There are two Digital 2xx4x drivers in the kernel (2.4.0-test1 on a
Compaq 1800T, Pentium III 450 MHz), one formerly tulip.c but now several
files in the drivers/net/tulip directory(0.9.4.3) , and the other
net/de4x5.c (V0.545 1999/11/28)
If I configure both of them to be built as modules, then tulip.o will
get loaded, but won't work. There are no error
messages, just no connectivity (I'm trying to talk through a cable
modem, getting set up with DHCP). I'd accidentally built my kernel
without virtual terminal support for the console, which makes the kernel
boot messages not display but helpfully saves them in /var/adm/messages:
tulip.o messages:
May 27 16:52:38 kingofswing kernel: Linux Tulip driver version 0.9.4.3
(Apr 14, 2000)
May 27 16:52:38 kingofswing kernel: eth0: Digital DS21143 Tulip rev 65
at 0xfc00, 00:08:C7:03:4A:37, IRQ 11.
May 27 16:52:38 kingofswing kernel: eth0: EEPROM default media type
Autosense.
May 27 16:52:38 kingofswing kernel: eth0: Index #0 - Media MII (#11)
described by a 21142 MII PHY (3) block.
May 27 16:52:38 kingofswing kernel: eth0: MII transceiver #0 config
0100 status 780d advertising 01e1.
I had thought that if I built both modules, I could just remove the
tulip.o file from /lib/modules/2.4.0-test1 and
restart to allow the other one to take effect, but that is not the case,
apparently there is some stub code or something that is still statically
linked. I'd still see the above message and de4x5.o would say something
about "resource in use" or some such.
But reconfiguring without the tulip driver included at all, but with
de4x5.o as a module, I get ethernet just fine. I don't have the boot
message (but it wasn't soo descriptive) but I have ifconfig eth0 with
generic driver:
eth0 Link encap:Ethernet HWaddr 00:08:C7:03:4A:37
inet addr:205.251.210.122 Bcast:205.251.211.255
Mask:255.255.252.0
UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1
RX packets:2481 errors:0 dropped:0 overruns:0 frame:0
TX packets:1733 errors:3 dropped:0 overruns:0 carrier:3
collisions:0 txqueuelen:100
Interrupt:11 Base address:0xfc00
I had basically the same behaviour with kernel 2.2.13. But
interestingly I _was_ able to get tulip.o to work with some development
version I got off the developer's website and stuck in the 2.2.13 kernel
replacing the one that is supplied.
I'm afraid I don't know which driver this was as I lost my old linux
installation when NT ate my disk (the installer doesn't understand large
hard drives); if there is an archive of old revs of tulip.c around, I
could try compiling various versions into my 2.2.13 kernel to see which
work and which don.t.
I describe this, and the configuration of my laptop and how to get it to
work with various OSes, at:
http://www.goingware.com/laptop
I will archive this mail and anything relevant to my Compaq at that
page.
So to be more clear about how I configured it, I used "make xconfig":
Network Device Support -
Ethernet (10 or 100 MBit) -
Generic DECchip & DIGITAL EtherWORKS PCI/EISA - built as module
DECchip Tulip (dc21x4x) PCI support - say no here on the Compaq
1800T
Based on the instructions in the config, this is not what you'd
expect. The de4x5 help should probably be updated to say that it can
support the tulip, and the tulip should be updated to say that you might
try the de4x5 if the tulip doesn't work for you.
Of course if anyone wants me to try anything to test fixes I'm happy to
oblige.
Michael D. Crawford
GoingWare, Inc. - Expert Software Development and Consulting
http://www.goingware.com
crawford@goingware.com
Tilting at Windmills for a Better Tomorrow.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat May 27 2000 - 17:56:14 EDT