TCLUG Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCLUG:2546] Configuring for 2 Network Adapters



Did the "Reset Configuration Data" bios trick work?

Peter Lukas

On Wed, 9 Dec 1998, Gordon Pedersen wrote:

> Peter/Richard/Chris:
> 
> Got down to nub of it last night and found that though the (stock Redhat
> 5.1) driver would load both old and new NIC, it somehow couldn't really
> use the new NIC.  I didn't try to figure out what did and didn't
> work--enough fish to fry!  The driver was the first fix I tried and it
> worked.  Thanks.  Cards load at irq 15 and 14 now; i'll test irq sharing
> another day--probably better get it working on NT first :-).  Guess I
> should check out the cards sheer bit-rattling performance after what
> some have written here but that will have to wait.
> 
> What a drag to be so close yet miss.  At least you have a good
> connection at work (I presume).
> 
> gp
> 
> Peter Lukas wrote:
> > 
> > You lucky dog, I'm about 5 feet over the limit for DSL.  :-).  IRQ sharing
> > is a driver issue.  Linux drivers seem to handle this without a problem,
> > as for NT, does it handle anything without a problem?  There's a few
> > things you can do.  If they're both PCI adapters, you should "Reset
> > Configuration Data" in your bios.  This will allow the BIOS to undo any
> > conflicting IRQ settings and hopefully reassign them to a non-conflicting
> > location.  Also, you might have better success loading each adapter driver
> > as a module.  The problem with using 2 similar cards is that it might be
> > tough to determine which one was loaded.  In any event, the DSL connection
> > isn't connected to a switch that does either 10/100 FD, so it shouldn't be
> > an issue.  To be safe, you can load the modules forcing the 905b to 10MB
> > half-duplex simply by passing no options to the module as you load it.
> > There are newer drivers usually making their way into kernels between
> > point releases, so chances are, the driver with 2.0.35 is older than the
> > one packaged with 2.0.36.  If all else fails, you can always pick up a
> > nice netgear PCI 10/100 adapter for about $30.
> > 
> > Peter Lukas
> > Math Systems Office
> > University of Minnesota
> > 
> > On Mon, 7 Dec 1998, Gordon Pedersen wrote:
> > 
> > > Thanks, Peter.  I had no idea these cards were extra finicky.  I won't
> > > be near the machine till tomorrow evening but here's a followup
> > > question:
> > >
> > >       I'm connecting it to one of the netspeed routers supplied by US West
> > > for their RADSL service.  Presumably if speeds stay below 768K as seems
> > > almost certain--I'm only buying the basic 256K service--then will the
> > > speed be an issue?  Can't quite tell from what you write.
> > >
> > > I'll check out the 905B driver.
> > >
> > >       Do you know what happened in the kernel between 2.0.34 and 2.0.36 that
> > > makes it work better for these drivers?
> > >
> > > gp
> > >
> > > Peter Lukas wrote:
> > > >
> > > > The 905 and 905B are both really fussy cards (the 905B in particular).
> > > > You'll need to upgrade to a more recent kernel that the stock RH5.1 (which
> > > > really only supports plain-vanilla 905).  Use a 2.0.36 or a 2.1.13x kernel
> > > > or download the 905B driver from the cesdis.gsfc.nasa.gov site.  Also,
> > > > some of these newer "autosensing" adapters get really confused depending
> > > > upon what type of switch they're hooked up to.  If you're connecting the
> > > > adapter to a 10/100 switch, you'll need to configure the port the 905B is
> > > > connected to forced 10 or 100 full or half-duplex and do the same for the
> > > > card as you load it.  Of the numerous 905's and 905B's I've worked with,
> > > > the only ones with the problem have been the adapters with the white WOL
> > > > connector on the card itself.  This problem has also exhibited itself on
> > > > the dec tulip-based adapters as well.  If you're not using an autosensing
> > > > switch, you'll probably just need to update the kernel 905B
> > > > Vortex/Boomarang driver.
> > > >
> > > > Peter Lukas
> > > > Math Systems Office
> > > > University of Minnesota
> > > >
> ...
> > > > > --
> > > > > Richard Seymour, Anarchy Software, Inc.
> > > > > anarchy@anarchysoftware.com
> > >
> > > Gordon Pedersen         1412 Portland Ave
> 
> -- 
> Gordon Pedersen         1412 Portland Ave
> info systems design     Saint Paul, Minnesota
> gordo@pclink.com        55104,  USA
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tclug-list-unsubscribe@listserv.real-time.com
> For additional commands, e-mail: tclug-list-help@listserv.real-time.com
> Try our website: http://tclug.real-time.com
> 
>