Date: Tue, 09 Jan 1996 19:32:45 -0800 From: "Amancio Hasty Jr." <hasty@rah.star-gate.com> To: Terry Lambert <terry@lambert.org> Cc: freebsd-hackers@FreeBSD.org Subject: Re: PnP problem... Message-ID: <199601100332.TAA00725@rah.star-gate.com> In-Reply-To: Your message of "Tue, 09 Jan 1996 19:47:37 MST." <199601100247.TAA13642@phaeton.artisoft.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>>> Terry Lambert said:
> > > > Grand Total: 4 Interrupts, 3 DMA Channels, 7 sets of I/O ports...
> > >
> > > One would think you could share interrupts... then ask where it was
> > > coming from.
>
> [ ... ]
>
> > Sorry Terry this does not make any sense. My gus does have different devic
es.
> > You can say that the GUS PnP has a poor architecture for handling differe
nt
> > devices -- that I would agree with.
> >
> > The fact remains that the GUS PnP is not an isolated PnP soundcard which i
s
> > going to offered similar configuration problems. Somehow we have to :
> >
> > a. know in advance which devices is the kernel configured for, if possibl
e
> > given that we have lkms
> >
> > b. then configured all PnP devices which we know we have driver for.
> >
> > c. continue with normal probe and attach driver initialization sequence.
>
>
> You have to preallocate locations for devices that can be neither
> moved nor disabled; your sequence order is incorrect.
>
>
> It makese sense that you would have one interrupt per card so you don't
> run out between card slots and onboard devices... it's stupid that the
> GUS doesn't have an interupt multiplex on board. You'll have to live
> with it until you buy a pure PCI system instead of a PCI bridged off
> of ISA. I recommend the Apple and Motorolla chipsets. 8-).
Well, Terry the super multimedia chipsets based on PCI are coming real
soon... Check out http://www.mpact.com for instance. However,
it is going to be a while before are able to support them.
> However, the PnP discovery ordering is:
>
> 1) disable all PnP
> 2) probe all non-PnP cards
> 3) Query PnP cards for where they may fit
> 4) Do a topological sort to fit them all
> 5) Make them pick one to disable if the sort results in a
> collision. Repeat as necessary.
> 6) Map the mappable locations
> 7) Enable the PnP cards that were not marked disabled
> 8) Attach drivers as available, loading them if necessary
Well, Terry it makes no sense to activate devices which I don't
have a device driver who knows I may even be able configure my
system 8)
> Exception: Treat ISA cards on a PnP ISA bus motherboard (slot disabling)
> as single address devices in terms of the topological sort.
>
> Note that you can still be screwed if you don't recognize a non-PnP
> card on an ISA bus, since you will not be able to infer its location
> for the sort, and it can't be disabled.
>
>
> Conclusion: If you have problems, it's because your GUS is too greedy.
Pretty lame Terry 8)
I do have full duplex audio over here and it works as for many of the
devices that the GUS has I don't need . For instance, the SB PRO emulation
and the CDROM interface which btw the normal board does have anyway.
Amancio
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199601100332.TAA00725>