On Mon, Oct 06, 2008 at 11:08:34AM -0700, Don O'Neil wrote:
> The hardware I have is the built in SATA controller on the motherboard,
> which is GIGABYTE GA-M61P-S3. With the NVIDIA GeForce 6100 / nForce 430 and
> Super I/O chip: ITE IT8716.  
>
> Oct  4 04:07:30 kermit kernel: atapci0: <GENERIC ATA controller> port 
> 0x1f0-0x1f7,0x3f6,0x170-0x177,0x376,0xf000-0xf00f at device 6.0 on pci0
> Oct  4 04:07:30 kermit kernel: ata0: <ATA channel 0> on atapci0
> Oct  4 04:07:30 kermit kernel: ata1: <ATA channel 1> on atapci0
> Oct  4 04:07:30 kermit kernel: pci0: <bridge> at device 7.0 (no driver 
> attached)
> Oct  4 04:07:30 kermit kernel: atapci1: <GENERIC ATA controller> port 
> 0x9f0-0x9f7,0xbf0-0xbf3,0x970-0x977,0xb70-0xb73,0xd000-0xd00f mem 
> 0xf7004000-0xf7004fff irq 20 at device 8.0 on pci0
> Oct  4 04:07:30 kermit kernel: ata2: <ATA channel 0> on atapci1
> Oct  4 04:07:30 kermit kernel: ata3: <ATA channel 1> on atapci1
> Oct  4 04:07:30 kermit kernel: atapci2: <GENERIC ATA controller> port 
> 0x9e0-0x9e7,0xbe0-0xbe3,0x960-0x967,0xb60-0xb63,0xe400-0xe40f mem 
> 0xf7000000-0xf7000fff irq 21 at device 8.1 on pci0
> Oct  4 04:07:30 kermit kernel: ata4: <ATA channel 0> on atapci2
> Oct  4 04:07:30 kermit kernel: ata5: <ATA channel 1> on atapci2
> Oct  4 04:07:30 kermit kernel: ad0: 76293MB <Maxtor 6L080P0 BAH41G10> at 
> ata0-master UDMA33
> Oct  4 04:07:30 kermit kernel: ad4: 953869MB <Seagate ST31000340AS SD15> at 
> ata2-master UDMA33

This motherboard uses the nForce 430, but the SATA portion is actually a
subset chip called the MCP61.  I've confirmed this by looking at PRs
116880 and 108830.

I can see two things from the dmesg:

1) FreeBSD has no idea what this controller is, or any "quirks"
surrounding the controller (meaning it's possible that disk or block
addressing is being done incorrectly),

2) The disks are seen as classic PATA disks and not SATA.  This could be
a result of there being no nForce 430 support in 6.1, but it could also
be due to a BIOS setting on that motherboard.

I'm looking at the User Manual for this motherboard, but I can't find
the BIOS option that I'm used to seeing on other nForce-based boards,
and Intel ICH-based boards:

A feature where you can change the way the OS sees the underlying SATA
controller; it's called "Emulated" or "Emulation" mode.  The controller
is able to interface with SATA disks, but the OS sees the controller
as a classic PATA/IDE controller.  This is often used for OSes which
lack SATA support or native SATA drivers, such as MS-DOS.

The only thing in the User Manual I see which sets off red flags is the
"Serial-ATA RAID Config" item under the Integrated Peripherals menu.  I
really hope the "NV SATA Raid Function" is set to Disabled on your box.

Looking at CVS commit logs for src/sys/dev/ata/ata-chipset.c, I can see
that MCP61 support was officially added to HEAD on on 2007/06/26.  I'm
having a difficult time determining what HEAD meant at that date.  I
can't figure out for the life of me if it was referring to RELENG_6
or RELENG_7.

Either way, point is, FreeBSD 6.1 flat out does not have support for
that chip, even a 6.1 dated August 2006.  I can't help but wonder if
that's what's causing the odd problem.

I also found another LBA48-related issue, dated 2007/10/04, labelled
"fix the LBA28/LBA48 crossover bug".  I'm still not sure what that is.

And I haven't even begun to look at GEOM changes/bugfixes, which might
be a more likely place.

> This is actually a FreeBSD-Stable install... From 08/2006.... I realize it's
> probably time to do an OS upgrade, but this is the ONLY issue I've run into
> running this code base. Some of the software I'm running hasn't been tested
> with 7.X, so I'm not comfortable going there yet.

What this means is that it's a 6.1-RELEASE install which follows the
RELENG_6 tag, and has been cvsup'd at least up until August 2006.

I understand you're not comfortable upgrading to FreeBSD 7, but it would
be worthwhile if you could download FreeBSD 7.1-PRERELEASE (specifically
disc 1 or a live CD), and see if that reports the same problem as 6.1.

I still can't explain why booting the 6.1 installer and using a fixit
image lets you work around the problem.  That is just flat out bizarre.

You have to understand: there's been a lot of evolution/bugfixes applied
between 6.1 and 7.1.  There's almost too much for me to try and track
down.  I'm trying very hard, but it's difficult.

-- 
| Jeremy Chadwick                                jdc at parodius.com |
| Parodius Networking                       http://www.parodius.com/ |
| UNIX Systems Administrator                  Mountain View, CA, USA |
| Making life hard for others since 1977.              PGP: 4BD6C0CB |

_______________________________________________
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to