Date: Wed, 14 Feb 2001 21:24:24 -0500 (EST) From: Andrew Gallatin <gallatin@cs.duke.edu> To: Mike Smith <msmith@FreeBSD.ORG> Cc: Peter Jeremy <peter.jeremy@alcatel.com.au>, alpha@FreeBSD.ORG Subject: Re: Accessing a PCI device by `physical' address Message-ID: <14987.15832.39211.2856@grasshopper.cs.duke.edu> In-Reply-To: <200102150200.f1F20ei02223@mass.dis.org> References: <20010215124001.A71053@gsmx07.alcatel.com.au> <200102150200.f1F20ei02223@mass.dis.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Mike Smith writes:
>
> Thinking about it some more, though, this probably doesn't integrate well
> with syscons. 8(
Exactly. Syscons has an easy time of it running early because it
knows VGA adapters will be at certain magic I/O ports. Ick.
> I'm kinda stumped, here. I don't think there's a clean way to do this,
> which is bad. We probably need to fix this. Thoughts?
Kinda gross, but you could add pcib_{read,write}_config() fields to
the platform struct & init them in the various chipset init functions.
Most platforms don't use the first arg (device_t). Those that do
(tsunami, mcpcia) could probably be hacked to treat a null device_t as
if it were the 0th hose. I don't think SRM supports console devices
anywhere but the 0th hose anyway.
I assume you're forward-porting Andrew Miklic's driver?
Drew
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14987.15832.39211.2856>