* [net-next 01/12] e1000e: additional error handling on PHY register accesses
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 02/12] e1000e: slow performance between two 82579 connected via 10Mbit hub Jeff Kirsher
` (11 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
PHY reads/writes via the MDIC register could potentially return results
from a previous PHY register access. If that happens, the offset in the
returned results will be that of the previous access and if that is
different from the expected offset, log a debug message and error out.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/defines.h | 1 +
drivers/net/ethernet/intel/e1000e/phy.c | 12 ++++++++++++
2 files changed, 13 insertions(+)
diff --git a/drivers/net/ethernet/intel/e1000e/defines.h b/drivers/net/ethernet/intel/e1000e/defines.h
index b7c664f..a6aeb19 100644
--- a/drivers/net/ethernet/intel/e1000e/defines.h
+++ b/drivers/net/ethernet/intel/e1000e/defines.h
@@ -788,6 +788,7 @@
GG82563_REG(194, 18) /* Inband Control */
/* MDI Control */
+#define E1000_MDIC_REG_MASK 0x001F0000
#define E1000_MDIC_REG_SHIFT 16
#define E1000_MDIC_PHY_SHIFT 21
#define E1000_MDIC_OP_WRITE 0x04000000
diff --git a/drivers/net/ethernet/intel/e1000e/phy.c b/drivers/net/ethernet/intel/e1000e/phy.c
index 60dbf02..cbb310f 100644
--- a/drivers/net/ethernet/intel/e1000e/phy.c
+++ b/drivers/net/ethernet/intel/e1000e/phy.c
@@ -178,6 +178,12 @@ s32 e1000e_read_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 *data)
e_dbg("MDI Error\n");
return -E1000_ERR_PHY;
}
+ if (((mdic & E1000_MDIC_REG_MASK) >> E1000_MDIC_REG_SHIFT) != offset) {
+ e_dbg("MDI Read offset error - requested %d, returned %d\n",
+ offset,
+ (mdic & E1000_MDIC_REG_MASK) >> E1000_MDIC_REG_SHIFT);
+ return -E1000_ERR_PHY;
+ }
*data = (u16)mdic;
/* Allow some time after each MDIC transaction to avoid
@@ -236,6 +242,12 @@ s32 e1000e_write_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 data)
e_dbg("MDI Error\n");
return -E1000_ERR_PHY;
}
+ if (((mdic & E1000_MDIC_REG_MASK) >> E1000_MDIC_REG_SHIFT) != offset) {
+ e_dbg("MDI Write offset error - requested %d, returned %d\n",
+ offset,
+ (mdic & E1000_MDIC_REG_MASK) >> E1000_MDIC_REG_SHIFT);
+ return -E1000_ERR_PHY;
+ }
/* Allow some time after each MDIC transaction to avoid
* reading duplicate data in the next MDIC transaction.
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 02/12] e1000e: slow performance between two 82579 connected via 10Mbit hub
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
2013-03-28 9:00 ` [net-next 01/12] e1000e: additional error handling on PHY register accesses Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 03/12] e1000e: fix LED blink logic for designs with LEDs driven by cathode Jeff Kirsher
` (10 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Two 82579 LOMs connected via a 10Mb hub experience extraordinarily low
performance. This is because 82579 is excessively aggressive on transmit
at 10Mb half-duplex and will not provide sufficient time for the link
partner to transmit. When the link partner is also 82579, the result is a
lot of collisions (and corresponding re-transmits) that cause the poor
performance. To work-around this issue, significantly increase the IPG in
the MAC to allow enough gap for the link partner to transmit and reduce the
Rx latency in the analog PHY to 0 to reduce the number of collisions.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/defines.h | 1 +
drivers/net/ethernet/intel/e1000e/ich8lan.c | 28 ++++++++++++++++++++++++++++
drivers/net/ethernet/intel/e1000e/ich8lan.h | 1 +
3 files changed, 30 insertions(+)
diff --git a/drivers/net/ethernet/intel/e1000e/defines.h b/drivers/net/ethernet/intel/e1000e/defines.h
index a6aeb19..351c94a 100644
--- a/drivers/net/ethernet/intel/e1000e/defines.h
+++ b/drivers/net/ethernet/intel/e1000e/defines.h
@@ -236,6 +236,7 @@
#define E1000_STATUS_FUNC_SHIFT 2
#define E1000_STATUS_FUNC_1 0x00000004 /* Function 1 */
#define E1000_STATUS_TXOFF 0x00000010 /* transmission paused */
+#define E1000_STATUS_SPEED_MASK 0x000000C0
#define E1000_STATUS_SPEED_10 0x00000000 /* Speed 10Mb/s */
#define E1000_STATUS_SPEED_100 0x00000040 /* Speed 100Mb/s */
#define E1000_STATUS_SPEED_1000 0x00000080 /* Speed 1000Mb/s */
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c
index 1cdec5f..4f2f0f6 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.c
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c
@@ -871,6 +871,34 @@ static s32 e1000_check_for_copper_link_ich8lan(struct e1000_hw *hw)
return ret_val;
}
+ /* When connected at 10Mbps half-duplex, 82579 parts are excessively
+ * aggressive resulting in many collisions. To avoid this, increase
+ * the IPG and reduce Rx latency in the PHY.
+ */
+ if ((hw->mac.type == e1000_pch2lan) && link) {
+ u32 reg;
+ reg = er32(STATUS);
+ if (!(reg & (E1000_STATUS_FD | E1000_STATUS_SPEED_MASK))) {
+ reg = er32(TIPG);
+ reg &= ~E1000_TIPG_IPGT_MASK;
+ reg |= 0xFF;
+ ew32(TIPG, reg);
+
+ /* Reduce Rx latency in analog PHY */
+ ret_val = hw->phy.ops.acquire(hw);
+ if (ret_val)
+ return ret_val;
+
+ ret_val =
+ e1000_write_emi_reg_locked(hw, I82579_RX_CONFIG, 0);
+
+ hw->phy.ops.release(hw);
+
+ if (ret_val)
+ return ret_val;
+ }
+ }
+
/* Work-around I218 hang issue */
if ((hw->adapter->pdev->device == E1000_DEV_ID_PCH_LPTLP_I218_LM) ||
(hw->adapter->pdev->device == E1000_DEV_ID_PCH_LPTLP_I218_V)) {
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.h b/drivers/net/ethernet/intel/e1000e/ich8lan.h
index 8bf4655..00ba6c9 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.h
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.h
@@ -211,6 +211,7 @@
#define I82579_MSE_THRESHOLD 0x084F /* 82579 Mean Square Error Threshold */
#define I82577_MSE_THRESHOLD 0x0887 /* 82577 Mean Square Error Threshold */
#define I82579_MSE_LINK_DOWN 0x2411 /* MSE count before dropping link */
+#define I82579_RX_CONFIG 0x3412 /* Receive configuration */
#define I82579_EEE_PCS_STATUS 0x182D /* IEEE MMD Register 3.1 >> 8 */
#define I82579_EEE_CAPABILITY 0x0410 /* IEEE MMD Register 3.20 */
#define I82579_EEE_ADVERTISEMENT 0x040E /* IEEE MMD Register 7.60 */
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 03/12] e1000e: fix LED blink logic for designs with LEDs driven by cathode
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
2013-03-28 9:00 ` [net-next 01/12] e1000e: additional error handling on PHY register accesses Jeff Kirsher
2013-03-28 9:00 ` [net-next 02/12] e1000e: slow performance between two 82579 connected via 10Mbit hub Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 04/12] e1000e: long access timeouts when I217/I218 MAC and PHY are out of sync Jeff Kirsher
` (9 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
When the LEDs are driven by cathode, the bit logic is reversed. Use the
LED Invert bit to invert the logic. Cleanup use of a magic number and
change the for loop increment to reduce the number of shifts.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/mac.c | 27 ++++++++++++++++++++-------
1 file changed, 20 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/mac.c b/drivers/net/ethernet/intel/e1000e/mac.c
index b25cc43..2480c10 100644
--- a/drivers/net/ethernet/intel/e1000e/mac.c
+++ b/drivers/net/ethernet/intel/e1000e/mac.c
@@ -1600,15 +1600,28 @@ s32 e1000e_blink_led_generic(struct e1000_hw *hw)
ledctl_blink = E1000_LEDCTL_LED0_BLINK |
(E1000_LEDCTL_MODE_LED_ON << E1000_LEDCTL_LED0_MODE_SHIFT);
} else {
- /* set the blink bit for each LED that's "on" (0x0E)
- * in ledctl_mode2
+ /* Set the blink bit for each LED that's "on" (0x0E)
+ * (or "off" if inverted) in ledctl_mode2. The blink
+ * logic in hardware only works when mode is set to "on"
+ * so it must be changed accordingly when the mode is
+ * "off" and inverted.
*/
ledctl_blink = hw->mac.ledctl_mode2;
- for (i = 0; i < 4; i++)
- if (((hw->mac.ledctl_mode2 >> (i * 8)) & 0xFF) ==
- E1000_LEDCTL_MODE_LED_ON)
- ledctl_blink |= (E1000_LEDCTL_LED0_BLINK <<
- (i * 8));
+ for (i = 0; i < 32; i += 8) {
+ u32 mode = (hw->mac.ledctl_mode2 >> i) &
+ E1000_LEDCTL_LED0_MODE_MASK;
+ u32 led_default = hw->mac.ledctl_default >> i;
+
+ if ((!(led_default & E1000_LEDCTL_LED0_IVRT) &&
+ (mode == E1000_LEDCTL_MODE_LED_ON)) ||
+ ((led_default & E1000_LEDCTL_LED0_IVRT) &&
+ (mode == E1000_LEDCTL_MODE_LED_OFF))) {
+ ledctl_blink &=
+ ~(E1000_LEDCTL_LED0_MODE_MASK << i);
+ ledctl_blink |= (E1000_LEDCTL_LED0_BLINK |
+ E1000_LEDCTL_MODE_LED_ON) << i;
+ }
+ }
}
ew32(LEDCTL, ledctl_blink);
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 04/12] e1000e: long access timeouts when I217/I218 MAC and PHY are out of sync
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (2 preceding siblings ...)
2013-03-28 9:00 ` [net-next 03/12] e1000e: fix LED blink logic for designs with LEDs driven by cathode Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 05/12] e1000e: EEE capability advertisement not set/disabled as required Jeff Kirsher
` (8 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
When the MAC and PHY are in two different modes (different power levels
and interconnect speeds), it could take a long time before a PHY register
access timed out using the existing MAC-PHY interconnect configuration
coded into the driver for ICH- and PCH-based LOMs. Introduce an I217/I218-
specific .setup_physical_interface operation which does not override the
interconnect configuration in the NVM.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/ich8lan.c | 29 ++++++++++++++++++++++++++++-
1 file changed, 28 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c
index 4f2f0f6..3b1c1a7 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.c
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c
@@ -142,6 +142,7 @@ static void e1000_rar_set_pch2lan(struct e1000_hw *hw, u8 *addr, u32 index);
static void e1000_rar_set_pch_lpt(struct e1000_hw *hw, u8 *addr, u32 index);
static s32 e1000_k1_workaround_lv(struct e1000_hw *hw);
static void e1000_gate_hw_phy_config_ich8lan(struct e1000_hw *hw, bool gate);
+static s32 e1000_setup_copper_link_pch_lpt(struct e1000_hw *hw);
static inline u16 __er16flash(struct e1000_hw *hw, unsigned long reg)
{
@@ -636,6 +637,8 @@ static s32 e1000_init_mac_params_ich8lan(struct e1000_hw *hw)
if (mac->type == e1000_pch_lpt) {
mac->rar_entry_count = E1000_PCH_LPT_RAR_ENTRIES;
mac->ops.rar_set = e1000_rar_set_pch_lpt;
+ mac->ops.setup_physical_interface =
+ e1000_setup_copper_link_pch_lpt;
}
/* Enable PCS Lock-loss workaround for ICH8 */
@@ -3788,7 +3791,6 @@ static s32 e1000_setup_copper_link_ich8lan(struct e1000_hw *hw)
break;
case e1000_phy_82577:
case e1000_phy_82579:
- case e1000_phy_i217:
ret_val = e1000_copper_link_setup_82577(hw);
if (ret_val)
return ret_val;
@@ -3824,6 +3826,31 @@ static s32 e1000_setup_copper_link_ich8lan(struct e1000_hw *hw)
}
/**
+ * e1000_setup_copper_link_pch_lpt - Configure MAC/PHY interface
+ * @hw: pointer to the HW structure
+ *
+ * Calls the PHY specific link setup function and then calls the
+ * generic setup_copper_link to finish configuring the link for
+ * Lynxpoint PCH devices
+ **/
+static s32 e1000_setup_copper_link_pch_lpt(struct e1000_hw *hw)
+{
+ u32 ctrl;
+ s32 ret_val;
+
+ ctrl = er32(CTRL);
+ ctrl |= E1000_CTRL_SLU;
+ ctrl &= ~(E1000_CTRL_FRCSPD | E1000_CTRL_FRCDPX);
+ ew32(CTRL, ctrl);
+
+ ret_val = e1000_copper_link_setup_82577(hw);
+ if (ret_val)
+ return ret_val;
+
+ return e1000e_setup_copper_link(hw);
+}
+
+/**
* e1000_get_link_up_info_ich8lan - Get current link speed and duplex
* @hw: pointer to the HW structure
* @speed: pointer to store current link speed
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 05/12] e1000e: EEE capability advertisement not set/disabled as required
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (3 preceding siblings ...)
2013-03-28 9:00 ` [net-next 04/12] e1000e: long access timeouts when I217/I218 MAC and PHY are out of sync Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 06/12] e1000e: enable EEE by default Jeff Kirsher
` (7 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Devices supported by the driver which support EEE (currently 82579, I217
and I218) are advertising EEE capabilities during auto-negotiation even
when EEE has been disabled. In addition to not acting as expected, this
also caused the EEE status reported by 'ethtool --show-eee' to be wrong
when two of these devices are connected back-to-back and EEE is disabled
on one. In addition to fixing this issue, the ability for the user to
specify which speeds (100 or 1000 full-duplex) to advertise EEE support
has been added.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/e1000.h | 3 ++
drivers/net/ethernet/intel/e1000e/ethtool.c | 63 +++++++++--------------------
drivers/net/ethernet/intel/e1000e/ich8lan.c | 55 ++++++++++++-------------
drivers/net/ethernet/intel/e1000e/ich8lan.h | 3 +-
drivers/net/ethernet/intel/e1000e/netdev.c | 36 +++++++++++++++++
5 files changed, 88 insertions(+), 72 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/e1000.h b/drivers/net/ethernet/intel/e1000e/e1000.h
index e371f77..82f1c84 100644
--- a/drivers/net/ethernet/intel/e1000e/e1000.h
+++ b/drivers/net/ethernet/intel/e1000e/e1000.h
@@ -46,6 +46,7 @@
#include <linux/ptp_clock_kernel.h>
#include <linux/ptp_classify.h>
#include <linux/mii.h>
+#include <linux/mdio.h>
#include "hw.h"
struct e1000_info;
@@ -350,6 +351,8 @@ struct e1000_adapter {
struct timecounter tc;
struct ptp_clock *ptp_clock;
struct ptp_clock_info ptp_clock_info;
+
+ u16 eee_advert;
};
struct e1000_info {
diff --git a/drivers/net/ethernet/intel/e1000e/ethtool.c b/drivers/net/ethernet/intel/e1000e/ethtool.c
index e835e7b..7c8ca65 100644
--- a/drivers/net/ethernet/intel/e1000e/ethtool.c
+++ b/drivers/net/ethernet/intel/e1000e/ethtool.c
@@ -35,7 +35,6 @@
#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/vmalloc.h>
-#include <linux/mdio.h>
#include <linux/pm_runtime.h>
#include "e1000.h"
@@ -2076,23 +2075,20 @@ static int e1000e_get_eee(struct net_device *netdev, struct ethtool_eee *edata)
{
struct e1000_adapter *adapter = netdev_priv(netdev);
struct e1000_hw *hw = &adapter->hw;
- u16 cap_addr, adv_addr, lpa_addr, pcs_stat_addr, phy_data, lpi_ctrl;
- u32 status, ret_val;
+ u16 cap_addr, lpa_addr, pcs_stat_addr, phy_data;
+ u32 ret_val;
- if (!(adapter->flags & FLAG_IS_ICH) ||
- !(adapter->flags2 & FLAG2_HAS_EEE))
+ if (!(adapter->flags2 & FLAG2_HAS_EEE))
return -EOPNOTSUPP;
switch (hw->phy.type) {
case e1000_phy_82579:
cap_addr = I82579_EEE_CAPABILITY;
- adv_addr = I82579_EEE_ADVERTISEMENT;
lpa_addr = I82579_EEE_LP_ABILITY;
pcs_stat_addr = I82579_EEE_PCS_STATUS;
break;
case e1000_phy_i217:
cap_addr = I217_EEE_CAPABILITY;
- adv_addr = I217_EEE_ADVERTISEMENT;
lpa_addr = I217_EEE_LP_ABILITY;
pcs_stat_addr = I217_EEE_PCS_STATUS;
break;
@@ -2111,10 +2107,7 @@ static int e1000e_get_eee(struct net_device *netdev, struct ethtool_eee *edata)
edata->supported = mmd_eee_cap_to_ethtool_sup_t(phy_data);
/* EEE Advertised */
- ret_val = e1000_read_emi_reg_locked(hw, adv_addr, &phy_data);
- if (ret_val)
- goto release;
- edata->advertised = mmd_eee_adv_to_ethtool_adv_t(phy_data);
+ edata->advertised = mmd_eee_adv_to_ethtool_adv_t(adapter->eee_advert);
/* EEE Link Partner Advertised */
ret_val = e1000_read_emi_reg_locked(hw, lpa_addr, &phy_data);
@@ -2132,25 +2125,11 @@ release:
if (ret_val)
return -ENODATA;
- e1e_rphy(hw, I82579_LPI_CTRL, &lpi_ctrl);
- status = er32(STATUS);
-
/* Result of the EEE auto negotiation - there is no register that
* has the status of the EEE negotiation so do a best-guess based
- * on whether both Tx and Rx LPI indications have been received or
- * base it on the link speed, the EEE advertised speeds on both ends
- * and the speeds on which EEE is enabled locally.
+ * on whether Tx or Rx LPI indications have been received.
*/
- if (((phy_data & E1000_EEE_TX_LPI_RCVD) &&
- (phy_data & E1000_EEE_RX_LPI_RCVD)) ||
- ((status & E1000_STATUS_SPEED_100) &&
- (edata->advertised & ADVERTISED_100baseT_Full) &&
- (edata->lp_advertised & ADVERTISED_100baseT_Full) &&
- (lpi_ctrl & I82579_LPI_CTRL_100_ENABLE)) ||
- ((status & E1000_STATUS_SPEED_1000) &&
- (edata->advertised & ADVERTISED_1000baseT_Full) &&
- (edata->lp_advertised & ADVERTISED_1000baseT_Full) &&
- (lpi_ctrl & I82579_LPI_CTRL_1000_ENABLE)))
+ if (phy_data & (E1000_EEE_TX_LPI_RCVD | E1000_EEE_RX_LPI_RCVD))
edata->eee_active = true;
edata->eee_enabled = !hw->dev_spec.ich8lan.eee_disable;
@@ -2167,19 +2146,10 @@ static int e1000e_set_eee(struct net_device *netdev, struct ethtool_eee *edata)
struct ethtool_eee eee_curr;
s32 ret_val;
- if (!(adapter->flags & FLAG_IS_ICH) ||
- !(adapter->flags2 & FLAG2_HAS_EEE))
- return -EOPNOTSUPP;
-
ret_val = e1000e_get_eee(netdev, &eee_curr);
if (ret_val)
return ret_val;
- if (eee_curr.advertised != edata->advertised) {
- e_err("Setting EEE advertisement is not supported\n");
- return -EINVAL;
- }
-
if (eee_curr.tx_lpi_enabled != edata->tx_lpi_enabled) {
e_err("Setting EEE tx-lpi is not supported\n");
return -EINVAL;
@@ -2190,16 +2160,21 @@ static int e1000e_set_eee(struct net_device *netdev, struct ethtool_eee *edata)
return -EINVAL;
}
- if (hw->dev_spec.ich8lan.eee_disable != !edata->eee_enabled) {
- hw->dev_spec.ich8lan.eee_disable = !edata->eee_enabled;
-
- /* reset the link */
- if (netif_running(netdev))
- e1000e_reinit_locked(adapter);
- else
- e1000e_reset(adapter);
+ if (edata->advertised & ~(ADVERTISE_100_FULL | ADVERTISE_1000_FULL)) {
+ e_err("EEE advertisement supports only 100TX and/or 1000T full-duplex\n");
+ return -EINVAL;
}
+ adapter->eee_advert = ethtool_adv_to_mmd_eee_adv_t(edata->advertised);
+
+ hw->dev_spec.ich8lan.eee_disable = !edata->eee_enabled;
+
+ /* reset the link */
+ if (netif_running(netdev))
+ e1000e_reinit_locked(adapter);
+ else
+ e1000e_reset(adapter);
+
return 0;
}
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c
index 3b1c1a7..174507c 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.c
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c
@@ -695,7 +695,7 @@ s32 e1000_read_emi_reg_locked(struct e1000_hw *hw, u16 addr, u16 *data)
*
* Assumes the SW/FW/HW Semaphore is already acquired.
**/
-static s32 e1000_write_emi_reg_locked(struct e1000_hw *hw, u16 addr, u16 data)
+s32 e1000_write_emi_reg_locked(struct e1000_hw *hw, u16 addr, u16 data)
{
return __e1000_access_emi_reg_locked(hw, addr, &data, false);
}
@@ -712,11 +712,22 @@ static s32 e1000_set_eee_pchlan(struct e1000_hw *hw)
{
struct e1000_dev_spec_ich8lan *dev_spec = &hw->dev_spec.ich8lan;
s32 ret_val;
- u16 lpi_ctrl;
+ u16 lpa, pcs_status, adv, adv_addr, lpi_ctrl, data;
- if ((hw->phy.type != e1000_phy_82579) &&
- (hw->phy.type != e1000_phy_i217))
+ switch (hw->phy.type) {
+ case e1000_phy_82579:
+ lpa = I82579_EEE_LP_ABILITY;
+ pcs_status = I82579_EEE_PCS_STATUS;
+ adv_addr = I82579_EEE_ADVERTISEMENT;
+ break;
+ case e1000_phy_i217:
+ lpa = I217_EEE_LP_ABILITY;
+ pcs_status = I217_EEE_PCS_STATUS;
+ adv_addr = I217_EEE_ADVERTISEMENT;
+ break;
+ default:
return 0;
+ }
ret_val = hw->phy.ops.acquire(hw);
if (ret_val)
@@ -731,34 +742,24 @@ static s32 e1000_set_eee_pchlan(struct e1000_hw *hw)
/* Enable EEE if not disabled by user */
if (!dev_spec->eee_disable) {
- u16 lpa, pcs_status, data;
-
/* Save off link partner's EEE ability */
- switch (hw->phy.type) {
- case e1000_phy_82579:
- lpa = I82579_EEE_LP_ABILITY;
- pcs_status = I82579_EEE_PCS_STATUS;
- break;
- case e1000_phy_i217:
- lpa = I217_EEE_LP_ABILITY;
- pcs_status = I217_EEE_PCS_STATUS;
- break;
- default:
- ret_val = -E1000_ERR_PHY;
- goto release;
- }
ret_val = e1000_read_emi_reg_locked(hw, lpa,
&dev_spec->eee_lp_ability);
if (ret_val)
goto release;
+ /* Read EEE advertisement */
+ ret_val = e1000_read_emi_reg_locked(hw, adv_addr, &adv);
+ if (ret_val)
+ goto release;
+
/* Enable EEE only for speeds in which the link partner is
- * EEE capable.
+ * EEE capable and for which we advertise EEE.
*/
- if (dev_spec->eee_lp_ability & I82579_EEE_1000_SUPPORTED)
+ if (adv & dev_spec->eee_lp_ability & I82579_EEE_1000_SUPPORTED)
lpi_ctrl |= I82579_LPI_CTRL_1000_ENABLE;
- if (dev_spec->eee_lp_ability & I82579_EEE_100_SUPPORTED) {
+ if (adv & dev_spec->eee_lp_ability & I82579_EEE_100_SUPPORTED) {
e1e_rphy_locked(hw, MII_LPA, &data);
if (data & LPA_100FULL)
lpi_ctrl |= I82579_LPI_CTRL_100_ENABLE;
@@ -770,13 +771,13 @@ static s32 e1000_set_eee_pchlan(struct e1000_hw *hw)
dev_spec->eee_lp_ability &=
~I82579_EEE_100_SUPPORTED;
}
-
- /* R/Clr IEEE MMD 3.1 bits 11:10 - Tx/Rx LPI Received */
- ret_val = e1000_read_emi_reg_locked(hw, pcs_status, &data);
- if (ret_val)
- goto release;
}
+ /* R/Clr IEEE MMD 3.1 bits 11:10 - Tx/Rx LPI Received */
+ ret_val = e1000_read_emi_reg_locked(hw, pcs_status, &data);
+ if (ret_val)
+ goto release;
+
ret_val = e1e_wphy_locked(hw, I82579_LPI_CTRL, lpi_ctrl);
release:
hw->phy.ops.release(hw);
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.h b/drivers/net/ethernet/intel/e1000e/ich8lan.h
index 00ba6c9..21d21b9 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.h
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.h
@@ -212,7 +212,7 @@
#define I82577_MSE_THRESHOLD 0x0887 /* 82577 Mean Square Error Threshold */
#define I82579_MSE_LINK_DOWN 0x2411 /* MSE count before dropping link */
#define I82579_RX_CONFIG 0x3412 /* Receive configuration */
-#define I82579_EEE_PCS_STATUS 0x182D /* IEEE MMD Register 3.1 >> 8 */
+#define I82579_EEE_PCS_STATUS 0x182E /* IEEE MMD Register 3.1 >> 8 */
#define I82579_EEE_CAPABILITY 0x0410 /* IEEE MMD Register 3.20 */
#define I82579_EEE_ADVERTISEMENT 0x040E /* IEEE MMD Register 7.60 */
#define I82579_EEE_LP_ABILITY 0x040F /* IEEE MMD Register 7.61 */
@@ -268,4 +268,5 @@ s32 e1000_configure_k1_ich8lan(struct e1000_hw *hw, bool k1_enable);
void e1000_copy_rx_addrs_to_phy_ich8lan(struct e1000_hw *hw);
s32 e1000_lv_jumbo_workaround_ich8lan(struct e1000_hw *hw, bool enable);
s32 e1000_read_emi_reg_locked(struct e1000_hw *hw, u16 addr, u16 *data);
+s32 e1000_write_emi_reg_locked(struct e1000_hw *hw, u16 addr, u16 data);
#endif /* _E1000E_ICH8LAN_H_ */
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index 142ca39..0459fe3 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -3875,6 +3875,38 @@ void e1000e_reset(struct e1000_adapter *adapter)
/* initialize systim and reset the ns time counter */
e1000e_config_hwtstamp(adapter);
+ /* Set EEE advertisement as appropriate */
+ if (adapter->flags2 & FLAG2_HAS_EEE) {
+ s32 ret_val;
+ u16 adv_addr;
+
+ switch (hw->phy.type) {
+ case e1000_phy_82579:
+ adv_addr = I82579_EEE_ADVERTISEMENT;
+ break;
+ case e1000_phy_i217:
+ adv_addr = I217_EEE_ADVERTISEMENT;
+ break;
+ default:
+ dev_err(&adapter->pdev->dev,
+ "Invalid PHY type setting EEE advertisement\n");
+ return;
+ }
+
+ ret_val = hw->phy.ops.acquire(hw);
+ if (ret_val) {
+ dev_err(&adapter->pdev->dev,
+ "EEE advertisement - unable to acquire PHY\n");
+ return;
+ }
+
+ e1000_write_emi_reg_locked(hw, adv_addr,
+ hw->dev_spec.ich8lan.eee_disable ?
+ 0 : adapter->eee_advert);
+
+ hw->phy.ops.release(hw);
+ }
+
if (!netif_running(adapter->netdev) &&
!test_bit(__E1000_TESTING, &adapter->state)) {
e1000_power_down_phy(adapter);
@@ -6540,6 +6572,10 @@ static int e1000_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
goto err_flashmap;
}
+ /* Set default EEE advertisement */
+ if (adapter->flags2 & FLAG2_HAS_EEE)
+ adapter->eee_advert = MDIO_EEE_100TX | MDIO_EEE_1000T;
+
/* construct the net_device struct */
netdev->netdev_ops = &e1000e_netdev_ops;
e1000e_set_ethtool_ops(netdev);
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 06/12] e1000e: enable EEE by default
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (4 preceding siblings ...)
2013-03-28 9:00 ` [net-next 05/12] e1000e: EEE capability advertisement not set/disabled as required Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 07/12] e1000e: add support for LTR on I217/I218 Jeff Kirsher
` (6 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Now that IEEE802.3az-2010 Energy Efficient Ethernet has been approved as
standard (September 2010) and the driver can enable and disable it via
ethtool, enable the feature by default on parts which support it.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/ich8lan.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c
index 174507c..56c4935 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.c
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c
@@ -1034,10 +1034,6 @@ static s32 e1000_get_variants_ich8lan(struct e1000_adapter *adapter)
(er32(FWSM) & E1000_ICH_FWSM_FW_VALID))
adapter->flags2 |= FLAG2_PCIM2PCI_ARBITER_WA;
- /* Disable EEE by default until IEEE802.3az spec is finalized */
- if (adapter->flags2 & FLAG2_HAS_EEE)
- adapter->hw.dev_spec.ich8lan.eee_disable = true;
-
return 0;
}
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 07/12] e1000e: add support for LTR on I217/I218
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (5 preceding siblings ...)
2013-03-28 9:00 ` [net-next 06/12] e1000e: enable EEE by default Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 08/12] e1000e: cleanup unused defines Jeff Kirsher
` (5 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Set the Latency Tolerance Reporting (LTR) values for the "PCIe-like"
GbE MAC in the Lynx Point PCH based on Rx buffer size and link speed
when link is up (which must not exceed the maximum latency supported
by the platform), otherwise specify there is no LTR requirement.
Unlike true-PCIe devices which set the LTR maximum snoop/no-snoop
latencies in the LTR Extended Capability Structure in the PCIe Extended
Capability register set, on this device LTR is set by writing the
equivalent snoop/no-snoop latencies in the LTRV register in the MAC and
set the SEND bit to send an Intel On-chip System Fabric sideband (IOSF-SB)
message to the PMC.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/ich8lan.c | 97 +++++++++++++++++++++++++++++
1 file changed, 97 insertions(+)
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c
index 56c4935..ad9d8f2 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.c
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c
@@ -839,6 +839,94 @@ release:
}
/**
+ * e1000_platform_pm_pch_lpt - Set platform power management values
+ * @hw: pointer to the HW structure
+ * @link: bool indicating link status
+ *
+ * Set the Latency Tolerance Reporting (LTR) values for the "PCIe-like"
+ * GbE MAC in the Lynx Point PCH based on Rx buffer size and link speed
+ * when link is up (which must not exceed the maximum latency supported
+ * by the platform), otherwise specify there is no LTR requirement.
+ * Unlike true-PCIe devices which set the LTR maximum snoop/no-snoop
+ * latencies in the LTR Extended Capability Structure in the PCIe Extended
+ * Capability register set, on this device LTR is set by writing the
+ * equivalent snoop/no-snoop latencies in the LTRV register in the MAC and
+ * set the SEND bit to send an Intel On-chip System Fabric sideband (IOSF-SB)
+ * message to the PMC.
+ **/
+static s32 e1000_platform_pm_pch_lpt(struct e1000_hw *hw, bool link)
+{
+ u32 reg = link << (E1000_LTRV_REQ_SHIFT + E1000_LTRV_NOSNOOP_SHIFT) |
+ link << E1000_LTRV_REQ_SHIFT | E1000_LTRV_SEND;
+ u16 lat_enc = 0; /* latency encoded */
+
+ if (link) {
+ u16 speed, duplex, scale = 0;
+ u16 max_snoop, max_nosnoop;
+ u16 max_ltr_enc; /* max LTR latency encoded */
+ s64 lat_ns; /* latency (ns) */
+ s64 value;
+ u32 rxa;
+
+ if (!hw->adapter->max_frame_size) {
+ e_dbg("max_frame_size not set.\n");
+ return -E1000_ERR_CONFIG;
+ }
+
+ hw->mac.ops.get_link_up_info(hw, &speed, &duplex);
+ if (!speed) {
+ e_dbg("Speed not set.\n");
+ return -E1000_ERR_CONFIG;
+ }
+
+ /* Rx Packet Buffer Allocation size (KB) */
+ rxa = er32(PBA) & E1000_PBA_RXA_MASK;
+
+ /* Determine the maximum latency tolerated by the device.
+ *
+ * Per the PCIe spec, the tolerated latencies are encoded as
+ * a 3-bit encoded scale (only 0-5 are valid) multiplied by
+ * a 10-bit value (0-1023) to provide a range from 1 ns to
+ * 2^25*(2^10-1) ns. The scale is encoded as 0=2^0ns,
+ * 1=2^5ns, 2=2^10ns,...5=2^25ns.
+ */
+ lat_ns = ((s64)rxa * 1024 -
+ (2 * (s64)hw->adapter->max_frame_size)) * 8 * 1000;
+ if (lat_ns < 0)
+ lat_ns = 0;
+ else
+ do_div(lat_ns, speed);
+
+ value = lat_ns;
+ while (value > PCI_LTR_VALUE_MASK) {
+ scale++;
+ value = DIV_ROUND_UP(value, (1 << 5));
+ }
+ if (scale > E1000_LTRV_SCALE_MAX) {
+ e_dbg("Invalid LTR latency scale %d\n", scale);
+ return -E1000_ERR_CONFIG;
+ }
+ lat_enc = (u16)((scale << PCI_LTR_SCALE_SHIFT) | value);
+
+ /* Determine the maximum latency tolerated by the platform */
+ pci_read_config_word(hw->adapter->pdev, E1000_PCI_LTR_CAP_LPT,
+ &max_snoop);
+ pci_read_config_word(hw->adapter->pdev,
+ E1000_PCI_LTR_CAP_LPT + 2, &max_nosnoop);
+ max_ltr_enc = max_t(u16, max_snoop, max_nosnoop);
+
+ if (lat_enc > max_ltr_enc)
+ lat_enc = max_ltr_enc;
+ }
+
+ /* Set Snoop and No-Snoop latencies the same */
+ reg |= lat_enc | (lat_enc << E1000_LTRV_NOSNOOP_SHIFT);
+ ew32(LTRV, reg);
+
+ return 0;
+}
+
+/**
* e1000_check_for_copper_link_ich8lan - Check for link (Copper)
* @hw: pointer to the HW structure
*
@@ -911,6 +999,15 @@ static s32 e1000_check_for_copper_link_ich8lan(struct e1000_hw *hw)
return ret_val;
}
+ if (hw->mac.type == e1000_pch_lpt) {
+ /* Set platform power management values for
+ * Latency Tolerance Reporting (LTR)
+ */
+ ret_val = e1000_platform_pm_pch_lpt(hw, link);
+ if (ret_val)
+ return ret_val;
+ }
+
/* Clear link partner's EEE ability */
hw->dev_spec.ich8lan.eee_lp_ability = 0;
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 08/12] e1000e: cleanup unused defines
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (6 preceding siblings ...)
2013-03-28 9:00 ` [net-next 07/12] e1000e: add support for LTR on I217/I218 Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 09/12] e1000e: increase driver version number Jeff Kirsher
` (4 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/ich8lan.h | 7 -------
1 file changed, 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.h b/drivers/net/ethernet/intel/e1000e/ich8lan.h
index 21d21b9..80034a2 100644
--- a/drivers/net/ethernet/intel/e1000e/ich8lan.h
+++ b/drivers/net/ethernet/intel/e1000e/ich8lan.h
@@ -250,13 +250,6 @@
/* Proprietary Latency Tolerance Reporting PCI Capability */
#define E1000_PCI_LTR_CAP_LPT 0xA8
-/* OBFF Control & Threshold Defines */
-#define E1000_SVCR_OFF_EN 0x00000001
-#define E1000_SVCR_OFF_MASKINT 0x00001000
-#define E1000_SVCR_OFF_TIMER_MASK 0xFFFF0000
-#define E1000_SVCR_OFF_TIMER_SHIFT 16
-#define E1000_SVT_OFF_HWM_MASK 0x0000001F
-
void e1000e_write_protect_nvm_ich8lan(struct e1000_hw *hw);
void e1000e_set_kmrn_lock_loss_workaround_ich8lan(struct e1000_hw *hw,
bool state);
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 09/12] e1000e: increase driver version number
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (7 preceding siblings ...)
2013-03-28 9:00 ` [net-next 08/12] e1000e: cleanup unused defines Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 10/12] e1000e: fix scheduling while atomic bugs Jeff Kirsher
` (3 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/netdev.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index 0459fe3..858d2a3 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -55,7 +55,7 @@
#define DRV_EXTRAVERSION "-k"
-#define DRV_VERSION "2.2.14" DRV_EXTRAVERSION
+#define DRV_VERSION "2.3.2" DRV_EXTRAVERSION
char e1000e_driver_name[] = "e1000e";
const char e1000e_driver_version[] = DRV_VERSION;
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 10/12] e1000e: fix scheduling while atomic bugs
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (8 preceding siblings ...)
2013-03-28 9:00 ` [net-next 09/12] e1000e: increase driver version number Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 11/12] ixgbe: Don't give VFs random MAC addresses Jeff Kirsher
` (2 subsequent siblings)
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, gospo, sassmann, Jeff Kirsher
From: Bruce Allan <bruce.w.allan@intel.com>
The previous commit ce43a2168c59bc47b5f0c1825fd5f9a2a9e3b447 (e1000e:
cleanup USLEEP_RANGE checkpatch checks) converted a number of delays and
sleeps as recommended in ./Documentation/timers/timers-howto.txt.
Unfortunately, a few of the udelay() to usleep_range() conversions are in
code paths that are in an atomic context in which usleep_range() should
not be used. Revert those specific changes.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/e1000e/phy.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/phy.c b/drivers/net/ethernet/intel/e1000e/phy.c
index cbb310f..59c76a6 100644
--- a/drivers/net/ethernet/intel/e1000e/phy.c
+++ b/drivers/net/ethernet/intel/e1000e/phy.c
@@ -165,7 +165,7 @@ s32 e1000e_read_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 *data)
* the lower time out
*/
for (i = 0; i < (E1000_GEN_POLL_TIMEOUT * 3); i++) {
- usleep_range(50, 100);
+ udelay(50);
mdic = er32(MDIC);
if (mdic & E1000_MDIC_READY)
break;
@@ -190,7 +190,7 @@ s32 e1000e_read_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 *data)
* reading duplicate data in the next MDIC transaction.
*/
if (hw->mac.type == e1000_pch2lan)
- usleep_range(100, 200);
+ udelay(100);
return 0;
}
@@ -229,7 +229,7 @@ s32 e1000e_write_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 data)
* the lower time out
*/
for (i = 0; i < (E1000_GEN_POLL_TIMEOUT * 3); i++) {
- usleep_range(50, 100);
+ udelay(50);
mdic = er32(MDIC);
if (mdic & E1000_MDIC_READY)
break;
@@ -253,7 +253,7 @@ s32 e1000e_write_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 data)
* reading duplicate data in the next MDIC transaction.
*/
if (hw->mac.type == e1000_pch2lan)
- usleep_range(100, 200);
+ udelay(100);
return 0;
}
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 11/12] ixgbe: Don't give VFs random MAC addresses
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (9 preceding siblings ...)
2013-03-28 9:00 ` [net-next 10/12] e1000e: fix scheduling while atomic bugs Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 9:00 ` [net-next 12/12] ixgbevf: Adjust to handle unassigned MAC address from PF Jeff Kirsher
2013-03-28 18:39 ` [net-next 00/12][pull request] Intel Wired LAN Driver Updates David Miller
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem
Cc: Greg Rose, netdev, gospo, sassmann, Andy Gospodarek,
Stefan Assmann, Jeff Kirsher
From: Greg Rose <gregory.v.rose@intel.com>
If the user has not assigned a MAC address to a VM, then don't give it a
random one. Instead, just give it zeros and let it figure out what to do
with them.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
CC: Andy Gospodarek <andy@greyhouse.net>
CC: Stefan Assmann <sassmann@kpanic.de>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 23 +++++++++++++----------
1 file changed, 13 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
index d44b4d2..b3e6530 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
@@ -661,13 +661,7 @@ int ixgbe_vf_configuration(struct pci_dev *pdev, unsigned int event_mask)
bool enable = ((event_mask & 0x10000000U) != 0);
if (enable) {
- eth_random_addr(vf_mac_addr);
- e_info(probe, "IOV: VF %d is enabled MAC %pM\n",
- vfn, vf_mac_addr);
- /*
- * Store away the VF "permananet" MAC address, it will ask
- * for it later.
- */
+ eth_zero_addr(vf_mac_addr);
memcpy(adapter->vfinfo[vfn].vf_mac_addresses, vf_mac_addr, 6);
}
@@ -688,7 +682,8 @@ static int ixgbe_vf_reset_msg(struct ixgbe_adapter *adapter, u32 vf)
ixgbe_vf_reset_event(adapter, vf);
/* set vf mac address */
- ixgbe_set_vf_mac(adapter, vf, vf_mac);
+ if (!is_zero_ether_addr(vf_mac))
+ ixgbe_set_vf_mac(adapter, vf, vf_mac);
vf_shift = vf % 32;
reg_offset = vf / 32;
@@ -729,8 +724,16 @@ static int ixgbe_vf_reset_msg(struct ixgbe_adapter *adapter, u32 vf)
IXGBE_WRITE_REG(hw, IXGBE_VMECM(reg_offset), reg);
/* reply to reset with ack and vf mac address */
- msgbuf[0] = IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_ACK;
- memcpy(addr, vf_mac, ETH_ALEN);
+ msgbuf[0] = IXGBE_VF_RESET;
+ if (!is_zero_ether_addr(vf_mac)) {
+ msgbuf[0] |= IXGBE_VT_MSGTYPE_ACK;
+ memcpy(addr, vf_mac, ETH_ALEN);
+ } else {
+ msgbuf[0] |= IXGBE_VT_MSGTYPE_NACK;
+ dev_warn(&adapter->pdev->dev,
+ "VF %d has no MAC address assigned, you may have to assign one manually\n",
+ vf);
+ }
/*
* Piggyback the multicast filter type so VF can compute the
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* [net-next 12/12] ixgbevf: Adjust to handle unassigned MAC address from PF
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (10 preceding siblings ...)
2013-03-28 9:00 ` [net-next 11/12] ixgbe: Don't give VFs random MAC addresses Jeff Kirsher
@ 2013-03-28 9:00 ` Jeff Kirsher
2013-03-28 18:39 ` [net-next 00/12][pull request] Intel Wired LAN Driver Updates David Miller
12 siblings, 0 replies; 14+ messages in thread
From: Jeff Kirsher @ 2013-03-28 9:00 UTC (permalink / raw)
To: davem
Cc: Greg Rose, netdev, gospo, sassmann, Andy Gospodarek,
Stefan Assmann, Jeff Kirsher
From: Greg Rose <gregory.v.rose@intel.com>
If the administrator has not assigned a MAC address to the VF via the
PF then handle it gracefully by generating a temporary MAC address.
This ensures that we always know when we have a random address and
udev won't get upset about it.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
CC: Andy Gospodarek <andy@greyhouse.net>
CC: Stefan Assmann <sassmann@kpanic.de>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 21 +++++++++++++++------
drivers/net/ethernet/intel/ixgbevf/vf.c | 7 ++++++-
2 files changed, 21 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
index 5563250..eeae934 100644
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
@@ -2052,6 +2052,7 @@ static int ixgbevf_sw_init(struct ixgbevf_adapter *adapter)
{
struct ixgbe_hw *hw = &adapter->hw;
struct pci_dev *pdev = adapter->pdev;
+ struct net_device *netdev = adapter->netdev;
int err;
/* PCI config space info */
@@ -2071,18 +2072,26 @@ static int ixgbevf_sw_init(struct ixgbevf_adapter *adapter)
err = hw->mac.ops.reset_hw(hw);
if (err) {
dev_info(&pdev->dev,
- "PF still in reset state, assigning new address\n");
- eth_hw_addr_random(adapter->netdev);
- memcpy(adapter->hw.mac.addr, adapter->netdev->dev_addr,
- adapter->netdev->addr_len);
+ "PF still in reset state. Is the PF interface up?\n");
} else {
err = hw->mac.ops.init_hw(hw);
if (err) {
pr_err("init_shared_code failed: %d\n", err);
goto out;
}
- memcpy(adapter->netdev->dev_addr, adapter->hw.mac.addr,
- adapter->netdev->addr_len);
+ err = hw->mac.ops.get_mac_addr(hw, hw->mac.addr);
+ if (err)
+ dev_info(&pdev->dev, "Error reading MAC address\n");
+ else if (is_zero_ether_addr(adapter->hw.mac.addr))
+ dev_info(&pdev->dev,
+ "MAC address not assigned by administrator.\n");
+ memcpy(netdev->dev_addr, hw->mac.addr, netdev->addr_len);
+ }
+
+ if (!is_valid_ether_addr(netdev->dev_addr)) {
+ dev_info(&pdev->dev, "Assigning random MAC address\n");
+ eth_hw_addr_random(netdev);
+ memcpy(hw->mac.addr, netdev->dev_addr, netdev->addr_len);
}
/* lock to protect mailbox accesses */
diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c
index 0c94557..387b526 100644
--- a/drivers/net/ethernet/intel/ixgbevf/vf.c
+++ b/drivers/net/ethernet/intel/ixgbevf/vf.c
@@ -109,7 +109,12 @@ static s32 ixgbevf_reset_hw_vf(struct ixgbe_hw *hw)
if (ret_val)
return ret_val;
- if (msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_ACK))
+ /* New versions of the PF may NACK the reset return message
+ * to indicate that no MAC address has yet been assigned for
+ * the VF.
+ */
+ if (msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_ACK) &&
+ msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_NACK))
return IXGBE_ERR_INVALID_MAC_ADDR;
memcpy(hw->mac.perm_addr, addr, ETH_ALEN);
--
1.7.11.7
^ permalink raw reply related [flat|nested] 14+ messages in thread* Re: [net-next 00/12][pull request] Intel Wired LAN Driver Updates
2013-03-28 9:00 [net-next 00/12][pull request] Intel Wired LAN Driver Updates Jeff Kirsher
` (11 preceding siblings ...)
2013-03-28 9:00 ` [net-next 12/12] ixgbevf: Adjust to handle unassigned MAC address from PF Jeff Kirsher
@ 2013-03-28 18:39 ` David Miller
12 siblings, 0 replies; 14+ messages in thread
From: David Miller @ 2013-03-28 18:39 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, sassmann
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Thu, 28 Mar 2013 02:00:03 -0700
> This series contains updates to e1000e, ixgbe and ixgbevf.
>
> Majority of the changes are against e1000e (from Bruce Allan).
> Bruce adds additional error handling on PHY register access, as
> well as improve slow performance on 82579 when connected to a
> 10Mbit hub. In addition, fixes LED blink logic for cathode
> LED design. Most notable is added EEE support which is enabled
> by default and the added support for LTR on I217/I218.
>
> The ixgbe and ixgbevf from Greg Rose changes the VM so that if a user
> does not assign a MAC address, the MAC address is set to all zeros
> instead of a random MAC address. This ensures that we always know when
> we have a random address and udev won't get upset about it.
>
> The following are changes since commit 8b49a4c75965ed157e21450d23dcadd6b27c1aa3:
> bnx2x: fix compilation without CONFIG_BNX2X_SRIOV
> and are available in the git repository at:
> git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master
Pulled, thanks Jeff.
^ permalink raw reply [flat|nested] 14+ messages in thread