[PATCH] Add 3c90x: 3Com 3C90xB/C network driver

40 views
Skip to first unread message

David van Moolenbroek

unread,
Feb 11, 2015, 6:17:17 AM2/11/15
to mini...@googlegroups.com
A new network driver, for hardware that I happen to have lying around. The card type used to be pretty popular I believe; more PCI IDs will have to be added in the future..

http://gerrit.minix3.org/#/c/2939/

David

Roc Vallès

unread,
Feb 12, 2015, 5:04:45 AM2/12/15
to mini...@googlegroups.com
Very welcome!

I have one of those in some drawer :).

David van Moolenbroek

unread,
Feb 12, 2015, 6:50:59 AM2/12/15
to mini...@googlegroups.com
Cool! If you get the chance to test the driver with it, please let me know how it goes :)

Regards,
David

James Sprinkle

unread,
Feb 17, 2015, 9:45:33 AM2/17/15
to mini...@googlegroups.com
Built it and tried it out on a Tyan Tiger 2466 and it panics.  I know the adapter works fine under Linux.

I'll hack on it a bit and see if I can get it working.

David van Moolenbroek

unread,
Feb 20, 2015, 11:02:52 AM2/20/15
to mini...@googlegroups.com
Hi James,


On Tuesday, February 17, 2015 at 3:45:33 PM UTC+1, James Sprinkle wrote:
Built it and tried it out on a Tyan Tiger 2466 and it panics.  I know the adapter works fine under Linux.

I'll hack on it a bit and see if I can get it working.

Thank you for trying the driver! If you post the details of the panic, I might be able to help..

Regards,
David

James Sprinkle

unread,
Feb 20, 2015, 9:10:15 PM2/20/15
to mini...@googlegroups.com
Hi David!

Here's where it panics:

3c90x.c:127: assert "!(XLBC_READ_16(XLBC_STATUS_REG) & XLBC_STATUS_IN_PROGRESS)" failed, function "xlbc_issue_cmd"
3c90x(65563): panic: assert failed
syslib:painc: stacktrace: 0x804e5cf 0x804e4e6 0x8048a2c 0x804b298 0x804dc72 0x804df52 0x804b161 0x8048394 0x8048273 0x8048198
Request 0x700 to RS failed: specified endpoint is not alive (error 215)

Then it hangs.

David van Moolenbroek

unread,
Feb 23, 2015, 7:34:27 AM2/23/15
to mini...@googlegroups.com
Hey James,

On Saturday, February 21, 2015 at 3:10:15 AM UTC+1, James Sprinkle wrote:
Here's where it panics: [..]

Hmm, that looks pretty bad. Somehow the previous command had not completed at the time that a new one got issued, and that should not be possible at all. I saw this problem once, but it turned out I didn't have the card properly plugged into the PCI slot. I suppose that cannot be the case for you, though. One thing you could try is setting XLBC_CMD_TIMEOUT to a larger value and seeing if that makes a difference. A stacktrace with symbols might also help. I'll try to get my hands on more of these cards, though..

Thanks!
David
Reply all
Reply to author
Forward
0 new messages