Date: Mon, 28 Jul 2003 13:31:57 +0930 From: Greg 'groggy' Lehey <grog@FreeBSD.org> To: "M. Warner Losh" <imp@bsdimp.com> Cc: freebsd-mobile@FreeBSD.org Subject: Re: Mapping Video BIOS? Message-ID: <20030728040157.GP45069@wantadilla.lemis.com> In-Reply-To: <20030727.214235.20045695.imp@bsdimp.com> References: <20030727015903.GJ45069@wantadilla.lemis.com> <20030726.221859.122847811.imp@bsdimp.com> <20030728032653.GA37499@wantadilla.lemis.com> <20030727.214235.20045695.imp@bsdimp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--k++jrJxptm7IAPkr
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
On Sunday, 27 July 2003 at 21:42:35 -0600, M. Warner Losh wrote:
> In message: <20030728032653.GA37499@wantadilla.lemis.com>
> "Greg 'groggy' Lehey" <grog@FreeBSD.org> writes:
>> On Saturday, 26 July 2003 at 22:18:59 -0600, M. Warner Losh wrote:
>>> In message: <20030727015903.GJ45069@wantadilla.lemis.com>
>>> "Greg 'groggy' Lehey" <grog@FreeBSD.org> writes:
>>>> Presuming that it's the ROM driver, I get this in the dmesg I posted:
>>>> pnpbios: Bad PnP BIOS data checksum
>>>
>>> That's likely the problem. However, PnP BIOS information isn't the
>>> same thing that the orm[sic] driver probes for.
>>
>> They look related. I've now found the orm output:
>>
>> orm0: <Option ROMs> at iomem 0xe0000-0xe3fff,0xdf800-0xdffff,0xd0000-0xd17ff,0xc0000-0xcefff on isa0
>>
>> The last one is the video BIOS. It's interesting to note that it
>> doesn't report the 4 kB BIOS at 0xcf000, which suggests that at this
>> point the 16 kB area is already unmapped.
>
> Hmmmm, The list comes from scanning the "ISA HOLE" for certain memory
> signatures. These signatures have a length in them that say "I'm a
> rom that's X long".
Sure. The data at offset 0xc0000 are:
C000:0000 55 AA 78 E9 44 06 00 00-00 00 00 00 00 00 00 00 U.x.D...........
The 0xaa55 is the BIOS signature ("Here be a BIOS"), and the 0x78 is
the length byte (120 sectors, or 60 kB). That's how orm0 knows the
end address.
> I don't think that it suggests that things are 'unmapped'...
If the area between 0xcc000 and 0xcffff had been mapped, orm0 would
have found this too:
C000:F000 55 AA 08 E8 6D 0B CB 11-FE 02 00 00 00 00 00 00 U...m...........
>> I've worked around the problem by compiling the video BIOS into the X
>> server and not trying to access the BIOS in the machine. Obviously
>> not a solution, but it works for the moment. I'd really like to track
>> down the problem. Does anybody have an idea?
>
> I don't, I'm sorry.
Understood. I was hoping that somebody else might have some ideas.
Greg
--
See complete headers for address and phone numbers
--k++jrJxptm7IAPkr
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (FreeBSD)
iD8DBQE/JKA1IubykFB6QiMRAm02AKCzcn4XoyGj8+KNmhmqq79ZDd29HACeN6nQ
/CBW5KZQ910VwfvK0ICvXMM=
=4oRx
-----END PGP SIGNATURE-----
--k++jrJxptm7IAPkr--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030728040157.GP45069>