[PATCH] AHCI powersaving and port-stopping (2.6.22-rc4)

[Török Edvin]

On 6/12/07, Török Edvin <edwintorok at gmail.com> wrote:
> It says: AHCI BAR not set - leaving.
> Here's my lspci output (1f.2 is the SATA controller). At offsets
> 24h-27h I have only 0s, AFAICT this was supposed to contain the
> address for AHCI BAR?
>

Update: I have succesfully turned on AHCI, and ALPM seems to be
available (/sys/class/scsi_host/host0/link_power_management_policy
says min_power).

I have modified Kristen's patch: I enabled the MSE bit, and also
changed to AHCI /non-combined mode in the MAP register. This still
didn't work, so I changed DECLARE_PCI_FIXUP_HEADER to
DECLARE_PCI_FIXUP_EARLY, and after replacing the iomap_region test....
I have AHCI! (I also had to change the ata_piix MAP to recognize MAP
value 0.)

AHCI relevant portion of dmesg:
AHCI BAR not set - try to enable
Quirked PIIX device to AHCI mode
ahci 0000:00:1f.2: version 2.2
ahci 0000:00:1f.2: PORTS_IMPL is zero, forcing 0x1
ahci 0000:00:1f.2: AHCI 0001.0100 32 slots 4 ports 1.5 Gbps 0xf impl SATA mode
ahci 0000:00:1f.2: flags: 64bit ncq ilck stag pm led clo pmp pio slum part
scsi0 : ahci
scsi1 : ahci
scsi2 : ahci
scsi3 : ahci

The modified patch is available here:
http://edwintorok.googlepages.com/ahci_quirk_.patch

What do you think?

Thanks,
Edwin