[9fans] ELNK3 and 3CSOHO100-TX

[9fans] ELNK3 and 3CSOHO100-TX

[Plan9]

Hi all,

Did anyone test the recent 3CSOHO100-TX
The elnk3 drv detects it but it seems that card registers make it crash (no debug output, violent rebbot)...
On the 9pcfs kernel, a "memory" error message appears on console before rebooting.

The chip is a :
	3Com	40-0483-004
	0201T  46919437
	AGERE  40-04834

Any experience ?

Re: [9fans] ELNK3 and 3CSOHO100-TX

[plan9]

As we needed more Plan9 machines, we also modified 
/sys/srv/boot/pc/9load  and /sys/src/9/pc/etherelnk3.c
accordingly to Jim's advices and it works fine : 

	* Now the 9load `powered' floppy initialized the 3CSOHO100-TX and get a 9pccpudisk via BOOTP

	* The 9pccpud also initialize and uses correctly the card

<plan9@itic.ca> a crit dans le message news:<FIEDLHOOAEHBJAPIKGJCCEKPCAAA.plan9@itic.ca>...
Many Thanks Jim,

It now works.

We have modified /sys/src/fs/plan9pc/etherelnk3.c

The 9pcfs serving through 3CSOHO100-TX worked fine while moving 876.15 MB in both directions (pccpud/kfs<->9fs) 

----------

Re: [9fans] ELNK3 and 3CSOHO100-TX

[plan9]

Many Thanks Jim,

It now works.

We have modified /sys/src/fs/plan9pc/etherelnk3.c

The 9pcfs serving through 3CSOHO100-TX worked fine while moving 876.15 MB in both directions (pccpud/kfs<->9fs) 

Re: [9fans] ELNK3 and 3CSOHO100-TX

[jmk]

On Wed Mar  6 22:13:13 EST 2002, plan9@itic.ca wrote:
> Hi all,
> 
> Did anyone test the recent 3CSOHO100-TX
> The elnk3 drv detects it but it seems that card registers make it crash (no debug output, violent rebbot)...
> On the 9pcfs kernel, a "memory" error message appears on console before rebooting.
> 
> The chip is a :
> 	3Com	40-0483-004
> 	0201T  46919437
> 	AGERE  40-04834
> 
> Any experience ?

No experience, but if it has a PCI device id of 0x7646 then try adding this
to the etherelnk3reset function as I think it is just a 3C905B with a different
device id:

1) in the switch on the device id
	switch(did = eepromdata(did, port, 0x03)){
   add
	case 0x7646:
   after the
	case 0x9055:

2) later on there is a test for the 905B:
	if(did == 0x9055 || did == 0x9200 || did == 0x5157){
   (your version may or may not have the tests for 0x9200 and 0x5157)
   add a test for 0x7646 here too, e.g.
	if(did == 0x9055 || did == 0x7646 || did == 0x9200 || did == 0x5157){

--jim