On 21/05/2025 10:18 pm, dm...@proton.me wrote: > From: Denis Mukhin <dmuk...@ford.com> > > The existing argo test depends on xenstore to retrieve the domain ID. > > Also, test does not perform peer-to-peer communication using Argo hypercall, > it > communicates with itself. > > Since xenstore currently runs in dom0, xenstore adds unnecessary dependency on > dom0 for the test in x86 QEMU environment. > > Use event channel to identify the domain ID which eliminates dom0 dependency > for the test. > > Signed-off-by: Denis Mukhin <dmuk...@ford.com> > --- > Changes since v1: > - dropped hack of hardcoding test's own domain ID to 0, reworked getting own > domain ID by relying on message channel. > > Thanks to Michal Orzel and Andrew Cooper for helping with that.
Thanks, although this needs integrating into the pending series I've already got. Notably, we want to use the CPUID if it's available. It also needs to be ahead of the argo test to avoid a bisection hazard. I'll pick this up and sort things out. ~Andrew
