mask Set */
#define E1000_IMS_TXDW E1000_ICR_TXDW /* Transmit desc written back */
@@ -449,6 +451,7 @@
#define E1000_IMS_RXSEQ E1000_ICR_RXSEQ /* rx sequence error */
#define E1000_IMS_RXDMT0 E1000_ICR_RXDMT0 /* rx desc min. threshold */
#define E1000_IMS_RXT0 E1000_ICR_RXT0 /* rx timer intr */
+#define E1000_IMS_DOUTSYNC E1000_ICR_DOUTSYNC /* NIC DMA out of sync */
/* Extended Interrupt Mask Set */
#define E1000_EIMS_TCP_TIMER E1000_EICR_TCP_TIMER /* TCP Timer */
@@ -457,6 +460,7 @@
/* Interrupt Cause Set */
#define E1000_ICS_LSC E1000_ICR_LSC /* Link Status Change */
#define E1000_ICS_RXDMT0 E1000_ICR_RXDMT0 /* rx desc min. threshold */
+#define E1000_ICS_DOUTSYNC E1000_ICR_DOUTSYNC /* NIC DMA out of sync */
/* Extended Interrupt Cause Set */
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index acb42a2..62ccd49 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -359,6 +359,7 @@ struct e1000_hw_stats {
u64 lenerrs;
u64 scvpc;
u64 hrmpc;
+ u64 doosync;
};
struct e1000_phy_stats {
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index 33c23a1..de09430 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -88,6 +88,7 @@ static const struct igb_stats igb_gstrings_stats[] = {
{ "rx_long_byte_count", IGB_STAT(stats.gorc) },
{ "rx_csum_offload_good", IGB_STAT(hw_csum_good) },
{ "rx_csum_offload_errors", IGB_STAT(hw_csum_err) },
+ { "tx_dma_out_of_sync", IGB_STAT(stats.doosync) },
{ "alloc_rx_buff_failed", IGB_STAT(alloc_rx_buff_failed) },
{ "tx_smbus", IGB_STAT(stats.mgptc) },
{ "rx_smbus", IGB_STAT(stats.mgprc) },
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 8c27e0a..c05ca34 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -689,7 +689,7 @@ static void igb_irq_enable(struct igb_adapter *adapter)
wr32(E1000_EIAC, adapter->eims_enable_mask);
wr32(E1000_EIAM, adapter->eims_enable_mask);
wr32(E1000_EIMS, adapter->eims_enable_mask);
- wr32(E1000_IMS, E1000_IMS_LSC);
+ wr32(E1000_IMS, E1000_IMS_LSC | E1000_IMS_DOUTSYNC);
} else {
wr32(E1000_IMS, IMS_ENABLE_MASK);
wr32(E1000_IAM, IMS_ENABLE_MASK);
@@ -3287,6 +3287,11 @@ static irqreturn_t igb_msix_other(int irq, void *data)
u32 icr = rd32(E1000_ICR);
/* reading ICR causes bit 31 of EICR to be cleared */
+
+ if(icr & E1000_ICR_DOUTSYNC) {
+ /* HW is reporting DMA is out of sync */
+ adapter->stats.doosync++;
+ }
if (!(icr & E1000_ICR_LSC))
goto no_link_interrupt;
hw->mac.get_link_status = 1;
@@ -3295,7 +3300,7 @@ static irqreturn_t igb_msix_other(int irq, void *data)
mod_timer(&adapter->watchdog_timer, jiffies + 1);
no_link_interrupt:
- wr32(E1000_IMS, E1000_IMS_LSC);
+ wr32(E1000_IMS, E1000_IMS_LSC | E1000_IMS_DOUTSYNC);
wr32(E1000_EIMS, adapter->eims_other);
return IRQ_HANDLED;
@@ -3499,6 +3504,11 @@ static irqreturn_t igb_intr_msi(int irq, void *data)
igb_write_itr(adapter->rx_ring);
+ if(icr & E1000_ICR_DOUTSYNC) {
+ /* HW is reporting DMA is out of sync */
+ adapter->stats.doosync++;
+ }
+
if (icr & (E1000_ICR_RXSEQ | E1000_ICR_LSC)) {
hw->mac.get_link_status = 1;
if (!test_bit(__IGB_DOWN, &adapter->state))
@@ -3534,6 +3544,11 @@ static irqreturn_t igb_intr(int irq, void *data)
if (!(icr & E1000_ICR_INT_ASSERTED))
return IRQ_NONE;
+ if(icr & E1000_ICR_DOUTSYNC) {
+ /* HW is reporting DMA is out of sync */
+ adapter->stats.doosync++;
+ }
+
eicr = rd32(E1000_EICR);
if (icr & (E1000_ICR_RXSEQ | E1000_ICR_LSC)) {
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 14/26] igb: cleanup igb_netpoll to be more friendly with napi & GRO
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (11 preceding siblings ...)
2009-02-07 9:19 ` [net-next PATCH 13/26] igb: add counter for dma out of sync errors Jeff Kirsher
@ 2009-02-07 9:19 ` Jeff Kirsher
2009-02-07 9:19 ` [net-next PATCH 15/26] igb: remove redundant timer updates and cleanup watchdog_task Jeff Kirsher
` (12 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch cleans up igb_netpoll so that it is more friendly with both the
current napi and newly introduced GRO features.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb.h | 5 ++---
drivers/net/igb/igb_main.c | 37 +++++++++++++++++++++----------------
2 files changed, 23 insertions(+), 19 deletions(-)
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 88fdfe4..0519ae4 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -280,9 +280,8 @@ struct igb_adapter {
#define IGB_FLAG_HAS_MSI (1 << 0)
#define IGB_FLAG_MSI_ENABLE (1 << 1)
#define IGB_FLAG_DCA_ENABLED (1 << 2)
-#define IGB_FLAG_IN_NETPOLL (1 << 3)
-#define IGB_FLAG_QUAD_PORT_A (1 << 4)
-#define IGB_FLAG_NEED_CTX_IDX (1 << 5)
+#define IGB_FLAG_QUAD_PORT_A (1 << 3)
+#define IGB_FLAG_NEED_CTX_IDX (1 << 4)
enum e1000_state_t {
__IGB_TESTING,
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index c05ca34..3bf560f 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -404,7 +404,7 @@ static void igb_configure_msix(struct igb_adapter *adapter)
/* Turn on MSI-X capability first, or our settings
* won't stick. And it will take days to debug. */
wr32(E1000_GPIE, E1000_GPIE_MSIX_MODE |
- E1000_GPIE_PBA | E1000_GPIE_EIAME |
+ E1000_GPIE_PBA | E1000_GPIE_EIAME |
E1000_GPIE_NSICR);
for (i = 0; i < adapter->num_tx_queues; i++) {
@@ -1629,7 +1629,7 @@ static int igb_setup_all_tx_resources(struct igb_adapter *adapter)
for (i = 0; i < IGB_MAX_TX_QUEUES; i++) {
r_idx = i % adapter->num_tx_queues;
adapter->multi_tx_table[i] = &adapter->tx_ring[r_idx];
- }
+ }
return err;
}
@@ -3298,7 +3298,7 @@ static irqreturn_t igb_msix_other(int irq, void *data)
/* guard against interrupt when we're going down */
if (!test_bit(__IGB_DOWN, &adapter->state))
mod_timer(&adapter->watchdog_timer, jiffies + 1);
-
+
no_link_interrupt:
wr32(E1000_IMS, E1000_IMS_LSC | E1000_IMS_DOUTSYNC);
wr32(E1000_EIMS, adapter->eims_other);
@@ -3751,7 +3751,7 @@ static bool igb_clean_tx_irq(struct igb_ring *tx_ring)
/**
* igb_receive_skb - helper function to handle rx indications
- * @ring: pointer to receive ring receving this packet
+ * @ring: pointer to receive ring receving this packet
* @status: descriptor status field as written by hardware
* @vlan: descriptor vlan field as written by hardware (no le/be conversion)
* @skb: pointer to sk_buff to be indicated to stack
@@ -4378,22 +4378,27 @@ static void igb_shutdown(struct pci_dev *pdev)
static void igb_netpoll(struct net_device *netdev)
{
struct igb_adapter *adapter = netdev_priv(netdev);
+ struct e1000_hw *hw = &adapter->hw;
int i;
- int work_done = 0;
- igb_irq_disable(adapter);
- adapter->flags |= IGB_FLAG_IN_NETPOLL;
-
- for (i = 0; i < adapter->num_tx_queues; i++)
- igb_clean_tx_irq(&adapter->tx_ring[i]);
+ if (!adapter->msix_entries) {
+ igb_irq_disable(adapter);
+ napi_schedule(&adapter->rx_ring[0].napi);
+ return;
+ }
- for (i = 0; i < adapter->num_rx_queues; i++)
- igb_clean_rx_irq_adv(&adapter->rx_ring[i],
- &work_done,
- adapter->rx_ring[i].napi.weight);
+ for (i = 0; i < adapter->num_tx_queues; i++) {
+ struct igb_ring *tx_ring = &adapter->tx_ring[i];
+ wr32(E1000_EIMC, tx_ring->eims_value);
+ igb_clean_tx_irq(tx_ring);
+ wr32(E1000_EIMS, tx_ring->eims_value);
+ }
- adapter->flags &= ~IGB_FLAG_IN_NETPOLL;
- igb_irq_enable(adapter);
+ for (i = 0; i < adapter->num_rx_queues; i++) {
+ struct igb_ring *rx_ring = &adapter->rx_ring[i];
+ wr32(E1000_EIMC, rx_ring->eims_value);
+ napi_schedule(&rx_ring->napi);
+ }
}
#endif /* CONFIG_NET_POLL_CONTROLLER */
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 15/26] igb: remove redundant timer updates and cleanup watchdog_task
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (12 preceding siblings ...)
2009-02-07 9:19 ` [net-next PATCH 14/26] igb: cleanup igb_netpoll to be more friendly with napi & GRO Jeff Kirsher
@ 2009-02-07 9:19 ` Jeff Kirsher
2009-02-07 9:20 ` [net-next PATCH 16/26] igb: rename igb_update_mc_addr_list_82575 to not include the 82575 Jeff Kirsher
` (11 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
The igb watchdog task is modifying the watchdog timer twice duing a single
run. It only needs to be called once to reschedule itself for 2 seconds from
the last time it ran.
In addition I removed the allocation of the mac_info structure since it is
only called twice and is easier to access via the e1000_hw struct.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 10 +++++-----
1 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 3bf560f..cbb3858 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2314,13 +2314,10 @@ static void igb_watchdog_task(struct work_struct *work)
struct igb_adapter *adapter = container_of(work,
struct igb_adapter, watchdog_task);
struct e1000_hw *hw = &adapter->hw;
-
struct net_device *netdev = adapter->netdev;
struct igb_ring *tx_ring = adapter->tx_ring;
- struct e1000_mac_info *mac = &adapter->hw.mac;
u32 link;
u32 eics = 0;
- s32 ret_val;
int i;
link = igb_has_link(adapter);
@@ -2365,6 +2362,7 @@ static void igb_watchdog_task(struct work_struct *work)
netif_carrier_on(netdev);
netif_tx_wake_all_queues(netdev);
+ /* link state has changed, schedule phy info update */
if (!test_bit(__IGB_DOWN, &adapter->state))
mod_timer(&adapter->phy_info_timer,
round_jiffies(jiffies + 2 * HZ));
@@ -2378,6 +2376,8 @@ static void igb_watchdog_task(struct work_struct *work)
netdev->name);
netif_carrier_off(netdev);
netif_tx_stop_all_queues(netdev);
+
+ /* link state has changed, schedule phy info update */
if (!test_bit(__IGB_DOWN, &adapter->state))
mod_timer(&adapter->phy_info_timer,
round_jiffies(jiffies + 2 * HZ));
@@ -2387,9 +2387,9 @@ static void igb_watchdog_task(struct work_struct *work)
link_up:
igb_update_stats(adapter);
- mac->tx_packet_delta = adapter->stats.tpt - adapter->tpt_old;
+ hw->mac.tx_packet_delta = adapter->stats.tpt - adapter->tpt_old;
adapter->tpt_old = adapter->stats.tpt;
- mac->collision_delta = adapter->stats.colc - adapter->colc_old;
+ hw->mac.collision_delta = adapter->stats.colc - adapter->colc_old;
adapter->colc_old = adapter->stats.colc;
adapter->gorc = adapter->stats.gorc - adapter->gorc_old;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 16/26] igb: rename igb_update_mc_addr_list_82575 to not include the 82575
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (13 preceding siblings ...)
2009-02-07 9:19 ` [net-next PATCH 15/26] igb: remove redundant timer updates and cleanup watchdog_task Jeff Kirsher
@ 2009-02-07 9:20 ` Jeff Kirsher
2009-02-07 9:20 ` [net-next PATCH 17/26] igb: remove unnecessary adapter->hw calls when just hw-> will do Jeff Kirsher
` (10 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:20 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
There isn't much point in having the _82575 hanging off the end of this
function since there aren't any other version of this function running
around within this driver. This also allows for a bit of whitespace
cleanup due to a shorter function name.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_82575.c | 8 ++++----
drivers/net/igb/e1000_82575.h | 2 +-
drivers/net/igb/igb_main.c | 7 +++----
3 files changed, 8 insertions(+), 9 deletions(-)
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index 527d4c8..68a9822 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -787,7 +787,7 @@ static void igb_init_rx_addrs_82575(struct e1000_hw *hw, u16 rar_count)
}
/**
- * igb_update_mc_addr_list_82575 - Update Multicast addresses
+ * igb_update_mc_addr_list - Update Multicast addresses
* @hw: pointer to the HW structure
* @mc_addr_list: array of multicast addresses to program
* @mc_addr_count: number of multicast addresses to program
@@ -799,9 +799,9 @@ static void igb_init_rx_addrs_82575(struct e1000_hw *hw, u16 rar_count)
* The parameter rar_count will usually be hw->mac.rar_entry_count
* unless there are workarounds that change this.
**/
-void igb_update_mc_addr_list_82575(struct e1000_hw *hw,
- u8 *mc_addr_list, u32 mc_addr_count,
- u32 rar_used_count, u32 rar_count)
+void igb_update_mc_addr_list(struct e1000_hw *hw,
+ u8 *mc_addr_list, u32 mc_addr_count,
+ u32 rar_used_count, u32 rar_count)
{
u32 hash_value;
u32 i;
diff --git a/drivers/net/igb/e1000_82575.h b/drivers/net/igb/e1000_82575.h
index c1928b5..e0a376f 100644
--- a/drivers/net/igb/e1000_82575.h
+++ b/drivers/net/igb/e1000_82575.h
@@ -28,7 +28,7 @@
#ifndef _E1000_82575_H_
#define _E1000_82575_H_
-void igb_update_mc_addr_list_82575(struct e1000_hw*, u8*, u32, u32, u32);
+void igb_update_mc_addr_list(struct e1000_hw*, u8*, u32, u32, u32);
extern void igb_shutdown_fiber_serdes_link_82575(struct e1000_hw *hw);
extern void igb_rx_fifo_flush_82575(struct e1000_hw *hw);
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index cbb3858..2d169a4 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2227,8 +2227,8 @@ static void igb_set_multi(struct net_device *netdev)
if (!netdev->mc_count) {
/* nothing to program, so clear mc list */
- igb_update_mc_addr_list_82575(hw, NULL, 0, 1,
- mac->rar_entry_count);
+ igb_update_mc_addr_list(hw, NULL, 0, 1,
+ mac->rar_entry_count);
return;
}
@@ -2245,8 +2245,7 @@ static void igb_set_multi(struct net_device *netdev)
memcpy(mta_list + (i*ETH_ALEN), mc_ptr->dmi_addr, ETH_ALEN);
mc_ptr = mc_ptr->next;
}
- igb_update_mc_addr_list_82575(hw, mta_list, i, 1,
- mac->rar_entry_count);
+ igb_update_mc_addr_list(hw, mta_list, i, 1, mac->rar_entry_count);
kfree(mta_list);
}
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 17/26] igb: remove unnecessary adapter->hw calls when just hw-> will do.
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (14 preceding siblings ...)
2009-02-07 9:20 ` [net-next PATCH 16/26] igb: rename igb_update_mc_addr_list_82575 to not include the 82575 Jeff Kirsher
@ 2009-02-07 9:20 ` Jeff Kirsher
2009-02-07 9:20 ` [net-next PATCH 18/26] igb: don't read eicr when responding to legacy interrupts Jeff Kirsher
` (9 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:20 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck
From: Alexander Duyck <alexander.h.duyck@intel.com>
There were several spots in the code making calls to adapter->hw when they
could have just been accessing hw-> directly. I cleaned up the spots where
this was visibly apparent.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
---
drivers/net/igb/igb_main.c | 21 ++++++++++-----------
1 files changed, 10 insertions(+), 11 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 2d169a4..e3f521c 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1096,8 +1096,8 @@ static int __devinit igb_probe(struct pci_dev *pdev,
mmio_len = pci_resource_len(pdev, 0);
err = -EIO;
- adapter->hw.hw_addr = ioremap(mmio_start, mmio_len);
- if (!adapter->hw.hw_addr)
+ hw->hw_addr = ioremap(mmio_start, mmio_len);
+ if (!hw->hw_addr)
goto err_ioremap;
netdev->netdev_ops = &igb_netdev_ops;
@@ -1357,9 +1357,7 @@ static void __devexit igb_remove(struct pci_dev *pdev)
{
struct net_device *netdev = pci_get_drvdata(pdev);
struct igb_adapter *adapter = netdev_priv(netdev);
-#ifdef CONFIG_IGB_DCA
struct e1000_hw *hw = &adapter->hw;
-#endif
int err;
/* flush_scheduled work may reschedule our watchdog task, so
@@ -1392,9 +1390,9 @@ static void __devexit igb_remove(struct pci_dev *pdev)
igb_free_queues(adapter);
- iounmap(adapter->hw.hw_addr);
- if (adapter->hw.flash_address)
- iounmap(adapter->hw.flash_address);
+ iounmap(hw->hw_addr);
+ if (hw->flash_address)
+ iounmap(hw->flash_address);
pci_release_selected_regions(pdev, pci_select_bars(pdev,
IORESOURCE_MEM));
@@ -1784,7 +1782,7 @@ static void igb_setup_rctl(struct igb_adapter *adapter)
rctl &= ~(E1000_RCTL_LBM_TCVR | E1000_RCTL_LBM_MAC);
rctl |= E1000_RCTL_EN | E1000_RCTL_BAM | E1000_RCTL_RDMTS_HALF |
- (adapter->hw.mac.mc_filter_type << E1000_RCTL_MO_SHIFT);
+ (hw->mac.mc_filter_type << E1000_RCTL_MO_SHIFT);
/*
* enable stripping of CRC. It's unlikely this will break BMC
@@ -2176,15 +2174,16 @@ static void igb_clean_all_rx_rings(struct igb_adapter *adapter)
static int igb_set_mac(struct net_device *netdev, void *p)
{
struct igb_adapter *adapter = netdev_priv(netdev);
+ struct e1000_hw *hw = &adapter->hw;
struct sockaddr *addr = p;
if (!is_valid_ether_addr(addr->sa_data))
return -EADDRNOTAVAIL;
memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
- memcpy(adapter->hw.mac.addr, addr->sa_data, netdev->addr_len);
+ memcpy(hw->mac.addr, addr->sa_data, netdev->addr_len);
- adapter->hw.mac.ops.rar_set(&adapter->hw, adapter->hw.mac.addr, 0);
+ hw->mac.ops.rar_set(hw, hw->mac.addr, 0);
return 0;
}
@@ -4140,7 +4139,7 @@ static void igb_vlan_rx_add_vid(struct net_device *netdev, u16 vid)
struct e1000_hw *hw = &adapter->hw;
u32 vfta, index;
- if ((adapter->hw.mng_cookie.status &
+ if ((hw->mng_cookie.status &
E1000_MNG_DHCP_COOKIE_STATUS_VLAN) &&
(vid == adapter->mng_vlan_id))
return;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 18/26] igb: don't read eicr when responding to legacy interrupts
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (15 preceding siblings ...)
2009-02-07 9:20 ` [net-next PATCH 17/26] igb: remove unnecessary adapter->hw calls when just hw-> will do Jeff Kirsher
@ 2009-02-07 9:20 ` Jeff Kirsher
2009-02-07 9:21 ` [net-next PATCH 19/26] igb: move get_hw_control within igb_resume Jeff Kirsher
` (8 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:20 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
The interrupt handler was reading eicr and then doing nothing with the
result. I have removed the variable and the register read since they
provide no value to the legacy interrupt handler.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 5 +----
1 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index e3f521c..56c1455 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -3519,7 +3519,7 @@ static irqreturn_t igb_intr_msi(int irq, void *data)
}
/**
- * igb_intr - Interrupt Handler
+ * igb_intr - Legacy Interrupt Handler
* @irq: interrupt number
* @data: pointer to a network interface device structure
**/
@@ -3531,7 +3531,6 @@ static irqreturn_t igb_intr(int irq, void *data)
/* Interrupt Auto-Mask...upon reading ICR, interrupts are masked. No
* need for the IMC write */
u32 icr = rd32(E1000_ICR);
- u32 eicr = 0;
if (!icr)
return IRQ_NONE; /* Not our interrupt */
@@ -3547,8 +3546,6 @@ static irqreturn_t igb_intr(int irq, void *data)
adapter->stats.doosync++;
}
- eicr = rd32(E1000_EICR);
-
if (icr & (E1000_ICR_RXSEQ | E1000_ICR_LSC)) {
hw->mac.get_link_status = 1;
/* guard against interrupt when we're going down */
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 19/26] igb: move get_hw_control within igb_resume.
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (16 preceding siblings ...)
2009-02-07 9:20 ` [net-next PATCH 18/26] igb: don't read eicr when responding to legacy interrupts Jeff Kirsher
@ 2009-02-07 9:21 ` Jeff Kirsher
2009-02-07 9:21 ` [net-next PATCH 20/26] igb: change pba size determination from if to switch statement Jeff Kirsher
` (7 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:21 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Move igb_get_hw_control up so that it is called just after the reset in
igb_resume. This notifies the HW sooner that the driver is reassuming
control of the device.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 9 +++++----
1 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 56c1455..accab3f 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -4341,6 +4341,11 @@ static int igb_resume(struct pci_dev *pdev)
/* e1000_power_up_phy(adapter); */
igb_reset(adapter);
+
+ /* let the f/w know that the h/w is now under the control of the
+ * driver. */
+ igb_get_hw_control(adapter);
+
wr32(E1000_WUS, ~0);
if (netif_running(netdev)) {
@@ -4351,10 +4356,6 @@ static int igb_resume(struct pci_dev *pdev)
netif_device_attach(netdev);
- /* let the f/w know that the h/w is now under the control of the
- * driver. */
- igb_get_hw_control(adapter);
-
return 0;
}
#endif
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 20/26] igb: change pba size determination from if to switch statement
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (17 preceding siblings ...)
2009-02-07 9:21 ` [net-next PATCH 19/26] igb: move get_hw_control within igb_resume Jeff Kirsher
@ 2009-02-07 9:21 ` Jeff Kirsher
2009-02-07 9:21 ` [net-next PATCH 21/26] igb: remove disable_av variable from mac_info struct Jeff Kirsher
` (6 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:21 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
As additional hardware is added to the igb driver it is easier to support
the expansion via switch statements instead of using nested ifs. For
this reason I am changing this to a switch statement.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 11 +++++++----
1 files changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index accab3f..d27e502 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -888,11 +888,14 @@ void igb_reset(struct igb_adapter *adapter)
/* Repartition Pba for greater than 9k mtu
* To take effect CTRL.RST is required.
*/
- if (mac->type != e1000_82576) {
- pba = E1000_PBA_34K;
- }
- else {
+ switch (mac->type) {
+ case e1000_82576:
pba = E1000_PBA_64K;
+ break;
+ case e1000_82575:
+ default:
+ pba = E1000_PBA_34K;
+ break;
}
if ((adapter->max_frame_size > ETH_FRAME_LEN + ETH_FCS_LEN) &&
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 21/26] igb: remove disable_av variable from mac_info struct
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (18 preceding siblings ...)
2009-02-07 9:21 ` [net-next PATCH 20/26] igb: change pba size determination from if to switch statement Jeff Kirsher
@ 2009-02-07 9:21 ` Jeff Kirsher
2009-02-07 9:22 ` [net-next PATCH 22/26] igb: remove redundant count set and err_hw_init Jeff Kirsher
` (5 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:21 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
The disable_av variable is never used by the driver and provides no value as
it is likely a leftover debugging variable. I have removed it and replaced
the one spot that checked for it with a check for a valid address.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/e1000_hw.h | 1 -
drivers/net/igb/e1000_mac.c | 3 ++-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index 62ccd49..6d1564f 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -484,7 +484,6 @@ struct e1000_mac_info {
bool asf_firmware_present;
bool autoneg;
bool autoneg_failed;
- bool disable_av;
bool disable_hw_init_bits;
bool get_link_status;
bool ifs_params_forced;
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index 6682206..e5d23e0 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -200,7 +200,8 @@ void igb_rar_set(struct e1000_hw *hw, u8 *addr, u32 index)
rar_high = ((u32) addr[4] | ((u32) addr[5] << 8));
- if (!hw->mac.disable_av)
+ /* If MAC address zero, no need to set the AV bit */
+ if (rar_low || rar_high)
rar_high |= E1000_RAH_AV;
wr32(E1000_RAL(index), rar_low);
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 22/26] igb: remove redundant count set and err_hw_init
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (19 preceding siblings ...)
2009-02-07 9:21 ` [net-next PATCH 21/26] igb: remove disable_av variable from mac_info struct Jeff Kirsher
@ 2009-02-07 9:22 ` Jeff Kirsher
2009-02-07 9:22 ` [net-next PATCH 23/26] igb: update stats before doing reset in igb_down Jeff Kirsher
` (4 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:22 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Remove the setting of ring->count variables from igb_probe as they are
duplicating the same configuration that is done igb_alloc_queues.
Remove the err_hw_init tag as it can be replaced by err_sw_init.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 15 ++++-----------
1 files changed, 4 insertions(+), 11 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index d27e502..3b79ad8 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1016,7 +1016,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
struct pci_dev *us_dev;
const struct e1000_info *ei = igb_info_tbl[ent->driver_data];
unsigned long mmio_start, mmio_len;
- int i, err, pci_using_dac, pos;
+ int err, pci_using_dac, pos;
u16 eeprom_data = 0, state = 0;
u16 eeprom_apme_mask = IGB_EEPROM_APME;
u32 part_num;
@@ -1128,8 +1128,9 @@ static int __devinit igb_probe(struct pci_dev *pdev,
/* Initialize skew-specific constants */
err = ei->get_invariants(hw);
if (err)
- goto err_hw_init;
+ goto err_sw_init;
+ /* setup the private structure */
err = igb_sw_init(adapter);
if (err)
goto err_sw_init;
@@ -1219,14 +1220,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
INIT_WORK(&adapter->reset_task, igb_reset_task);
INIT_WORK(&adapter->watchdog_task, igb_watchdog_task);
- /* Initialize link & ring properties that are user-changeable */
- adapter->tx_ring->count = 256;
- for (i = 0; i < adapter->num_tx_queues; i++)
- adapter->tx_ring[i].count = adapter->tx_ring->count;
- adapter->rx_ring->count = 256;
- for (i = 0; i < adapter->num_rx_queues; i++)
- adapter->rx_ring[i].count = adapter->rx_ring->count;
-
+ /* Initialize link properties that are user-changeable */
adapter->fc_autoneg = true;
hw->mac.autoneg = true;
hw->phy.autoneg_advertised = 0x2f;
@@ -1334,7 +1328,6 @@ err_eeprom:
igb_free_queues(adapter);
err_sw_init:
-err_hw_init:
iounmap(hw->hw_addr);
err_ioremap:
free_netdev(netdev);
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 23/26] igb: update stats before doing reset in igb_down
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (20 preceding siblings ...)
2009-02-07 9:22 ` [net-next PATCH 22/26] igb: remove redundant count set and err_hw_init Jeff Kirsher
@ 2009-02-07 9:22 ` Jeff Kirsher
2009-02-07 9:22 ` [net-next PATCH 24/26] igb: fix two minor items found during code review Jeff Kirsher
` (3 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:22 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
It was seen with repeated interface up/down testing that there was a large
stray between the stats reported by the queues and the stats reported by the
HW. It was found to be an issue in that hw stats were being reset without
first being recorded. This change records the stats before wiping them from
the system via the reset.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 3b79ad8..88f135f 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -858,6 +858,10 @@ void igb_down(struct igb_adapter *adapter)
netdev->tx_queue_len = adapter->tx_queue_len;
netif_carrier_off(netdev);
+
+ /* record the stats before reset*/
+ igb_update_stats(adapter);
+
adapter->link_speed = 0;
adapter->link_duplex = 0;
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 24/26] igb: fix two minor items found during code review
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (21 preceding siblings ...)
2009-02-07 9:22 ` [net-next PATCH 23/26] igb: update stats before doing reset in igb_down Jeff Kirsher
@ 2009-02-07 9:22 ` Jeff Kirsher
2009-02-07 9:23 ` [net-next PATCH 25/26] igb: update version number and copyright dates Jeff Kirsher
` (2 subsequent siblings)
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:22 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch addresses two minor items I found while cleaning up the igb
driver for our sourceforge version.
The first clears the context index if we don't flag that we need it.
The second item is that eims_other should be used instead of bit defines
when setting all of the EICS bits prior to reset.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 6 ++++--
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 88f135f..6713840 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2771,6 +2771,8 @@ static inline bool igb_tx_csum_adv(struct igb_adapter *adapter,
if (adapter->flags & IGB_FLAG_NEED_CTX_IDX)
context_desc->mss_l4len_idx =
cpu_to_le32(tx_ring->queue_index << 4);
+ else
+ context_desc->mss_l4len_idx = 0;
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
@@ -3040,8 +3042,8 @@ static void igb_tx_timeout(struct net_device *netdev)
/* Do the reset outside of interrupt context */
adapter->tx_timeout_count++;
schedule_work(&adapter->reset_task);
- wr32(E1000_EICS, adapter->eims_enable_mask &
- ~(E1000_EIMS_TCP_TIMER | E1000_EIMS_OTHER));
+ wr32(E1000_EICS,
+ (adapter->eims_enable_mask & ~adapter->eims_other));
}
static void igb_reset_task(struct work_struct *work)
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 25/26] igb: update version number and copyright dates
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (22 preceding siblings ...)
2009-02-07 9:22 ` [net-next PATCH 24/26] igb: fix two minor items found during code review Jeff Kirsher
@ 2009-02-07 9:23 ` Jeff Kirsher
2009-02-07 9:23 ` [net-next PATCH 26/26] igb: remove dead code in transmit routine Jeff Kirsher
2009-02-07 10:46 ` [net-next PATCH 01/26] igb: optomize/refactor receive path David Miller
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:23 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Alexander Duyck, Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
Update the version number to 1.3.16 and update copyright dates for 2009.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/Makefile | 2 +-
drivers/net/igb/e1000_82575.c | 2 +-
drivers/net/igb/e1000_82575.h | 2 +-
drivers/net/igb/e1000_defines.h | 2 +-
drivers/net/igb/e1000_hw.h | 2 +-
drivers/net/igb/e1000_mac.c | 2 +-
drivers/net/igb/e1000_mac.h | 2 +-
drivers/net/igb/e1000_nvm.c | 2 +-
drivers/net/igb/e1000_phy.c | 2 +-
drivers/net/igb/e1000_phy.h | 2 +-
drivers/net/igb/e1000_regs.h | 2 +-
drivers/net/igb/igb.h | 2 +-
drivers/net/igb/igb_ethtool.c | 2 +-
drivers/net/igb/igb_main.c | 6 +++---
14 files changed, 16 insertions(+), 16 deletions(-)
diff --git a/drivers/net/igb/Makefile b/drivers/net/igb/Makefile
index 1927b3f..cda3ad5 100644
--- a/drivers/net/igb/Makefile
+++ b/drivers/net/igb/Makefile
@@ -1,7 +1,7 @@
################################################################################
#
# Intel 82575 PCI-Express Ethernet Linux driver
-# Copyright(c) 1999 - 2007 Intel Corporation.
+# Copyright(c) 1999 - 2009 Intel Corporation.
#
# This program is free software; you can redistribute it and/or modify it
# under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index 68a9822..7f43e25 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 - 2008 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_82575.h b/drivers/net/igb/e1000_82575.h
index e0a376f..dd50237 100644
--- a/drivers/net/igb/e1000_82575.h
+++ b/drivers/net/igb/e1000_82575.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 - 2008 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_defines.h b/drivers/net/igb/e1000_defines.h
index bff62dd..5342e23 100644
--- a/drivers/net/igb/e1000_defines.h
+++ b/drivers/net/igb/e1000_defines.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 - 2008 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h
index 6d1564f..bd86ceb 100644
--- a/drivers/net/igb/e1000_hw.h
+++ b/drivers/net/igb/e1000_hw.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index e5d23e0..5c249e2 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_mac.h b/drivers/net/igb/e1000_mac.h
index 4ef40d5..91461de 100644
--- a/drivers/net/igb/e1000_mac.h
+++ b/drivers/net/igb/e1000_mac.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_nvm.c b/drivers/net/igb/e1000_nvm.c
index 3379864..a88bfe2 100644
--- a/drivers/net/igb/e1000_nvm.c
+++ b/drivers/net/igb/e1000_nvm.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_phy.c b/drivers/net/igb/e1000_phy.c
index d73ea71..ff0050e 100644
--- a/drivers/net/igb/e1000_phy.c
+++ b/drivers/net/igb/e1000_phy.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_phy.h b/drivers/net/igb/e1000_phy.h
index 7499e5d..3228a86 100644
--- a/drivers/net/igb/e1000_phy.h
+++ b/drivers/net/igb/e1000_phy.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/e1000_regs.h b/drivers/net/igb/e1000_regs.h
index bdf5d83..5038b73 100644
--- a/drivers/net/igb/e1000_regs.h
+++ b/drivers/net/igb/e1000_regs.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 0519ae4..e507449 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index de09430..bd050b1 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 6713840..8bc1ffa 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1,7 +1,7 @@
/*******************************************************************************
Intel(R) Gigabit Ethernet Linux driver
- Copyright(c) 2007 Intel Corporation.
+ Copyright(c) 2007-2009 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
@@ -48,12 +48,12 @@
#endif
#include "igb.h"
-#define DRV_VERSION "1.2.45-k2"
+#define DRV_VERSION "1.3.16-k2"
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) 2008 Intel Corporation.";
+static const char igb_copyright[] = "Copyright (c) 2007-2009 Intel Corporation.";
static const struct e1000_info *igb_info_tbl[] = {
[board_82575] = &e1000_82575_info,
^ permalink raw reply related [flat|nested] 27+ messages in thread* [net-next PATCH 26/26] igb: remove dead code in transmit routine
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (23 preceding siblings ...)
2009-02-07 9:23 ` [net-next PATCH 25/26] igb: update version number and copyright dates Jeff Kirsher
@ 2009-02-07 9:23 ` Jeff Kirsher
2009-02-07 10:46 ` [net-next PATCH 01/26] igb: optomize/refactor receive path David Miller
25 siblings, 0 replies; 27+ messages in thread
From: Jeff Kirsher @ 2009-02-07 9:23 UTC (permalink / raw)
To: davem; +Cc: netdev, jeff, gospo, Jesse Brandeburg, Jeff Kirsher
From: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 3 ---
1 files changed, 0 insertions(+), 3 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 8bc1ffa..f8c2919 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2951,12 +2951,9 @@ static int igb_xmit_frame_ring_adv(struct sk_buff *skb,
struct igb_adapter *adapter = netdev_priv(netdev);
unsigned int first;
unsigned int tx_flags = 0;
- unsigned int len;
u8 hdr_len = 0;
int tso = 0;
- len = skb_headlen(skb);
-
if (test_bit(__IGB_DOWN, &adapter->state)) {
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
^ permalink raw reply related [flat|nested] 27+ messages in thread* Re: [net-next PATCH 01/26] igb: optomize/refactor receive path
2009-02-07 9:15 [net-next PATCH 01/26] igb: optomize/refactor receive path Jeff Kirsher
` (24 preceding siblings ...)
2009-02-07 9:23 ` [net-next PATCH 26/26] igb: remove dead code in transmit routine Jeff Kirsher
@ 2009-02-07 10:46 ` David Miller
25 siblings, 0 replies; 27+ messages in thread
From: David Miller @ 2009-02-07 10:46 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, jeff, gospo, alexander.h.duyck
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Sat, 07 Feb 2009 01:15:04 -0800
> While cleaning up the skb_over panic with small frames I found there was
> room for improvement in the ordering of operations within the rx receive
> flow. These changes will place the prefetch for the next descriptor to a
> point earlier in the rx path.
>
> Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
"Optomize"? Is that something like labotomize? :-)
Anyways, I fixed up that typo and applied all 26 patches
to net-next-2.6
Thanks!
^ permalink raw reply [flat|nested] 27+ messages in thread