* On 2004-12-12 at 18:45 GMT, Jonathan Perkin wrote:

> Hmm, still broken :/
> 
>   GENERIC32:
>   [..]
>   ## Disable UDMA 4 which causes data corruption on the Acer Labs
>   ## chipset on Sun Blade 100 and Netra X1 machines.  
>   #wd*     at atabus? drive ? flags 0x0000
>   wd0     at atabus0 drive 0 flags 0xfac
>   wd1     at atabus1 drive 0 flags 0xfac
> 
>   wd1 at atabus1 drive 0: <ST3120026A>
>   wd1: drive supports 16-sector PIO transfers, LBA48 addressing
>   wd1: 111 GB, 232581 cyl, 16 head, 63 sec, 512 bytes/sect x 234441648 sectors
>   wd1: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 5 (Ultra/100)
>   wd1(aceride0:1:0): using PIO mode 4, DMA mode 2 (using DMA data transfers)
>   [..]
>   wd1a: DMA error writing fsbn 22634336 of 22634336-22634367 (wd1 bn 22634336; cn 22454 tn 11 sn 11), retrying
>   wd1: soft error (corrected)
> 
> I may just get some cables for good measure anyway.

Well, I tried two cables lying around, same results.  Interestingly,
however, I tried disabling one of the drives after reading about power
related issues driving both disks simultaneously as would happen in a
RAID1 configuration.  Since doing that, I've not had a single problem
even after a pkgsrc checkout which was spewing lots of errors
previously.  I'm hoping though that this isn't a PSU problem (not
likely to find a more powerful one) and is some kind of bus contention
with aceride(4) which can be worked around..

I also noticed the atabus changes bouyer made last year seem to have
removed the setting for disabling UDMA4; was this intentional?  I was
wondering why the comment in GENERIC32 didn't make any sense.

  -> cvs diff -r1.70 -r1.71 GENERIC32
  [..]
  ## Disable UDMA 4 which causes data corruption on the Acer Labs
  ## chipset on Sun Blade 100 and Netra X1 machines.
  -wd*    at pciide? channel ? drive ? flags 0x0a00 # Disable UDMA 4
  +wd*     at atabus? drive ? flags 0x0000

Either we need to have the 0x0a00 setting back or, if the problem has
been fixed and it auto detects correctly, remove the comment.

Thanks,

-- 
Jonathan Perkin                                     The NetBSD Project
http://www.perkin.org.uk/                       http://www.netbsd.org/