Re: 2.6.33: libata issues with VIA SATA/PATA controller

Re: 2.6.33: libata issues with VIA SATA/PATA controller

[Jeff Garzik]

On 04/08/2010 08:43 PM, Jonathan Woithe wrote:
> Hi
>
> I have a machine equipped with the following SATA/PATA controllers:
>
>   * 00:0f.1 IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/
>     VT823x/A/C PIPC Bus Master IDE (rev 06)
>
>   * 00:0f.0 RAID bus controller: VIA Technologies, Inc. VIA VT6420 SATA RAID
>     Controller (rev 80)
>
> Two hard drives are attached to the SATA controller (messages from 2.6.24.2):
>
>    scsi 0:0:0:0: Direct-Access     ATA      WDC WD2502ABYS-0 02.0 PQ: 0 ANSI: 5
>    scsi 1:0:0:0: Direct-Access     ATA      WDC WD2502ABYS-0 02.0 PQ: 0 ANSI: 5
>
> Both drives are jumpered to force 1.5 Gbps SATA speed.  This is needed
> because these WD drives won't automatically fall back to 1.5 Gbps if 3.0
> Gbps isn't requested, and the VIA controller in use is only capable of
> 1.5 Gbps.
>
> The PATA interface has a tape drive attached as primary master and a CD
> writer attached as secondary master.  Under 2.6.24.2 these show up as
>
>    hda: SONY SDX-260V, ATAPI TAPE drive
>    hdc: ATAPI 40X CD-ROM CD-R/RW drive, 2048kB Cache
>
> The problem I have is that I am unable to boot 2.6.33 (or 2.6.33.2).  The
> interfaces are probed but it seems the system has trouble communicating
> reliably over the SATA links.  Sometimes both drives are identified
> correctly, only to encounter I/O errors when the partition table is read.
> Other times failures occur during the drive detection phase.  The problem
> persists even when both PATA devices have been unplugged, so the PATA
> devices themselves are not the cause.  In all cases the end result is the
> same: the rootfs can't be found, the kernel panics.
>
> The 2.6.33.2 I have tried to boot today is configured to use libata VIA PATA
> support.  The 2.6.24.2 kernel (which mostly works, except for some issues
> with ide-tape which might be hardware related) instead uses the old ide
> driver for the PATA interface.  I have not yet been able to test whether
> disabling the VIA PATA component of libata works around the problem.  The
> reason for enabling libata VIA PATA support is due to ongoing issues with
> ide-tape in recent kernels (I've tried from 2.6.29 on).  The suggestion was
> made to try libata since its tape driver was in better shape.
>
> Unfortunately the machine in question is a production machine and I have
> limited opportunities to reboot and test kernels.  I also didn't have a
> camera handy to take screen shots this morning during testing.  The
> following are messages I noted by hand before they were scrolled away by the
> panic.
>
>    ata1: lost interrupt status 0x50
>    unhandled error code
>    Result: hostbyte=0x00, driverbyte=0x06
>    CDB: cdb[0] = 0x28 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x08 0x00
>    end request: I/O error, dev sda, sector 0
>
> The same was repeated for ata2 / sdb.
>
> Another group of messages which crops up during drive detection is:
>
>    qc timeout cmd 0x27
>    failed to read native max address (error mask 0x4)
>    HPA suport seems broken, skipping HPA handling
>
> I would like to try to get to the bottom of what's happening here.  This
> problem does not occur in 2.6.24.2 (using the VIA ide PATA driver).  Both
> 2.6.33 and 2.6.33.2 (with the VIA libata PATA driver) suffer from the
> problem.  If there are further tests I can do please let me know and I'll do
> my best to schedule them (given that the machine in question is a production
> system).  Also ask if you require more information about the machine.
>
> Finally, there's a complete dmsg output I get from 2.6.24.2 at the end of
> this email in case it contains additional useful information on the hardware
> configuration of this system.  Please CC me any replies to ensure I see them
> (I monitor lkml via web gateways, so it's easy to miss followups at times).

(adding linux-ide to CC)

A complete dmesg is definitely useful.  Posting one from the failing 
kernel would be preferred, though.  That will show us libata boot 
messages as well as the failures you are seeing, in full detail.

Also, please try the latest 2.6.34-rc kernel, as that has several fixes 
for both pata_via and sata_via which did not make 2.6.33.

	Jeff

Re: 2.6.33: libata issues with VIA SATA/PATA controller

[Jonathan Woithe]

Hi Jeff

> A complete dmesg is definitely useful.

It's at the end of my previous post.

> Posting one from the failing kernel would be preferred, though.

Any ideas as to how I can capture it?  Since the SATA disc interfaces stop
working long before they are mounted none of the early boot messages make it
to disc.  All the interesting bits are therefore scrolled into the bit
bucket.

A serial console might be doable (is there documentation about how to set
one up?) but that would take some time to arrange - I'm only physically at
the machine in question sporadically and I'd have to rustle up another PC
ready for my next visit.  It could be done, but might take some time.

> That will show us libata boot messages as well as the failures you are
> seeing, in full detail.

For sure.  In terms of the detailed failure messages what I posted
previously is about all that was shown.

> Also, please try the latest 2.6.34-rc kernel, as that has several fixes 
> for both pata_via and sata_via which did not make 2.6.33.

I'll see if I can squeeze this in this afternoon and report back.

Regards
  jonathan

Re: 2.6.33: libata issues with VIA SATA/PATA controller

[James Andrewartha]

On 9/04/2010 12:36 PM, Jonathan Woithe wrote:
> Hi Jeff
> 
>> A complete dmesg is definitely useful.
> 
> It's at the end of my previous post.
> 
>> Posting one from the failing kernel would be preferred, though.
> 
> Any ideas as to how I can capture it?  Since the SATA disc interfaces stop
> working long before they are mounted none of the early boot messages make it
> to disc.  All the interesting bits are therefore scrolled into the bit
> bucket.
> 
> A serial console might be doable (is there documentation about how to set
> one up?) but that would take some time to arrange - I'm only physically at
> the machine in question sporadically and I'd have to rustle up another PC
> ready for my next visit.  It could be done, but might take some time.

Netconsole is the easiest way to get dmesg when disks aren't working - see
Documentation/networking/netconsole.txt for details on how to set it up.

-- 
James Andrewartha

Re: 2.6.33: libata issues with VIA SATA/PATA controller

[Jonathan Woithe]

Hi Jeff

Following up on my previous reply:

> Also, please try the latest 2.6.34-rc kernel, as that has several fixes 
> for both pata_via and sata_via which did not make 2.6.33.

I was able to try this on the machine just before I left.  Unfortunately
2.6.34-rc3 does not fix the problem - it behaves exactly like 2.6.33.  In
terms of error messages, there is nothing more displayed beyond that which I
reported in the initial email except for this:

  SStatus 113, SControl 300

As far as I can tell, that (plus the previously reported details) is all
that's printed by the kernel in connection with the failure.

I'll look into the alternative console methods to see if I can capture a
full output from the failing kernel.  However, it will be at least two weeks
before I can do this because that's the earliest I'll be physically back at
the machine for an extended period of time (I could probably run quick tests
sooner).

Regards
  jonathan

Re: 2.6.33: libata issues with VIA SATA/PATA controller

[Jeff Garzik]

On 04/09/2010 07:14 AM, Jonathan Woithe wrote:
> I'll look into the alternative console methods to see if I can capture a
> full output from the failing kernel.  However, it will be at least two weeks
> before I can do this because that's the earliest I'll be physically back at
> the machine for an extended period of time (I could probably run quick tests
> sooner).


We won't be able to help without the full dmesg from the failing kernel...

	Jeff