On Mon, 2009-06-08 at 16:48 +0800, Avi Kivity wrote:
> Huang Ying wrote:
> > From: Andi Kleen <ak@xxxxxxxxxxxxxxx>
> >
> > VT-x needs an explicit MC vector intercept to handle machine checks in the
> > hyper visor.
> >
> > It also has a special option to catch machine checks that happen
> > during VT entry.
> >
> > Do these interceptions and forward them to the Linux machine check
> > handler. Make it always look like user space is interrupted because
> > the machine check handler treats kernel/user space differently.
> >
> > Thanks to Jiang Yunhong for help and testing.
> >
> >
>
> Noticed an additional problem. See below.
>
> > static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
> > {
> > struct vcpu_vmx *vmx = to_vmx(vcpu);
> > @@ -2616,6 +2640,10 @@ static int handle_exception(struct kvm_v
> > vect_info = vmx->idt_vectoring_info;
> > intr_info = vmcs_read32(VM_EXIT_INTR_INFO);
> >
> > + ex_no = intr_info & INTR_INFO_VECTOR_MASK;
> > + if (ex_no == MC_VECTOR)
> > + return handle_machine_check(vcpu, kvm_run);
> > +
> >
>
> It isn't sufficient to check the vector number; you also need to check
> the interrupt type. See for example is_page_fault() and is_no_device().
>
> I suggest you add an is_machine_check() along similar lines. Later we
> can collapse all those functions to use a single helper.
OK. I will do this.
> > --- a/arch/x86/include/asm/kvm_host.h
> > +++ b/arch/x86/include/asm/kvm_host.h
> > @@ -382,6 +382,8 @@ struct kvm_vcpu_arch {
> > u64 mcg_status;
> > u64 mcg_ctl;
> > u64 *mce_banks;
> > +
> > + u32 exit_reason;
> > };
> >
>
> Please move this into struct vcpu_vmx.
OK.
Best Regards,
Huang Ying
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
