[ANNOUNCEMENT] Wiki for discussing P35/IHC9(R)/SATA issues set up

[Sean C. Farley]

On Tue, 6 Nov 2007, Scott Long wrote:

> Jeremy Chadwick wrote:
>> On Tue, Nov 06, 2007 at 08:37:27AM -0600, Sean C. Farley wrote:
>>> 1. Dell Optiplex 745:
>>> ATA channel 2:
>>>     Master:  ad4 <WDC WD800JD-75MSA3/10.01E04> Serial ATA II
>>>     Slave:       no device present
>> 
>> Intel Q965 + ICH8, which does SATA300.  atacontrol output proves the
>> disk is capable of SATA300 (re: "Serial ATA II"), but for some reason
>> it's not.
>
> SATA II does not automatically mean 3GHz link speed.  It's just an
> umbrella term meaning that the drive implements one or more features
> defined in SATA-2.  It's similar to the mis-representation in the USB
> world of what USB 2.0 means.

In this case, it is does support SATA/300:
http://www.westerndigital.com/en/products/Products.asp?DriveID=83

>> This one is indeed odd and needs investigation.
>> 
>>> 2. Asus P5K-E/WiFi:
>>> ad10: 152627MB <Seagate ST3160827AS 3.42> at ata5-master SATA150
>>> ad12: 152627MB <Seagate ST3160827AS 3.42> at ata6-master SATA150
>>> ATA channel 5:
>>>     Master: ad10 <ST3160827AS/3.42> Serial ATA v1.0
>>>     Slave:       no device present
>>> ATA channel 6:
>>>     Master: ad12 <ST3160827AS/3.42> Serial ATA v1.0
>>>     Slave:       no device present
>> 
>> Intel P35 + ICH9.  The chipset does SATA300 but your disks do not --
>> they are Barracuda 7200.7 disks, which support SATA150 maximum.
>> However, the motherboard also has a JMicron controller on it.

Correct.  This system only has SATA/150 drives (both HD's and DVD).

>> After looking at
>> http://www.farley.org/freebsd/tmp/SATA/dmesg-asus.boot, it's easy to
>> determine that your SATA150 disks are connected to the ICH9
>> controller (which is being reported at ICH8 for some reason; maybe a
>> chipset ID thing).  They are not connected to the JMicron, so the
>> SATA150 limitation here is caused by the disks.

I did not apply my patch to this build, otherwise, it would be in the
log.  I just do not know if the patch ID's ICH9R only.

>> The SATA ATAPI errors you see are something I can't really help with,
>> many see these kinds of messages on all sorts of hardware.
>
> There seems to be a lot of differing technical opinions on SATA-ATAPI
> support in FreeBSD.  I'll be digging more into it shortly, but
> probably not in time for the 7.0 release.

Unfortunately, these timeouts cause problem for installs from SATA ATAPI
drives.  Maybe a release note would be warranted?  I had to fall back to
an NFS install.

Sean
-- 
scf at FreeBSD.org