ATA DMA timeouts

[Andrew MacIntyre]

Tony Byrne wrote:
> Martin,
> 
> TB> We're now at 2 hours uptime and there is no sign of the ATA
> TB> timeouts. With the newer kernels the first timeout would appear
> TB> within 5 to 10 minutes of rebooting. I can't see any changes to
> TB> the ATA code in the interim so I'm curious as to what could have
> TB> caused this issue.
> 
> TB> My hardware is Intel ICH5 based with a Western Digital SATA hard-disk.
> 
> Well it was too good to last. I came in this morning to find a handful
> of "ad0: TIMEOUT - ..." in the messages file. It looks like the
> nightly cron jobs trigger it even for kernel and world prior to May
> 9th. I get the sense that there are less timeouts with the older
> kernel, but it could be wishful thinking. If your timeouts have really
> gone away then it's possible that our problem has a different cause to
> yours.
> 
> Can anyone else shed some light on this issue?

I've just been through something like this, eventually traced to one
rail (+5V) on the power supply sagging out of tolerance.

-------------------------------------------------------------------------
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  (pref) | Snail: PO Box 370
        andymac at pcug.org.au             (alt) |        Belconnen ACT 2616
Web:    http://www.andymac.org/               |        Australia