[LinuxBIOS] [PATCH] 440bx RAM calculation fix

[Ceri Coburn]

Here is a patch for the 440bx RAM calculation size.  The top of RAM was
being calculated incorrectly.

Ceri.

[roger]

On Wed, 2007-05-09 at 13:54 +0100, Ceri Coburn wrote:
> Here is a patch for the 440bx RAM calculation size.  The top of RAM
> was being calculated incorrectly.

This patch didn't take correctly here, so I applied it manually and
created another patch.

(I can boot as far as "Northbridge prior to SDRAM init:" and then I
still get random hangs about this point.  Patch doesn't seem to break
anything here.  Since I'm hacking on my DFI P2XBL w/ w83977tf, I can't
easily test on my Tyan S1832DL with S1846 profile right now.)

Wed May 9 07:49:17 PDT 2007

[Ceri Coburn]

roger wrote:
> On Wed, 2007-05-09 at 13:54 +0100, Ceri Coburn wrote:
> > Here is a patch for the 440bx RAM calculation size.  The top of RAM
> > was being calculated incorrectly.
>
> This patch didn't take correctly here, so I applied it manually and
> created another patch.
>
> (I can boot as far as "Northbridge prior to SDRAM init:" and then I
> still get random hangs about this point.  Patch doesn't seem to break
> anything here.  Since I'm hacking on my DFI P2XBL w/ w83977tf, I can't
> easily test on my Tyan S1832DL with S1846 profile right now.)

I only noticed issues when elfboot was trying to load the payload.  That 
is how I found the bug.  It was trying to copy the payload to top of 
memory which was 4 times bigger that it should have been
> Wed May 9 07:49:17 PDT 2007

[Uwe Hermann]

On Wed, May 09, 2007 at 07:58:52AM -0700, roger wrote:
> Fixed a bug within the 440bx RAM size calculation.  Since the DRB values
> on the 440bx are 8MB units we need to shift left by 13 to get it into kb.
>
> Signed-off-by: Ceri Coburn 
> Signed-off-by: Roger Zauner

Thanks, fixed in r2655.

The file needs more fixing, it's copied from another target and not
adapted yet (I'll fix most of that in a few minutes).

Uwe.