Re: [PATCH v2 1/3] KVM: x86: dump_vmcs should not assume GUEST_IA32_EFER is valid

[Jim Mattson]

On Fri, Feb 19, 2021 at 6:46 AM David Edmondson
<david.edmondson@xxxxxxxxxx> wrote:
>
> If the VM entry/exit controls for loading/saving MSR_EFER are either
> not available (an older processor or explicitly disabled) or not
> used (host and guest values are the same), reading GUEST_IA32_EFER
> from the VMCS returns an inaccurate value.
>
> Because of this, in dump_vmcs() don't use GUEST_IA32_EFER to decide
> whether to print the PDPTRs - do so if the EPT is in use and CR4.PAE
> is set.
>
> Fixes: 4eb64dce8d0a ("KVM: x86: dump VMCS on invalid entry")
> Signed-off-by: David Edmondson <david.edmondson@xxxxxxxxxx>
> ---
>  arch/x86/kvm/vmx/vmx.c | 8 +++-----
>  1 file changed, 3 insertions(+), 5 deletions(-)
>
> diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> index eb69fef57485..818051c9fa10 100644
> --- a/arch/x86/kvm/vmx/vmx.c
> +++ b/arch/x86/kvm/vmx/vmx.c
> @@ -5759,7 +5759,6 @@ void dump_vmcs(void)
>         u32 vmentry_ctl, vmexit_ctl;
>         u32 cpu_based_exec_ctrl, pin_based_exec_ctrl, secondary_exec_control;
>         unsigned long cr4;
> -       u64 efer;
>
>         if (!dump_invalid_vmcs) {
>                 pr_warn_ratelimited("set kvm_intel.dump_invalid_vmcs=1 to dump internal KVM state.\n");
> @@ -5771,7 +5770,6 @@ void dump_vmcs(void)
>         cpu_based_exec_ctrl = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
>         pin_based_exec_ctrl = vmcs_read32(PIN_BASED_VM_EXEC_CONTROL);
>         cr4 = vmcs_readl(GUEST_CR4);
> -       efer = vmcs_read64(GUEST_IA32_EFER);
>         secondary_exec_control = 0;
>         if (cpu_has_secondary_exec_ctrls())
>                 secondary_exec_control = vmcs_read32(SECONDARY_VM_EXEC_CONTROL);
> @@ -5784,8 +5782,7 @@ void dump_vmcs(void)
>                cr4, vmcs_readl(CR4_READ_SHADOW), vmcs_readl(CR4_GUEST_HOST_MASK));
>         pr_err("CR3 = 0x%016lx\n", vmcs_readl(GUEST_CR3));
>         if ((secondary_exec_control & SECONDARY_EXEC_ENABLE_EPT) &&
> -           (cr4 & X86_CR4_PAE) && !(efer & EFER_LMA))
> -       {
> +           (cr4 & X86_CR4_PAE)) {

Assuming that we really want to restrict the printing of the PDPTEs, I
think you also need to test "cr0 & CR0.PG" (cf. section 26.3.1.6 of
the SDM, volume 3).