Messages in this thread | | | Date | Mon, 22 Dec 2008 18:54:57 +0900 | From | Tejun Heo <> | Subject | Re: Help interpreting AHCI failure messages |
| |
Hello,
Andrew Lyon wrote: >>>>> Furthermore, ata2 - which is sharing the same interrupt - seems to have >>>>> managed to read the DVD drive's information, so something must be >>>>> working. But why is it failing for the disk? >>>> Can you try irqpoll on it? >>> Here's a log of a run where everything works initially, but then it >>> starts to fail (about half way into the log). ATA_DEBUG is enabled, so >>> its pretty verbose. >>> >>> http://kraxel.fedorapeople.org/for-jeremy/xeni.log >>> >>> Does anything stand out? >> Can you please exclude pata_acpi from the initrd and see whether >> anything changes? > > I have already tried using a kernel with pata_acpi disabled, it does not help.
There's a BUG messages in the log and several IRQ enable/disable messages && once timeout happens on one of the ata port, the other one begins to time out too. I'm out of ideas and think this really needs a look from someone who knows xen. It doesn't seem to be a libata problem at this point.
Thanks.
-- tejun
| |