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

Re: [Xen-devel] Issue in MPX support in Xen



> From: Jan Beulich [mailto:JBeulich@xxxxxxxx]
> Sent: Friday, August 26, 2016 4:09 PM
> 
> >>> On 25.08.16 at 18:53, <dmitry.rockosov@xxxxxxxxx> wrote:
> > We are working on enabling XEN with WindRiver Simics, which is Intel
> > reference functional simulator for servers.
> >
> > We found the issue in XEN with MPX using.
> > If MPX is supported by CPUID, but MPX is not supported by VMX, XEN is
> > failing on store CPU MSR GUEST_BNDCFGS (file
> > xen-4.7.0/xen/arch/x86/hvm/vmx/vmx.c:798).
> >
> > SDM says in 24.4.1 Guest Register State:
> > IA32_BNDCFGS (64 bits). This field is supported only on processors that
> > support either the 1-setting of the
> > "load IA32_BNDCFGS" VM-entry control or that of the "clear IA32_BNDCFGS"
> > VM-exit control.
> >
> > Looks like XEN doesn't consult VM-entry control or VM-exit control listed in
> > SDM.
> 
> Could you give the attached patch a try? It's against -unstable and
> will require trivial adjustment to two of the vmx.c hunks in order to
> apply to 4.7.
> 
> Andrew, Kevin, Jun - please take a look as well, in particular regarding
> the non-commit-message remark.
> 
> Jan

My feeling is below:

- Looks checking hvm_cpuid is a natural condition since MPX in hvm_cpuid 
is set only when both situations are true (CPU MPX and VMX MPX)

- For save/load_msr, it's purely about VMCS structure. So checking VMX
MPX alone should be enough.

Thanks
Kevin

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

 


Rackspace

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