[PATCH] MS7135 fixes and updates

[Jonathan A. Kollasch]

Fix various issues on MSI MS7135 board.

- w83627thf is strapped to 0x4e, not 0x2e
 - there's no device 9 on PCI-E x1 bus, it should be device 0
 - add mptable entries for AGR slot, based on info in user manual
 - enable floppy drive controller so that some legacy VGA ROMs will work

Signed-off-by: Jonathan A. Kollasch 
---

[Joe]

On Wed, 30 Apr 2008 15:21:13 +0000, "Jonathan A. Kollasch"
 wrote:
> Fix various issues on MSI MS7135 board.
> enable floppy drive controller so that some legacy VGA ROMs will work

Little confused by this. Why does the floppy drive controller have to be
enabled for VGA??

[Ward Vandewege]

On Wed, Apr 30, 2008 at 01:03:21PM -0400, Joe wrote:
> On Wed, 30 Apr 2008 15:21:13 +0000, "Jonathan A. Kollasch"
>  wrote:
> > Fix various issues on MSI MS7135 board.
> > enable floppy drive controller so that some legacy VGA ROMs will work
>
> Little confused by this. Why does the floppy drive controller have to be
> enabled for VGA??

I think this might be a 'feature' of the proprietary VGA ROM... Who knows
why - there's no source...

Thanks,
Ward.

[Jonathan A. Kollasch]

On Wed, Apr 30, 2008 at 01:09:56PM -0400, Ward Vandewege wrote:
> On Wed, Apr 30, 2008 at 01:03:21PM -0400, Joe wrote:
> > On Wed, 30 Apr 2008 15:21:13 +0000, "Jonathan A. Kollasch"
> >  wrote:
> > > Fix various issues on MSI MS7135 board.
> > > enable floppy drive controller so that some legacy VGA ROMs will work
> >
> > Little confused by this. Why does the floppy drive controller have to be
> > enabled for VGA??
>
> I think this might be a 'feature' of the proprietary VGA ROM... Who knows
> why - there's no source...

Yeah, I found that a RV370 board I have was working when the
superio address was wrong, because the floppy controller
was in the power-on-reset state of enabled.  When I corrected
the superio address, and the FDC got disabled as instructed,
the VGA BIOS did weird things I don't exactly recall and
didn't work.  I then tried enabling the FDC again and
found that this made the VGA BIOS work again.  I have
no idea why it wants to touch the floppy controller,
but I have to let it or it won't work.

Jonathan Kollasch

[Uwe Hermann]

On Wed, Apr 30, 2008 at 03:21:13PM +0000, Jonathan A. Kollasch wrote:
> Fix various issues on MSI MS7135 board.
> w83627thf is strapped to 0x4e, not 0x2e
> there's no device 9 on PCI-E x1 bus, it should be device 0
> add mptable entries for AGR slot, based on info in user manual
> enable floppy drive controller so that some legacy VGA ROMs will work
>
> Signed-off-by: Jonathan A. Kollasch

Acked-by: Uwe Hermann
>      device pnp 2e.6 off end		# non-existant or undocumented
>      device pnp 2e.7 off end		# Game, MIDI, GPIO 1, GPIO 5
>      device pnp 2e.8 off end		# GPIO 2
>      device pnp 2e.9 off end		# GPIO 3, GPIO 4
>      device pnp 2e.a off end		# ACPI
>          device pnp 2e.b on		# env monitor
>          device pnp 4e.6 off end		# unknown

LDN 6 is undocumented in the datasheet, might even not exist.
Did you test if removing the line has problematic effects? Or do you
have a reason to believe there's something on that LDN?

Uwe.

[Jonathan A. Kollasch]

On Thu, May 01, 2008 at 02:22:25AM +0200, Uwe Hermann wrote:
> On Wed, Apr 30, 2008 at 03:21:13PM +0000, Jonathan A. Kollasch wrote:
> > Fix various issues on MSI MS7135 board.
> > w83627thf is strapped to 0x4e, not 0x2e
> > there's no device 9 on PCI-E x1 bus, it should be device 0
> > add mptable entries for AGR slot, based on info in user manual
> > enable floppy drive controller so that some legacy VGA ROMs will work
> >
> > Signed-off-by: Jonathan A. Kollasch
>
> Acked-by: Uwe Hermann
> >      device pnp 2e.6 off end		# non-existant or undocumented
> >      device pnp 2e.7 off end		# Game, MIDI, GPIO 1, GPIO 5
> >      device pnp 2e.8 off end		# GPIO 2
> >      device pnp 2e.9 off end		# GPIO 3, GPIO 4
> >      device pnp 2e.a off end		# ACPI
> >          device pnp 2e.b on		# env monitor
> >          device pnp 4e.6 off end		# unknown
>
> LDN 6 is undocumented in the datasheet, might even not exist.
> Did you test if removing the line has problematic effects? Or do you
> have a reason to believe there's something on that LDN?

I don't think I've really tested it. Does having a LDN unlisted
in Config.lb mean it doesn't get touched?  If that's the case
that LDN should just be omitted there.  It probably is unimplemented.
(It is implemented in the W83627EHG, apparently. But i have a THG
on this board.)

Jonathan Kollasch

[Rudolf Marek]

Just a side note, if the superIO is similar to W83627EHG then you could use 
virtual LDNs for the different functionality in same LDN which are sharing same 
enable byte. (I did it for the  W83627EHG).

http://www.coreboot.org/SuperIO_port_guide

Maybe this page is not linked from anywhere?

Rudolf

[Jonathan A. Kollasch]

On Thu, May 01, 2008 at 12:37:28PM +0200, Rudolf Marek wrote:
> Just a side note, if the superIO is similar to W83627EHG then you could use 
> virtual LDNs for the different functionality in same LDN which are sharing 
> same enable byte. (I did it for the  W83627EHG).
>
> http://www.coreboot.org/SuperIO_port_guide

Huh, but I don't think this is the case here.
> Maybe this page is not linked from anywhere?

That's what it looks like:
http://www.coreboot.org/Special:Whatlinkshere/SuperIO_port_guide

Jonathan Kollasch

[Rudolf Marek]

Jonathan A. Kollasch napsal(a):
> On Thu, May 01, 2008 at 12:37:28PM +0200, Rudolf Marek wrote:
> > Just a side note, if the superIO is similar to W83627EHG then you could use 
> > virtual LDNs for the different functionality in same LDN which are sharing 
> > same enable byte. (I did it for the  W83627EHG).
> >
> > http://www.coreboot.org/SuperIO_port_guide
>
> Huh, but I don't think this is the case here.

In logical device 7 and 9 yes. If you want to enable at least something, you 
need to use virtual LDNs.
> > Maybe this page is not linked from anywhere?
>
> That's what it looks like:
> http://www.coreboot.org/Special:Whatlinkshere/SuperIO_port_guide

Hmm I added it at least to devel info.

Rudolf

[Uwe Hermann]

On Fri, May 02, 2008 at 10:15:17PM +0200, Rudolf Marek wrote:
> Jonathan A. Kollasch napsal(a):
> > On Thu, May 01, 2008 at 12:37:28PM +0200, Rudolf Marek wrote:
> > > Just a side note, if the superIO is similar to W83627EHG then you could use 
> > > virtual LDNs for the different functionality in same LDN which are sharing 
> > > same enable byte. (I did it for the  W83627EHG).
> > >
> > > http://www.coreboot.org/SuperIO_port_guide
> >
> > Huh, but I don't think this is the case here.
>
> In logical device 7 and 9 yes. If you want to enable at least something, you 
> need to use virtual LDNs.

Committed current patch in r3279 for now, as it fixes a few issues.
Please post another patch for additional fixes and improvements.

Note that I dropped LDN 6, it's not available in *THF either, I checked
the datasheet.

Uwe.

[Uwe Hermann]

On Thu, May 01, 2008 at 12:37:28PM +0200, Rudolf Marek wrote:
> Just a side note, if the superIO is similar to W83627EHG then you could 
> use virtual LDNs for the different functionality in same LDN which are 
> sharing same enable byte. (I did it for the  W83627EHG).
>
> http://www.coreboot.org/SuperIO_port_guide

Can you merge that page somewhere into
http://www.coreboot.org/Developer_Manual#Serial_output_and_the_Super_I.2FO
please (if you agree to license the content under the GPLv2-or-later, as
the rest of the manual)? That's the correct place for that type of info, IMHO.

I'll drop the SuperIO_port_guide then afterwards.

Uwe.