summaryrefslogtreecommitdiffstats
path: root/drivers/net/igb/e1000_mac.c
diff options
context:
space:
mode:
authorAuke Kok <auke-jan.h.kok@intel.com>2008-06-27 20:00:18 +0200
committerJeff Garzik <jgarzik@redhat.com>2008-07-04 14:46:59 +0200
commit652fff321490fc3fcc8e8d302826a9c2379f03d2 (patch)
tree00e9675980cbee174305533dd3d0274715afa5e6 /drivers/net/igb/e1000_mac.c
parentigb: fix parameter options (diff)
downloadlinux-652fff321490fc3fcc8e8d302826a9c2379f03d2.tar.xz
linux-652fff321490fc3fcc8e8d302826a9c2379f03d2.zip
igb: eliminate hw from the hw_dbg macro arguments
Various cosmetic cleanups. Comment fixes. Eliminate the hw part out of the hw_dbg macro since it's always used. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Diffstat (limited to 'drivers/net/igb/e1000_mac.c')
-rw-r--r--drivers/net/igb/e1000_mac.c75
1 files changed, 37 insertions, 38 deletions
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index 0dadcfdfa176..47ad2c4277c3 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -158,12 +158,12 @@ void igb_init_rx_addrs(struct e1000_hw *hw, u16 rar_count)
u32 i;
/* Setup the receive address */
- hw_dbg(hw, "Programming MAC Address into RAR[0]\n");
+ hw_dbg("Programming MAC Address into RAR[0]\n");
hw->mac.ops.rar_set(hw, hw->mac.addr, 0);
/* Zero out the other (rar_entry_count - 1) receive addresses */
- hw_dbg(hw, "Clearing RAR[1-%u]\n", rar_count-1);
+ hw_dbg("Clearing RAR[1-%u]\n", rar_count-1);
for (i = 1; i < rar_count; i++) {
array_wr32(E1000_RA, (i << 1), 0);
wrfl();
@@ -193,7 +193,7 @@ s32 igb_check_alt_mac_addr(struct e1000_hw *hw)
ret_val = hw->nvm.ops.read_nvm(hw, NVM_ALT_MAC_ADDR_PTR, 1,
&nvm_alt_mac_addr_offset);
if (ret_val) {
- hw_dbg(hw, "NVM Read Error\n");
+ hw_dbg("NVM Read Error\n");
goto out;
}
@@ -209,7 +209,7 @@ s32 igb_check_alt_mac_addr(struct e1000_hw *hw)
offset = nvm_alt_mac_addr_offset + (i >> 1);
ret_val = hw->nvm.ops.read_nvm(hw, offset, 1, &nvm_data);
if (ret_val) {
- hw_dbg(hw, "NVM Read Error\n");
+ hw_dbg("NVM Read Error\n");
goto out;
}
@@ -336,7 +336,7 @@ void igb_update_mc_addr_list(struct e1000_hw *hw,
}
/* Clear the old settings from the MTA */
- hw_dbg(hw, "Clearing MTA\n");
+ hw_dbg("Clearing MTA\n");
for (i = 0; i < hw->mac.mta_reg_count; i++) {
array_wr32(E1000_MTA, i, 0);
wrfl();
@@ -345,7 +345,7 @@ void igb_update_mc_addr_list(struct e1000_hw *hw,
/* Load any remaining multicast addresses into the hash table. */
for (; mc_addr_count > 0; mc_addr_count--) {
hash_value = igb_hash_mc_addr(hw, mc_addr_list);
- hw_dbg(hw, "Hash value = 0x%03X\n", hash_value);
+ hw_dbg("Hash value = 0x%03X\n", hash_value);
igb_mta_set(hw, hash_value);
mc_addr_list += ETH_ALEN;
}
@@ -540,7 +540,7 @@ s32 igb_check_for_copper_link(struct e1000_hw *hw)
*/
ret_val = igb_config_fc_after_link_up(hw);
if (ret_val)
- hw_dbg(hw, "Error configuring flow control\n");
+ hw_dbg("Error configuring flow control\n");
out:
return ret_val;
@@ -578,7 +578,7 @@ s32 igb_setup_link(struct e1000_hw *hw)
*/
hw->fc.original_type = hw->fc.type;
- hw_dbg(hw, "After fix-ups FlowControl is now = %x\n", hw->fc.type);
+ hw_dbg("After fix-ups FlowControl is now = %x\n", hw->fc.type);
/* Call the necessary media_type subroutine to configure the link. */
ret_val = hw->mac.ops.setup_physical_interface(hw);
@@ -591,8 +591,7 @@ s32 igb_setup_link(struct e1000_hw *hw)
* control is disabled, because it does not hurt anything to
* initialize these registers.
*/
- hw_dbg(hw,
- "Initializing the Flow Control address, type and timer regs\n");
+ hw_dbg("Initializing the Flow Control address, type and timer regs\n");
wr32(E1000_FCT, FLOW_CONTROL_TYPE);
wr32(E1000_FCAH, FLOW_CONTROL_ADDRESS_HIGH);
wr32(E1000_FCAL, FLOW_CONTROL_ADDRESS_LOW);
@@ -689,7 +688,7 @@ static s32 igb_set_default_fc(struct e1000_hw *hw)
&nvm_data);
if (ret_val) {
- hw_dbg(hw, "NVM Read Error\n");
+ hw_dbg("NVM Read Error\n");
goto out;
}
@@ -740,7 +739,7 @@ s32 igb_force_mac_fc(struct e1000_hw *hw)
* 3: Both Rx and TX flow control (symmetric) is enabled.
* other: No other values should be possible at this point.
*/
- hw_dbg(hw, "hw->fc.type = %u\n", hw->fc.type);
+ hw_dbg("hw->fc.type = %u\n", hw->fc.type);
switch (hw->fc.type) {
case e1000_fc_none:
@@ -758,7 +757,7 @@ s32 igb_force_mac_fc(struct e1000_hw *hw)
ctrl |= (E1000_CTRL_TFCE | E1000_CTRL_RFCE);
break;
default:
- hw_dbg(hw, "Flow control param set incorrectly\n");
+ hw_dbg("Flow control param set incorrectly\n");
ret_val = -E1000_ERR_CONFIG;
goto out;
}
@@ -801,7 +800,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
}
if (ret_val) {
- hw_dbg(hw, "Error forcing flow control settings\n");
+ hw_dbg("Error forcing flow control settings\n");
goto out;
}
@@ -827,7 +826,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
goto out;
if (!(mii_status_reg & MII_SR_AUTONEG_COMPLETE)) {
- hw_dbg(hw, "Copper PHY and Auto Neg "
+ hw_dbg("Copper PHY and Auto Neg "
"has not completed.\n");
goto out;
}
@@ -893,11 +892,11 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
*/
if (hw->fc.original_type == e1000_fc_full) {
hw->fc.type = e1000_fc_full;
- hw_dbg(hw, "Flow Control = FULL.\r\n");
+ hw_dbg("Flow Control = FULL.\r\n");
} else {
hw->fc.type = e1000_fc_rx_pause;
- hw_dbg(hw, "Flow Control = "
- "RX PAUSE frames only.\r\n");
+ hw_dbg("Flow Control = "
+ "RX PAUSE frames only.\r\n");
}
}
/*
@@ -913,7 +912,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
(mii_nway_lp_ability_reg & NWAY_LPAR_PAUSE) &&
(mii_nway_lp_ability_reg & NWAY_LPAR_ASM_DIR)) {
hw->fc.type = e1000_fc_tx_pause;
- hw_dbg(hw, "Flow Control = TX PAUSE frames only.\r\n");
+ hw_dbg("Flow Control = TX PAUSE frames only.\r\n");
}
/*
* For transmitting PAUSE frames ONLY.
@@ -928,7 +927,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
!(mii_nway_lp_ability_reg & NWAY_LPAR_PAUSE) &&
(mii_nway_lp_ability_reg & NWAY_LPAR_ASM_DIR)) {
hw->fc.type = e1000_fc_rx_pause;
- hw_dbg(hw, "Flow Control = RX PAUSE frames only.\r\n");
+ hw_dbg("Flow Control = RX PAUSE frames only.\r\n");
}
/*
* Per the IEEE spec, at this point flow control should be
@@ -955,10 +954,10 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
hw->fc.original_type == e1000_fc_tx_pause) ||
hw->fc.strict_ieee) {
hw->fc.type = e1000_fc_none;
- hw_dbg(hw, "Flow Control = NONE.\r\n");
+ hw_dbg("Flow Control = NONE.\r\n");
} else {
hw->fc.type = e1000_fc_rx_pause;
- hw_dbg(hw, "Flow Control = RX PAUSE frames only.\r\n");
+ hw_dbg("Flow Control = RX PAUSE frames only.\r\n");
}
/*
@@ -968,7 +967,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
*/
ret_val = hw->mac.ops.get_speed_and_duplex(hw, &speed, &duplex);
if (ret_val) {
- hw_dbg(hw, "Error getting link speed and duplex\n");
+ hw_dbg("Error getting link speed and duplex\n");
goto out;
}
@@ -981,7 +980,7 @@ s32 igb_config_fc_after_link_up(struct e1000_hw *hw)
*/
ret_val = igb_force_mac_fc(hw);
if (ret_val) {
- hw_dbg(hw, "Error forcing flow control settings\n");
+ hw_dbg("Error forcing flow control settings\n");
goto out;
}
}
@@ -1007,21 +1006,21 @@ s32 igb_get_speed_and_duplex_copper(struct e1000_hw *hw, u16 *speed,
status = rd32(E1000_STATUS);
if (status & E1000_STATUS_SPEED_1000) {
*speed = SPEED_1000;
- hw_dbg(hw, "1000 Mbs, ");
+ hw_dbg("1000 Mbs, ");
} else if (status & E1000_STATUS_SPEED_100) {
*speed = SPEED_100;
- hw_dbg(hw, "100 Mbs, ");
+ hw_dbg("100 Mbs, ");
} else {
*speed = SPEED_10;
- hw_dbg(hw, "10 Mbs, ");
+ hw_dbg("10 Mbs, ");
}
if (status & E1000_STATUS_FD) {
*duplex = FULL_DUPLEX;
- hw_dbg(hw, "Full Duplex\n");
+ hw_dbg("Full Duplex\n");
} else {
*duplex = HALF_DUPLEX;
- hw_dbg(hw, "Half Duplex\n");
+ hw_dbg("Half Duplex\n");
}
return 0;
@@ -1051,7 +1050,7 @@ s32 igb_get_hw_semaphore(struct e1000_hw *hw)
}
if (i == timeout) {
- hw_dbg(hw, "Driver can't access device - SMBI bit is set.\n");
+ hw_dbg("Driver can't access device - SMBI bit is set.\n");
ret_val = -E1000_ERR_NVM;
goto out;
}
@@ -1071,7 +1070,7 @@ s32 igb_get_hw_semaphore(struct e1000_hw *hw)
if (i == timeout) {
/* Release semaphores */
igb_put_hw_semaphore(hw);
- hw_dbg(hw, "Driver can't access the NVM\n");
+ hw_dbg("Driver can't access the NVM\n");
ret_val = -E1000_ERR_NVM;
goto out;
}
@@ -1117,7 +1116,7 @@ s32 igb_get_auto_rd_done(struct e1000_hw *hw)
}
if (i == AUTO_READ_DONE_TIMEOUT) {
- hw_dbg(hw, "Auto read by HW from NVM has not completed.\n");
+ hw_dbg("Auto read by HW from NVM has not completed.\n");
ret_val = -E1000_ERR_RESET;
goto out;
}
@@ -1140,7 +1139,7 @@ static s32 igb_valid_led_default(struct e1000_hw *hw, u16 *data)
ret_val = hw->nvm.ops.read_nvm(hw, NVM_ID_LED_SETTINGS, 1, data);
if (ret_val) {
- hw_dbg(hw, "NVM Read Error\n");
+ hw_dbg("NVM Read Error\n");
goto out;
}
@@ -1322,7 +1321,7 @@ s32 igb_disable_pcie_master(struct e1000_hw *hw)
}
if (!timeout) {
- hw_dbg(hw, "Master requests are pending.\n");
+ hw_dbg("Master requests are pending.\n");
ret_val = -E1000_ERR_MASTER_REQUESTS_PENDING;
goto out;
}
@@ -1342,7 +1341,7 @@ void igb_reset_adaptive(struct e1000_hw *hw)
struct e1000_mac_info *mac = &hw->mac;
if (!mac->adaptive_ifs) {
- hw_dbg(hw, "Not in Adaptive IFS mode!\n");
+ hw_dbg("Not in Adaptive IFS mode!\n");
goto out;
}
@@ -1372,7 +1371,7 @@ void igb_update_adaptive(struct e1000_hw *hw)
struct e1000_mac_info *mac = &hw->mac;
if (!mac->adaptive_ifs) {
- hw_dbg(hw, "Not in Adaptive IFS mode!\n");
+ hw_dbg("Not in Adaptive IFS mode!\n");
goto out;
}
@@ -1413,7 +1412,7 @@ s32 igb_validate_mdi_setting(struct e1000_hw *hw)
s32 ret_val = 0;
if (!hw->mac.autoneg && (hw->phy.mdix == 0 || hw->phy.mdix == 3)) {
- hw_dbg(hw, "Invalid MDI setting detected\n");
+ hw_dbg("Invalid MDI setting detected\n");
hw->phy.mdix = 1;
ret_val = -E1000_ERR_CONFIG;
goto out;
@@ -1452,7 +1451,7 @@ s32 igb_write_8bit_ctrl_reg(struct e1000_hw *hw, u32 reg,
break;
}
if (!(regvalue & E1000_GEN_CTL_READY)) {
- hw_dbg(hw, "Reg %08x did not indicate ready\n", reg);
+ hw_dbg("Reg %08x did not indicate ready\n", reg);
ret_val = -E1000_ERR_PHY;
goto out;
}