On Thu, Jul 12, 2001 at 11:57:25AM -0300, Paulo Afonso Graner Fessel wrote:
> Hello.
> 
> I'm trying to use a Performa 6360 as a firewall. After going into some
> frustration with Linux, I've decided to give NetBSD a try.
> 
> The 6360 has only one PCI slot available; OTOH, I had the chance to buy
> a Compaq Netelligent Dual board by a very reasonable price; this way I'd
> be able to use the 6360 as a router/firewall.
> 
> The problem is: NetBSD does find the board, but appears not to be sure
> about what PHY it should use. Here's an excerpt of the boot log:
> 
> pchb0 at pci0 dev 11 function
> 0                                                 
> pchb0: Apple Computer Bandit Host-PCI Bridge (rev.
> 0x03)                        
> ppb0 at pci0 dev 13 function 0: IBM 82351 PCI-PCI Bridge (rev.
> 0x01)            
> pci1 at ppb0 bus
> 1                                                              
> pci1: i/o space, memory space
> enabled                                           
> tl0 at pci1 dev 0 function
> 0                                                    
> tl0: Compaq Dual Port Netelligent 10/100
> TX                                     
> tl0: Ethernet address
> 00:80:5f:a7:b2:3f                                         
> tl0: interrupting at irq
> 23                                                     
> ukphy0 at tl0 phy 1: Generic IEEE 802.3u media
> interface                        
> ukphy0: DP83840 10/100 media interface (OUI 0x1000e8, model 0x0000),
> rev. 1     
> ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX,
> auto                    
> tlphy0 at tl0 phy 31: ThunderLAN 10baseT media interface, rev.
> 5                
> tlphy0:
> 10base5                                                                 
> tl1 at pci1 dev 1 function
> 0                                                    
> tl1: Compaq Dual Port Netelligent 10/100
> TX                                     
> tl1: Ethernet address
> 00:80:5f:a7:b2:bf                                         
> tl1: interrupting at irq
> 23                                                     
> ukphy1 at tl1 phy 1: Generic IEEE 802.3u media
> interface                        
> ukphy1: DP83840 10/100 media interface (OUI 0x1000e8, model 0x0000),
> rev. 1     
> ukphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX,
> auto                    
> tlphy1 at tl1 phy 31: ThunderLAN 10baseT media interface, rev.
> 5                
> tlphy1: 10base5
> 
> Seems that the PHY being used is tlphy, right? But if I compile NetBSD

Yes, and this is perfectly fine. Some of the netelligent boards have 2
PHY (in fact you can have up to 32 PHY on the MII bus).
You can select which PHY you want to use with the 'instance' keyword of
ifconfig.

> 
> Also, when I use the ukphy I get even a link, as ifconfig -a reports
> (only tl0 is attached to my hub):
> 
> tl0: flags=8822<BROADCAST,NOTRAILERS,SIMPLEX,MULTICAST> mtu
> 1500                
>         address:
> 00:80:5f:a7:b2:3f                                              
>         media: Ethernet autoselect
> (10baseT)                                    
>         status:
> active                                                          
> tl1: flags=8822<BROADCAST,NOTRAILERS,SIMPLEX,MULTICAST> mtu
> 1500                
>         address:
> 00:80:5f:a7:b2:bf                                              
>         media: Ethernet autoselect
> (none)                                       
>         status: no
> carrier                                                      
> lo0: flags=8009<UP,LOOPBACK,MULTICAST> mtu
> 33220                                
>         inet 127.0.0.1 netmask 0xff000000
> 
> But as soon as I try to ifconfig tl0 up I get these messages and the
> link goes down:
> 
> Jul 12 10:26:18 performa /netbsd: tl0: check condition,
> intvect=0x200000, ch_par
> am=0x180005
> Jul 12 10:26:24 performa last message repeated 11 times
> Jul 12 10:26:24 performa /netbsd: tl0: device timeout
> 
> I've found now that the PHY of this board is a NS83840 chip, but I still
> get some strange messages in the log and the board does not work,
> exactly as before.

Well, it's possible that this driver has some endian issue. Also it doesn't
properly uses bus-dma.
Give me a few days and I'll try to look at this (I just got a brand new
Ultra/5 on which I'll be able to test my PCI drivers :)

--
Manuel Bouyer <bouyer@antioche.eu.org>
--