Re: [PATCH v3 1/2] KVM: x86: always stop emulation on page fault

[Sean Christopherson]

On Thu, Aug 29, 2019 at 08:23:18AM +0000, Jan Dakinevich wrote:
> inject_emulated_exception() returns true if and only if nested page
> fault happens. However, page fault can come from guest page tables
> walk, either nested or not nested. In both cases we should stop an
> attempt to read under RIP and give guest to step over its own page
> fault handler.
> 
> Fixes: 6ea6e84 ("KVM: x86: inject exceptions produced by x86_decode_insn")

Commit ids should be at least 12 chars, the full tag should be

  Fixes: 6ea6e84309ca ("KVM: x86: inject exceptions produced by x86_decode_insn")

You can force this in your git config, e.g.

  git config --global core.abbrev 12

Both patches in this series should probably have

  Cc: <stable@xxxxxxxxxxxxxxx>


Reviewed-and-tested-by: Sean Christopherson <sean.j.christopherson@xxxxxxxxx>

> Cc: Denis Lunev <den@xxxxxxxxxxxxx>
> Cc: Roman Kagan <rkagan@xxxxxxxxxxxxx>
> Cc: Denis Plotnikov <dplotnikov@xxxxxxxxxxxxx>
> Signed-off-by: Jan Dakinevich <jan.dakinevich@xxxxxxxxxxxxx>
> ---
>  arch/x86/kvm/x86.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index b4cfd78..6bf7b55 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -6537,8 +6537,13 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu,
>  			if (reexecute_instruction(vcpu, cr2, write_fault_to_spt,
>  						emulation_type))
>  				return EMULATE_DONE;
> -			if (ctxt->have_exception && inject_emulated_exception(vcpu))
> +			if (ctxt->have_exception) {
> +				WARN_ON_ONCE(ctxt->exception.vector == UD_VECTOR);
> +				WARN_ON_ONCE(exception_type(ctxt->exception.vector)
> +					== EXCPT_TRAP);
> +				inject_emulated_exception(vcpu);
>  				return EMULATE_DONE;
> +			}
>  			if (emulation_type & EMULTYPE_SKIP)
>  				return EMULATE_FAIL;
>  			return handle_emulation_failure(vcpu, emulation_type);
> -- 
> 2.1.4
>