Expensive timeout caused by uma_timeout

Hi,

I have a fb not booting up in xen environment[ becomes very slow]. I tried getting the ps status from rc script and noticed that "100.00% swi4: clock sio" . On researching i gathered that its because of the callout resets taking more time and I enabled diagnostic to find out that uma_timeout was causing "Expensive timeout". I am not sure how to progress after this. any help on how to avoid this. This happens when I add em as PCI passthrough device, if I remove the pci config, things seems to be fine.

Thanks
Arvind
 
Back
Top