ixgbe: Enable another bit for flow control operation
Subsystems (2)
PATCH 1/10 ixgbe: Enable another bit for flow control operation
From: PJ Waskiewicz <[email protected]>
The discard pause frames bit was not enabled, so flow control frames could be passed up to the host. Enabled DPF to keep the frames off the stack.
---
drivers/net/ixgbe/ixgbe_82598.c | 1 + drivers/net/ixgbe/ixgbe_common.c | 1 + 2 files changed, 2 insertions(+), 0 deletions(-)
drivers/net/ixgbe/ixgbe_82598.c +1 -0
@@ -322,6 +322,7 @@ static s32 ixgbe_fc_enable_82598(struct ixgbe_hw *hw, s32 packetbuf_num) }/* Enable 802.3x based flow control settings. */ + fctrl_reg |=IXGBE_FCTRL_DPF ;IXGBE_WRITE_REG (hw,IXGBE_FCTRL , fctrl_reg);IXGBE_WRITE_REG (hw,IXGBE_RMCS , rmcs_reg);
drivers/net/ixgbe/ixgbe_common.c +1 -0
@@ -1700,6 +1700,7 @@ s32 ixgbe_fc_enable(struct ixgbe_hw *hw, s32 packetbuf_num) }/* Enable 802.3x based flow control settings. */ + mflcn_reg |=IXGBE_MFLCN_DPF ;IXGBE_WRITE_REG (hw,IXGBE_MFLCN , mflcn_reg);IXGBE_WRITE_REG (hw,IXGBE_FCCFG , fccfg_reg);
PATCH 2/10 ixgbe: fix IXGBE_MTQC values
From: Don Skidmore <[email protected]>
IXGBE_MTQC_64VF was wrong and 32VF not defined at all. This patch corrects that.
---
drivers/net/ixgbe/ixgbe_type.h | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-)
drivers/net/ixgbe/ixgbe_type.h +2 -1
@@ -1553,7 +1553,8 @@#define IXGBE_MTQC_RT_ENA 0x1/* DCB Enable */ #define IXGBE_MTQC_VT_ENA 0x2/* VMDQ2 Enable */ #define IXGBE_MTQC_64Q_1PB 0x0/* 64 queues 1 pack buffer */ -#define IXGBE_MTQC_64VF 0x8/* 2 TX Queues per pool w/64VF's */ +#define IXGBE_MTQC_32VF 0x8/* 4 TX Queues per pool w/32VF's */ +#define IXGBE_MTQC_64VF 0x4/* 2 TX Queues per pool w/64VF's */ #define IXGBE_MTQC_8TC_8TQ 0xC/* 8 TC if RT_ENA or 8 TQ if VT_ENA */ /* Receive Descriptor bit definitions */
Re: [PATCH 02/10] ixgbe: fix IXGBE_MTQC values
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:26:40 -0700
Quoted 6 lines
IXGBE_MTQC_64VFwas wrong and 32VF not defined at all. This patch corrects that.Signed-off-by Don Skidmore <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 3/10 ixgbe: add support for addition VT_CTL defines
In reply to Jeff KirsherFrom: Don Skidmore <[email protected]>
These defines allow for definition of target pool for packets that pass L2 filtering but didn't pass any of the pool filters. They are needed to reset the default pool.
---
drivers/net/ixgbe/ixgbe_type.h | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)
drivers/net/ixgbe/ixgbe_type.h +2 -0
@@ -958,6 +958,8 @@#define IXGBE_VT_CTL_DIS_DEFPL 0x20000000/* disable default pool */ #define IXGBE_VT_CTL_REPLEN 0x40000000/* replication enabled */ #define IXGBE_VT_CTL_VT_ENABLE 0x00000001/* Enable VT Mode */ +#define IXGBE_VT_CTL_POOL_SHIFT 7 +#define IXGBE_VT_CTL_POOL_MASK (0x3F << IXGBE_VT_CTL_POOL_SHIFT)/* VMOLR bitmasks */ #define IXGBE_VMOLR_AUPE 0x01000000/* accept untagged packets */
Re: [PATCH 03/10] ixgbe: add support for addition VT_CTL defines
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:27:00 -0700
Quoted 7 lines
These defines allow for definition of target pool for packets that pass L2 filtering but didn't pass any of the pool filters. They are needed to reset the default pool.
Signed-off-by Don Skidmore <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 4/10 ixgbe: add define to support 82599 64 IVAR registers
In reply to Jeff KirsherFrom: Don Skidmore <[email protected]>
82599 supports 64 IVAR registers this patch adds a define to allow us to access them.
---
drivers/net/ixgbe/ixgbe_type.h | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)
drivers/net/ixgbe/ixgbe_type.h +1 -0
@@ -1150,6 +1150,7 @@/* Interrupt Vector Allocation Registers */ #define IXGBE_IVAR_REG_NUM 25 +#define IXGBE_IVAR_REG_NUM_82599 64#define IXGBE_IVAR_TXRX_ENTRY 96#define IXGBE_IVAR_RX_ENTRY 64#define IXGBE_IVAR_RX_QUEUE (_i) (0 + (_i))
Re: [PATCH 04/10] ixgbe: add define to support 82599 64 IVAR registers
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:27:19 -0700
Quoted 6 lines
82599 supports 64 IVAR registers this patch adds a define to allow us to access them.
Signed-off-by Don Skidmore <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 5/10 ixgbe: Don't return error in flow control configuration if FC is off
In reply to Jeff KirsherFrom: PJ Waskiewicz <[email protected]>
When flow control is disabled, an invalid low/high watermark configuration should not matter.
---
drivers/net/ixgbe/ixgbe_82598.c | 8 +++++--- drivers/net/ixgbe/ixgbe_common.c | 8 +++++--- 2 files changed, 10 insertions(+), 6 deletions(-)
drivers/net/ixgbe/ixgbe_82598.c +5 -3
@@ -381,9 +381,11 @@ static s32 ixgbe_setup_fc_82598(struct ixgbe_hw *hw, s32 packetbuf_num)* because it causes the controller to just blast out fc packets. */ if (!hw->fc.low_water || !hw->fc.high_water || !hw->fc.pause_time) { -hw_dbg (hw,"Invalid water mark configuration\n" ); - ret_val =IXGBE_ERR_INVALID_LINK_SETTINGS ; - goto out; +if (hw->fc.requested_mode != ixgbe_fc_none) { +hw_dbg (hw,"Invalid water mark configuration\n" ); + ret_val =IXGBE_ERR_INVALID_LINK_SETTINGS ; + goto out; + } }/*
drivers/net/ixgbe/ixgbe_common.c +5 -3
@@ -1907,9 +1907,11 @@ s32 ixgbe_setup_fc_generic(struct ixgbe_hw *hw, s32 packetbuf_num)* because it causes the controller to just blast out fc packets. */ if (!hw->fc.low_water || !hw->fc.high_water || !hw->fc.pause_time) { -hw_dbg (hw,"Invalid water mark configuration\n" ); - ret_val =IXGBE_ERR_INVALID_LINK_SETTINGS ; - goto out; +if (hw->fc.requested_mode != ixgbe_fc_none) { +hw_dbg (hw,"Invalid water mark configuration\n" ); + ret_val =IXGBE_ERR_INVALID_LINK_SETTINGS ; + goto out; + } }/*
Re: [PATCH 05/10] ixgbe: Don't return error in flow control configuration if FC is off
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:27:39 -0700
Quoted 5 lines
When flow control is disabled, an invalid low/high watermark configuration should not matter.
Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 6/10 ixgbe: Move PHY ops initialization to centralize bus accesses
In reply to Jeff KirsherFrom: PJ Waskiewicz <[email protected]>
When PHY operations are determined, the PHY must be identified. This identification causes bus access, and should be contained within its own routines. This also helps the 82599 PHY init paths for both SFP+ and KX/KX4 devices to be easier to maintain.
---
drivers/net/ixgbe/ixgbe_82598.c | 77 ++++++++++++++++++++++++++------------- drivers/net/ixgbe/ixgbe_82599.c | 70 +++++++++++++++++++++++------------ drivers/net/ixgbe/ixgbe_main.c | 9 ++++- drivers/net/ixgbe/ixgbe_type.h | 1 + 4 files changed, 107 insertions(+), 50 deletions(-)
drivers/net/ixgbe/ixgbe_82598.c +52 -25
@@ -75,18 +75,49 @@ static u16 ixgbe_get_pcie_msix_count_82598(struct ixgbe_hw *hw)static s32 ixgbe_get_invariants_82598 (struct ixgbe_hw *hw) {struct ixgbe_mac_info *mac = &hw->mac; + +/* Call PHY identify routine to get the phy type */ +ixgbe_identify_phy_generic (hw); + + mac->mcft_size =IXGBE_82598_MC_TBL_SIZE ; + mac->vft_size =IXGBE_82598_VFT_TBL_SIZE ; + mac->num_rar_entries =IXGBE_82598_RAR_ENTRIES ; + mac->max_rx_queues =IXGBE_82598_MAX_RX_QUEUES ; + mac->max_tx_queues =IXGBE_82598_MAX_TX_QUEUES ; + mac->max_msix_vectors =ixgbe_get_pcie_msix_count_82598 (hw); + +return 0 ; +} + +/** +* ixgbe_init_phy_ops_82598 - PHY/SFP specific init +* @hw: pointer to hardware structure +* +* Initialize any function pointers that were not able to be +* set during get_invariants because the PHY/SFP type was +* not known. Perform the SFP init if necessary. +* +**/ +s32 ixgbe_init_phy_ops_82598 (struct ixgbe_hw *hw) +{ +struct ixgbe_mac_info *mac = &hw->mac;struct ixgbe_phy_info *phy = &hw->phy;s32 ret_val =0 ;u16 list_offset, data_offset; -/* Set the bus information prior to PHY identification */ - mac->ops.get_bus_info (hw); +/* Identify the PHY */ + phy->ops.identify (hw); -/* Call PHY identify routine to get the phy type */ -ixgbe_identify_phy_generic (hw); +/* Overwrite the link function pointers if copper PHY */ +if (mac->ops.get_media_type (hw) == ixgbe_media_type_copper) { + mac->ops.setup_link = &ixgbe_setup_copper_link_82598; + mac->ops.setup_link_speed = + &ixgbe_setup_copper_link_speed_82598; + mac->ops.get_link_capabilities = + &ixgbe_get_copper_link_capabilities_82598; + } -/* PHY Init */ -switch (phy->type) { +switch (hw->phy.type) {case ixgbe_phy_tn: phy->ops.check_link = &ixgbe_check_phy_link_tnx; phy->ops.get_firmware_version =
@@ -106,8 +137,8 @@ static s32 ixgbe_get_invariants_82598(struct ixgbe_hw *hw)/* Check to see if SFP+ module is supported */ ret_val =ixgbe_get_sfp_init_sequence_offsets (hw, - &list_offset, - &data_offset); + &list_offset, + &data_offset);if (ret_val !=0 ) { ret_val =IXGBE_ERR_SFP_NOT_SUPPORTED ; goto out;
@@ -117,21 +148,6 @@ static s32 ixgbe_get_invariants_82598(struct ixgbe_hw *hw)break ; } -if (mac->ops.get_media_type (hw) == ixgbe_media_type_copper) { - mac->ops.setup_link = &ixgbe_setup_copper_link_82598; - mac->ops.setup_link_speed = - &ixgbe_setup_copper_link_speed_82598; - mac->ops.get_link_capabilities = - &ixgbe_get_copper_link_capabilities_82598; - } - - mac->mcft_size =IXGBE_82598_MC_TBL_SIZE ; - mac->vft_size =IXGBE_82598_VFT_TBL_SIZE ; - mac->num_rar_entries =IXGBE_82598_RAR_ENTRIES ; - mac->max_rx_queues =IXGBE_82598_MAX_RX_QUEUES ; - mac->max_tx_queues =IXGBE_82598_MAX_TX_QUEUES ; - mac->max_msix_vectors =ixgbe_get_pcie_msix_count_82598 (hw); - out:return ret_val; }
@@ -719,14 +735,23 @@ static s32 ixgbe_reset_hw_82598(struct ixgbe_hw *hw) }/* Reset PHY */ -if (hw->phy.reset_disable == false) +if (hw->phy.reset_disable == false) { +/* PHY ops must be identified and initialized prior to reset */ + +/* Init PHY and function pointers, perform SFP setup */ + status = hw->phy.ops.init (hw); +if (status ==IXGBE_ERR_SFP_NOT_SUPPORTED ) + goto reset_hw_out; + hw->phy.ops.reset (hw); + }/* * Prevent the PCI-E bus from from hanging by disabling PCI-E master * access and verify no pending requests before reset */ -if (ixgbe_disable_pcie_master (hw) !=0 ) { + status =ixgbe_disable_pcie_master (hw); +if (status !=0 ) { status =IXGBE_ERR_MASTER_REQUESTS_PENDING ;hw_dbg (hw,"PCI-E Master disable polling has failed.\n" ); }
@@ -773,6 +798,7 @@ static s32 ixgbe_reset_hw_82598(struct ixgbe_hw *hw)/* Store the permanent mac address */ hw->mac.ops.get_mac_addr (hw, hw->mac.perm_addr); +reset_hw_out:return status; }
@@ -1102,6 +1128,7 @@ static struct ixgbe_eeprom_operations eeprom_ops_82598 = {static struct ixgbe_phy_operations phy_ops_82598 = { .identify = &ixgbe_identify_phy_generic, .identify_sfp = &ixgbe_identify_sfp_module_generic, + .init = &ixgbe_init_phy_ops_82598, .reset = &ixgbe_reset_phy_generic, .read_reg = &ixgbe_read_phy_reg_generic, .write_reg = &ixgbe_write_phy_reg_generic,
drivers/net/ixgbe/ixgbe_82599.c +46 -24
122 lines not on this page · raw message
drivers/net/ixgbe/ixgbe_main.c +8 -1
@@ -4630,7 +4630,11 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,/* reset_hw fills in the perm_addr as well */ err = hw->mac.ops.reset_hw (hw); -if (err) { +if (err ==IXGBE_ERR_SFP_NOT_SUPPORTED ) { +dev_err (&adapter->pdev->dev,"failed to load because an " +"unsupported SFP+ module type was detected.\n" ); + goto err_sw_init; + }else if (err) {dev_err (&adapter->pdev->dev,"HW Init failed: %d\n" , err); goto err_sw_init; }
@@ -4703,6 +4707,9 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,device_init_wakeup (&adapter->pdev->dev, true);device_set_wakeup_enable (&adapter->pdev->dev, adapter->wol); +/* pick up the PCI bus settings for reporting later */ + hw->mac.ops.get_bus_info (hw); +/* print bus type/speed/width info */ dev_info(&pdev->dev, "(PCI Express:%s:%s) %pM\n", ((hw->bus.speed == ixgbe_bus_speed_5000) ? "5.0Gb/s":
drivers/net/ixgbe/ixgbe_type.h +1 -0
@@ -2150,6 +2150,7 @@ struct ixgbe_mac_operations {struct ixgbe_phy_operations {s32 (*identify)(struct ixgbe_hw *);s32 (*identify_sfp)(struct ixgbe_hw *); +s32 (*init)(struct ixgbe_hw *);s32 (*reset)(struct ixgbe_hw *);s32 (*read_reg)(struct ixgbe_hw *,u32 ,u32 ,u16 *);s32 (*write_reg)(struct ixgbe_hw *,u32 ,u32 ,u16 );
Re: [PATCH 06/10] ixgbe: Move PHY ops initialization to centralize bus accesses
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:27:57 -0700
Quoted 7 lines
When PHY operations are determined, the PHY must be identified. This identification causes bus access, and should be contained within its own routines. This also helps the 82599 PHY init paths for both SFP+ and KX/KX4 devices to be easier to maintain.
Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 7/10 ixgbe: Remove unnecessary PHY reset, properly identify multispeed fiber modules
In reply to Jeff KirsherFrom: PJ Waskiewicz <[email protected]>
This patch does two things:
1) On 82599, the PHY is emedded in the MAC. On 82598, the SFP+ NIC has an external PHY. The reset in the SFP+ setup patch for 82598 is unnecessary on 82599, and adds extra dead time to device initialization. This removes that PHY reset for 82599 only.
2) On 82599, the SFP+ modules are multispeed fiber modules (10G/1G). We need to make sure to identify them properly for the remaining init sections to properly set them up.
---
drivers/net/ixgbe/ixgbe_82599.c | 24 ++++++++++++++++-------- drivers/net/ixgbe/ixgbe_phy.c | 11 +++++++++++ drivers/net/ixgbe/ixgbe_phy.h | 1 + drivers/net/ixgbe/ixgbe_type.h | 1 + 4 files changed, 29 insertions(+), 8 deletions(-)
drivers/net/ixgbe/ixgbe_82599.c +16 -8
@@ -100,6 +100,9 @@ s32 ixgbe_setup_sfp_modules_82599(struct ixgbe_hw *hw)if (hw->phy.sfp_type != ixgbe_sfp_type_unknown) {ixgbe_init_mac_link_ops_82599 (hw); + + hw->phy.ops.reset =NULL ; + ret_val =ixgbe_get_sfp_init_sequence_offsets (hw, &list_offset, &data_offset);
@@ -714,19 +717,24 @@ s32 ixgbe_reset_hw_82599(struct ixgbe_hw *hw)/* Call adapter stop to disable tx/rx and clear interrupts */ hw->mac.ops.stop_adapter (hw); -/* Reset PHY */ -if (hw->phy.reset_disable == false) { -/* PHY ops must be identified and initialized prior to reset */ +/* PHY ops must be identified and initialized prior to reset */ -/* Init PHY and function pointers, perform SFP setup */ - status = hw->phy.ops.init (hw); +/* Init PHY and function pointers, perform SFP setup */ + status = hw->phy.ops.init (hw); -if (status ==IXGBE_ERR_SFP_NOT_SUPPORTED ) - goto reset_hw_out; +if (status ==IXGBE_ERR_SFP_NOT_SUPPORTED ) + goto reset_hw_out; - hw->phy.ops.reset (hw); +/* Setup SFP module if there is one present. */ +if (hw->phy.sfp_setup_needed) { + status = hw->mac.ops.setup_sfp (hw); + hw->phy.sfp_setup_needed = false; } +/* Reset PHY */ +if (hw->phy.reset_disable == false && hw->phy.ops.reset !=NULL ) + hw->phy.ops.reset (hw); +/* * Prevent the PCI-E bus from from hanging by disabling PCI-E master * access and verify no pending requests before reset
drivers/net/ixgbe/ixgbe_phy.c +11 -0
@@ -552,6 +552,7 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw) {s32 status =IXGBE_ERR_PHY_ADDR_INVALID ;u32 vendor_oui =0 ; +enum ixgbe_sfp_type stored_sfp_type = hw->phy.sfp_type;u8 identifier =0 ;u8 comp_codes_1g =0 ;u8 comp_codes_10g =0 ;
@@ -620,6 +621,16 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw) hw->phy.sfp_type = ixgbe_sfp_type_unknown; } +if (hw->phy.sfp_type != stored_sfp_type) + hw->phy.sfp_setup_needed = true; + +/* Determine if the SFP+ PHY is dual speed or not. */ +if (((comp_codes_1g &IXGBE_SFF_1GBASESX_CAPABLE ) && + (comp_codes_10g &IXGBE_SFF_10GBASESR_CAPABLE )) || + ((comp_codes_1g &IXGBE_SFF_1GBASELX_CAPABLE ) && + (comp_codes_10g &IXGBE_SFF_10GBASELR_CAPABLE ))) + hw->phy.multispeed_fiber = true; +/* Determine PHY vendor */ if (hw->phy.type == ixgbe_phy_unknown) { hw->phy.id = identifier;
drivers/net/ixgbe/ixgbe_phy.h +1 -0
@@ -44,6 +44,7 @@/* Bitmasks */ #define IXGBE_SFF_TWIN_AX_CAPABLE 0x80#define IXGBE_SFF_1GBASESX_CAPABLE 0x1 +#define IXGBE_SFF_1GBASELX_CAPABLE 0x2#define IXGBE_SFF_10GBASESR_CAPABLE 0x10#define IXGBE_SFF_10GBASELR_CAPABLE 0x20#define IXGBE_I2C_EEPROM_READ_MASK 0x100
drivers/net/ixgbe/ixgbe_type.h +1 -0
@@ -2198,6 +2198,7 @@ struct ixgbe_phy_info {u32 addr;u32 id;enum ixgbe_sfp_type sfp_type; +bool sfp_setup_needed;u32 revision;enum ixgbe_media_type media_type;bool reset_disable;
Re: [PATCH 07/10] ixgbe: Remove unnecessary PHY reset, properly identify multispeed fiber modules
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:28:15 -0700
Quoted 7 lines
This patch does two things:
1) On 82599, the PHY is emedded in the MAC. On 82598, the SFP+ NIC has an external PHY. The reset in the SFP+ setup patch for 82598 is unnecessary on 82599, and adds extra dead time to device initialization. This removes that PHY reset for 82599 only.2) On 82599, the SFP+ modules are multispeed fiber modules (10G/1G). We need to make sure to identify them properly for the remaining init sections to properly set them up.
Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 8/10 ixgbe: Update the usage of orig_autoc to be more consistent
In reply to Jeff KirsherFrom: PJ Waskiewicz <[email protected]>
The orig_autoc variable tracks the original setting of the autonegotiate state prior to trying a new speed. The usage is inconsistent and not very maintainable. This patch updates the usage to make it more consistent.
---
drivers/net/ixgbe/ixgbe_82598.c | 15 +++++++++++---- drivers/net/ixgbe/ixgbe_82599.c | 37 ++++++++++++++++++++++++++++--------- 2 files changed, 39 insertions(+), 13 deletions(-)
drivers/net/ixgbe/ixgbe_82598.c +11 -4
@@ -165,12 +165,19 @@ static s32 ixgbe_get_link_capabilities_82598(struct ixgbe_hw *hw, bool *autoneg) {s32 status =0 ; +u32 autoc =0 ;/* * Determine link capabilities based on the stored value of AUTOC, -* which represents EEPROM defaults. +* which represents EEPROM defaults. If AUTOC value has not been +* stored, use the current register value. */ -switch (hw->mac.orig_autoc &IXGBE_AUTOC_LMS_MASK ) { +if (hw->mac.orig_link_settings_stored) + autoc = hw->mac.orig_autoc; +else + autoc =IXGBE_READ_REG (hw,IXGBE_AUTOC ); + +switch (autoc &IXGBE_AUTOC_LMS_MASK ) {case IXGBE_AUTOC_LMS_1G_LINK_NO_AN : *speed =IXGBE_LINK_SPEED_1GB_FULL ; *autoneg = false;
@@ -189,9 +196,9 @@ static s32 ixgbe_get_link_capabilities_82598(struct ixgbe_hw *hw,case IXGBE_AUTOC_LMS_KX4_AN :case IXGBE_AUTOC_LMS_KX4_AN_1G_AN : *speed =IXGBE_LINK_SPEED_UNKNOWN ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX4_SUPP ) +if (autoc &IXGBE_AUTOC_KX4_SUPP ) *speed |=IXGBE_LINK_SPEED_10GB_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX_SUPP ) +if (autoc &IXGBE_AUTOC_KX_SUPP ) *speed |=IXGBE_LINK_SPEED_1GB_FULL ; *autoneg = true;break ;
drivers/net/ixgbe/ixgbe_82599.c +28 -9
@@ -219,8 +219,19 @@ s32 ixgbe_get_link_capabilities_82599(struct ixgbe_hw *hw, bool *negotiation) {s32 status =0 ; +u32 autoc =0 ; -switch (hw->mac.orig_autoc &IXGBE_AUTOC_LMS_MASK ) { +/* +* Determine link capabilities based on the stored value of AUTOC, +* which represents EEPROM defaults. If AUTOC value has not been +* stored, use the current register value. +*/ +if (hw->mac.orig_link_settings_stored) + autoc = hw->mac.orig_autoc; +else + autoc =IXGBE_READ_REG (hw,IXGBE_AUTOC ); + +switch (autoc &IXGBE_AUTOC_LMS_MASK ) {case IXGBE_AUTOC_LMS_1G_LINK_NO_AN : *speed =IXGBE_LINK_SPEED_1GB_FULL ; *negotiation = false;
@@ -244,22 +255,22 @@ s32 ixgbe_get_link_capabilities_82599(struct ixgbe_hw *hw,case IXGBE_AUTOC_LMS_KX4_KX_KR :case IXGBE_AUTOC_LMS_KX4_KX_KR_1G_AN : *speed =IXGBE_LINK_SPEED_UNKNOWN ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KR_SUPP ) +if (autoc &IXGBE_AUTOC_KR_SUPP ) *speed |=IXGBE_LINK_SPEED_10GB_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX4_SUPP ) +if (autoc &IXGBE_AUTOC_KX4_SUPP ) *speed |=IXGBE_LINK_SPEED_10GB_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX_SUPP ) +if (autoc &IXGBE_AUTOC_KX_SUPP ) *speed |=IXGBE_LINK_SPEED_1GB_FULL ; *negotiation = true;break ;case IXGBE_AUTOC_LMS_KX4_KX_KR_SGMII : *speed =IXGBE_LINK_SPEED_100_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KR_SUPP ) +if (autoc &IXGBE_AUTOC_KR_SUPP ) *speed |=IXGBE_LINK_SPEED_10GB_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX4_SUPP ) +if (autoc &IXGBE_AUTOC_KX4_SUPP ) *speed |=IXGBE_LINK_SPEED_10GB_FULL ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX_SUPP ) +if (autoc &IXGBE_AUTOC_KX_SUPP ) *speed |=IXGBE_LINK_SPEED_1GB_FULL ; *negotiation = true;break ;
@@ -570,6 +581,7 @@ s32 ixgbe_setup_mac_link_speed_82599(struct ixgbe_hw *hw,s32 status =0 ;u32 autoc =IXGBE_READ_REG (hw,IXGBE_AUTOC );u32 autoc2 =IXGBE_READ_REG (hw,IXGBE_AUTOC2 ); +u32 orig_autoc =0 ;u32 link_mode = autoc &IXGBE_AUTOC_LMS_MASK ;u32 pma_pmd_1g = autoc &IXGBE_AUTOC_1G_PMA_PMD_MASK ;u32 pma_pmd_10g_serial = autoc2 &IXGBE_AUTOC2_10G_SERIAL_PMA_PMD_MASK ;
@@ -581,6 +593,13 @@ s32 ixgbe_setup_mac_link_speed_82599(struct ixgbe_hw *hw, hw->mac.ops.get_link_capabilities (hw, &link_capabilities, &autoneg); speed &= link_capabilities; +/* Use stored value (EEPROM defaults) of AUTOC to find KR/KX4 support*/ +if (hw->mac.orig_link_settings_stored) + orig_autoc = hw->mac.orig_autoc; +else + orig_autoc = autoc; + +if (speed ==IXGBE_LINK_SPEED_UNKNOWN ) { status =IXGBE_ERR_LINK_SETUP ; } else if (link_mode == IXGBE_AUTOC_LMS_KX4_KX_KR ||
@@ -589,9 +608,9 @@ s32 ixgbe_setup_mac_link_speed_82599(struct ixgbe_hw *hw,/* Set KX4/KX/KR support according to speed requested */ autoc &= ~(IXGBE_AUTOC_KX4_KX_SUPP_MASK |IXGBE_AUTOC_KR_SUPP );if (speed &IXGBE_LINK_SPEED_10GB_FULL ) -if (hw->mac.orig_autoc &IXGBE_AUTOC_KX4_SUPP ) +if (orig_autoc &IXGBE_AUTOC_KX4_SUPP ) autoc |=IXGBE_AUTOC_KX4_SUPP ; -if (hw->mac.orig_autoc &IXGBE_AUTOC_KR_SUPP ) +if (orig_autoc &IXGBE_AUTOC_KR_SUPP ) autoc |=IXGBE_AUTOC_KR_SUPP ;if (speed &IXGBE_LINK_SPEED_1GB_FULL ) autoc |=IXGBE_AUTOC_KX_SUPP ;
Re: [PATCH 08/10] ixgbe: Update the usage of orig_autoc to be more consistent
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:28:33 -0700
Quoted 6 lines
The
orig_autocvariable tracks the original setting of the autonegotiate state prior to trying a new speed. The usage is inconsistent and not very maintainable. This patch updates the usage to make it more consistent.Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 9/10 ixgbe: Update get_physical_layer() calls, plus a version bump
In reply to Jeff KirsherFrom: Peter P Waskiewicz Jr <[email protected]>
Not all physical connection types are being correctly identified. This fixes that issue, and cleans up the logic to make it more maintainable.
Also clean up the code for device capabilities from the EEPROM to support multiple SKUs of the same hardware.
Bump the version to reflect all the updates since the 82599 merge.
---
drivers/net/ixgbe/ixgbe_82598.c | 85 +++++++++++++++++-------- drivers/net/ixgbe/ixgbe_82599.c | 134 +++++++++++++++++++++++++++++---------- drivers/net/ixgbe/ixgbe_main.c | 2 - drivers/net/ixgbe/ixgbe_phy.c | 8 +- drivers/net/ixgbe/ixgbe_type.h | 10 ++- 5 files changed, 169 insertions(+), 70 deletions(-)
drivers/net/ixgbe/ixgbe_82598.c +59 -26
@@ -1034,35 +1034,56 @@ out:static u32 ixgbe_get_supported_physical_layer_82598 (struct ixgbe_hw *hw) {u32 physical_layer =IXGBE_PHYSICAL_LAYER_UNKNOWN ; +u32 autoc =IXGBE_READ_REG (hw,IXGBE_AUTOC ); +u32 pma_pmd_10g = autoc &IXGBE_AUTOC_10G_PMA_PMD_MASK ; +u32 pma_pmd_1g = autoc &IXGBE_AUTOC_1G_PMA_PMD_MASK ; +u16 ext_ability =0 ; + + hw->phy.ops.identify (hw); + +/* Copper PHY must be checked before AUTOC LMS to determine correct +* physical layer because 10GBase-T PHYs use LMS = KX4/KX */ +if (hw->phy.type == ixgbe_phy_tn || + hw->phy.type == ixgbe_phy_cu_unknown) { + hw->phy.ops.read_reg (hw,IXGBE_MDIO_PHY_EXT_ABILITY , +IXGBE_MDIO_PMA_PMD_DEV_TYPE , &ext_ability); +if (ext_ability &IXGBE_MDIO_PHY_10GBASET_ABILITY ) + physical_layer |=IXGBE_PHYSICAL_LAYER_10GBASE_T ; +if (ext_ability &IXGBE_MDIO_PHY_1000BASET_ABILITY ) + physical_layer |=IXGBE_PHYSICAL_LAYER_1000BASE_T ; +if (ext_ability &IXGBE_MDIO_PHY_100BASETX_ABILITY ) + physical_layer |=IXGBE_PHYSICAL_LAYER_100BASE_TX ; + goto out; + } -switch (hw->device_id) { -case IXGBE_DEV_ID_82598 : -/* Default device ID is mezzanine card KX/KX4 */ - physical_layer = (IXGBE_PHYSICAL_LAYER_10GBASE_KX4 | -IXGBE_PHYSICAL_LAYER_1000BASE_KX ); -break ; -case IXGBE_DEV_ID_82598_BX : - physical_layer =IXGBE_PHYSICAL_LAYER_1000BASE_BX ; -case IXGBE_DEV_ID_82598EB_CX4 : -case IXGBE_DEV_ID_82598_CX4_DUAL_PORT : - physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_CX4 ; -break ; -case IXGBE_DEV_ID_82598_DA_DUAL_PORT : - physical_layer =IXGBE_PHYSICAL_LAYER_SFP_PLUS_CU ; +switch (autoc &IXGBE_AUTOC_LMS_MASK ) { +case IXGBE_AUTOC_LMS_1G_AN : +case IXGBE_AUTOC_LMS_1G_LINK_NO_AN : +if (pma_pmd_1g ==IXGBE_AUTOC_1G_KX ) + physical_layer =IXGBE_PHYSICAL_LAYER_1000BASE_KX ; +else + physical_layer =IXGBE_PHYSICAL_LAYER_1000BASE_BX ;break ; -case IXGBE_DEV_ID_82598AF_DUAL_PORT : -case IXGBE_DEV_ID_82598AF_SINGLE_PORT : -case IXGBE_DEV_ID_82598_SR_DUAL_PORT_EM : - physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_SR ; +case IXGBE_AUTOC_LMS_10G_LINK_NO_AN : +if (pma_pmd_10g ==IXGBE_AUTOC_10G_CX4 ) + physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_CX4 ; +else if (pma_pmd_10g ==IXGBE_AUTOC_10G_KX4 ) + physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_KX4 ; +else /* XAUI */ + physical_layer =IXGBE_PHYSICAL_LAYER_UNKNOWN ;break ; -case IXGBE_DEV_ID_82598EB_XF_LR : - physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_LR ; +case IXGBE_AUTOC_LMS_KX4_AN : +case IXGBE_AUTOC_LMS_KX4_AN_1G_AN : +if (autoc &IXGBE_AUTOC_KX_SUPP ) + physical_layer |=IXGBE_PHYSICAL_LAYER_1000BASE_KX ; +if (autoc &IXGBE_AUTOC_KX4_SUPP ) + physical_layer |=IXGBE_PHYSICAL_LAYER_10GBASE_KX4 ;break ; -case IXGBE_DEV_ID_82598AT : - physical_layer = (IXGBE_PHYSICAL_LAYER_10GBASE_T | -IXGBE_PHYSICAL_LAYER_1000BASE_T ); +default :break ; -case IXGBE_DEV_ID_82598EB_SFP_LOM : + } + +if (hw->phy.type == ixgbe_phy_nl) { hw->phy.ops.identify_sfp (hw);switch (hw->phy.sfp_type) {
@@ -1079,13 +1100,25 @@ static u32 ixgbe_get_supported_physical_layer_82598(struct ixgbe_hw *hw) physical_layer =IXGBE_PHYSICAL_LAYER_UNKNOWN ;break ; } -break ; + } +switch (hw->device_id) { +case IXGBE_DEV_ID_82598_DA_DUAL_PORT : + physical_layer =IXGBE_PHYSICAL_LAYER_SFP_PLUS_CU ; +break ; +case IXGBE_DEV_ID_82598AF_DUAL_PORT : +case IXGBE_DEV_ID_82598AF_SINGLE_PORT : +case IXGBE_DEV_ID_82598_SR_DUAL_PORT_EM : + physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_SR ; +break ; +case IXGBE_DEV_ID_82598EB_XF_LR : + physical_layer =IXGBE_PHYSICAL_LAYER_10GBASE_LR ; +break ;default : - physical_layer =IXGBE_PHYSICAL_LAYER_UNKNOWN ;break ; } +out:return physical_layer; }
drivers/net/ixgbe/ixgbe_82599.c +98 -36
163 lines not on this page · raw message
drivers/net/ixgbe/ixgbe_main.c +1 -1
@@ -47,7 +47,7 @@ char ixgbe_driver_name[] = "ixgbe";static const char ixgbe_driver_string[] ="Intel(R) 10 Gigabit PCI Express Network Driver" ; -#define DRV_VERSION "2.0.8-k2" +#define DRV_VERSION "2.0.16-k2"const char ixgbe_driver_version[] =DRV_VERSION ;static char ixgbe_copyright[] ="Copyright (c) 1999-2009 Intel Corporation." ;
drivers/net/ixgbe/ixgbe_phy.c +4 -4
@@ -632,7 +632,7 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw) hw->phy.multispeed_fiber = true;/* Determine PHY vendor */ -if (hw->phy.type == ixgbe_phy_unknown) { +if (hw->phy.type != ixgbe_phy_nl) { hw->phy.id = identifier; hw->phy.ops.read_i2c_eeprom(hw, IXGBE_SFF_VENDOR_OUI_BYTE0,
@@ -682,9 +682,9 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw) goto out; } - hw->eeprom.ops.read (hw,IXGBE_PHY_ENFORCE_INTEL_SFP_OFFSET , - &enforce_sfp); -if (!(enforce_sfp &IXGBE_PHY_ALLOW_ANY_SFP )) { +/* This is guaranteed to be 82599, no need to check for NULL */ + hw->mac.ops.get_device_caps (hw, &enforce_sfp); +if (!(enforce_sfp &IXGBE_DEVICE_CAPS_ALLOW_ANY_SFP )) {/* Make sure we're a supported PHY type */ if (hw->phy.type == ixgbe_phy_sfp_intel) { status =0 ;
drivers/net/ixgbe/ixgbe_type.h +7 -3
@@ -862,6 +862,7 @@#define IXGBE_MDIO_PHY_EXT_ABILITY 0xB/* Ext Ability Reg */ #define IXGBE_MDIO_PHY_10GBASET_ABILITY 0x0004/* 10GBaseT capable */ #define IXGBE_MDIO_PHY_1000BASET_ABILITY 0x0020/* 1000BaseT capable */ +#define IXGBE_MDIO_PHY_100BASETX_ABILITY 0x0080/* 100BaseTX capable */ #define IXGBE_MDIO_PMA_PMD_SDA_SCL_ADDR 0xC30A/* PHY_XS SDA/SCL Addr Reg */ #define IXGBE_MDIO_PMA_PMD_SDA_SCL_DATA 0xC30B/* PHY_XS SDA/SCL Data Reg */
@@ -898,8 +899,6 @@#define IXGBE_CONTROL_NL 0x000F#define IXGBE_CONTROL_EOL_NL 0x0FFF#define IXGBE_CONTROL_SOL_NL 0x0000 -#define IXGBE_PHY_ENFORCE_INTEL_SFP_OFFSET 0x002C -#define IXGBE_PHY_ALLOW_ANY_SFP 0x1/* General purpose Interrupt Enable */ #define IXGBE_SDP0_GPIEN 0x00000001/* SDP0 */
@@ -1385,6 +1384,7 @@#define IXGBE_FW_PTR 0x0F#define IXGBE_PBANUM0_PTR 0x15#define IXGBE_PBANUM1_PTR 0x16 +#define IXGBE_DEVICE_CAPS 0x2C#define IXGBE_PCIE_MSIX_82599_CAPS 0x72#define IXGBE_PCIE_MSIX_82598_CAPS 0x62
@@ -1428,6 +1428,8 @@#define IXGBE_EERD_ATTEMPTS 100000#endif +#define IXGBE_DEVICE_CAPS_ALLOW_ANY_SFP 0x1 +/* PCI Bus Info */ #define IXGBE_PCI_LINK_STATUS 0xB2#define IXGBE_PCI_LINK_WIDTH 0x3F0
@@ -1865,7 +1867,7 @@ typedef u32 ixgbe_physical_layer;#define IXGBE_PHYSICAL_LAYER_UNKNOWN 0#define IXGBE_PHYSICAL_LAYER_10GBASE_T 0x0001#define IXGBE_PHYSICAL_LAYER_1000BASE_T 0x0002 -#define IXGBE_PHYSICAL_LAYER_100BASE_T 0x0004 +#define IXGBE_PHYSICAL_LAYER_100BASE_TX 0x0004#define IXGBE_PHYSICAL_LAYER_SFP_PLUS_CU 0x0008#define IXGBE_PHYSICAL_LAYER_10GBASE_LR 0x0010#define IXGBE_PHYSICAL_LAYER_10GBASE_LRM 0x0020
@@ -1874,6 +1876,7 @@ typedef u32 ixgbe_physical_layer;#define IXGBE_PHYSICAL_LAYER_10GBASE_CX4 0x0100#define IXGBE_PHYSICAL_LAYER_1000BASE_KX 0x0200#define IXGBE_PHYSICAL_LAYER_1000BASE_BX 0x0400 +#define IXGBE_PHYSICAL_LAYER_10GBASE_KR 0x0800enum ixgbe_eeprom_type { ixgbe_eeprom_uninitialized =0 ,
@@ -2105,6 +2108,7 @@ struct ixgbe_mac_operations {enum ixgbe_media_type (*get_media_type)(struct ixgbe_hw *);u32 (*get_supported_physical_layer)(struct ixgbe_hw *);s32 (*get_mac_addr)(struct ixgbe_hw *,u8 *); +s32 (*get_device_caps)(struct ixgbe_hw *,u16 *);s32 (*stop_adapter)(struct ixgbe_hw *);s32 (*get_bus_info)(struct ixgbe_hw *);void (*set_lan_id)(struct ixgbe_hw *);
Re: [PATCH 09/10] ixgbe: Update get_physical_layer() calls, plus a version bump
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:28:50 -0700
Quoted 10 lines
Not all physical connection types are being correctly identified. This fixes that issue, and cleans up the logic to make it more maintainable.
Also clean up the code for device capabilities from the EEPROM to support multiple SKUs of the same hardware.
Bump the version to reflect all the updates since the 82599 merge.
Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
PATCH 10/10 Be explict with what we are !'ing in ixgbe_sfp_config_module_task()
In reply to Jeff KirsherFrom: Tony Breeds <[email protected]>
GCC warns:drivers/net/ixgbe/ixgbe_main.c: In function 'ixgbe_sfp_config_module_task':drivers/net/ixgbe/ixgbe_main.c:3920: warning: suggest parantheses around operand of '!' or change '&' to '&&' or '!' to '~'
Which I think is right. Bracket to remove ambiguity.
---
drivers/net/ixgbe/ixgbe_main.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
drivers/net/ixgbe/ixgbe_main.c +1 -1
@@ -3917,7 +3917,7 @@ static void ixgbe_sfp_config_module_task(struct work_struct *work) } hw->mac.ops.setup_sfp (hw); -if (!adapter->flags &IXGBE_FLAG_IN_SFP_LINK_TASK ) +if (!(adapter->flags &IXGBE_FLAG_IN_SFP_LINK_TASK ))/* This will also work for DA Twinax connections */ schedule_work (&adapter->multispeed_fiber_task); adapter->flags &= ~IXGBE_FLAG_IN_SFP_MOD_TASK ;
Re: [PATCH 10/10] Be explict with what we are !'ing in ixgbe_sfp_config_module_task()
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:29:10 -0700
Quoted 9 lines
GCC warns:
drivers/net/ixgbe/ixgbe_main.c: In function 'ixgbe_sfp_config_module_task':drivers/net/ixgbe/ixgbe_main.c:3920: warning: suggest parantheses around operand of '!' or change '&' to '&&' or '!' to '~'Which I think is right. Bracket to remove ambiguity.
Signed-off-by Tony Breeds <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6
From: Jeff Kirsher <[email protected]>
Date: Fri, 10 Apr 2009 01:26:21 -0700
Quoted 7 lines
From: PJ Waskiewicz <[email protected]>
The discard pause frames bit was not enabled, so flow control frames could be passed up to the host. Enabled DPF to keep the frames off the stack.
Signed-off-by Peter P Waskiewicz Jr <[email protected]>Signed-off-by Jeff Kirsher <[email protected]>
Applied to net-next-2.6