Hi Eric, > From: Peter Xu <peterx@xxxxxxxxxx> > Sent: Monday, March 30, 2020 10:47 PM > To: Auger Eric <eric.auger@xxxxxxxxxx> > Subject: Re: [PATCH v2 00/22] intel_iommu: expose Shared Virtual Addressing to > VMs > > On Mon, Mar 30, 2020 at 12:36:23PM +0200, Auger Eric wrote: > > I think in general, as long as the kernel dependencies are not > > resolved, the QEMU series is supposed to stay in RFC state. > > Yeah I agree. I think the subject is not extremely important, but we definitely should > wait for the kernel part to be ready before merging the series. > > Side note: I offered quite a few r-bs for the series (and I still plan to move on > reading it this week since there's a new version, and try to offer more r-bs when I > still have some context in my brain-cache), however they're mostly only for myself > to avoid re-reading the whole series again in the future especially because it's > huge... :) Agreed. I'll rename the next version as RFCv6 then. BTW. although there is dependency on kernel side, but I think we'd get agreement on the interaction mechanism between vfio and vIOMMU within QEMU. Also, for the VT-d specific changes (e.g. the pasid cache invalidation patches and the pasid-based-iotlb invalidations), we can actually get them ready as they have no dependency on kernel side change. Please help. :-) Regards, Yi Liu
