Re: [Xen-devel] [PATCH 06/10] nested vmx: enable IA32E mode while do VM entry

["Xu, Dongxiao" <dongxiao.xu@xxxxxxxxx>]

> -----Original Message-----
> From: Jan Beulich [mailto:JBeulich@xxxxxxxx]
> Sent: Tuesday, December 04, 2012 6:03 PM
> To: Xu, Dongxiao
> Cc: xen-devel@xxxxxxxxxxxxx
> Subject: Re: [Xen-devel] [PATCH 06/10] nested vmx: enable IA32E mode while
> do VM entry
> 
> >>> On 04.12.12 at 06:53, Dongxiao Xu <dongxiao.xu@xxxxxxxxx> wrote:
> 
> How did things work without this, or if it worked, what does this fix?
> 
> Jan

For current Xen, it doesn't check the VM_ENTRY_IA32E_MODE bit in related MSR 
but directly enable this bit in VMCS if guest supports long mode.
Therefore Xen on Xen doesn't have problem.

However other VMMs may detect this bit in MSR and then set value to related 
VMCS fields.

Thanks,
Dongxiao

> 
> > Signed-off-by: Dongxiao Xu <dongxiao.xu@xxxxxxxxx>
> > ---
> >  xen/arch/x86/hvm/vmx/vvmx.c |    3 ++-
> >  1 files changed, 2 insertions(+), 1 deletions(-)
> >
> > diff --git a/xen/arch/x86/hvm/vmx/vvmx.c b/xen/arch/x86/hvm/vmx/vvmx.c
> > index 0ac78af..1304636 100644
> > --- a/xen/arch/x86/hvm/vmx/vvmx.c
> > +++ b/xen/arch/x86/hvm/vmx/vvmx.c
> > @@ -1388,7 +1388,8 @@ int nvmx_msr_read_intercept(unsigned int msr,
> > u64
> > *msr_content)
> >              tmp = 0x11fb;
> >          data = VM_ENTRY_LOAD_GUEST_PAT |
> >                 VM_ENTRY_LOAD_GUEST_EFER |
> > -               VM_ENTRY_LOAD_PERF_GLOBAL_CTRL;
> > +               VM_ENTRY_LOAD_PERF_GLOBAL_CTRL |
> > +               VM_ENTRY_IA32E_MODE;
> >          data = ((data | tmp) << 32) | tmp;
> >          break;
> >
> > --
> > 1.7.1
> >
> >
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@xxxxxxxxxxxxx
> > http://lists.xen.org/xen-devel
> 
> 


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