ad8: TIMEOUT - WRITE_DMA errors UFS 7.0-RC1

[Josh Carroll]

> Can anyone else using 7.0 who hasn't already (especially those using ZFS)
> check his/her /var/log/messages for disk TIMEOUTs or other disk error
> messages?  If this is widespread, I think the chances re slim that it is a
> hardware problem in every case.

I noticed this week that I was getting DMA WRITE timeouts on two of my
disks on my 7.0-RC1 box. The strange thing in my case was that only 2
of my 3 drives were exhibiting the behavior. The two having the
problem were both SATA300 drives, while the third was SATA150. I
jumpered the two "bad" drives to SATA150, and the timeouts went away.
I replaced the motherboard, set the two drives back to SATA300 and
have not had a timeout since. I was thinking maybe it was related, but
in my case it was truly a hardware failure.

Here were the messages, for what it's worth:

Jan 23 10:05:01 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=1525567
Jan 23 10:05:09 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=3828159
Jan 23 10:07:27 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=3450111
Jan 23 10:07:33 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=4561983
Jan 23 11:30:27 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=4178303
Jan 23 11:30:33 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=5660287
Jan 23 11:30:39 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=6805951
Jan 23 11:30:48 pflog kernel: ad10: TIMEOUT - WRITE_DMA retrying (1
retry left) LBA=9856959

Josh