Re: 2.6.27.19 + 28.7: network timeouts for r8169 and 8139too

From: David Dillow
Date: Wed May 13 2009 - 22:38:56 EST


On Tue, 2009-05-12 at 22:29 +0200, Michael Riepe wrote:
> Hi!
>
> David Dillow wrote:
>
> > I was saying that I don't think the timeouts are necessarily the NIC
> > chipset -- or the bridge chip for that matter -- having issues with
> > MSI. There were some substantial IRQ handling changes in 2.6.28 and my
> > bisection of the problem seem to lead into that code. I'll try this
> > later tonight hopefully, but can you try to run 2.6.27 with the current
> > r8169 driver and see if it is solid for you? That way it is using the
> > same driver code, but avoids the IRQ changes.
>
> Unfortunately, 2.6.27 won't build with r8169.c copied from 2.6.29.

You are correct, and I should have thought about that. The following
patch reverts the following commits:

288379 net: Remove redundant NAPI functions
908a7a net: Remove unused netdev arg from some NAPI interfaces.
008298 netdev: add more functions to netdevice ops
8b4ab2 r8169: convert to net_device_ops
babcda drivers/net: Kill now superfluous ->last_rx stores.

The patched driver runs on 2.6.27 and survives my 5 minutes 'dd
if=/dev/zero bs=1024k | nc target 9000' test which usually dies in less
than 90 seconds on 2.6.28+.

To use the patch, just copy r8169.c from HEAD and use patch -p1.


diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 0b6e8c8..2d751bf 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -490,7 +490,6 @@ struct rtl8169_private {
void (*hw_start)(struct net_device *);
unsigned int (*phy_reset_pending)(void __iomem *);
unsigned int (*link_ok)(void __iomem *);
- int (*do_ioctl)(struct rtl8169_private *tp, struct mii_ioctl_data *data, int cmd);
int pcie_cap;
struct delayed_work task;
unsigned features;
@@ -1850,11 +1849,9 @@ static int rtl8169_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
struct rtl8169_private *tp = netdev_priv(dev);
struct mii_ioctl_data *data = if_mii(ifr);

- return netif_running(dev) ? tp->do_ioctl(tp, data, cmd) : -ENODEV;
-}
+ if (!netif_running(dev))
+ return -ENODEV;

-static int rtl_xmii_ioctl(struct rtl8169_private *tp, struct mii_ioctl_data *data, int cmd)
-{
switch (cmd) {
case SIOCGMIIPHY:
data->phy_id = 32; /* Internal PHY */
@@ -1873,11 +1870,6 @@ static int rtl_xmii_ioctl(struct rtl8169_private *tp, struct mii_ioctl_data *dat
return -EOPNOTSUPP;
}

-static int rtl_tbi_ioctl(struct rtl8169_private *tp, struct mii_ioctl_data *data, int cmd)
-{
- return -EOPNOTSUPP;
-}
-
static const struct rtl_cfg_info {
void (*hw_start)(struct net_device *);
unsigned int region;
@@ -1943,26 +1935,6 @@ static void rtl_disable_msi(struct pci_dev *pdev, struct rtl8169_private *tp)
}
}

-static const struct net_device_ops rtl8169_netdev_ops = {
- .ndo_open = rtl8169_open,
- .ndo_stop = rtl8169_close,
- .ndo_get_stats = rtl8169_get_stats,
- .ndo_start_xmit = rtl8169_start_xmit,
- .ndo_tx_timeout = rtl8169_tx_timeout,
- .ndo_validate_addr = eth_validate_addr,
- .ndo_change_mtu = rtl8169_change_mtu,
- .ndo_set_mac_address = rtl_set_mac_address,
- .ndo_do_ioctl = rtl8169_ioctl,
- .ndo_set_multicast_list = rtl_set_rx_mode,
-#ifdef CONFIG_R8169_VLAN
- .ndo_vlan_rx_register = rtl8169_vlan_rx_register,
-#endif
-#ifdef CONFIG_NET_POLL_CONTROLLER
- .ndo_poll_controller = rtl8169_netpoll,
-#endif
-
-};
-
static int __devinit
rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
{
@@ -1989,7 +1961,6 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
}

SET_NETDEV_DEV(dev, &pdev->dev);
- dev->netdev_ops = &rtl8169_netdev_ops;
tp = netdev_priv(dev);
tp->dev = dev;
tp->pci_dev = pdev;
@@ -2126,7 +2097,6 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
tp->phy_reset_enable = rtl8169_tbi_reset_enable;
tp->phy_reset_pending = rtl8169_tbi_reset_pending;
tp->link_ok = rtl8169_tbi_link_ok;
- tp->do_ioctl = rtl_tbi_ioctl;

tp->phy_1000_ctrl_reg = ADVERTISE_1000FULL; /* Implied by TBI */
} else {
@@ -2135,7 +2105,8 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
tp->phy_reset_enable = rtl8169_xmii_reset_enable;
tp->phy_reset_pending = rtl8169_xmii_reset_pending;
tp->link_ok = rtl8169_xmii_link_ok;
- tp->do_ioctl = rtl_xmii_ioctl;
+
+ dev->do_ioctl = rtl8169_ioctl;
}

spin_lock_init(&tp->lock);
@@ -2147,15 +2118,28 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
dev->dev_addr[i] = RTL_R8(MAC0 + i);
memcpy(dev->perm_addr, dev->dev_addr, dev->addr_len);

+ dev->open = rtl8169_open;
+ dev->hard_start_xmit = rtl8169_start_xmit;
+ dev->get_stats = rtl8169_get_stats;
SET_ETHTOOL_OPS(dev, &rtl8169_ethtool_ops);
+ dev->stop = rtl8169_close;
+ dev->tx_timeout = rtl8169_tx_timeout;
+ dev->set_multicast_list = rtl_set_rx_mode;
dev->watchdog_timeo = RTL8169_TX_TIMEOUT;
dev->irq = pdev->irq;
dev->base_addr = (unsigned long) ioaddr;
+ dev->change_mtu = rtl8169_change_mtu;
+ dev->set_mac_address = rtl_set_mac_address;

netif_napi_add(dev, &tp->napi, rtl8169_poll, R8169_NAPI_WEIGHT);

#ifdef CONFIG_R8169_VLAN
dev->features |= NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX;
+ dev->vlan_rx_register = rtl8169_vlan_rx_register;
+#endif
+
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ dev->poll_controller = rtl8169_netpoll;
#endif

tp->intr_mask = 0xffff;
@@ -3513,6 +3497,7 @@ static int rtl8169_rx_interrupt(struct net_device *dev,
if (rtl8169_rx_vlan_skb(tp, desc, skb) < 0)
netif_receive_skb(skb);

+ dev->last_rx = jiffies;
dev->stats.rx_bytes += pkt_size;
dev->stats.rx_packets++;
}
@@ -3594,8 +3579,8 @@ static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance)
RTL_W16(IntrMask, tp->intr_event & ~tp->napi_event);
tp->intr_mask = ~tp->napi_event;

- if (likely(napi_schedule_prep(&tp->napi)))
- __napi_schedule(&tp->napi);
+ if (likely(netif_rx_schedule_prep(dev, &tp->napi)))
+ __netif_rx_schedule(dev, &tp->napi);
else if (netif_msg_intr(tp)) {
printk(KERN_INFO "%s: interrupt %04x in poll\n",
dev->name, status);
@@ -3616,7 +3601,7 @@ static int rtl8169_poll(struct napi_struct *napi, int budget)
rtl8169_tx_interrupt(dev, tp, ioaddr);

if (work_done < budget) {
- napi_complete(napi);
+ netif_rx_complete(dev, napi);
tp->intr_mask = 0xffff;
/*
* 20040426: the barrier is not strictly required but the


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/