I too have the 3c905C. My problem is not that it doesn't work, it does
(after a fashion). The problem I have is that during bootup, when dhcpcd
sets out to acquire an IP for my computer, it takes FOREVER. It eventually
gets around to it and bootup continues and my connection is fine (and the
proper module is loaded for the card).
This didn't happen after the inital 7.2 install - getting an IP during bootup
happened almost immediately. Later, I built my own kernels, as I am wont to
do, and I select the proper 3c90x vortex/boomerang module during xconfig.
This is apparently repeatable - after the initial install (which I have done
multiple times in the past due to occasionally REALLY screwing my system up
and finding it easier just to reinstall and go from there) all is well and I
get an IP address almost immediately. After building a kernel, however, even
if I do not end up using it, appears to dick it up so that during the bootup,
even though the module is identical, SLOOOOOW IP acquisition.
I am not using the kernels I have built myself right now and am back to the
2.2.17-21mdk kernel (and ITS modules), but the situation remains.
How can I fix this? What settings for the card might I mess with to get
things rolling so that getting an IP is almost immediate. The only thing I
an think that could be different is some card/network setting is slightly
altered from the default when a new kernel is built and installed.
I use dhcpcd-1.3.18pl3-4mdk, have a DSL connection through a Cisco 675
modem/router.
On Friday 09 February 2001 16:44, [EMAIL PROTECTED] you wrote:
> My brother has a 3c905B and I have a 3c905C. We are both using the
> 3C59x driver that mdk 7.2 auto selects and neither of us have had any
> problems in 100baseTX full duplex mode under mdk 7.2.
>
> My brother did once have a problem with very slow x-fer rates under an
> earlier version of Mandrake when he once switched the card from a
> 100baseTX switched port to a 10baseT port on a hub on the same (logical)
> subnet without bringing the machine or even the interface down. To the
> best that I could figure it looked as though the card had automatically
> switched down to 10baseT, but stayed in full duplex mode.
[...]
--
Against stupidity, the gods themselves contend in vain.
