Fabien Salvi wrote:
>
> Bernd Eckenfels wrote:
> >
> > In article <3D5140D0.813D27A9@cri74.org> you wrote:
> > > Almost everything works fine with a 2.4.18 kernel
> > > There are some little problems about media status that can "setfault"
> > > ifconfig
> >
> > can you provide me with more info, so i can fix this? What was the
> > segfaulting command line?
>
> Hello,
> The problem occurs in 2.4.18 kernel, I can't say if it's already the
> case in 2.4.19 because kernel freeze before I can try this :)
>
> But, according to Chat Tindel, it shoud have been fixed in the latest
> patches :
> ftp://ftp.sourceforge.net/pub/sourceforge/bonding/bonding-2.4.18-20020729
>
> If I'm not wrong, the problem is bout media reporting that not all
> network drivers support.
> I've forgotten to say that it freezes ifconfig with version 1.42
> With an older version, for example 1.39, I've got this :
>
> eth1 Link encap:Ethernet HWaddr 00:50:DA:33:BE:50 Media:net-tools
> 1.54
> inet addr:192.168.0.6 Bcast:192.168.0.127
> Mask:255.255.255.128
> UP BROADCAST RUNNING PROMISC ALLMULTI SLAVE MULTICAST
> MTU:1500 Metric:1
> RX packets:1309 errors:0 dropped:0 overruns:0 frame:0
> TX packets:623 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:100
> Interrupt:9 Base address:0x4400
>
> You can notice that the Media value is very strange...
>
> But the main problem is not this, it's the 2.4.19 freeze with bonding.
Thanks to Chat Tindel help, I've tried the latest bonding patches for
2.4.18 :
ftp://ftp.sourceforge.net//pub/sourceforge/bonding/bonding-2.4.18-20020729
And it works fine, it corrected freeze problems that have the previous
patch 20020617
So, I port the 2.4.18 patch for 2.4.19
The only change is about bonding.c in drivers/net/bonding.c
I have attached the correct patch for 2.4.19 to this email.
(md5sum : baae6a5551bb2cbbc1c2beaccb52790a)
I've carefully tested it and it solves all problems...
B.R.
-------------
Fabien SALVI Centre de Ressources Informatiques
Archamps, France -- http://www.cri74.org
PingOO GNU/linux distribution : http://www.pingoo.org
diff -Naur linux-2.4.19_orig/drivers/net/bonding.c linux-2.4.19/drivers/net/bonding.c
--- linux-2.4.19_orig/drivers/net/bonding.c Sat Aug 3 02:39:44 2002
+++ linux-2.4.19/drivers/net/bonding.c Thu Aug 8 11:58:09 2002
@@ -176,6 +176,16 @@
* Steve Mead <steve.mead at comdev dot cc>
* - Port Gleb Natapov's multicast support patchs from 2.4.12
* to 2.4.18 adding support for multicast.
+ *
+ * 2002/06/17 - Tony Cureington <tony.cureington * hp_com>
+ * - corrected uninitialized pointer (ifr.ifr_data) in bond_check_dev_link;
+ * actually changed function to use ETHTOOL, then MIIPHY, and finally
+ * MIIREG to determine the link status
+ * - fixed bad ifr_data pointer assignments in bond_ioctl
+ * - corrected mode 1 being reported as active-backup in bond_get_info;
+ * also added text to distinguish type of load balancing (rr or xor)
+ * - change arp_ip_target module param from "1-12s" (array of 12 ptrs)
+ * to "s" (a single ptr)
*/
#include <linux/config.h>
@@ -210,6 +220,9 @@
#include <linux/smp.h>
#include <linux/if_ether.h>
#include <linux/if_arp.h>
+#include <linux/mii.h>
+#include <linux/ethtool.h>
+
/* monitor all links that often (in milliseconds). <=0 disables monitoring */
#ifndef BOND_LINK_MON_INTERV
@@ -253,7 +266,7 @@
MODULE_PARM(mode, "i");
MODULE_PARM(arp_interval, "i");
MODULE_PARM_DESC(arp_interval, "arp interval in milliseconds");
-MODULE_PARM(arp_ip_target, "1-12s");
+MODULE_PARM(arp_ip_target, "s");
MODULE_PARM_DESC(arp_ip_target, "arp target in n.n.n.n form");
MODULE_PARM_DESC(mode, "Mode of operation : 0 for round robin, 1 for active-backup, 2 for xor");
MODULE_PARM(updelay, "i");
@@ -386,21 +399,54 @@
{
static int (* ioctl)(struct net_device *, struct ifreq *, int);
struct ifreq ifr;
- u16 *data = (u16 *)&ifr.ifr_data;
-
- /* data[0] automagically filled by the ioctl */
- data[1] = 1; /* MII location 1 reports Link Status */
+ struct mii_ioctl_data *mii;
+ struct ethtool_value etool;
+
+ if ((ioctl = dev->do_ioctl) != NULL) { /* ioctl to access MII */
+ /* TODO: set pointer to correct ioctl on a per team member */
+ /* bases to make this more efficient. that is, once */
+ /* we determine the correct ioctl, we will always */
+ /* call it and not the others for that team */
+ /* member. */
+
+ /* try SIOCETHTOOL ioctl, some drivers cache ETHTOOL_GLINK */
+ /* for a period of time; we need to encourage link status */
+ /* be reported by network drivers in real time; if the */
+ /* value is cached, the mmimon module parm may have no */
+ /* effect... */
+ etool.cmd = ETHTOOL_GLINK;
+ ifr.ifr_data = (char*)&etool;
+ if (ioctl(dev, &ifr, SIOCETHTOOL) == 0) {
+ if (etool.data == 1) {
+ return(MII_LINK_READY);
+ }
+ else {
+ return(0);
+ }
+ }
+
+ /*
+ * We cannot assume that SIOCGMIIPHY will also read a
+ * register; not all network drivers support that.
+ */
+
+ /* Yes, the mii is overlaid on the ifreq.ifr_ifru */
+ mii = (struct mii_ioctl_data *)&ifr.ifr_data;
+ if (ioctl(dev, &ifr, SIOCGMIIPHY) != 0) {
+ return MII_LINK_READY; /* can't tell */
+ }
+
+ mii->reg_num = 1;
+ if (ioctl(dev, &ifr, SIOCGMIIREG) == 0) {
+ /*
+ * mii->val_out contains MII reg 1, BMSR
+ * 0x0004 means link established
+ */
+ return mii->val_out;
+ }
- if (((ioctl = dev->do_ioctl) != NULL) && /* ioctl to access MII */
- (ioctl(dev, &ifr, SIOCGMIIPHY) == 0)) {
- /* now, data[3] contains info about link status :
- - data[3] & 0x04 means link up
- - data[3] & 0x20 means end of auto-negociation
- */
- return data[3];
- } else {
- return MII_LINK_READY; /* spoof link up ( we can't check it) */
}
+ return MII_LINK_READY; /* spoof link up ( we can't check it) */
}
static u16 bond_check_mii_link(bonding_t *bond)
@@ -598,12 +644,6 @@
*/
write_lock_irqsave(&bond->lock, flags);
- /*
- * Lock the master device so that noone trys to transmit
- * while we're changing things
- */
- spin_lock_bh(&master->xmit_lock);
-
/* set promiscuity flag to slaves */
if ( (master->flags & IFF_PROMISC) && !(bond->flags & IFF_PROMISC) )
bond_set_promiscuity(bond, 1);
@@ -637,7 +677,6 @@
bond_mc_list_destroy (bond);
bond_mc_list_copy (master->mc_list, bond, GFP_KERNEL);
- spin_unlock_bh(&master->xmit_lock);
write_unlock_irqrestore(&bond->lock, flags);
}
@@ -1707,7 +1746,7 @@
switch (cmd) {
case SIOCGMIIPHY:
- data = (u16 *)&ifr->ifr_data;
+ data = (u16 *)ifr->ifr_data;
if (data == NULL) {
return -EINVAL;
}
@@ -1718,7 +1757,7 @@
* We do this again just in case we were called by SIOCGMIIREG
* instead of SIOCGMIIPHY.
*/
- data = (u16 *)&ifr->ifr_data;
+ data = (u16 *)ifr->ifr_data;
if (data == NULL) {
return -EINVAL;
}
@@ -2035,7 +2074,28 @@
link = bond_check_mii_link(bond);
len += sprintf(buf + len, "Bonding Mode: ");
- len += sprintf(buf + len, "%s\n", mode ? "active-backup" : "load balancing");
+
+ switch (mode) {
+ case BOND_MODE_ACTIVEBACKUP:
+ len += sprintf(buf + len, "%s\n",
+ "active-backup");
+ break;
+
+ case BOND_MODE_ROUNDROBIN:
+ len += sprintf(buf + len, "%s\n",
+ "load balancing (round-robin)");
+ break;
+
+ case BOND_MODE_XOR:
+ len += sprintf(buf + len, "%s\n",
+ "load balancing (xor)");
+ break;
+
+ default:
+ len += sprintf(buf + len, "%s\n",
+ "unknown");
+ break;
+ }
if (mode == BOND_MODE_ACTIVEBACKUP) {
read_lock_irqsave(&bond->lock, flags);
@@ -2138,7 +2198,9 @@
}
static struct notifier_block bond_netdev_notifier = {
- notifier_call: bond_event,
+ bond_event,
+ NULL,
+ 0
};
static int __init bond_init(struct net_device *dev)
@@ -2282,7 +2344,32 @@
}
memset(dev_bonds, 0, max_bonds*sizeof(struct net_device));
+ if (updelay < 0) {
+ printk(KERN_WARNING
+ "bonding_init(): updelay module parameter (%d), "
+ "not in range 0-%d, so it was reset to 0\n",
+ updelay, INT_MAX);
+ updelay = 0;
+ }
+
+ if (downdelay < 0) {
+ printk(KERN_WARNING
+ "bonding_init(): downdelay module parameter (%d), "
+ "not in range 0-%d, so it was reset to 0\n",
+ downdelay, INT_MAX);
+ downdelay = 0;
+ }
+
+ if (arp_interval < 0) {
+ printk(KERN_WARNING
+ "bonding_init(): arp_interval module parameter (%d), "
+ "not in range 0-%d, so it was reset to %d\n",
+ arp_interval, INT_MAX, BOND_LINK_ARP_INTERV);
+ arp_interval = BOND_LINK_ARP_INTERV;
+ }
+
if (arp_ip_target) {
+ /* TODO: check and log bad ip address */
if (my_inet_aton(arp_ip_target, &arp_target) == 0) {
arp_interval = 0;
}
