Re: [avrdude-dev] Blank skipping, multi-byte writes, notes on parallel ports

[E. Weddington]

On 22 Apr 2003 at 11:51, Joerg Wunsch wrote:

> As Lou Cyphre wrote:
> 
> > <silly>
> > Is it viable for avrdude to access I/O directly on PCI address...?
> > </silly>
> 
> If you know the exact port address, why not?  It needs a rewrite of
> the command-line parser though.  Maybe the parser should understand
> the usual names (lptN), but failing that, try to parse the argument as
> a number (using strtoul(), so 0xf00 will work).
> 

IIRC, you and I talked briefly about having this as a possible 
feature. It wouldn't have to be in the parser. The Windows code that 
validates the port would look at a fixed list of port names and if 
that fails then try to convert to a port address.

Right now the Windows parallel port names are mapped as:
lpt1 = 0x378
lpt2 = 0x278
lpt3 = 0x3BC

I would assume that this feature would allow the user to specify any 
parallel port base address in hex for -P. But this would definitely 
be caveat emptor: if the base address supplied by the user is 
incorrect, behaviour would be *very* unspecified.

This is why I originally wanted to see if there was a way to get from 
Windows all available parallel ports and their base addresses. So far 
I haven't found a way to do this. If anybody has thoughts in this 
direction, I'm open.

Eric