back from x-mas & new year vacation ...
>> It fails notably faster now, early in the init scripts. Some race
>> maybe? The timing changes due to the excessive debug output makes it
>> happen more likely?
Ok, seems to be pretty much random. Updated to latest kernel, ATA_DEBUG
is off again, a bunch of boots later: sometimes it dies at random places
in the init scripts, sometimes it boots up fine, lets me do stuff and
dies later on ...
> Well, it doesn't explain why it happens instantly for everyone else,
> unless there's some really odd timing issue going on. Is your xen the
> xen-unstable tip? Do you have DEBUG=y enabled?
Rawhide packages (3.3.0 right now, should be 3.3.1 soon).
>>> Also, does swiotlb=0 make any difference?
>>
>> It doesn't. Hmm, from the boot messages it looks like it doesn't
>> actually disable the swiotb ...
>
> Oh, maybe "iommu=off" then?
iommu=off seems to make no difference.
cheers
Gerd
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home