Re: [PATCH v2 2/4] x86: Introduce MSR_UNHANDLED

[Roger Pau Monné <roger.pau@xxxxxxxxxx>]

On Tue, Feb 23, 2021 at 08:57:00AM +0100, Jan Beulich wrote:
> On 22.02.2021 22:19, Boris Ostrovsky wrote:
> > 
> > On 2/22/21 6:08 AM, Roger Pau Monné wrote:
> >> On Fri, Feb 19, 2021 at 09:56:32AM -0500, Boris Ostrovsky wrote:
> >>> On 2/18/21 5:51 AM, Roger Pau Monné wrote:
> >>>> On Wed, Jan 20, 2021 at 05:49:10PM -0500, Boris Ostrovsky wrote:
> >>>>> When toolstack updates MSR policy, this MSR offset (which is the last
> >>>>> index in the hypervisor MSR range) is used to indicate hypervisor
> >>>>> behavior when guest accesses an MSR which is not explicitly emulated.
> >>>> It's kind of weird to use an MSR to store this. I assume this is done
> >>>> for migration reasons?
> >>>
> >>> Not really. It just seemed to me that MSR policy is the logical place to 
> >>> do that. Because it *is* a policy of how to deal with such accesses, 
> >>> isn't it?
> >> I agree that using the msr_policy seems like the most suitable place
> >> to convey this information between the toolstack and Xen. I wonder if
> >> it would be fine to have fields in msr_policy that don't directly
> >> translate into an MSR value?
> > 
> > 
> > We have xen_msr_entry_t.flags that we can use when passing policy array 
> > back and forth. Then we can ignore xen_msr_entry_t.idx for that entry 
> > (although in earlier version of this series Jan preferred to use idx and 
> > leave flags alone).
> 
> Which, just to clarify, was not the least because of the flags
> field being per-entry, i.e. per MSR, while here we want a global
> indicator.

We could exploit a bit the xen_msr_entry_t structure and use it
like:

typedef struct xen_msr_entry {
    uint32_t idx;
#define XEN_MSR_IGNORE (1u << 0)
    uint32_t flags;
    uint64_t val;
} xen_msr_entry_t;

Then use the idx and val fields to signal the start and size of the
range to ignore accesses to when XEN_MSR_IGNORE is set in the flags
field?

xen_msr_entry_t = {
    .idx = 0,
    .val = 0xffffffff,
    .flags = XEN_MSR_IGNORE,
};

Would be equivalent to ignoring accesses to the whole MSR range.

This would allow selectively selecting which MSRs to ignore, while
not wasting a MSR itself to convey the information?

It would still need to be stored somewhere in the Xen internal domain
structure using a rangeset I think, which could be translated back and
forth into this xen_msr_entry_t format.

Roger.