Re: 2.4.0 test 7 has pcmcia problem
From: Mike Phillips <phillim2@home.com>
Subject: Re: 2.4.0 test 7 has pcmcia problem
Date: Tue, 29 Aug 2000 05:55:39 -0400
Message-ID: <[🔎] 39AB889B.58A5585E@home.com>
> > ah, i hadn't thought of trying that -- i really like the in-kernel
> > support because it seems to have done away w/ the "dhcpcd comes up
> > before the interface is ready during the boot process" problem, so
> > i don't want to go back ;-)
>
> Hmmm, you should be calling dhcp from the pcmcia scripts, my config
> uses pump to bring up the pcmcia adapter and get the interface
> information.
there is a timing problem w/ the current set up (2.2.x series) -- the
interface doesn't come up fast enough so that things like arpwatch
fail to start up properly. if i type control-s at an appropriate
moment and subsequently type control-q after waiting for the interface
to come up, the problems in question vanish. silly, isn't it?
this doesn't appear to be a problem when using pcmcia network cards
under 2.4.0-testX.
i think it is reasonable to assume that when one installs and
configures the dhcpcd package (or comparable), the user should not
have to write or modify any scripts/code -- i used to heavily tweak my
pcmcia scripts, but maintenance and transition to new machines were
painful and i realized other people were doing it too, so i stopped.
now i do everything manually every time to remind myself how painful
it is ;-)
when i was motivated to work on improving the situation before (if you
look in the archives of this list, you'll likely find a number of
threads i partcipcated in), i concluded after little success that the
pcmcia subsystem not being part of the kernel was the biggest problem.
so now i am content to wait as that seems to be changing -- once that
happens, i think i'll find my motivation returning.
> (In fact the interface opens and then pump closes it and then
> re-opens it - but that is a quirk of the token ring driver - but hey
> can't complain as I wrote it :)
;-)
Reply to: