[PATCH] ich7 & kontron update

[Stefan Reinauer]

see patches

[Patrick Georgi]

> From: "Stefan Reinauer"
> use new features of the ich7 update
> move rambase above 1M to avoid memory trashing through SMM relocation
> enable superio HWM
>
> Signed-off-by: Stefan Reinauer

Acked-by: Patrick Georgi
> Update ICH7 driver
> minor smi cosmetics (in progress)
> add real ac97 driver
> add real azalia driver
> fix some interrupt issues
> fix some sata issues
>
> Signed-off-by: Stefan Reinauer

This one has a small buglet that likely comes from splitting out the patches 
(see #endifs in i82801gx_lpc.c).
With that fixed,
Acked-by: Patrick Georgi

Regards,
Patrick Georgi

[Corey Osgood]

On Tue, Jan 20, 2009 at 5:21 PM, Stefan Reinauer wrote:
> see patches
> use new features of the ich7 update
> move rambase above 1M to avoid memory trashing through SMM relocation
> enable superio HWM
>
> Signed-off-by: Stefan Reinauer

Index: src/mainboard/kontron/986lcd-m/Config.lb
> ===================================================================
> --- src/mainboard/kontron/986lcd-m/Config.lb    (revision 3882)
> +++ src/mainboard/kontron/986lcd-m/Config.lb    (working copy)
> @@ -82,7 +82,7 @@
>      object acpi_tables.o
>      makerule dsdt.c
>          depends "$(MAINBOARD)/dsdt.dsl"
>    action  "iasl -p $(PWD)/dsdt -tc $(MAINBOARD)/dsdt.dsl"
>    action  "iasl -p dsdt -tc $(MAINBOARD)/dsdt.dsl"

Is this right, or a local change that snuck in?

Acked-by: Corey Osgood

Update ICH7 driver
> minor smi cosmetics (in progress)
> add real ac97 driver
> add real azalia driver
> fix some interrupt issues
> fix some sata issues
>
> Signed-off-by: Stefan Reinauer

There are a bunch of instances of uintX_t and a few of device_t, can you
update them to the newer style? Other then that:

Acked-by: Corey Osgood

[Stefan Reinauer]

Corey Osgood wrote:
> On Tue, Jan 20, 2009 at 5:21 PM, Stefan Reinauer
> <[e-mail] > wrote:
> see patches
>
>
>
> - use new features of the ich7 update
> - move rambase above 1M to avoid memory trashing through SMM
> relocation
> - enable superio HWM
>
> Signed-off-by: Stefan Reinauer <[e-mail]
> <mailto:[e-mail]>>
>
>
> Index: src/mainboard/kontron/986lcd-m/Config.lb
> ===================================================================
> --- src/mainboard/kontron/986lcd-m/Config.lb    (revision 3882)
> +++ src/mainboard/kontron/986lcd-m/Config.lb    (working copy)
> @@ -82,7 +82,7 @@
>      object acpi_tables.o
>      makerule dsdt.c
>          depends "$(MAINBOARD)/dsdt.dsl"
> -        action  "iasl -p $(PWD)/dsdt -tc $(MAINBOARD)/dsdt.dsl"
> +        action  "iasl -p dsdt -tc $(MAINBOARD)/dsdt.dsl"
>
>
> Is this right, or a local change that snuck in?

That's due to a bug in iasl not parsing long command lines correctly. If
you're working in a deep enough subdirectory, you might see it just
cutting off the path and producing bogus files. This is a workaround by
dramatically shortening the command line.
> Acked-by: Corey Osgood <[e-mail]
> >

Thanks!
> Update ICH7 driver
>
> - minor smi cosmetics (in progress)
> - add real ac97 driver
> - add real azalia driver
> - fix some interrupt issues
> - fix some sata issues
>
> Signed-off-by: Stefan Reinauer <[e-mail]
> <mailto:[e-mail]>>
>
>
> There are a bunch of instances of uintX_t and a few of device_t, can
> you update them to the newer style? Other then that:

What's the newer style? u32 instead of uint32_t? and struct device *
instead of device_t?

I'll go for this in another patch once my queue is a bit saner.
> Acked-by: Corey Osgood <[e-mail]
> >

Thanks.

[Corey Osgood]

On Tue, Jan 20, 2009 at 5:58 PM, Stefan Reinauer wrote:
> Corey Osgood wrote:
> > On Tue, Jan 20, 2009 at 5:21 PM, Stefan Reinauer
> > <[e-mail] > wrote:
> > see patches
> >
> >
> >
> > - use new features of the ich7 update
> > - move rambase above 1M to avoid memory trashing through SMM
> > relocation
> > - enable superio HWM
> >
> > Signed-off-by: Stefan Reinauer <[e-mail]
> > <mailto:[e-mail]>>
> >
> >
> > Index: src/mainboard/kontron/986lcd-m/Config.lb
> > ===================================================================
> > --- src/mainboard/kontron/986lcd-m/Config.lb    (revision 3882)
> > +++ src/mainboard/kontron/986lcd-m/Config.lb    (working copy)
> > @@ -82,7 +82,7 @@
> >      object acpi_tables.o
> >      makerule dsdt.c
> >          depends "$(MAINBOARD)/dsdt.dsl"
> > -        action  "iasl -p $(PWD)/dsdt -tc $(MAINBOARD)/dsdt.dsl"
> > +        action  "iasl -p dsdt -tc $(MAINBOARD)/dsdt.dsl"
> >
> >
> > Is this right, or a local change that snuck in?
>
> That's due to a bug in iasl not parsing long command lines correctly. If
> you're working in a deep enough subdirectory, you might see it just
> cutting off the path and producing bogus files. This is a workaround by
> dramatically shortening the command line.
> > Acked-by: Corey Osgood <[e-mail]
> > >
>
> Thanks!
> > Update ICH7 driver
> >
> > - minor smi cosmetics (in progress)
> > - add real ac97 driver
> > - add real azalia driver
> > - fix some interrupt issues
> > - fix some sata issues
> >
> > Signed-off-by: Stefan Reinauer <[e-mail]
> > <mailto:[e-mail]>>
> >
> >
> > There are a bunch of instances of uintX_t and a few of device_t, can
> > you update them to the newer style? Other then that:
>
> What's the newer style? u32 instead of uint32_t? and struct device *
> instead of device_t?

Yeah, that's what I meant.
> I'll go for this in another patch once my queue is a bit saner.

In that case, just commit it as-is and I'll take care of my own nitpicks
later ;)

-Corey
> > Acked-by: Corey Osgood <[e-mail]
> > >
>
> Thanks.