>I've recently installed a Archive FT-60 tape drive on my Linux box. It
>works fine under DOS but not at all under Linux. I have editted the
>tpqic02.h file. I uncommented the tapedrive & ifc sections for this
>drive as well as the irq(5),dma(1) and port settings(200). Everything
>else in the user config section is commented except the autodynam
>config which is ifnodef (or something like that). When I make config
>I answer no to the question pertaining to the autodynamic config.
>Everything compiles ok and I have installed the new kernal. MAKEDEV qic
>makes the rmt* devices. When I 'tar cvf /dev/rmt8' the tape spins for
>a couple of seconds and comes back I/O error. I've triple checked my
>settings on the controller and the tpqic02.h file. Is a hardware conflict
>possible even though this box runs the tape ok in DOS? Are there ways to
>test this drive in Linux? Any help would be greatly appreciated.
Quite possible, especially if you have a soundcard. One of the things I
have noticed about the Archive VP-402 board is that it does not seem to
get along too well with jumpless PnP sound boards. If you have a joystick
port in your system (or one unknowlingly enabled via a soundboard) this may
be the cause, port 200h is in use on my system so I ended up having to
reference my Phoenix AT BIOS reference manual for all commonly used
port addresses and moved it from 200h to 240h since it is used for some
obscure piece of IBM hardware.
I put my QIC-02 board at 240h, DMA 3, IRQ 5. I gave my soundboard DMA 1
since that is used for SB compatability. For some reason I have not
been able to find, the VP-402 board will not work at IRQ 2 (9), even in
MS-DOS. This may be because of my Logitech Soundman 16. This is the
reason I am looking to sell it and just go with a SCSI tape drive. I
have always run into the problem of running out of IRQ lines for 8-bit
ISA cards even back when I was using SCO Xenix and ISC 386/ix 2.0.2.
Other than the hardware config quirks under Unix, it is quite a reliable
tape drive system to use under any OS with driver and software support.
While sharing a port address and/or IRQ is alright to do in MS-DOS, it is an
absolute no no in any derivative of Unix. All testing under MS-DOS gets you
is just seeing if the hardware works. It isn't going to test for a possible
hardware conflict under Unix. It's not really indicative of if it's an
actual working configuration under any given OS that has driver support for
the hardware in your system. Unfortunately, there's no way to test the
drive for functionality under Linux. You need to find a working hardware
configuration and Linux isn't really going to do it for you. Unix, like
the C language, will let you hang yourself if you don't know what you're
doing. But I know that the Archive VP-402 board works, I've been using it
with venerable Linux 0.99pl15 (Slackware 2.2.1) so you don't have to worry
about it being a non-supported device.
It may benefit you to log in a notebook and log in writing all system
hardware resources that are in use by your system.