grant@torque.net
Wed, 22 Jul 1998 20:28:54 -0400 (EDT)
> > Which seems to suggest that the (natsemi) program didn't actually do
> > anything.
>
> yess, I agree, but hopefully this has been explained by the National Semi's
> datasheet :(
Perhaps, but do you also have the wiring schematic for your laptop's
motherboard ? This really is nowhere nearly as straightforward as you
imagine. There are lots of us here who have the datasheet you keep
referring to, by the way, it is not the entire story.
Please try working with David to get the natsemi program working. If,
however, the MSI program is unable to force the port into EPP mode, your
port may be locked by the BIOS in such a way that only a new BIOS could
fix this.
(Yes, such configurations exist.)
> > You might have to allow access to a whole range of oddball ioports before
> > the backpack will be able to "unlock" your port.
>
> it unlocked the port fine, however, it just seemed to make no difference
No, by unlocking the port, I mean enabling access to the configuration
registers so that the mode can be changed into something useful. The
MSI driver pokes at several dozen random addresses while attempting to
access all the different possible configuration registers. You'd need
to "enable" access to all those ports in your DOSemu config file before
it would actually unlock the port.
> i know that the versions are different, but i cant believe that the dosemu
> team changed tracing all that much. {quick check through the initialisation
> logfile}. Hurm, surely the info you need _is_ there, there is def.
> reads to port 3f0 and 3f1 > (parts of my parallel port range) and also
> the reads to / from ports before that...
Yes, DOSemu 0.67 is very different. No, there is no useful information
there. An actual SPP mode initialisation trace will be more than 500K
uncompressed using my +T trace code, using the new 0.67 code it will be
more like 3M.
> > and then be sure not to use the
> > "fast" attribute in your ports clauses of the configuration file, "fast"
> > defeats the tracing logic.
>
> fast wasnt on during the logs
Probably they've changed the terminology or syntax or something. "fast"
is an attribute of a port range in 0.66. If you declare fast access to a
port, DOSemu will write to it directly, rather than trapping the i/o
instructions through the emulator. Obviously, if the i/o is not being
trapped, it can't be traced. That's why your trace log contained no
references to ports 378, 379 or 37a - the actual SPP ports.
--------------------------------------------------------------------------
Grant R. Guenther grant@torque.net
--------------------------------------------------------------------------
-- To unsubscribe, send mail to: linux-parport-request@torque.net --
-- with the single word "unsubscribe" in the body of the message. --
This archive was generated by hypermail 2.0b3 on Wed 30 Dec 1998 - 10:18:02 EST