Re-4: [PARPORT] Parallel Port Mode switching via normal C-code possible??

From: sven.luebke@auerswald.de
Date: Mon Mar 11 2002 - 04:12:52 EST

  • Next message: Gunther Mayer: "Re: Re-4: [PARPORT] Parallel Port Mode switching via normal C-code possible??"

    Gunther.Mayer@t-online.de wrote:

    Hi,

    > > >>>>> "sven" == sven luebke <sven.luebke@auerswald.de> writes:
    > >
    > > sven> Sorry, it's not for a Linux driver. Just MS-DOS environment... I
    > > sven> know, I'm in the wrong mailinglist, but I didn't find a MS-DOS
    > > sven> parallel port "discussion board"...
    > >
    > > If you are in DOS, maybe there are some BIOS Interrupts that could do that
    > > in a transparent way. ACPI comes also to mind, but probably will work, if
    > > ever, on only recent MBs.
    > > Otherwise, in DOS, you may also write directly to the Super-I/O controller,
    > > without the need for a driver. If however you want a solution that should
    > > work with many Super-I/O chips, you have to get their datasheets and/or
    > > programming manual and have to implement code that identies the chip and
    > > the
    > > switching with the appropriate sequence of cookies. Tedious works...
    > >
    > > I know of some work for the Linux kernel to identify the Super-IO chip.
    >
    > See http://home.t-online.de/home/gunther.mayer/lssuperio-0.63.tar.gz
    > for a start (it identifies as many chipset as known (by me)).
    >
    > You only need DOS equivalents of outb() and inb() to make this work.

    Oh, thanks for this advice. Nevertheless I want to know if really no GENERAL
    bios address is responsible for that "mode switching"!? I read the following
    in www.beyondlogic.org (Standard Parallel Port Manual):

    ---qoute on---
    The above modes are configurable via BIOS. You can reconfigure them by using your own software, but this is not recommended. These software registers, typically found at 0x2FA, 0x3F0, 0x3F1 etc are only intended to be accessed by BIOS. There is no set standard for these configuration registers, thus if you were to use these registers, your software would not be very portable. With today's multitasking operating systems, its also not a good idea to change them when it suits you.
    ---quote off---

    So, could anyone confirm this statement? Or is this limited to only a few
    chipsets?

    Does your lssuperio package identify even old chipsets in 486/DX100 notebooks or
    is this limited to new chipsets?

    Thanx for your help...

    Regards,

    Sven

    To: Gunther.Mayer@t-online.de
        bon@elektron.ikp.physik.tu-darmstadt.de
    Cc: linux-parport@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 2b29 : Mon Mar 11 2002 - 04:20:03 EST