Re: [PATCH v8 2/8] KVM: Extend the memslot to support fd-based private memory

[Sean Christopherson]

On Fri, Oct 07, 2022, Jarkko Sakkinen wrote:
> On Thu, Oct 06, 2022 at 03:34:58PM +0000, Sean Christopherson wrote:
> > On Thu, Oct 06, 2022, Jarkko Sakkinen wrote:
> > > On Thu, Oct 06, 2022 at 05:58:03PM +0300, Jarkko Sakkinen wrote:
> > > > On Thu, Sep 15, 2022 at 10:29:07PM +0800, Chao Peng wrote:
> > > > > This new extension, indicated by the new flag KVM_MEM_PRIVATE, adds two
> > > > > additional KVM memslot fields private_fd/private_offset to allow
> > > > > userspace to specify that guest private memory provided from the
> > > > > private_fd and guest_phys_addr mapped at the private_offset of the
> > > > > private_fd, spanning a range of memory_size.
> > > > > 
> > > > > The extended memslot can still have the userspace_addr(hva). When use, a
> > > > > single memslot can maintain both private memory through private
> > > > > fd(private_fd/private_offset) and shared memory through
> > > > > hva(userspace_addr). Whether the private or shared part is visible to
> > > > > guest is maintained by other KVM code.
> > > > 
> > > > What is anyway the appeal of private_offset field, instead of having just
> > > > 1:1 association between regions and files, i.e. one memfd per region?
> > 
> > Modifying memslots is slow, both in KVM and in QEMU (not sure about Google's VMM).
> > E.g. if a vCPU converts a single page, it will be forced to wait until all other
> > vCPUs drop SRCU, which can have severe latency spikes, e.g. if KVM is faulting in
> > memory.  KVM's memslot updates also hold a mutex for the entire duration of the
> > update, i.e. conversions on different vCPUs would be fully serialized, exacerbating
> > the SRCU problem.
> > 
> > KVM also has historical baggage where it "needs" to zap _all_ SPTEs when any
> > memslot is deleted.
> > 
> > Taking both a private_fd and a shared userspace address allows userspace to convert
> > between private and shared without having to manipulate memslots.
> 
> Right, this was really good explanation, thank you.
> 
> Still wondering could this possibly work (or not):
> 
> 1. Union userspace_addr and private_fd.

No, because userspace needs to be able to provide both userspace_addr (shared
memory) and private_fd (private memory) for a single memslot.

> 2. Instead of introducing private_offset, use guest_phys_addr as the
>    offset.

No, because that would force userspace to use a single private_fd for all of guest
memory since it effectively means private_offset=0.  And userspace couldn't skip
over holes in guest memory, i.e. the size of the memfd would need to follow the
max guest gpa.  In other words, dropping private_offset could work, but it'd be
quite kludgy and not worth saving 8 bytes.