Skip to contentSkip to conversation

sky2: tx pause bug fix

PATCH In review Stephen Hemminger 4 participants 11 messages Sep 6, 2006 – Sep 22, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/3 sky2: tx pause bug fix

Fix problems with transmit pause frames. The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

This maybe the cause of the lockup under load.

	http://bugzilla.kernel.org/show_bug.cgi?id=6839

Patch against netdev-2.6 git tree

---

 drivers/net/sky2.c |  123 ++++++++++++++++++----------------------------------
 drivers/net/sky2.h |    2 -
 2 files changed, 43 insertions(+), 82 deletions(-)

1419c8ab49f8fd56bad1ac0d3dcbaf830cd5a5d6

drivers/net/sky2.c +42 -81
index 7ce0663..a3a4ab2 100644
@@ -289,7 +289,7 @@ static void sky2_gmac_reset(struct sky2_
 static void sky2_phy_init(struct sky2_hw *hw, unsigned port)
 {
 	struct sky2_port *sky2 = netdev_priv(hw->dev[port]);
-	u16 ctrl, ct1000, adv, pg, ledctrl, ledover;
+	u16 ctrl, ct1000, adv, pg, ledctrl, ledover, reg;
 
 	if (sky2->autoneg == AUTONEG_ENABLE &&
 	    !(hw->chip_id == CHIP_ID_YUKON_XL || hw->chip_id == CHIP_ID_YUKON_EC_U)) {
@@ -358,6 +358,7 @@ static void sky2_phy_init(struct sky2_hw
 	ctrl = 0;
 	ct1000 = 0;
 	adv = PHY_AN_CSMA;
+	reg = 0;
 
 	if (sky2->autoneg == AUTONEG_ENABLE) {
 		if (hw->copper) {
@@ -390,21 +391,46 @@ static void sky2_phy_init(struct sky2_hw
 		/* forced speed/duplex settings */
 		ct1000 = PHY_M_1000C_MSE;
 
-		if (sky2->duplex == DUPLEX_FULL)
-			ctrl |= PHY_CT_DUP_MD;
+		/* Disable auto update for duplex flow control and speed */
+		reg |= GM_GPCR_AU_ALL_DIS;
 
 		switch (sky2->speed) {
 		case SPEED_1000:
 			ctrl |= PHY_CT_SP1000;
+			reg |= GM_GPCR_SPEED_1000;
 			break;
 		case SPEED_100:
 			ctrl |= PHY_CT_SP100;
+			reg |= GM_GPCR_SPEED_100;
 			break;
 		}
 
+		if (sky2->duplex == DUPLEX_FULL) {
+			reg |= GM_GPCR_DUP_FULL;
+			ctrl |= PHY_CT_DUP_MD;
+		} else if (sky2->speed != SPEED_1000 && hw->chip_id != CHIP_ID_YUKON_EC_U) {
+			/* Turn off flow control for 10/100mbps */
+			sky2->rx_pause = 0;
+			sky2->tx_pause = 0;
+		}
+
+		if (!sky2->rx_pause)
+			reg |= GM_GPCR_FC_RX_DIS;
+
+		if (!sky2->tx_pause)
+			reg |= GM_GPCR_FC_TX_DIS;
+
+		/* Forward pause packets to GMAC? */
+		if (sky2->tx_pause || sky2->rx_pause)
+			sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_ON);
+		else
+			sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
+
 		ctrl |= PHY_CT_RESET;
 	}
 
+	gma_write16(hw, port, GM_GP_CTRL, reg);
+
 	if (hw->chip_id != CHIP_ID_YUKON_FE)
 		gm_phy_write(hw, port, PHY_MARV_1000T_CTRL, ct1000);
 
@@ -508,6 +534,7 @@ static void sky2_phy_init(struct sky2_hw
 			gm_phy_write(hw, port, PHY_MARV_LED_OVER, ledover);
 
 	}
+
 	/* Enable phy interrupt on auto-negotiation complete (or link up) */
 	if (sky2->autoneg == AUTONEG_ENABLE)
 		gm_phy_write(hw, port, PHY_MARV_INT_MASK, PHY_M_IS_AN_COMPL);
@@ -570,49 +597,11 @@ static void sky2_mac_init(struct sky2_hw
 			 gm_phy_read(hw, 1, PHY_MARV_INT_MASK) != 0);
 	}
 
-	if (sky2->autoneg == AUTONEG_DISABLE) {
-		reg = gma_read16(hw, port, GM_GP_CTRL);
-		reg |= GM_GPCR_AU_ALL_DIS;
-		gma_write16(hw, port, GM_GP_CTRL, reg);
-		gma_read16(hw, port, GM_GP_CTRL);
-
-		switch (sky2->speed) {
-		case SPEED_1000:
-			reg &= ~GM_GPCR_SPEED_100;
-			reg |= GM_GPCR_SPEED_1000;
-			break;
-		case SPEED_100:
-			reg &= ~GM_GPCR_SPEED_1000;
-			reg |= GM_GPCR_SPEED_100;
-			break;
-		case SPEED_10:
-			reg &= ~(GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100);
-			break;
-		}
-
-		if (sky2->duplex == DUPLEX_FULL)
-			reg |= GM_GPCR_DUP_FULL;
-
-		/* turn off pause in 10/100mbps half duplex */
-		else if (sky2->speed != SPEED_1000 &&
-			 hw->chip_id != CHIP_ID_YUKON_EC_U)
-			sky2->tx_pause = sky2->rx_pause = 0;
-	} else
-		reg = GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100 | GM_GPCR_DUP_FULL;
-
-	if (!sky2->tx_pause && !sky2->rx_pause) {
-		sky2_write32(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
-		reg |=
-		    GM_GPCR_FC_TX_DIS | GM_GPCR_FC_RX_DIS | GM_GPCR_AU_FCT_DIS;
-	} else if (sky2->tx_pause && !sky2->rx_pause) {
-		/* disable Rx flow-control */
-		reg |= GM_GPCR_FC_RX_DIS | GM_GPCR_AU_FCT_DIS;
-	}
-
-	gma_write16(hw, port, GM_GP_CTRL, reg);
-
 	sky2_read16(hw, SK_REG(port, GMAC_IRQ_SRC));
 
+	/* Enable Transmit FIFO Underrun */
+	sky2_write8(hw, SK_REG(port, GMAC_IRQ_MSK), GMAC_DEF_MSK);
+
 	spin_lock_bh(&sky2->phy_lock);
 	sky2_phy_init(hw, port);
 	spin_unlock_bh(&sky2->phy_lock);
@@ -1529,40 +1518,10 @@ static void sky2_link_up(struct sky2_por
 	unsigned port = sky2->port;
 	u16 reg;
 
-	/* Enable Transmit FIFO Underrun */
-	sky2_write8(hw, SK_REG(port, GMAC_IRQ_MSK), GMAC_DEF_MSK);
-
-	reg = gma_read16(hw, port, GM_GP_CTRL);
-	if (sky2->autoneg == AUTONEG_DISABLE) {
-		reg |= GM_GPCR_AU_ALL_DIS;
-
-		/* Is write/read necessary?  Copied from sky2_mac_init */
-		gma_write16(hw, port, GM_GP_CTRL, reg);
-		gma_read16(hw, port, GM_GP_CTRL);
-
-		switch (sky2->speed) {
-		case SPEED_1000:
-			reg &= ~GM_GPCR_SPEED_100;
-			reg |= GM_GPCR_SPEED_1000;
-			break;
-		case SPEED_100:
-			reg &= ~GM_GPCR_SPEED_1000;
-			reg |= GM_GPCR_SPEED_100;
-			break;
-		case SPEED_10:
-			reg &= ~(GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100);
-			break;
-		}
-	} else
-		reg &= ~GM_GPCR_AU_ALL_DIS;
-
-	if (sky2->duplex == DUPLEX_FULL || sky2->autoneg == AUTONEG_ENABLE)
-		reg |= GM_GPCR_DUP_FULL;
-
 	/* enable Rx/Tx */
+	reg = gma_read16(hw, port, GM_GP_CTRL);
 	reg |= GM_GPCR_RX_ENA | GM_GPCR_TX_ENA;
 	gma_write16(hw, port, GM_GP_CTRL, reg);
-	gma_read16(hw, port, GM_GP_CTRL);
 
 	gm_phy_write(hw, port, PHY_MARV_INT_MASK, PHY_M_DEF_MSK);
 
@@ -1616,7 +1575,6 @@ static void sky2_link_down(struct sky2_p
 	reg = gma_read16(hw, port, GM_GP_CTRL);
 	reg &= ~(GM_GPCR_RX_ENA | GM_GPCR_TX_ENA);
 	gma_write16(hw, port, GM_GP_CTRL, reg);
-	gma_read16(hw, port, GM_GP_CTRL);	/* PCI post */
 
 	if (sky2->rx_pause && !sky2->tx_pause) {
 		/* restore Asymmetric Pause bit */
@@ -1633,6 +1591,7 @@ static void sky2_link_down(struct sky2_p
 
 	if (netif_msg_link(sky2))
 		printk(KERN_INFO PFX "%s: Link is down.\n", sky2->netdev->name);
+
 	sky2_phy_init(hw, port);
 }
 
@@ -1673,8 +1632,11 @@ static int sky2_autoneg_done(struct sky2
 	sky2->rx_pause = (aux & PHY_M_PS_RX_P_EN) != 0;
 	sky2->tx_pause = (aux & PHY_M_PS_TX_P_EN) != 0;
 
-	if ((sky2->tx_pause || sky2->rx_pause)
-	    && !(sky2->speed < SPEED_1000 && sky2->duplex == DUPLEX_HALF))
+	if (sky2->duplex == DUPLEX_HALF && sky2->speed != SPEED_1000
+	    && hw->chip_id != CHIP_ID_YUKON_EC_U)
+		sky2->rx_pause = sky2->tx_pause = 0;
+
+	if (sky2->rx_pause || sky2->tx_pause)
 		sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_ON);
 	else
 		sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
@@ -1700,7 +1662,7 @@ static void sky2_phy_intr(struct sky2_hw
 		printk(KERN_INFO PFX "%s: phy interrupt status 0x%x 0x%x\n",
 		       sky2->netdev->name, istatus, phystat);
 
-	if (istatus & PHY_M_IS_AN_COMPL) {
+	if (sky2->autoneg == AUTONEG_ENABLE && (istatus & PHY_M_IS_AN_COMPL)) {
 		if (sky2_autoneg_done(sky2, phystat) == 0)
 			sky2_link_up(sky2);
 		goto out;
@@ -2890,7 +2852,6 @@ static int sky2_set_pauseparam(struct ne
 			       struct ethtool_pauseparam *ecmd)
 {
 	struct sky2_port *sky2 = netdev_priv(dev);
-	int err = 0;
 
 	sky2->autoneg = ecmd->autoneg;
 	sky2->tx_pause = ecmd->tx_pause != 0;
@@ -2898,7 +2859,7 @@ static int sky2_set_pauseparam(struct ne
 
 	sky2_phy_reinit(sky2);
 
-	return err;
+	return 0;
 }
 
 static int sky2_get_coalesce(struct net_device *dev,
drivers/net/sky2.h +1 -1
index fa8af9f..a27194c 100644
@@ -1566,7 +1566,7 @@ enum {
 
 	GMR_FS_ANY_ERR	= GMR_FS_RX_FF_OV | GMR_FS_CRC_ERR |
 			  GMR_FS_FRAGMENT | GMR_FS_LONG_ERR |
-		  	  GMR_FS_MII_ERR | GMR_FS_BAD_FC | GMR_FS_GOOD_FC |
+		  	  GMR_FS_MII_ERR | GMR_FS_BAD_FC |
 			  GMR_FS_UN_SIZE | GMR_FS_JABBER,
 };
 
Signature · 1 line
1.2.4
#1 lore raw

Stephen,

After some serious testing, this patch seems to fix the lockup issue completely. I manually applied these changes against the 2.6.17.13 release.

----- Original Message ----- From: "Stephen Hemminger" <[email protected]>
To: "Jeff Garzik" <[email protected]>
Cc: <[email protected]>
Sent: Thursday, September 07, 2006 5:44 AM
Subject: [PATCH 1/3] sky2: tx pause bug fix

Quoted 270 lines

Fix problems with transmit pause frames. The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

This maybe the cause of the lockup under load.
http://bugzilla.kernel.org/show_bug.cgi?id=6839

Patch against netdev-2.6 git tree

---

drivers/net/sky2.c | 123
++++++++++++++++++----------------------------------
drivers/net/sky2.h | 2 -

2 files changed, 43 insertions(+), 82 deletions(-)

1419c8ab49f8fd56bad1ac0d3dcbaf830cd5a5d6

drivers/net/sky2.c +1 -1
index 7ce0663..a3a4ab2 100644
@@ -289,7 +289,7 @@ static void sky2_gmac_reset(struct sky2_
 static void sky2_phy_init(struct sky2_hw *hw, unsigned port)
 {
 struct sky2_port *sky2 = netdev_priv(hw->dev[port]);
- u16 ctrl, ct1000, adv, pg, ledctrl, ledover;
+ u16 ctrl, ct1000, adv, pg, ledctrl, ledover, reg;
 
 if (sky2->autoneg == AUTONEG_ENABLE &&
     !(hw->chip_id == CHIP_ID_YUKON_XL || hw->chip_id == 

CHIP_ID_YUKON_EC_U)) {

drivers/net/sky2.c +41 -80
@@ -358,6 +358,7 @@ static void sky2_phy_init(struct sky2_hw
 ctrl = 0;
 ct1000 = 0;
 adv = PHY_AN_CSMA;
+ reg = 0;
 
 if (sky2->autoneg == AUTONEG_ENABLE) {
 if (hw->copper) {
@@ -390,21 +391,46 @@ static void sky2_phy_init(struct sky2_hw
 /* forced speed/duplex settings */
 ct1000 = PHY_M_1000C_MSE;
 
- if (sky2->duplex == DUPLEX_FULL)
- ctrl |= PHY_CT_DUP_MD;
+ /* Disable auto update for duplex flow control and speed */
+ reg |= GM_GPCR_AU_ALL_DIS;
 
 switch (sky2->speed) {
 case SPEED_1000:
 ctrl |= PHY_CT_SP1000;
+ reg |= GM_GPCR_SPEED_1000;
 break;
 case SPEED_100:
 ctrl |= PHY_CT_SP100;
+ reg |= GM_GPCR_SPEED_100;
 break;
 }
 
+ if (sky2->duplex == DUPLEX_FULL) {
+ reg |= GM_GPCR_DUP_FULL;
+ ctrl |= PHY_CT_DUP_MD;
+ } else if (sky2->speed != SPEED_1000 && hw->chip_id != 
 CHIP_ID_YUKON_EC_U) {
+ /* Turn off flow control for 10/100mbps */
+ sky2->rx_pause = 0;
+ sky2->tx_pause = 0;
+ }
+
+ if (!sky2->rx_pause)
+ reg |= GM_GPCR_FC_RX_DIS;
+
+ if (!sky2->tx_pause)
+ reg |= GM_GPCR_FC_TX_DIS;
+
+ /* Forward pause packets to GMAC? */
+ if (sky2->tx_pause || sky2->rx_pause)
+ sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_ON);
+ else
+ sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
+
 ctrl |= PHY_CT_RESET;
 }
 
+ gma_write16(hw, port, GM_GP_CTRL, reg);
+
 if (hw->chip_id != CHIP_ID_YUKON_FE)
 gm_phy_write(hw, port, PHY_MARV_1000T_CTRL, ct1000);
@@ -508,6 +534,7 @@ static void sky2_phy_init(struct sky2_hw
 gm_phy_write(hw, port, PHY_MARV_LED_OVER, ledover);
 
 }
+
 /* Enable phy interrupt on auto-negotiation complete (or link up) */
 if (sky2->autoneg == AUTONEG_ENABLE)
 gm_phy_write(hw, port, PHY_MARV_INT_MASK, PHY_M_IS_AN_COMPL);
@@ -570,49 +597,11 @@ static void sky2_mac_init(struct sky2_hw
 gm_phy_read(hw, 1, PHY_MARV_INT_MASK) != 0);
 }
 
- if (sky2->autoneg == AUTONEG_DISABLE) {
- reg = gma_read16(hw, port, GM_GP_CTRL);
- reg |= GM_GPCR_AU_ALL_DIS;
- gma_write16(hw, port, GM_GP_CTRL, reg);
- gma_read16(hw, port, GM_GP_CTRL);
-
- switch (sky2->speed) {
- case SPEED_1000:
- reg &= ~GM_GPCR_SPEED_100;
- reg |= GM_GPCR_SPEED_1000;
- break;
- case SPEED_100:
- reg &= ~GM_GPCR_SPEED_1000;
- reg |= GM_GPCR_SPEED_100;
- break;
- case SPEED_10:
- reg &= ~(GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100);
- break;
- }
-
- if (sky2->duplex == DUPLEX_FULL)
- reg |= GM_GPCR_DUP_FULL;
-
- /* turn off pause in 10/100mbps half duplex */
- else if (sky2->speed != SPEED_1000 &&
- hw->chip_id != CHIP_ID_YUKON_EC_U)
- sky2->tx_pause = sky2->rx_pause = 0;
- } else
- reg = GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100 | GM_GPCR_DUP_FULL;
-
- if (!sky2->tx_pause && !sky2->rx_pause) {
- sky2_write32(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
- reg |=
-     GM_GPCR_FC_TX_DIS | GM_GPCR_FC_RX_DIS | GM_GPCR_AU_FCT_DIS;
- } else if (sky2->tx_pause && !sky2->rx_pause) {
- /* disable Rx flow-control */
- reg |= GM_GPCR_FC_RX_DIS | GM_GPCR_AU_FCT_DIS;
- }
-
- gma_write16(hw, port, GM_GP_CTRL, reg);
-
 sky2_read16(hw, SK_REG(port, GMAC_IRQ_SRC));
 
+ /* Enable Transmit FIFO Underrun */
+ sky2_write8(hw, SK_REG(port, GMAC_IRQ_MSK), GMAC_DEF_MSK);
+
 spin_lock_bh(&sky2->phy_lock);
 sky2_phy_init(hw, port);
 spin_unlock_bh(&sky2->phy_lock);
@@ -1529,40 +1518,10 @@ static void sky2_link_up(struct sky2_por
 unsigned port = sky2->port;
 u16 reg;
 
- /* Enable Transmit FIFO Underrun */
- sky2_write8(hw, SK_REG(port, GMAC_IRQ_MSK), GMAC_DEF_MSK);
-
- reg = gma_read16(hw, port, GM_GP_CTRL);
- if (sky2->autoneg == AUTONEG_DISABLE) {
- reg |= GM_GPCR_AU_ALL_DIS;
-
- /* Is write/read necessary?  Copied from sky2_mac_init */
- gma_write16(hw, port, GM_GP_CTRL, reg);
- gma_read16(hw, port, GM_GP_CTRL);
-
- switch (sky2->speed) {
- case SPEED_1000:
- reg &= ~GM_GPCR_SPEED_100;
- reg |= GM_GPCR_SPEED_1000;
- break;
- case SPEED_100:
- reg &= ~GM_GPCR_SPEED_1000;
- reg |= GM_GPCR_SPEED_100;
- break;
- case SPEED_10:
- reg &= ~(GM_GPCR_SPEED_1000 | GM_GPCR_SPEED_100);
- break;
- }
- } else
- reg &= ~GM_GPCR_AU_ALL_DIS;
-
- if (sky2->duplex == DUPLEX_FULL || sky2->autoneg == AUTONEG_ENABLE)
- reg |= GM_GPCR_DUP_FULL;
-
 /* enable Rx/Tx */
+ reg = gma_read16(hw, port, GM_GP_CTRL);
 reg |= GM_GPCR_RX_ENA | GM_GPCR_TX_ENA;
 gma_write16(hw, port, GM_GP_CTRL, reg);
- gma_read16(hw, port, GM_GP_CTRL);
 
 gm_phy_write(hw, port, PHY_MARV_INT_MASK, PHY_M_DEF_MSK);
 
@@ -1616,7 +1575,6 @@ static void sky2_link_down(struct sky2_p
 reg = gma_read16(hw, port, GM_GP_CTRL);
 reg &= ~(GM_GPCR_RX_ENA | GM_GPCR_TX_ENA);
 gma_write16(hw, port, GM_GP_CTRL, reg);
- gma_read16(hw, port, GM_GP_CTRL); /* PCI post */
 
 if (sky2->rx_pause && !sky2->tx_pause) {
 /* restore Asymmetric Pause bit */
@@ -1633,6 +1591,7 @@ static void sky2_link_down(struct sky2_p
 
 if (netif_msg_link(sky2))
 printk(KERN_INFO PFX "%s: Link is down.\n", sky2->netdev->name);
+
 sky2_phy_init(hw, port);
 }
 
@@ -1673,8 +1632,11 @@ static int sky2_autoneg_done(struct sky2
 sky2->rx_pause = (aux & PHY_M_PS_RX_P_EN) != 0;
 sky2->tx_pause = (aux & PHY_M_PS_TX_P_EN) != 0;
 
- if ((sky2->tx_pause || sky2->rx_pause)
-     && !(sky2->speed < SPEED_1000 && sky2->duplex == DUPLEX_HALF))
+ if (sky2->duplex == DUPLEX_HALF && sky2->speed != SPEED_1000
+     && hw->chip_id != CHIP_ID_YUKON_EC_U)
+ sky2->rx_pause = sky2->tx_pause = 0;
+
+ if (sky2->rx_pause || sky2->tx_pause)
 sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_ON);
 else
 sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
@@ -1700,7 +1662,7 @@ static void sky2_phy_intr(struct sky2_hw
 printk(KERN_INFO PFX "%s: phy interrupt status 0x%x 0x%x\n",
        sky2->netdev->name, istatus, phystat);
 
- if (istatus & PHY_M_IS_AN_COMPL) {
+ if (sky2->autoneg == AUTONEG_ENABLE && (istatus & PHY_M_IS_AN_COMPL)) {
 if (sky2_autoneg_done(sky2, phystat) == 0)
 sky2_link_up(sky2);
 goto out;
@@ -2890,7 +2852,6 @@ static int sky2_set_pauseparam(struct ne
        struct ethtool_pauseparam *ecmd)
 {
 struct sky2_port *sky2 = netdev_priv(dev);
- int err = 0;
 
 sky2->autoneg = ecmd->autoneg;
 sky2->tx_pause = ecmd->tx_pause != 0;
@@ -2898,7 +2859,7 @@ static int sky2_set_pauseparam(struct ne
 
 sky2_phy_reinit(sky2);
 
- return err;
+ return 0;
 }
 
 static int sky2_get_coalesce(struct net_device *dev,
drivers/net/sky2.h +1 -1
index fa8af9f..a27194c 100644
@@ -1566,7 +1566,7 @@ enum {
 
 GMR_FS_ANY_ERR = GMR_FS_RX_FF_OV | GMR_FS_CRC_ERR |
   GMR_FS_FRAGMENT | GMR_FS_LONG_ERR |
-     GMR_FS_MII_ERR | GMR_FS_BAD_FC | GMR_FS_GOOD_FC |
+     GMR_FS_MII_ERR | GMR_FS_BAD_FC |
   GMR_FS_UN_SIZE | GMR_FS_JABBER,
 };
 
Signature · 6 lines
1.2.4

-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html 
#6 lore raw
In reply to Stephen Hemminger
Quoted 10 lines · Stephen Hemminger

Fix problems with transmit pause frames. The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

This maybe the cause of the lockup under load.

	http://bugzilla.kernel.org/show_bug.cgi?id=6839

Patch against netdev-2.6 git tree

Is this for #upstream (2.6.19) or #upstream-fixes (2.6.18-rc)?

It looks OK for #upstream, but too intrusive for #upstream-fixes.

Jeff

#2 lore raw
In reply to Andrew Hall
Quoted 4 lines · On Mon, 11 Sep 2006 11:35:11 +1000 "Andrew Hall"…

Stephen,

After some serious testing, this patch seems to fix the lockup issue completely. I manually applied these changes against the 2.6.17.13 release.

Any more problems?

#7 lore raw

PATCH 1/1 sky2: process tx pause frames.

In reply to Jeff Garzik

This patch already is in 2.6.17 stable, but the bigger version was pushed off till 2.6.19. Here is a less intrusive version that needs to go into 2.6.18 (or I'll end up sending it for 2.6.18.1). The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

drivers/net/sky2.h +1 -1
@@ -1566,7 +1566,7 @@
 
 	GMR_FS_ANY_ERR	= GMR_FS_RX_FF_OV | GMR_FS_CRC_ERR |
 			  GMR_FS_FRAGMENT | GMR_FS_LONG_ERR |
-		  	  GMR_FS_MII_ERR | GMR_FS_BAD_FC | GMR_FS_GOOD_FC |
+		  	  GMR_FS_MII_ERR | GMR_FS_BAD_FC |
 			  GMR_FS_UN_SIZE | GMR_FS_JABBER,
 };
 
#3 lore raw
In reply to Stephen Hemminger

Not that I've been able to uncover.

Thanks for fixing this..

----- Original Message ----- From: "Stephen Hemminger" <[email protected]>
To: "Andrew Hall" <[email protected]>
Cc: <[email protected]>
Sent: Wednesday, September 20, 2006 6:31 AM
Subject: Re: [PATCH 1/3] sky2: tx pause bug fix

Quoted 13 lines
Quoted 4 lines · On Mon, 11 Sep 2006 11:35:11 +1000 "Andrew Hall"…

Stephen,

After some serious testing, this patch seems to fix the lockup issue completely. I manually applied these changes against the 2.6.17.13 release.

Any more problems?
-
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#8 lore raw

Re: [PATCH] sky2: process tx pause frames.

In reply to Stephen Hemminger
Quoted 7 lines · Stephen Hemminger

This patch already is in 2.6.17 stable, but the bigger version was pushed off till 2.6.19. Here is a less intrusive version that needs to go into 2.6.18 (or I'll end up sending it for 2.6.18.1). The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

You'll need to send this to [email protected]...

Jeff

#9 lore raw

Re: [PATCH] sky2: process tx pause frames.

Quoted 7 lines · On Wed, 20 Sep 2006 17:49:41 -0400 Jeff Garzik <[email protected]> wrote:
Quoted 6 lines · Stephen Hemminger

This patch already is in 2.6.17 stable, but the bigger version was pushed off till 2.6.19. Here is a less intrusive version that needs to go into 2.6.18 (or I'll end up sending it for 2.6.18.1). The driver was telling the GMAC to flush (not process) pause frames. Manually disabling pause wasn't working because of problems in the setup.

Quoted 5 lines
Quoted 1 line

You'll need to send this to [email protected]...

Jeff

I did already thanks.

Signature · 1 line
Stephen Hemminger <[email protected]>
#4 lore raw
In reply to Stephen Hemminger

Unfortunately I spoke too soon.. :-(

The driver no longer appears to fail under load but still fails randomly with relatively light load:

Sep 22 07:04:38 localhost syslog.info -- MARK --
Sep 22 07:24:38 localhost syslog.info -- MARK --
Sep 22 07:31:52 localhost user.info kernel: NETDEV WATCHDOG: eth1: transmit timed out
Sep 22 07:31:52 localhost user.err kernel: sky2 eth1: tx timeout
Sep 22 07:31:52 localhost user.debug kernel: sky2 eth1: transmit ring 27 .. 498 report=27 done=27
Sep 22 07:31:52 localhost user.info kernel: sky2 hardware hung? flushing
Sep 22 07:39:27 localhost user.info kernel: NETDEV WATCHDOG: eth1: transmit timed out
Sep 22 07:39:27 localhost user.err kernel: sky2 eth1: tx timeout
Sep 22 07:39:27 localhost user.debug kernel: sky2 eth1: transmit ring 498 .. 457 report=27 done=27
Sep 22 07:39:27 localhost user.info kernel: sky2 status report lost?
Sep 22 07:40:07 localhost user.info kernel: NETDEV WATCHDOG: eth1: transmit timed out
Sep 22 07:40:07 localhost user.err kernel: sky2 eth1: tx timeout
Sep 22 07:40:07 localhost user.debug kernel: sky2 eth1: transmit ring 27 .. 498 report=27 done=27
Sep 22 07:40:07 localhost user.info kernel: sky2 hardware hung? flushing

The driver version that I am using is 1.6.1 with the tx pause bug fix manually applied. This is compiled against 2.6.17.13 stock.

This fault (since releasing this version into production) has now occured in 3 sites since upgrading them yesterday. The previous version of kernel that was being used was 2.6.12 (based on the old sk98lin driver) which didn't manifest any of these problems.

If there's anything I can do to help isolate this issue please let me know.

----- Original Message ----- From: "Stephen Hemminger" <[email protected]>
To: "Andrew Hall" <[email protected]>
Cc: <[email protected]>
Sent: Wednesday, September 20, 2006 6:31 AM
Subject: Re: [PATCH 1/3] sky2: tx pause bug fix

Quoted 9 lines
Quoted 4 lines · On Mon, 11 Sep 2006 11:35:11 +1000 "Andrew Hall"…

Stephen,

After some serious testing, this patch seems to fix the lockup issue completely. I manually applied these changes against the 2.6.17.13 release.

Any more problems?

#5 lore raw
Quoted 4 lines · Andrew Hall

Unfortunately I spoke too soon.. :-(

The driver no longer appears to fail under load but still fails randomly with relatively light load:

I don't know if this will help, but NETDEV watchdog timeouts in the bcm43xx driver were fixed by changing a netif_stop_queue call into a netif_tx_disable. Why wrapping the stop_queue inside netif_tx_lock_bh/netif_tx_unlock_bh pairs works is beyond me, but ...

Larry