Chris Radek wrote:
> On Wed, Aug 20, 2008 at 11:35:27AM -0500, Jon Elson wrote:
> 
>>OHHHhhhh. HIGHLY dangerous.  It writes 0x80 to the I/O port 
>>0x402 above that address you specify.  If you don't specify 
>>anything, it probably defaults to zero, so it writes to I/O port 
> 
> 
> This write should be added to the driver.  If it needs to be optional,
> it should be a load-time parameter for the driver so the user can
> change it in the hal file.
> 
> There is absolutely no reason for this to be a separate program
> anymore.
Well, the problem is it only works for parallel ports that 
follow the insane scheme of the multi-I/O chips used on 
motherboards that had floppy, parallel port and serial on one 
chip.  The block of config registers are 0x400 higher than the 
par port address.  Just blindly writing 0x80 to an address 0x402 
higher than the specified parallel port address seems VERY 
dangerous, especially for PCI plug-in boards.  Maybe putting a 
check for standard par port addresses such as 0x378, 0x278 and 
only doing it on those might be better.  But, still, writing to
address 0x77a on every PC out there might not be a great 
solution, either.  Umm, what happens on a PC when you write to a 
non-existant I/O port address?  On the old ISA systems it was 
just ignored, but I don't know on all PCI systems.

I'm rather reluctant to put this feature into the PPMC driver 
without at least some comment on the implications.  I don't 
actually KNOW of any bad interactions, but blindly writing to 
discontinuous addresses not actually part of the standard par 
port address block seems more hazardous than doing anything that 
stays within the assigned block of addresses for the par port.
Any comments?

Ken Lerman's GX150 is probably about a 10 year-old computer.

Jon

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to