[PATCH] Mixed PCI/ISA device name conflicts

[PATCH] Mixed PCI/ISA device name conflicts

[Stephen Hemminger]

PCI is getting initialized before the old ISA probing code, so on mixed
ISA/PCI systems a PCI card will get eth0 before the ISA probing starts.
Then when the first ISA card is probed it will (unsuccessfully) try and claim
eth0.

This fixes it by just having the ISA code skip eth0.  It relies on the convention
that if the ioaddr is one, the device should be skipped.

Patch against net-drivers-2.5-exp

diff -Nru a/net/core/dev.c b/net/core/dev.c
--- a/net/core/dev.c	Fri Dec 19 14:47:46 2003
+++ b/net/core/dev.c	Fri Dec 19 14:47:46 2003
@@ -382,6 +382,7 @@
  *	The found settings are set for the device to be used
  *	later in the device probing.
  *	Returns 0 if no settings found.
+ *	        1 if device already exists
  */
 unsigned long netdev_boot_base(const char *prefix, int unit)
 {
@@ -390,6 +391,9 @@
 	int i;
 
 	sprintf(name, "%s%d", prefix, unit);
+	if (__dev_get_by_name(name))
+		return 1;
+
 	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++)
 		if (!strcmp(name, s[i].name))
 			return s[i].map.base_addr;

Re: [PATCH] Mixed PCI/ISA device name conflicts

[Jeff Garzik]

Stephen Hemminger wrote:
> PCI is getting initialized before the old ISA probing code, so on mixed
> ISA/PCI systems a PCI card will get eth0 before the ISA probing starts.
> Then when the first ISA card is probed it will (unsuccessfully) try and claim
> eth0.
> 
> This fixes it by just having the ISA code skip eth0.  It relies on the convention
> that if the ioaddr is one, the device should be skipped.
> 
> Patch against net-drivers-2.5-exp
> 
> diff -Nru a/net/core/dev.c b/net/core/dev.c
> --- a/net/core/dev.c	Fri Dec 19 14:47:46 2003
> +++ b/net/core/dev.c	Fri Dec 19 14:47:46 2003
> @@ -382,6 +382,7 @@
>   *	The found settings are set for the device to be used
>   *	later in the device probing.
>   *	Returns 0 if no settings found.
> + *	        1 if device already exists
>   */
>  unsigned long netdev_boot_base(const char *prefix, int unit)
>  {
> @@ -390,6 +391,9 @@
>  	int i;
>  
>  	sprintf(name, "%s%d", prefix, unit);
> +	if (__dev_get_by_name(name))
> +		return 1;
> +
>  	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++)
>  		if (!strcmp(name, s[i].name))
>  			return s[i].map.base_addr;


Is this patch still needed?

	Jeff

[PATCH] Mixed PCI/ISA device name conflicts

[Stephen Hemminger]

In systems with mixed network cards, and all drivers compiled into
the kernel; the PCI device (eth0) will get probed first, before the ISA.

The problem is that the ISA device can mistakenly try to probe
for eth0.  The problem is that the ISA driver will not detect the failure
until it goes to call register_netdevice, and not all drivers have
perfect error unwind code. 

This patch short circuits the device probe, so it won't bother
looking for devices that already are registered.

diff -Nru a/net/core/dev.c b/net/core/dev.c
--- a/net/core/dev.c	Fri Apr 16 09:37:20 2004
+++ b/net/core/dev.c	Fri Apr 16 09:37:20 2004
@@ -432,6 +432,14 @@
 	int i;
 
 	sprintf(name, "%s%d", prefix, unit);
+
+	/*
+	 * If device already registered then return base of 1
+	 * to indicate not to probe for this interface
+	 */
+	if (__dev_get_by_name(name))
+		return 1;
+
 	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++)
 		if (!strcmp(name, s[i].name))
 			return s[i].map.base_addr;

Re: [PATCH] Mixed PCI/ISA device name conflicts

[Jeff Garzik]

applied, thanks for the resend