ata_piix broken in 2.6.22

ata_piix broken in 2.6.22

[Felipe Contreras]

Hi,

Since I updated to 2.6.22 everything has been slower and I think this
is the reason:

ata_piix 0000:00:1f.2: version 2.11
ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
ata_piix 0000:00:1f.2: 0x170 IDE port busy
ata_piix 0000:00:1f.2: no available legacy port

I tried to apply the patch d583bc18812f8da52bf25eef9cd111e5fd46a6ab,
but I get a similar error:

ata_piix 0000:00:1f.2: version 2.11
ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
PCI: Unable to reserve I/O region #1:8@1f0 for device 0000:00:1f.2
ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 0 (errno=-16)
PCI: Unable to reserve I/O region #3:8@170 for device 0000:00:1f.2
ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 1 (errno=-16)
ata_piix 0000:00:1f.2: no available native port

I think I'm not the only one with that problem:
http://www.dellcommunity.com/supportforums/board/message?board.id=sw_linux&thread.id=10825

Any hints?

-- 
Felipe Contreras

Re: ata_piix broken in 2.6.22

[Alan Cox]

O> ata_piix 0000:00:1f.2: version 2.11
> ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
> ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
> ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
> ata_piix 0000:00:1f.2: 0x170 IDE port busy
> ata_piix 0000:00:1f.2: no available legacy port

You appear to have old IDE compiled in as well and finding the device
first

Re: ata_piix broken in 2.6.22

[Gabriel C]

Felipe Contreras wrote:
> Hi,

Hi ,

> 
> Since I updated to 2.6.22 everything has been slower and I think this
> is the reason:
> 
> ata_piix 0000:00:1f.2: version 2.11
> ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
> ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
> ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
> ata_piix 0000:00:1f.2: 0x170 IDE port busy
> ata_piix 0000:00:1f.2: no available legacy port
> 
> I tried to apply the patch d583bc18812f8da52bf25eef9cd111e5fd46a6ab,
> but I get a similar error:
> 
> ata_piix 0000:00:1f.2: version 2.11
> ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
> ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
> PCI: Unable to reserve I/O region #1:8@1f0 for device 0000:00:1f.2
> ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 0 (errno=-16)
> PCI: Unable to reserve I/O region #3:8@170 for device 0000:00:1f.2
> ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 1 (errno=-16)
> ata_piix 0000:00:1f.2: no available native port
> 
> I think I'm not the only one with that problem:
> http://www.dellcommunity.com/supportforums/board/message?board.id=sw_linux&thread.id=10825
> 
> Any hints?
> 

Guessing you are using and SATA/PATA controller ?

Combined mode got removed in 2.6.22 and depending on your config things will break.

As example CONFIG_IDE=y  BLK_DEV_PIIX=y , ATA=y , ATA_PIIX=Y worked in <2.6.22 but won't on .22.

If you want to use libata try to boot with ideX=noprobe ideY=noprobe


Gabriel

Re: ata_piix broken in 2.6.22

[Gabriel C]

Alan Cox wrote:
> O> ata_piix 0000:00:1f.2: version 2.11
>> ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
>> ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
>> ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
>> ata_piix 0000:00:1f.2: 0x170 IDE port busy
>> ata_piix 0000:00:1f.2: no available legacy port
> 
> You appear to have old IDE compiled in as well and finding the device
> first

Alan ,

But this is allowed now and for me, after removing CONFIG_SATA_INTEL_COMBINED option
some other options combos may be illegal and removing it broke a lot working setups.

On <2.6.22 you got on your SATA/PATA controller with CONFIG_IDE=y , ATA=y ( ATA_X=y and even BLK_DEV_X=y) the

SATA disk as SATA and PATA one as IDE now you get the SATA disk as 'IDE' too which gives you almost 3-4MB/s speed.

Is such a config legall for you for an SATA/IDE combo controller in >=2.6.22 ?

CONFIG_BLK_DEV_IDE_SATA=n
CONFIG_IDE=y
ATA=y
BLK_DEV_PIIX=y(or m doesn't matter )
ATA_PIIX=y

IMO after removing CONFIG_SATA_INTEL_COMBINED you guys introduced an real conflict with the IDE subsystem.

The only way to get libata working with such a config is to boot ideX=noprobe.


Regards,

Gabriel

Re: ata_piix broken in 2.6.22

[Felipe Contreras]

On 7/20/07, Gabriel C <nix.or.die@googlemail.com> wrote:
> Felipe Contreras wrote:
> > Hi,
>
> Hi ,
>
> >
> > Since I updated to 2.6.22 everything has been slower and I think this
> > is the reason:
> >
> > ata_piix 0000:00:1f.2: version 2.11
> > ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
> > ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
> > ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
> > ata_piix 0000:00:1f.2: 0x170 IDE port busy
> > ata_piix 0000:00:1f.2: no available legacy port
> >
> > I tried to apply the patch d583bc18812f8da52bf25eef9cd111e5fd46a6ab,
> > but I get a similar error:
> >
> > ata_piix 0000:00:1f.2: version 2.11
> > ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
> > ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 16
> > PCI: Unable to reserve I/O region #1:8@1f0 for device 0000:00:1f.2
> > ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 0 (errno=-16)
> > PCI: Unable to reserve I/O region #3:8@170 for device 0000:00:1f.2
> > ata_piix 0000:00:1f.2: failed to request/iomap BARs for port 1 (errno=-16)
> > ata_piix 0000:00:1f.2: no available native port
> >
> > I think I'm not the only one with that problem:
> > http://www.dellcommunity.com/supportforums/board/message?board.id=sw_linux&thread.id=10825
> >
> > Any hints?
> >
>
> Guessing you are using and SATA/PATA controller ?
>
> Combined mode got removed in 2.6.22 and depending on your config things will break.
>
> As example CONFIG_IDE=y  BLK_DEV_PIIX=y , ATA=y , ATA_PIIX=Y worked in <2.6.22 but won't on .22.
>
> If you want to use libata try to boot with ideX=noprobe ideY=noprobe

Oh, that's it!

I tried CONFIG_IDE=n and now ata_piix is working just fine.

Would a note in CONFIG_ATA be pertinent? Or did I miss some piece of
documentation?

Thanks.

-- 
Felipe Contreras