On Tue, 26 May 1998, Brian D Chase wrote:

> 
> On Tue, 26 May 1998, Bertram Barth wrote:
> 
> > KA43 mustn't access the memory shared with LANCE via "normal" addresses,
> > but has to use this address mapped into machines diagnostic memory, ie.
> > le-mem | 0x28000000. Using the "normal" address yields machine checks 
> > (which would explain SCB2NDINT, IIRC when receiving) and/or non-working
> > network-code (ie. timeouts when trying to send, "transmit buffer full").
> > 
> > I've had a short look into arch/vax/boot/if_le.c on multivac and this
> > detail seems to be missing in the standalone le-driver.
> 
> Soooo... are you saying that perhaps one of the older bootloaders might
> work out in this situation, because they did incorporate the correct code
> for the KA43's lance device?  If so, I'm pretty sure that I've got mopable
> copies of the older bootloaders lying around in my directories *somewhere*. 
> 
> -brian.
> ---
> Brian "JARAI" Chase | http://world.std.com/~bdc/ | VAXZilla LIVES!!!
> 
> 
> 

the string 'le%d: output buffer busy' resides in MOPBOOT.SYS that I had
downloaded. It has this problem when trying to run w/ 32 MB RAM as opposed
to 16 MB RAM in which it runs with no problem.

P.S.A