On Wed, Sep 21, 2016 at 01:49:10PM +1000, David Gibson wrote: > On Tue, Sep 20, 2016 at 03:28:08PM +0800, Peter Xu wrote: > > On Tue, Sep 20, 2016 at 04:14:09PM +1000, David Gibson wrote: > > > On Wed, Sep 14, 2016 at 04:25:48PM +0800, Peter Xu wrote: > > > > Intel vIOMMU is still lacking of a complete IOMMU notifier mechanism. > > > > Before that is achieved, let's open a door for vhost DMAR support, which > > > > only requires cache invalidations (UNMAP operations). > > > > > > > > Meanwhile, converting hw_error() to error_report() and exit(1), to make > > > > the error messages clean and obvious (so no CPU registers will be > > > > dumped). > > > > > > Erm... AIUI currently the intel iommu driver doesn't do any > > > notifications. Surely it's only valid to allow this once you've > > > implemented unmap side notifications. > > > > Yes, I suppose vhost DMAR patches will be based upon this one. I can > > postpone this patch until Jason wants to pick it up, but it actually > > does not hurt if we just enable it now, anyway no one is using it. > > Even so, I think it's misleading to implicitly advertise a capability > that's not yet implemented. Changing this warning should happen after > or at the same time as implementing the notifications in the vIOMMU > model.
Sure. Then I'll post v6, and I'm okay with either way to merge it. Thanks. -- peterx