PATA vs. SATA

PATA vs. SATA

[Randy.Dunlap]

It's confusing enough using libata for PATA/IDE instead of
drivers/ide/* and using a PIIX driver for SATA.  :)

Yes, it Works For Me.  However, I see this one little thing
that still seems odd to me.


Intel SATA controller in IDE mode:  libata-core says:

ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0x18B0 irq 14
ata2: SATA max UDMA/133 cmd 0x170 ctl 0x376 bmdma 0x18B8 irq 15
      ~~~~
 This ^^^^ is the confusing part to me.

Intel SATA controller in AHCI (SATA) mode:  libata-core says:

ata1: SATA max UDMA/133 cmd 0x30C8 ctl 0x30E6 bmdma 0x30A0 irq 193
ata2: SATA max UDMA/133 cmd 0x30C0 ctl 0x30E2 bmdma 0x30A8 irq 193

Yes, ata_piix.c sets host_flags: ATA_FLAG_SATA in this case.
What does it do?  What is it for?  (this flag)

Yes, lspci reports SATA controller in IDE mode.

Thanks,
-- 
~Randy

Re: PATA vs. SATA

[Tejun]

Randy.Dunlap wrote:
> It's confusing enough using libata for PATA/IDE instead of
> drivers/ide/* and using a PIIX driver for SATA.  :)
> 
> Yes, it Works For Me.  However, I see this one little thing
> that still seems odd to me.
> 
> 
> Intel SATA controller in IDE mode:  libata-core says:
> 
> ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0x18B0 irq 14
> ata2: SATA max UDMA/133 cmd 0x170 ctl 0x376 bmdma 0x18B8 irq 15
>       ~~~~
>  This ^^^^ is the confusing part to me.

I'm not sure whether I understood your problem correctly but if your 
ata_piix is reporting SATA for a PATA port and the controller is ICH6 or 
7, there's a patch which recently made into the devel tree which fixes 
MAP register value interpretation on ICH6/7's (which BTW was slightly 
broken and fixed again by Jeff's patch).  Can you try the latest 
upstream branch?

-- 
tejun

Re: PATA vs. SATA

[Randy.Dunlap]

On Tue, 31 Jan 2006 11:47:26 +0900 Tejun wrote:

> Randy.Dunlap wrote:
> > It's confusing enough using libata for PATA/IDE instead of
> > drivers/ide/* and using a PIIX driver for SATA.  :)
> > 
> > Yes, it Works For Me.  However, I see this one little thing
> > that still seems odd to me.
> > 
> > 
> > Intel SATA controller in IDE mode:  libata-core says:
> > 
> > ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0x18B0 irq 14
> > ata2: SATA max UDMA/133 cmd 0x170 ctl 0x376 bmdma 0x18B8 irq 15
> >       ~~~~
> >  This ^^^^ is the confusing part to me.
> 
> I'm not sure whether I understood your problem correctly but if your 
> ata_piix is reporting SATA for a PATA port and the controller is ICH6 or 
> 7, there's a patch which recently made into the devel tree which fixes 
> MAP register value interpretation on ICH6/7's (which BTW was slightly 
> broken and fixed again by Jeff's patch).  Can you try the latest 
> upstream branch?

The patch(es) sound like they may fix this small nit.

I may try the upstream branch if I ever become a gitter.

---
~Randy