Re: SiI3531 SATA port multipler

Nick Hilliard wrote:
> I've been poking around with a SiI3531 sata300 host card, connected to
> an external disk enclosure with an internal SiI4726 card.  There are 3 x
> SATA disks hanging off it at the moment:
> 
>> http://www.startech.com/item/SAT3540ERGB-35in-4-Drive-eSATA-Multi-RAID-External-Hard-Drive-Enclosure.aspx
> 
> I have assumed that the SiI3531 is command-compatible with the SiI3112,
> and put in the attached patch to see what might happen.  The kernel
> appears to recognise the sata card, but dies before initialising it
> properly:
> 
>> ata2: Portmultiplier (id=47261095 rev=1f06) with 7 ports
>> ata2: error writing PM port
>> ata2: p6: writing ATA_SC_DET_RESET failed
>> ata2: Portmultiplier (id=47261095 rev=1f06) with 7 ports
>> ata2: error writing PM port
>> ata2: p4: writing ATA_SC_DET_RESET failed
>> ata2: error writing PM port
>> ata2: p5: writing ATA_SC_DET_RESET failed
>> ata2: error writing PM port
>> ata2: p6: writing ATA_SC_DET_RESET failed
>> ata2: Portmultiplier (id=47261095 rev=1f06) with 7 ports
>> ata2: error writing PM port
>> ata2: p6: writing ATA_SC_DET_RESET failed
>> ata2: Portmultiplier (id=47261095 rev=1f06) with 7 ports
>> ata2: error writing PM port
>> ata2: p6: writing ATA_SC_DET_RESET failed

Have you tried to enable verbose boot messages? It may give some
additional info.

> I've attached a full dmesg.   This setup is running on an Asus P5Q SE2
> motherboard, in amd64 mode with a -current kernel pulled today.
> 
> Any suggestions about what could be done to make this work?

Can you try to use AHCI controller present in your system instead of
SiI? It may allow to differentiate SiI controller related problems from
PM related ones.

I am expecting to receive some PM hardware in a week or so, so I will be
able to test it's support, at least with AHCI.

-- 
Alexander Motin