Re: timeout errors on sata_sil24 with port multiplier & PCI-X

[Jim Paris <jim@jtan.com>]

I wrote:
> I have a SiI3124 PCI-X controller
..
> I have been using this configuration with a DS-1220 external
> enclosure + port multiplier:
..
> Occasionally, I get an error like this:
> 
> Mar  9 12:53:18 localhost kernel: ata8.00: exception Emask 0x0 SAct 0x1 SErr 0x0 action 0x2 frozen
> Mar  9 12:53:18 localhost kernel: ata8.00: tag 0 cmd 0x61 Emask 0x4 stat 0x40 err 0x0 (timeout)
> Mar  9 12:53:18 localhost kernel: ata8.15: hard resetting port

After moving the port multiplier from the PCI-X SiI3124 to a PCIe
SiI3132, the timeouts are gone.  No errors at all while moving 1T back
and forth and resyncing the RAID.  I wonder if it's a bad card
(Norco 4618, came with the enclosure) or a driver issue.


Robin H. Johnson wrote:
> Also, afaik there has been no resolution of the performance issue - while the
> controller shows 3Gb, the actual combined speeds I get are only at the 1.5Gb
> level.

I agree, on a single port combined speeds are right around 150MB/sec.
On two ports it hits 180MB/sec.  I guess that's approaching the PCIe x1
limit.

One interesting note is that the original SiI3124 controller always
reported the host->PMP link as 1.5Gb.  The SiI3132 reports it as
3.0Gb.

-jim