libata acpi issue on gateway laptop

libata acpi issue on gateway laptop

[Kyle McMartin]

Hi Jeff,

A user is reporting an issue with 2.6.34.6 on a Gateway laptop that
booting with libata.noacpi=1 resolves.

Symptom is that their device dumps a lot of messages like this:

ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata1.00: ACPI cmd 00/00:00:00:00:00:a0 (NOP) rejected by device
(Stat=0x51 Err=0x04)
ata1.00: ACPI cmd ef/10:01:00:00:00:a0 (SET FEATURES) succeeded
ata1.00: ACPI cmd ef/10:02:00:00:00:a0 (SET FEATURES) succeeded
ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
ata1.00: ACPI cmd ef/10:04:00:00:00:a0 (SET FEATURES) succeeded
ata1.00: ACPI cmd ef/10:05:00:00:00:a0 (SET FEATURES) rejected by device
(Stat=0x51 Err=0x04)
ata1.00: ATA-8: Hitachi HTS545050B9A300, PB4OC60F, max UDMA/133
ata1.00: 976773168 sectors, multi 0: LBA48 NCQ (depth 31/32), AA
ata1.00: ACPI cmd 00/00:00:00:00:00:a0 (NOP) rejected by device
(Stat=0x51 Err=0x04)
ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
ata1.00: ACPI cmd ef/10:05:00:00:00:a0 (SET FEATURES) rejected by device
(Stat=0x51 Err=0x04)
ata1.00: configured for UDMA/133

on boot and resume from suspend.

Dmesg is at https://bugzilla.redhat.com/attachment.cgi?id=442529 and the
full bug report at https://bugzilla.redhat.com/show_bug.cgi?id=629315.

Any ideas? They say the issue exists all the way back to 2.6.31, so it
seems like a pretty long-lived bug.

Thanks!
 Kyle

Re: libata acpi issue on gateway laptop

[Tejun Heo]

On 09/06/2010 09:35 PM, Kyle McMartin wrote:
> Hi Jeff,
> 
> A user is reporting an issue with 2.6.34.6 on a Gateway laptop that
> booting with libata.noacpi=1 resolves.
> 
> Symptom is that their device dumps a lot of messages like this:
> 
> ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
> ata1.00: ACPI cmd 00/00:00:00:00:00:a0 (NOP) rejected by device
> (Stat=0x51 Err=0x04)
> ata1.00: ACPI cmd ef/10:01:00:00:00:a0 (SET FEATURES) succeeded
> ata1.00: ACPI cmd ef/10:02:00:00:00:a0 (SET FEATURES) succeeded
> ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
> ata1.00: ACPI cmd ef/10:04:00:00:00:a0 (SET FEATURES) succeeded
> ata1.00: ACPI cmd ef/10:05:00:00:00:a0 (SET FEATURES) rejected by device
> (Stat=0x51 Err=0x04)
> ata1.00: ATA-8: Hitachi HTS545050B9A300, PB4OC60F, max UDMA/133
> ata1.00: 976773168 sectors, multi 0: LBA48 NCQ (depth 31/32), AA
> ata1.00: ACPI cmd 00/00:00:00:00:00:a0 (NOP) rejected by device
> (Stat=0x51 Err=0x04)
> ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
> ata1.00: ACPI cmd ef/10:05:00:00:00:a0 (SET FEATURES) rejected by device
> (Stat=0x51 Err=0x04)
> ata1.00: configured for UDMA/133
> 
> on boot and resume from suspend.
> 
> Dmesg is at https://bugzilla.redhat.com/attachment.cgi?id=442529 and the
> full bug report at https://bugzilla.redhat.com/show_bug.cgi?id=629315.
> 
> Any ideas? They say the issue exists all the way back to 2.6.31, so it
> seems like a pretty long-lived bug.

These are debug/info messages.  Just ignore them.

-- 
tejun

Re: libata acpi issue on gateway laptop

[Kyle McMartin]

On Mon, Sep 06, 2010 at 09:46:10PM +0200, Tejun Heo wrote:
> These are debug/info messages.  Just ignore them.
> 

The user claims they are causing long delays waiting for the timeouts to
occur that go away if booting with noacpi...

--Kyle

Re: libata acpi issue on gateway laptop

[Tejun Heo]

Hello,

On 09/06/2010 09:51 PM, Kyle McMartin wrote:
> On Mon, Sep 06, 2010 at 09:46:10PM +0200, Tejun Heo wrote:
>> These are debug/info messages.  Just ignore them.
>>
> 
> The user claims they are causing long delays waiting for the timeouts to
> occur that go away if booting with noacpi...

Hmm... I don't see any timeout in the log you posted.  Ah, okay, the
interesting part is in the full log, the ACPI GTF sequence itself is
finishing fine but the next NCQ READ command is timing out.  The BIOS
is enabling non-zero buffer offsets and DMA setup FIS auto-activate.
It's likely that the controller isn't being configured correctly
leading to mismatch between the behavior expected by the controller
and the actual drive behavior.  Hmm.... I'll look further into it
tomorrow.  For the record, can you please attach the full log, lspci
-nn output and dmidecode output here?

Thanks.

-- 
tejun

Re: libata acpi issue on gateway laptop

[Kyle McMartin]

On Mon, Sep 06, 2010 at 10:05:27PM +0200, Tejun Heo wrote:
> Hmm... I don't see any timeout in the log you posted.  Ah, okay, the
> interesting part is in the full log, the ACPI GTF sequence itself is
> finishing fine but the next NCQ READ command is timing out.  The BIOS
> is enabling non-zero buffer offsets and DMA setup FIS auto-activate.
> It's likely that the controller isn't being configured correctly
> leading to mismatch between the behavior expected by the controller
> and the actual drive behavior.  Hmm.... I'll look further into it
> tomorrow.  For the record, can you please attach the full log, lspci
> -nn output and dmidecode output here?
> 

Thanks Tejun, I've requested this from the original reporter, and I'll
attach it here when I hear back.

(Oops, sorry for fudging linux-ide->linux-ata, thanks for fixing that.)

regards, Kyle