Problems with ULI 526x driver

Problems with ULI 526x driver

[Timur Tabi]

Hi,

I'm trying to get the ULI 526x driver working on a Freescale 8610 reference 
board, and it can't connect.  When I turn on debug, I see this:

Configuring network interfaces...uli526x: uli526x_open 0
uli526x: uli526x_init() 0
uli526x: uli526x_descriptor_init() 0
uli526x: send_filter_frame() 0
uli526x: uli526x_set_filter_mode() 0
uli526x: Set multicast address 1
uli526x: send_filter_frame() 0
uli526x: uli526x_get_stats 0
uli526x: uli526x_set_filter_mode() 0
uli526x: Set multicast address 2
uli526x: send_filter_frame() 0
ADDRCONF(NETDEV_UP): eth0: link is not ready
uli526x: uli526x_set_filter_mode() 0
uli526x: Set multicast address 2
uli526x: send_filter_frame() 0
uli526x: uli526x_get_stats 0
uli526x: uli526x_set_filter_mode() 0
uli526x: Set multicast address 2
uli526x: send_filter_frame() 0
done.

and then some time later, I see thisL

uli526x: Link link OK 3
uli526x: eth0 NIC Link is Up 100 Mbps Full duplex
ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
uli526x: uli526x_set_filter_mode() 0
uli526x: Set multicast address 3
uli526x: send_filter_frame() 0
uli526x: uli526x_get_stats 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0

If I try to ping, I get this:

  ping 10.82.48.106
PING 10.82.48.106 (10.82.48.106) 56(84) bytes of data.
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0
 From 10.82.19.159 icmp_seq=2 Destination Host Unreachable
 From 10.82.19.159 icmp_seq=3 Destination Host Unreachable
 From 10.82.19.159 icmp_seq=4 Destination Host Unreachable
uli526x: uli526x_start_xmit 0
uli526x: uli526x_start_xmit 0


A long time ago, there was a patch for the Tulip driver that added support for 
the ULI 526x.  This support was removed in commit 
ea8f400c98ec9ae0604bc5a6721174ef68635815:

http://git.kernel.org/?p=linux/kernel/git/jgarzik/netdev-2.6.git;a=commit;h=ea8f400c98ec9ae0604bc5a6721174ef68635815

Now, if I revert those changes, the Tulip driver works for me.  So there's 
something that the tulip+uli526x driver does that the stand-alone uli526x driver 
doesn't do.

Unfortunately, I don't really know anything about the tulip or uli526x drivers. 
  Is there anyone who can give me a clue as to what could be wrong?

-- 
Timur Tabi
Linux kernel developer at Freescale

Re: Problems with ULI 526x driver

[Timur Tabi]

Timur Tabi wrote:
> Hi,
> 
> I'm trying to get the ULI 526x driver working on a Freescale 8610 
> reference board, and it can't connect.  When I turn on debug, I see this:

Update: I can ping anything on my subnet, but nothing else.  Also, it looks like 
my MAC address is hosed:

eth0      Link encap:Ethernet  HWaddr 00:00:00:00:00:00
           inet addr:10.82.19.159  Bcast:10.82.19.255  Mask:255.255.255.0
           inet6 addr: fe80::200:ff:fe00:0/64 Scope:Link
           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
           RX packets:5 errors:0 dropped:0 overruns:0 frame:0
           TX packets:26 errors:0 dropped:0 overruns:0 carrier:0
           collisions:0 txqueuelen:1000
           RX bytes:414 (414.0 b)  TX bytes:1224 (1.1 KiB)
           Interrupt:18 Base address:0x1000

It turns out that the driver is getting the MAC address from the ID table:

	if(((u16 *) db->srom)[0] == 0xffff || ((u16 *) db->srom)[0] == 0)		/* SROM 
absent, so read MAC address from ID Table */
	{


In my case, the address that is returned is all zeros.  I don't know anything 
about the ULI 526x hardware.  Is it possible that there's a missing EEPROM on my 
board that's supposed to hold the MAC address?  Are these registers user-writable?

-- 
Timur Tabi
Linux kernel developer at Freescale

Re: Problems with ULI 526x driver

[Francois Romieu]

Timur Tabi <timur@freescale.com> :
[...]
> Update: I can ping anything on my subnet, but nothing else.  Also, it looks 
> like my MAC address is hosed:
[...]
> It turns out that the driver is getting the MAC address from the ID table:
> 
> 	if(((u16 *) db->srom)[0] == 0xffff || ((u16 *) db->srom)[0] == 0) /* 
> 	SROM absent, so read MAC address from ID Table */
> 	{
> 
> 
> In my case, the address that is returned is all zeros.  I don't know 
> anything about the ULI 526x hardware.  Is it possible that there's a 
> missing EEPROM on my board that's supposed to hold the MAC address?  Are 
> these registers user-writable?

You previously said that the tulip driver + reverted change works with
your card (TULIP_MMIO was not set, right ?). I guess it did correctly set
the HW address, did it not ?

-- 
Ueimor