Re: [MiNT] lp driver patch for the Firebee to extend the parallel port strobe duration

[Alan Hourihane <alanh@fairlite.co.uk>]

Hi Markus,

Why can't this problem be entirely fixed within the FPGA and not resort
to these additional requirements ?

Alan.

On 03/10/15 21:35, markus@mubf.de wrote:
> Hi all,
>
> please find attached a patch to MiNT's lp driver.
>
> It fixes the problem that with the Firebee's Coldfire processor, the
> parallel port strobe signal is not kept active long enough to be
> recognised by most printers. The original delay was two NOP's between
> toggling the PSG ports which is much too small for the Firebee. I have
> replaced this (only if __mcoldfire__ is defined) with a simple wait
> loop around a nop.
>
> Loop count has been adjusted to meet the minimum requirement of 750 ns
> for strobe activation according to IEEE 1284.
>
> This is not entirely clean (if we'll ever have another machine with
> different clock speed, it may fail again), but since the Firebee is
> the only known Coldfire machine with a PSG chip, I'd consider it good
> enough for now.
>
> Note that this alone is not sufficient to make the Firebee print,
> you'll need a (hopefully soon to be released) FPGA fix as well. It
> also won't work (for yet unknown reasons) with MiNT running on top of
> FireTOS.
>
> But it will make (with the FPGA fix) MiNT + EmuTOS print on the
> parallel port which I definitively consider an improvement.
>
> Commit text:
>
> lp driver patch to extend the parallel port strobe duration on the
> Firebee
> Contributed by Markus Fröschle
>
>
> Cheers,
> Markus