Re: [RFC PATCH v2 00/19] virtual-bus

[Gregory Haskins <ghaskins@novell.com>]

[-- Attachment #1: Type: text/plain, Size: 1701 bytes --]

Avi,

Gregory Haskins wrote:
>
> Todo:
> *) Develop some kind of hypercall registration mechanism for KVM so that
>    we can use that as an integration point instead of directly hooking
>    kvm hypercalls
>   

What would you like to see here?  I now remember why I removed the
original patch I had for registration...it requires some kind of
discovery mechanism on its own.  Note that this is hard, but I figured
it would make the overall series simpler if I didn't go this route and
instead just integrated with a statically allocated vector.  That being
said, I have no problem adding this back in but figure we should discuss
the approach so I don't go down a rat-hole ;)

So, one thing we could do is use a string-identifier to discover
hypercall resources.  In this model, we would have one additional
hypercall registered with kvm (in addition to the mmu-ops, etc) called
KVM_HC_DYNHC or something like that.  The support for DYNHC could be
indicated in the cpuid (much like I do with the RESET, DYNIRQ, and VBUS
support today.   When hypercall provides register, the could provide a
string such as "vbus", and they would be allocated a hypercall id. 
Likewise, the HC_DYNHC interface would allow a guest to query the cpuid
for the DYNHC feature, and then query the HC_DYNHC vector for a string
to hc# translation.  If the provider is not present, we return -1 for
the hc#, otherwise we return the one that was allocated.

I know how you feel about string-ids in general, but I am not quite sure
how to design this otherwise without it looking eerily similar to what I
already have (which is registering a new HC vector in kvm_para.h)

Thoughts?

-Greg


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 257 bytes --]