Re: [coreboot] SimNOW VGA int 1a

[Myles Watson]

> -----Original Message-----
> From: ron minnich [mailto:[e-mail]]
> Sent: Thursday, October 16, 2008 1:34 PM
> To: Marc Jones
> Cc: Myles Watson; Jordan Crouse
> Subject: Re: [coreboot] SimNOW VGA int 1a
>
> On Thu, Oct 16, 2008 at 12:32 PM, Marc Jones  wrote:
> > That gets us back to relocating or mutilple copies in the stages. I am
>
> ok
> > with multiple copies in the ROM but it should be on version in the
>
> source.
> > The problem with v2 is that there was ROMCC versions and main coreboot
> > versions.
>
> The beauty of carl-daniel's redesign of the "call ROM" stuff is that
> we can have the same printk code in stage2 as in ROM and there is no
> symbol conflict. But we now have to pretty much put all stage0 stuff
> in stage2 as well. no choice.
>
> Myles, caching won't help.

I'm showing my ignorance here.  Conceptually couldn't you have a small area
set up like CAR for the shared portions?  As long as it was all loaded
before probing wouldn't it always be accessed before the ROM?

Thanks,
Myles

SimNOW VGA int 1a

[ron minnich]

On Thu, Oct 16, 2008 at 12:41 PM, Myles Watson  wrote:
> I'm showing my ignorance here.  Conceptually couldn't you have a small area
> set up like CAR for the shared portions?  As long as it was all loaded
> before probing wouldn't it always be accessed before the ROM?

It's possible but it is implementation and architecture dependent and
I would hate to count on it ....

I think marc is right. Let's get that code out of ROM.

ron

SimNOW VGA int 1a

[Jordan Crouse]

On 16/10/08 12:46 -0700, ron minnich wrote:
> On Thu, Oct 16, 2008 at 12:41 PM, Myles Watson  wrote:
> > I'm showing my ignorance here.  Conceptually couldn't you have a small area
> > set up like CAR for the shared portions?  As long as it was all loaded
> > before probing wouldn't it always be accessed before the ROM?
>
> It's possible but it is implementation and architecture dependent and
> I would hate to count on it ....
>
> I think marc is right. Let's get that code out of ROM.

I don't think this is really that big a deal.   I don't think shared
functions hold any value for us other then the fact that it "feels"
right.  With the right Makefile fu, we can make this painless for
the builder and the user.

Jordan

SimNOW VGA int 1a

[ron minnich]

On Thu, Oct 16, 2008 at 1:06 PM, Jordan Crouse  wrote:
> I don't think this is really that big a deal.   I don't think shared
> functions hold any value for us other then the fact that it "feels"
> right.  With the right Makefile fu, we can make this painless for
> the builder and the user.

agree. I will try to fix the makefiles today.
ron