[LinuxBIOS] [PATCH] K8T890/VT8237 /a8v-se preview

[Rudolf Marek]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello all,

I'm attaching quite big patch which adds support for the K8T890/VT8237R chipset
and for my motherboard Asus A8V-E SE.

It works fine except for reboot, I'm using it daily ;) Now I need help with the
release.

I have done some modifications to raminit.c for K8, fixing unbuffered mem stuff,
adding some tables instead of ifs... I added support for MCFG ACPI table
(mmconfig for PCIe) but I need to somehow figure out how to tell FILO/LXBtables
to reserve some mem region. I will put this into separate patches. What I need
now is to discuss technical aspects of patches rather than codingstyle etc etc.

I will prepare some more tidy patches later, but I really want to show something
 now ;)

The patch is on http://assembler.cz/download/preview.patch

I dont know if I can pass 150KB to the list and also Wifi sucks.

It is 2am in morning so please forgive some errors and omissions in this mail.

Thanks,

Rudolf
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFG/ZWW3J9wPJqZRNURAo4WAJ0XR4sz81kzRk1wWMechwfzLiB4swCfeKKV
lGfQrL6KWyuHZaeSjULc438=
=CBIn
-----END PGP SIGNATURE-----

[Carl-Daniel Hailfinger]

On 29.09.2007 02:19, Rudolf Marek wrote:
> Hello all,
>
> I'm attaching quite big patch which adds support for the K8T890/VT8237R chipset
> and for my motherboard Asus A8V-E SE.
>
> I have done some modifications to raminit.c for K8, fixing unbuffered mem stuff,
> adding some tables instead of ifs... I added support for MCFG ACPI table
> (mmconfig for PCIe) but I need to somehow figure out how to tell FILO/LXBtables
> to reserve some mem region. I will put this into separate patches. What I need
> now is to discuss technical aspects of patches rather than codingstyle etc etc.
>
> I will prepare some more tidy patches later, but I really want to show something
>  now ;)
>
> The patch is on http://assembler.cz/download/preview.patch

Very nice! I think that especially the raminit.c parts are a step
forward. Maybe this also fixes the RAM bank population issues some of us
were seeing.

Unfortunately, I don't have any K8 platform which suppports LB to check,
my only systems running LB are all Geode GX (to be exact, all of them
are OLPC prototypes at various stages in development).

Regards,
Carl-Daniel