[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 3/7] vm-event: introduce vm_event_vcpu_enter



On 6/16/2016 7:17 PM, Tamas K Lengyel wrote:
On Thu, Jun 16, 2016 at 8:08 AM, Corneliu ZUZU <czuzu@xxxxxxxxxxxxxxx> wrote:
In an effort to improve on the vm-event interface, we introduce a new function
called vm_event_vcpu_enter. Its significance is that of a "final touch" vCPU
function - i.e. it should be called by implementing architectures just before
re-entering vCPUs.
On X86 for example, it is called on the scheduling tail (hvm_do_resume) and just
before reentering the guest world after a hypervisor trap (vmx_vmenter_helper).

I don't think this belongs to the vm_event system. It should probably
be in the monitor subsystem as it has nothing to do with passing
messages, which is essentially what vm_event is for.

Tamas


I figured it's usage is Xen-internal and it's meant to be of eventual use - whenever the need arises to do something just before a vCPU is reentered from the vm-event subsystem, you can simply adjust the implementation of that function. I thought non-monitor vm-events could also be 'eventual users'. But then again by this reasoning I could broaden its significance even more, outside the vm-event world and make it 'usable' by any parts of code in Xen. Shall I then rename it to vm_event_monitor_vcpu_enter and put it alongside the other vm_event_monitor_* functions?

Corneliu.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.