[coreboot] alix1c PATA

[Jordan Crouse]

On 05/05/08 18:32 +0200, Peter Stuge wrote:
> On Mon, May 05, 2008 at 09:26:39AM -0600, Jordan Crouse wrote:
> > > > pata_cs5536: disabled by BIOS
> > > 
> > > Hmm, I wonder if this is something coreboot has disabled or not
> > > enabled?
> > > drivers/ata/pata_cs5536.c:244 tests a bit in the IDE config
> > > register of CS5536.
> > 
> > Can you try the legacy driver (ide/pci/amd74xx.c)?
> 
> Will try later.
> 
> 
> > It sounds like the libata port brought over something distasteful.
> > I've been using the old CS5536 driver for a very long time no with
> > no problems.
> 
> Regardless, no bit should indicate that PATA is disabled.
> 
> Particularly surprising since IDE is enabled from reset.

The enable bits for the AMD74XX controller (which the 5536 uses) are
in PCI config byte 0x40.  bit 1 is the enable bit for port 0 and bit 0
is the enable bit for port 1.

I have just tried both db800 and norwich in v2 with the legacy AMD5536 driver
and both worked (well, db800 puked, but for other reasons).  Here's the
boot log:

AMD5536: IDE controller (0x1022:0x209a rev 0x01) at  PCI slot 0000:00:0f.2      
AMD5536: not 100% native mode: will probe irqs later                            
AMD5536: 0000:00:0f.2 (rev 01) UDMA100 controller                               
    ide0: BM-DMA at 0x1ca0-0x1ca7, BIOS settings: hda:pio, hdb:pio              
AMD5536: IDE port disabled       

The last line is the secondary bus being disabled.

I looked at the libata driver, and it seems to be doing the right thing -
reading 0x40 and checking bit 1.  

So I checked the code: in v2, src/southbridge/amd/cs5536/cs5536_ide.c
handles the honors.

So long story short - this is possibly something that just never got 
brought over to v3?

Jordan

-- 
Jordan Crouse
Systems Software Development Engineer 
Advanced Micro Devices, Inc.