Hello,
I've recently re-subscribed to this list. I'm glad to see you all again.

I read the chain about Cabletron EA41x (SCSI to Ethernet adaptor).
My EA412/19 works fine with my 030 Macs by changing one kernel
configuration.  I'd like to tell "how to" and  to ask about setting  
PDMA.

Subject: Re: configuring se0 device
Date: 09/13/2006 06:16:25
Thomas Carlson wrote:
 >I am not really in a position to do that with this old machine.  I
 >think I will delve back into history and try out various kernels in  
the
 >archives to see if they would play more nicely with the se0 device
 >driver.  At one point, according to the manual, the thing must have
 >worked (1997?).

Yes, the device driver for Cabletron EA41x is sys/dev/scsipi/if_se.c,
the first version was written by Ian W. Dall in 1997.
Now we have new 1.62 2006/09/07.
(Thanks a lot to Ian W. Dall for keeping the driver up to date.)


Cabletron EA412 with Quadra800 (NetBSD 4.99.3/ncrscsi)
ifconfig -a
sn0: flags=8822<BROADCAST,NOTRAILERS,SIMPLEX,MULTICAST> mtu 1500
         address: 08:00:07:1c:dd:0d

se0: flags=8863<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST>  
mtu 1500
         address: 00:00:1d:0c:c5:c8
         inet 192.168.1.28 netmask 0xffffff00 broadcast 192.168.1.255
         inet6 fe80::200:1dff:fe0c:c5c8%se0 prefixlen 64 scopeid 0x3

EA412 (se0)
ftp get NetBSD-3.0.1/Mac68k base.tgz from ftp.iij.ad.jp
150 Opening BINARY mode data connection for 'base.tgz' (22037614 bytes).
100% |****************************| 21521 KB  257.22 KB/s    00:00 ETA
22037614 bytes received in 01:23 (257.18 KB/s)

Quadra800 onBoard (sn0)
22037614 bytes received in 00:38 (560.37 KB/s)

Cabletron EA412 works with GENERIC(ncrscsi) kernel / Quadra800
without any special setting. It's transfer speed is almost a half of
onBoard sn0.  I saw "se0: not queued, error 12" more than 20 times
in this ftp session, but I don't care that because I think no one use
Cabletron EA41x with Quadra. It's very useful for PB1xx, ClassicII
and some LC which has an accelerator on PDS slot.

Unfortunatly, for those 030 Macs (sbc kernel), Cabletron EA41x does
NOT work. Thomas Carlson wrote that details correctly in the chain;
Subject: configuring se0 device   Date: 09/08/2006 05:05:32.


GENERICSBC kernel users see the next line in the dmesg;

 >sbc0 at obio0 addr 0: options=1<PDMA>

This 'options=1<PDMA>' let Cabletron EA41x fall asleep.
I don't know the reason why, anyway, I rudely changed this option
in my GENERICSBC kernel config file;

 ># SCSI controllers
 ># XXX - use only one of ncrscsi or sbc
 >#ncrscsi0 at obio? addr 0               # SCSI NCR 5380
 >sbc0    at obio? addr 0 flags 0x0       # MI SCSI NCR 5380
 >esp0    at obio? addr 0                 # SCSI NCR 53C9x
 >esp1    at obio? addr 1                 # SCSI NCR 53C9x

With changing sbc0 flags 0x1 to 0x0, yes, that's all, very simple and  
easy,
Cabletron EA412/19 wakes up and works lively with my ClassicII,
LC (serial console / 030 + FPU accelerator) and LCII. :)

NetBSD 4.0_BETA (GENERICSBC_LC2) #0: Sun Oct 15 01:18:42 JST 2006
         kazu@bop:/mac68k-4/usr/obj/sys/arch/mac68k/compile/ 
GENERICSBC_LC2
Apple Macintosh LC II  (68030)
cpu: delay factor 331
total memory = 10240 KB
avail memory = 6492 KB
(snip)
sbc0 at obio0 addr 0      <-- no options
scsibus0 at sbc0: 8 targets, 8 luns per target
(snip)
scsibus0: waiting 2 seconds for devices to settle...
sd0 at scsibus0 target 0 lun 0: <QUANTUM, TRB850S, 0405> disk fixed
sd0: 810 MB, 3653 cyl, 4 head, 113 sec, 512 bytes/sect x 1660299 sectors
sd0: async, 8-bit transfers
se0 at scsibus0 target 5 lun 0: <CABLETRN, EA412/14/19  D00, 1.00>  
processor fixed
se0: ethernet address 00:00:1d:08:47:d4
se0: async, 8-bit transfers
(snip)
Configuring network interfaces: se0se0: device enable failed