[FreeBSD] Fix for ServerWorks HT1000 in upcoming 7.1?

[Mike Tancsa]

At 05:34 PM 9/30/2008, Fernan Aguero wrote:

>I've been following the "HT1000 chipset errata saga" thread, and the
>commits by sos@ to CVS (around Jan 2008), but have not seen other more
>recent posts about this issue ... is it because it's already fixed and
>working fine for everyone?
>
>http://lists.freebsd.org/pipermail/freebsd-current/2008-March/084272.html


Yes, we ran into this yesterday on a fresh install using the 7.1 beta 
CD as it was set to PATA mode by accident.  Also, on some earlier 
BIOS revs, we had to turn off "enable USB legacy mode" as well as 
"EHCI handoff".  By default we set those to disabled as it seems to 
sometimes create a high interrupt load on the USB bus if its enabled.

If you forget to set the mode to SATA, the dmesg will look like

Sep 30 13:37:08 dev2 kernel: ad4: DMA limited to UDMA33, device found 
non-ATA66 cable
Sep 30 13:37:08 dev2 kernel: ad4: 152627MB <Seagate ST3160815AS 
3.AAD> at ata2-master UDMA33

and you will indeed get corruption

Turning onto normal SATA mode in the BIOS, you should see

Sep 30 16:14:15 dev2 kernel: ad4: 152627MB <Seagate ST3160815AS 
3.AAD> at ata2-master SATA150

... And everything works great.

atapci0 at pci0:1:14:0:    class=0x010405 card=0x024a1166 
chip=0x024a1166 rev=0x00 hdr=0x00
     vendor     = 'ServerWorks (Was: Reliance Computer Corp)'
     device     = 'BCM5785 (HT1000) SATA Native SATA Mode'
     class      = mass storage
     subclass   = RAID
     cap 07[60] = PCI-X 64-bit supports 133MHz, 512 burst read, 8 
split transactions
     cap 01[90] = powerspec 2  supports D0 D3  current D0
     cap 05[a0] = MSI supports 1 message

Start with 
ftp://ftp.freebsd.org//pub/FreeBSD/releases/i386/ISO-IMAGES/7.1/7.1-BETA-i386-disc1.iso

         ---Mike