Subject: Re: yamaha YMF715 (cone of silence) chip
To: Wolfgang Rupprecht <wolfgang@wsrcc.com>
From: Bill Studenmund <skippy@macro.stanford.edu>
List: port-i386
Date: 03/03/1998 12:51:16
On 3 Mar 1998, Wolfgang Rupprecht wrote:

> 
> A while ago there was a thread about the Yamaha YMF715 sound chip.
> Intel appears to have chosen this chip for their new motherboards
> (such as the P-II AL440LX board).
> 
> The first problem is that on a genenic 1.3 kernel this chip probes for
> both the "sb at isa" and "sb at isapnp" probes.

I don't think it's the chip's fault. The problem is that our PNP
implimentataion assumes that the BIOS did NOT do PNP setup, and wipes out
all the PNP initialization in the cards. It then proceeds to re-init all
the PNP cards.

What we should probably do is scan the PNP setup early to see what, if
anything, is set up. If a card's set up, do the NetBSD attach setup. If a
card's not setup, wait for later. Then we set up all the ISA cards. Then
we set up whatever's left.

I think the original problem was that we got PNP before the bus space
mapping went in. I.E. in 1.2, if there was a PNP sound blaster which the
PNP drivers set up, there was no way to make the normal ISA sb driver NOT
also attach the hardware. That's why we do PNP last.

For now, can you turn off PNP in the BIOS?

Take care,

Bill