| View previous topic :: View next topic |
| Author |
Message |
Gerk
Retired Dev
Joined: 07 May 2002
Posts: 435
|
 Posted: Mon Sep 09, 2002 4:33 pm Post subject: Netra X1 sucesses and failures |
 |
|
Hello all
Just wanted to post some progress I have made getting 1.4 stuffs onto my netra X1.
It was a long painful journey, but I did manage to get a base system on it via the debian netinst.iso for woody (3.0). It involved many cdroms being burnt (mostly due to the fact that I screwed up and/or my cdrom I attached to this machine sucks at reading burned cd's)
The problem I am now stopped at is the onboard ethernet. I read in several places that the build on davicom NIC's now work with tulip drivers in recent kernels. Using a vanialla 2.4.19 kernel I managed to get the machine booting from serial console, but am unable ot use onboard eth as of yet. It was suggested I try latest 2.4.20-pre5 patched kernel, so I am doing so right now.
seeing very strange IRQ settings for onboard nic's presently which leads me to believe the 2.4.19 is not too sparc64 happy...
| Code: |
Bus 0, device 5, function 0:
Class 0200: PCI device 1282:9102 (rev 49).
IRQ 6529856.
Master Capable. Latency=160. Min Gnt=20.Max Lat=40.
I/O at 0x1fe02010100 [0x1fe020101ff].
Non-prefetchable 32 bit memory at 0x1ff00002000 [0x1ff000020ff].
Bus 0, device 12, function 0:
Class 0200: PCI device 1282:9102 (rev 49).
IRQ 6529152.
Master Capable. Latency=160. Min Gnt=20.Max Lat=40.
I/O at 0x1fe02010000 [0x1fe020100ff].
Non-prefetchable 32 bit memory at 0x1ff00000000 [0x1ff000000ff].
|
Note the IRQ values in the millions  All my pci devices show up like this. If anyone can offer further advice I'm all ears  I'll post my progress with patched kernel.
Gerk |
|
| Back to top |
|
 |
Gerk
Retired Dev
Joined: 07 May 2002
Posts: 435
|
| Posted: Mon Sep 09, 2002 5:38 pm Post subject: |
|
|
OK.. using 2.4.20-pre5 I think I am getting closer to seeing where the problem lies...
| Code: |
PCI: Address space collision on region 6 [000001ff00080000:000001ff000bffff] of device Davicom Semiconductor, Inc. Ethernet 100/10 MBit
|
The previous kernel's didn't seem to either have this problem or detect it. Also the dmfe driver doesn't seem tobe able to set the MAC address of the cards, this was true with both 2.4.19 and 2.4.20-pre5:
| Code: |
eth0: Davicom DM9102 at pci00:05.0, 00:00:00:00:00:00, irq 6751040.
eth1: Davicom DM9102 at pci00:0c.0, 00:00:00:00:00:00, irq 6750336.
|
Using the tulip drivers seem toget a MAC address, but they don't function.
| Code: |
Linux Tulip driver version 0.9.15-pre12 (Aug 9, 2002)
tulip0: Old style EEPROM with no media selection information.
tulip0: MII transceiver #1 config 1100 status 7809 advertising 01e1.
eth0: Davicom DM9102/DM9102A rev 49 at 0x1fe02010100, EEPROM not present, 00:03:
BA:04:D1:49, IRQ 6751040.
tulip1: Old style EEPROM with no media selection information.
tulip1: MII transceiver #1 config 1000 status 782d advertising 01e1.
eth1: Davicom DM9102/DM9102A rev 49 at 0x1fe02010000, EEPROM not present, 00:03:
BA:04:D1:48, IRQ 6750336.
|
I got to ping the outside world 3 times before ethernet died with tulip.. still fighting
Gerk |
|
| Back to top |
|
|
Gerk
Retired Dev
Joined: 07 May 2002
Posts: 435
|
| Posted: Mon Sep 09, 2002 5:49 pm Post subject: |
|
|
ok, its finally working... here's the trick
modprobe tulip options 0
(which forces it to single duplex 10Bt mode) seems to allow me to get some things going. It's been up hill but worth it now that I can finally emerge rsync and get to things here
I will investigate further with the tulip module to see how things pan out for better ethernet support. |
|
| Back to top |
|
|
kluzz
n00b
Joined: 29 Jun 2002
Posts: 18
Location: Bømlo, Norway
|
| Posted: Tue Sep 10, 2002 11:21 pm Post subject: Sun Fire V100 (next gen X1, I think) |
|
|
After jerking around with the old gentoo 1.1a boot image (which works nicely for sparc32 installs) i finally ended up using the same netinst image you mentioned. I located the tulip module somewhere debian-related, and after using insmod tulip.o full_duplex=1 it networked like a charm. Now I'm bootstrapping the sparc64-2.0 profile, and it seems to be going very well...
I'll post again when there's progress.
If anyone requires it, I'll post a more detailed description of what I did.
_________________
--
Jan Fredrik Leversund
Dark Overlord for Hire |
|
| Back to top |
|
|
howax
n00b
Joined: 21 Jan 2004
Posts: 29
Location: /home/howax
|
| Posted: Thu Jan 20, 2005 8:24 pm Post subject: |
|
|
Hi
I have the same problem, however i cant seem to find anyway to work around it.
iam booting the 2004.3 directly from cdrom, after trying about 400 cdrom drives i finally found one that works correctly with the netra x1.
anyways, now to thise wired network problems. iam tryin to modprobe tulip options 0 to force 10Mbit. however, the live cd loads the tulip driver automaticly, so, i dosent seem to happen anything then i try loading it again.
so, how do i "unload" tulip, so i can load it in again with the right parameters?
Regards
PS. This is kinda a sweet machine, ive just found it laying around at work, for like 3 years. |
|
| Back to top |
|
|
howax
n00b
Joined: 21 Jan 2004
Posts: 29
Location: /home/howax
|
| Posted: Thu Jan 20, 2005 9:27 pm Post subject: |
|
|
after browsing the "google internet" ive found this page:
http://www.cise.ufl.edu/~jnw/SparcLinux/
it talks about;
The ethernet interfaces (as numbered on the back of the box) are numbered in the wrong order for Linux.
We plugged our cable into the interface labelled 1 on the back, rather than 0
Can anyone confirm this ?
ill try this tomorrow as i cant get in touch with the box right now... |
|
| Back to top |
|
|
howax
n00b
Joined: 21 Jan 2004
Posts: 29
Location: /home/howax
|
| Posted: Wed Jan 26, 2005 9:57 am Post subject: |
|
|
okay, indeed the eth0 and eth1 is orderd wrong at the back of the netra x1, as net0 is eth1 and net1 is eth0
however.. im getting alot of errors at the eth0.
eth0 Link encap:Ethernet HWaddr 00:*:BA:*:F3:*
inet addr:*.192.*.254 Bcast:*.192.*.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1489856 errors:0 dropped:0 overruns:0 frame:0
TX packets:208 errors:217435 dropped:0 overruns:0 carrier:217435
collisions:0 txqueuelen:1000
RX bytes:163662959 (156.0 Mb) TX bytes:15032 (14.6 Kb)
Interrupt:64 Base address:0x100
any suggestions ? |
|
| Back to top |
|
|
twos
n00b
Joined: 13 Aug 2004
Posts: 2
|
| Posted: Sat Jan 29, 2005 12:02 pm Post subject: |
|
|
| Quote: | | however.. im getting alot of errors at the eth0 |
Sounds like duplex mis-match. Might try forcing a speed and duplex. Look here for all the tulip drver options:
http://www.scyld.com/tulip.html |
|
| Back to top |
|
|
|
Gentoo on Sparc
