Home » Mailing list archives » linux-netdev » 2008 » April » 23

[PATCH 2/8] e1000e: Add interrupt moderation run-time ethtool interface

Previous thread: [PATCH 2/2] [NIU] Add support for Neptune FEM/NEM cards for C10 server blades by Matheos Worku on Wednesday, April 23, 2008 - 11:16 am. (1 message)

Next thread: Будет так-же ? by irina on Wednesday, April 23, 2008 - 11:50 am. (1 message)
[view original message]
From: Auke Kok
Subject: [PATCH 1/8] e1000e: cleanup several stats issues
Date: Wednesday, April 23, 2008 - 11:09 am

From: Bruce Allan <bruce.w.allan@intel.com>

Several stats registers are completely unused and we just waste pci
bus time reading them. We also omit using the high 32 bits of the GORC/
GOTC counters. We can just read clear them and only read the low registers.

Mii-tool can also break es2lan if it executes a MII PHY register
ioctl while the device is in autonegotiation. Unfortunately it seems
that several applications and installations still perform this ioctl
call periodically and especially in this crucial startup time. We
can fool the ioctl by providing fail safe information that mimics
the "down" link state and only perform the dangerous PHY reads once
after link comes up to fill in the real values. As long as link
stays up the information will not change.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/82571.c   |    6 --
 drivers/net/e1000e/defines.h |    2 +
 drivers/net/e1000e/e1000.h   |   25 +++++--
 drivers/net/e1000e/es2lan.c  |    2 -
 drivers/net/e1000e/ethtool.c |    6 +-
 drivers/net/e1000e/hw.h      |   12 +--
 drivers/net/e1000e/netdev.c  |  155 +++++++++++++++++++++++++-----------------
 7 files changed, 122 insertions(+), 86 deletions(-)

diff --git a/drivers/net/e1000e/82571.c b/drivers/net/e1000e/82571.c
index 01c8866..462351c 100644
--- a/drivers/net/e1000e/82571.c
+++ b/drivers/net/e1000e/82571.c
@@ -1326,12 +1326,10 @@ struct e1000_info e1000_82571_info = {
 	.mac			= e1000_82571,
 	.flags			= FLAG_HAS_HW_VLAN_FILTER
 				  | FLAG_HAS_JUMBO_FRAMES
-				  | FLAG_HAS_STATS_PTC_PRC
 				  | FLAG_HAS_WOL
 				  | FLAG_APME_IN_CTRL3
 				  | FLAG_RX_CSUM_ENABLED
 				  | FLAG_HAS_CTRLEXT_ON_LOAD
-				  | FLAG_HAS_STATS_ICR_ICT
 				  | FLAG_HAS_SMART_POWER_DOWN
 				  | FLAG_RESET_OVERWRITES_LAA /* errata */
 				  | FLAG_TARC_SPEED_MODE_BIT /* ...
[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 3/8] e1000e: Fix HW Error on es2lan, ARP capture issue by BMC
Date: Wednesday, April 23, 2008 - 11:09 am

From: David Graham <david.graham@intel.com>

Several components to this complex fix. The es2lan cards occasionally
gave a "HW Error" especially when forcing speed. Some users also
reported that the BMC stole ARP packets.

The fixes include setting the proper SW_FW bits to tell the BMC
that we're active and not do any un-initialization at all, so the
setup routine is largely changed.

Signed-off-by: David Graham <david.graham@intel.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/defines.h |    1 
 drivers/net/e1000e/e1000.h   |    2 +
 drivers/net/e1000e/es2lan.c  |  127 +++++++++++++++++++++++++++++++-----------
 drivers/net/e1000e/phy.c     |   73 +++++++++++++-----------
 4 files changed, 137 insertions(+), 66 deletions(-)

diff --git a/drivers/net/e1000e/defines.h b/drivers/net/e1000e/defines.h
index 4fb9d87..2a53875 100644
--- a/drivers/net/e1000e/defines.h
+++ b/drivers/net/e1000e/defines.h
@@ -184,6 +184,7 @@
 #define E1000_SWFW_EEP_SM   0x1
 #define E1000_SWFW_PHY0_SM  0x2
 #define E1000_SWFW_PHY1_SM  0x4
+#define E1000_SWFW_CSR_SM   0x8
 
 /* Device Control */
 #define E1000_CTRL_FD       0x00000001  /* Full duplex.0=half; 1=full */
diff --git a/drivers/net/e1000e/e1000.h b/drivers/net/e1000e/e1000.h
index 79a426f..f6835c3 100644
--- a/drivers/net/e1000e/e1000.h
+++ b/drivers/net/e1000e/e1000.h
@@ -449,6 +449,8 @@ extern s32 e1000e_read_kmrn_reg(struct e1000_hw *hw, u32 offset, u16 *data);
 extern s32 e1000e_phy_has_link_generic(struct e1000_hw *hw, u32 iterations,
 			       u32 usec_interval, bool *success);
 extern s32 e1000e_phy_reset_dsp(struct e1000_hw *hw);
+extern s32 e1000e_read_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 *data);
+extern s32 e1000e_write_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 data);
 extern s32 e1000e_check_downshift(struct e1000_hw *hw);
 
 static inline s32 e1000_phy_hw_reset(struct e1000_hw *hw)
diff --git a/drivers/net/e1000e/es2lan.c b/drivers/net/e1000e/es2lan.c
index ...
[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 4/8] e1000e: lower ring minimum size to 64
Date: Wednesday, April 23, 2008 - 11:09 am

The lower limit of 80 descriptors in the ring is only valid for
one older 8254x chipset. All e1000e devices can use as low as
64 descriptors.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/e1000.h |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/e1000e/e1000.h b/drivers/net/e1000e/e1000.h
index f6835c3..38bfd0d 100644
--- a/drivers/net/e1000e/e1000.h
+++ b/drivers/net/e1000e/e1000.h
@@ -64,11 +64,11 @@ struct e1000_info;
 /* Tx/Rx descriptor defines */
 #define E1000_DEFAULT_TXD		256
 #define E1000_MAX_TXD			4096
-#define E1000_MIN_TXD			80
+#define E1000_MIN_TXD			64
 
 #define E1000_DEFAULT_RXD		256
 #define E1000_MAX_RXD			4096
-#define E1000_MIN_RXD			80
+#define E1000_MIN_RXD			64
 
 #define E1000_MIN_ITR_USECS		10 /* 100000 irq/sec */
 #define E1000_MAX_ITR_USECS		10000 /* 100    irq/sec */

--


[ message continues ]
[view original message]
From: Kok, Auke
Subject: Re: [E1000-devel] [PATCH 4/8] e1000e: lower ring minimum size to 64
Date: Wednesday, April 23, 2008 - 11:12 am

Jeff,

patch #1-4 are RESENDS, where I omitted the VLAN/TSO fix that PJ will respin.
patch 5-8 are trivial patches (partial RESENDS) that I did not see merged yet and
I could not find in any of your trees ATM. apologies if I sound pushy - just
making sure these don't fall through the cracks ;)

Cheers

Auke
--


[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 6/8] e1000e: save and restore pcie/msi state to support EEH recovery
Date: Wednesday, April 23, 2008 - 11:09 am

From: Wendy Xiong <wendyx@us.ibm.com>

To enable EEH support for pci-express network adapters, pcie/msi state
needs to be saved and restored for that adapter.

Tested this EEH patch with 2ports and 4ports pci-express e1000e
adapters.

Signed-off-by: Wendy Xiong <wendyx@us.ibm.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/netdev.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index 146edf7..7282751 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -3815,6 +3815,7 @@ static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev)
 		return PCI_ERS_RESULT_DISCONNECT;
 	}
 	pci_set_master(pdev);
+	pci_restore_state(pdev);
 
 	pci_enable_wake(pdev, PCI_D3hot, 0);
 	pci_enable_wake(pdev, PCI_D3cold, 0);
@@ -3941,6 +3942,7 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
 		goto err_pci_reg;
 
 	pci_set_master(pdev);
+	pci_save_state(pdev);
 
 	err = -ENOMEM;
 	netdev = alloc_etherdev(sizeof(struct e1000_adapter));

--


[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 7/8] [RESEND] igb: save and restore pcie/msi state to support EEH recovery
Date: Wednesday, April 23, 2008 - 11:09 am

To enable EEH support for pci-express network adapters, pcie/msi state
needs to be saved and restored for that adapter.

[after similar patches for ixgbe and e1000e from Wendy Xiong]

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Cc: Wendy Xiong <wendyx@us.ibm.com>
---

 drivers/net/igb/igb_main.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index aaee02e..ae398f0 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -871,6 +871,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
 		goto err_pci_reg;
 
 	pci_set_master(pdev);
+	pci_save_state(pdev);
 
 	err = -ENOMEM;
 	netdev = alloc_etherdev(sizeof(struct igb_adapter));
@@ -4079,6 +4080,7 @@ static pci_ers_result_t igb_io_slot_reset(struct pci_dev *pdev)
 		return PCI_ERS_RESULT_DISCONNECT;
 	}
 	pci_set_master(pdev);
+	pci_restore_state(pdev);
 
 	pci_enable_wake(pdev, PCI_D3hot, 0);
 	pci_enable_wake(pdev, PCI_D3cold, 0);

--


[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 8/8] e1000e: Increment version to 0.2.1
Date: Wednesday, April 23, 2008 - 11:09 am

From: Jesse Brandeburg <jesse.brandeburg@intel.com>

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/netdev.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index 7282751..5110382 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -46,7 +46,7 @@
 
 #include "e1000.h"
 
-#define DRV_VERSION "0.2.0"
+#define DRV_VERSION "0.2.1"
 char e1000e_driver_name[] = "e1000e";
 const char e1000e_driver_version[] = DRV_VERSION;
 

--


[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 5/8] ixgbe: save and restore pcie/msi state to support EEH recovery
Date: Wednesday, April 23, 2008 - 11:09 am

From: Wendy Xiong <wendyx@us.ibm.com>

To enable EEH support for pci-express network adapters, pcie/msi state
needs to be saved and restored for that adapter.

Tested this EEH patch with Intel 10G pci-express ixgbe adapter.

Signed-off-by: Wendy Xiong <wendyx@us.ibm.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/ixgbe/ixgbe_main.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index cb371a8..7b85922 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -3431,6 +3431,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	}
 
 	pci_set_master(pdev);
+	pci_save_state(pdev);
 
 #ifdef CONFIG_NETDEVICES_MULTIQUEUE
 	netdev = alloc_etherdev_mq(sizeof(struct ixgbe_adapter), MAX_TX_QUEUES);
@@ -3721,6 +3722,7 @@ static pci_ers_result_t ixgbe_io_slot_reset(struct pci_dev *pdev)
 		return PCI_ERS_RESULT_DISCONNECT;
 	}
 	pci_set_master(pdev);
+	pci_restore_state(pdev);
 
 	pci_enable_wake(pdev, PCI_D3hot, 0);
 	pci_enable_wake(pdev, PCI_D3cold, 0);

--


[ message continues ]
[view original message]
From: Auke Kok
Subject: [PATCH 2/8] e1000e: Add interrupt moderation run-time ethtool interface
Date: Wednesday, April 23, 2008 - 11:09 am

The ethtool -c / -C interface can now be used to modify the
irq moderation algorithm. This change does not require an
adapter reset and can thus be used at all times. The adapter
only supports changing/reading rx-usecs which has special
values for 0, 1 and 3:

0 - no irq moderation whatsoever
1 - normal moderation favoring regular mixed traffic (default)
3 - best attempt at low latency possible at cost of CPU

For values between 10 and 10000 the rx-usecs defines "the minimum
time between successive irqs" in usec, unlike the module parameter.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
---

 drivers/net/e1000e/e1000.h   |    3 +++
 drivers/net/e1000e/ethtool.c |   43 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 46 insertions(+), 0 deletions(-)

diff --git a/drivers/net/e1000e/e1000.h b/drivers/net/e1000e/e1000.h
index 4d3d1c2..79a426f 100644
--- a/drivers/net/e1000e/e1000.h
+++ b/drivers/net/e1000e/e1000.h
@@ -70,6 +70,9 @@ struct e1000_info;
 #define E1000_MAX_RXD			4096
 #define E1000_MIN_RXD			80
 
+#define E1000_MIN_ITR_USECS		10 /* 100000 irq/sec */
+#define E1000_MAX_ITR_USECS		10000 /* 100    irq/sec */
+
 /* Early Receive defines */
 #define E1000_ERT_2048			0x100
 
diff --git a/drivers/net/e1000e/ethtool.c b/drivers/net/e1000e/ethtool.c
index c894a6f..ce045ac 100644
--- a/drivers/net/e1000e/ethtool.c
+++ b/drivers/net/e1000e/ethtool.c
@@ -1770,6 +1770,47 @@ static int e1000_phys_id(struct net_device *netdev, u32 data)
 	return 0;
 }
 
+static int e1000_get_coalesce(struct net_device *netdev,
+			      struct ethtool_coalesce *ec)
+{
+	struct e1000_adapter *adapter = netdev_priv(netdev);
+
+	if (adapter->itr_setting <= 3)
+		ec->rx_coalesce_usecs = adapter->itr_setting;
+	else
+		ec->rx_coalesce_usecs = 1000000 / adapter->itr_setting;
+
+	return 0;
+}
+
+static int e1000_set_coalesce(struct net_device *netdev,
+			      struct ethtool_coalesce *ec)
+{
+	struct e1000_adapter *adapter = netdev_priv(netdev);
+	struct ...
[ message continues ]
[view original message]
From: Jeff Garzik
Subject: Re: [PATCH 1/8] e1000e: cleanup several stats issues
Date: Thursday, April 24, 2008 - 11:01 pm

applied 1-8

I was mainly waiting for the discussion of the abuse of the ethtool 
interface, which I chose to allow.

Now, send me a patch to the ethtool man page for this new behavior :)


--


[ message continues ]
[view original message]
From: Kok, Auke
Subject: Re: [PATCH 1/8] e1000e: cleanup several stats issues
Date: Friday, April 25, 2008 - 9:39 am

agreed - I'll come up with an appropriate explanation to cover this issue.

Thanks.

Auke


--


[ message continues ]
[view original message]
From: Kok, Auke
Subject: Re: [PATCH 1/8] e1000e: cleanup several stats issues
Date: Monday, April 28, 2008 - 3:13 pm

it almost seems more appropriate to send a patch to the e1000 README and explain
the parameter use in there since this is a per-driver modification and there's
nothing special about the ethtool call itself.

Would you be OK with a general patch to ethtool that states that "drivers may
interpret values differently and users should consult the driver's source code or
README for proper use of the parameter." ?

This seems the logical solution.

Auke

--


[ message continues ]
Previous thread: [PATCH 2/2] [NIU] Add support for Neptune FEM/NEM cards for C10 server blades by Matheos Worku on Wednesday, April 23, 2008 - 11:16 am. (1 message)

Next thread: Будет так-же ? by irina on Wednesday, April 23, 2008 - 11:50 am. (1 message)