Re: [V0 PATCH] pvh: expose feature flags from tools for domUs

[Mukesh Rathor <mukesh.rathor@oracle.com>]

On Tue, 28 Jan 2014 13:24:15 +0100
Roger Pau Monné <roger.pau@citrix.com> wrote:

> On 25/01/14 02:13, Mukesh Rathor wrote:
> > Expose features for pvh domUs from tools.
> > 
> > Signed-off-by: Mukesh Rathor <mukesh.rathor@oracle.com>
> > ---
> >  tools/libxc/xc_cpuid_x86.c |   26 ++++++++++++++++----------
> >  tools/libxc/xc_domain.c    |    1 +
> >  tools/libxc/xenctrl.h      |    2 +-
> >  3 files changed, 18 insertions(+), 11 deletions(-)
> > 
> > diff --git a/tools/libxc/xc_cpuid_x86.c b/tools/libxc/xc_cpuid_x86.c
> > index bbbf9b8..33f6829 100644
> > --- a/tools/libxc/xc_cpuid_x86.c
> > +++ b/tools/libxc/xc_cpuid_x86.c
> > @@ -433,7 +433,7 @@ static void xc_cpuid_hvm_policy(
> >  
> >  static void xc_cpuid_pv_policy(
> >      xc_interface *xch, domid_t domid,
> > -    const unsigned int *input, unsigned int *regs)
> > +    const unsigned int *input, unsigned int *regs, int is_pvh)
> >  {
> >      DECLARE_DOMCTL;
> >      unsigned int guest_width;
> > @@ -455,13 +455,16 @@ static void xc_cpuid_pv_policy(
> >  
> >      if ( (input[0] & 0x7fffffff) == 0x00000001 )
> >      {
> > -        clear_bit(X86_FEATURE_VME, regs[3]);
> > -        clear_bit(X86_FEATURE_PSE, regs[3]);
> > -        clear_bit(X86_FEATURE_PGE, regs[3]);
> > -        clear_bit(X86_FEATURE_MCE, regs[3]);
> > -        clear_bit(X86_FEATURE_MCA, regs[3]);
> > +        if ( !is_pvh )
> > +        {
> > +            clear_bit(X86_FEATURE_VME, regs[3]);
> > +            clear_bit(X86_FEATURE_PSE, regs[3]);
> > +            clear_bit(X86_FEATURE_PGE, regs[3]);
> > +            clear_bit(X86_FEATURE_MCE, regs[3]);
> > +            clear_bit(X86_FEATURE_MCA, regs[3]);
> 
> Should we enable MCA/MCE flags for PVH DomUs? It looks to me like Dom0
> is the only domain that can make use of MCE/MCA.

Yes, PVH, like any other guest, may setup MCE handlers if they are
enabled, and xen would inject any MCE to guest if it belongs to the
guest.

thanks
Mukesh



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