When I compiled my kernel CCF_PYXISBUG was not defined.  Was this patch
for current or 1.4.1? I patched against 1.4.1. grep -r only shows
CCF_PYXISBUG being used in your patch.. so what am I missing here?  It
doesn't look like cia.c sets this config parameter anywhere.  My original
cia.c revision was 1.47, so it looks like your patch was applied to newer
sources.  It worked after I replaced the if with if (1), so I'm not so
worried about it.  The system seems to be a bit faster now, although I
haven't done any benchmarking to test it. I apreciate the help. :-)

Thanks,
Jeff

On Mon, 7 Feb 2000, Jason Thorpe wrote:

> Date: Mon, 07 Feb 2000 09:40:08 -0800
> From: Jason Thorpe <thorpej@nas.nasa.gov>
> To: Jeff Roberson <nomad@nop.aliensystems.com>
> Cc: port-alpha@netbsd.org
> Subject: Re: DPWS500/Pyxis bugs, corrupt memory etc. 
> 
> On Sun, 6 Feb 2000 22:41:33 -0800 (PST) 
>  Jeff Roberson <nomad@nop.aliensystems.com> wrote:
> 
>  > How do you properly define CCF_PYXISBUG? I'm probably not as familiar with
>  > config as I should be, so I don't know where this is set.  I tried options
>  > PYXISBUG, but obviously I'm missing something so that didn't work.  For
>  > now I just removed that conditional and replaced it with if (1) to test
>  > the patch.  I'll let you know how it goes.
>  > 
>  > Thanks,
>  > Jeff
>  > 
>  > > + 	if ((ccp->cc_flags & CCF_PYXISBUG) != 0 &&
> 
> you don't need to worry about it.  If your kernel prints out:
> 
> 	cia0: WARNING: Pyxis pass 1 DMA bug; no bets...
> 
> ...than CCF_PYXISBUG is set already :-)
> 
> It's set in cia.c, when the core logic chipset is attached, based on
> the revision of your workstation's mainboard.  I.e. if it's < Miata 1.5
> (which is the first rev to have the bug fixed), then it gets set.
> 
>         -- Jason R. Thorpe <thorpej@nas.nasa.gov>
> 
>