stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andy Lutomirski <luto@amacapital.net>
To: Paolo Bonzini <pbonzini@redhat.com>,
	linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc: "Michael S. Tsirkin" <mst@redhat.com>, stable@vger.kernel.org
Subject: Re: [PATCH 08/14] kvm: x86: don't kill guest on unknown exit reason
Date: Fri, 24 Oct 2014 10:57:52 -0700	[thread overview]
Message-ID: <544A9320.6010102@amacapital.net> (raw)
In-Reply-To: <1414163245-18555-9-git-send-email-pbonzini@redhat.com>

On 10/24/2014 08:07 AM, Paolo Bonzini wrote:
> From: "Michael S. Tsirkin" <mst@redhat.com>
> 
> KVM_EXIT_UNKNOWN is a kvm bug, we don't really know whether it was
> triggered by a priveledged application.  Let's not kill the guest: WARN
> and inject #UD instead.

This scares me a bit.  For guest CPL3, it's probably okay.  For guest
CPL0, on the other hand, #UD might not use IST (or a task switch on
32-bit guests), resulting in possible corruption if unprivileged code
controls SP.  Admittedly, there aren't that many contexts from which
that should happen (on Linux, at least), but something like #DF (or even
a triple fault) might be safer if the guest is at CPL0 when this happens.

--Andy

> 
> Cc: stable@vger.kernel.org
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  arch/x86/kvm/svm.c | 6 +++---
>  arch/x86/kvm/vmx.c | 6 +++---
>  2 files changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> index 00bed2c5e948..7527cefc5a43 100644
> --- a/arch/x86/kvm/svm.c
> +++ b/arch/x86/kvm/svm.c
> @@ -3551,9 +3551,9 @@ static int handle_exit(struct kvm_vcpu *vcpu)
>  
>  	if (exit_code >= ARRAY_SIZE(svm_exit_handlers)
>  	    || !svm_exit_handlers[exit_code]) {
> -		kvm_run->exit_reason = KVM_EXIT_UNKNOWN;
> -		kvm_run->hw.hardware_exit_reason = exit_code;
> -		return 0;
> +		WARN_ONCE(1, "vmx: unexpected exit reason 0x%x\n", exit_code);
> +		kvm_queue_exception(vcpu, UD_VECTOR);
> +		return 1;
>  	}
>  
>  	return svm_exit_handlers[exit_code](svm);
> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
> index cf3cd079ec52..a8b76c4c95e2 100644
> --- a/arch/x86/kvm/vmx.c
> +++ b/arch/x86/kvm/vmx.c
> @@ -7174,10 +7174,10 @@ static int vmx_handle_exit(struct kvm_vcpu *vcpu)
>  	    && kvm_vmx_exit_handlers[exit_reason])
>  		return kvm_vmx_exit_handlers[exit_reason](vcpu);
>  	else {
> -		vcpu->run->exit_reason = KVM_EXIT_UNKNOWN;
> -		vcpu->run->hw.hardware_exit_reason = exit_reason;
> +		WARN_ONCE(1, "vmx: unexpected exit reason 0x%x\n", exit_reason);
> +		kvm_queue_exception(vcpu, UD_VECTOR);
> +		return 1;
>  	}
> -	return 0;
>  }
>  
>  static void update_cr8_intercept(struct kvm_vcpu *vcpu, int tpr, int irr)
> 


  reply	other threads:[~2014-10-24 17:57 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1414163245-18555-1-git-send-email-pbonzini@redhat.com>
2014-10-24 15:07 ` [PATCH 01/14] KVM: x86: Check non-canonical addresses upon WRMSR Paolo Bonzini
2014-10-24 15:07 ` [PATCH 02/14] KVM: x86: Prevent host from panicking on shared MSR writes Paolo Bonzini
2014-10-24 15:07 ` [PATCH 03/14] KVM: x86: Improve thread safety in pit Paolo Bonzini
2014-10-24 15:07 ` [PATCH 04/14] KVM: x86: Fix wrong masking on relative jump/call Paolo Bonzini
2014-10-24 15:07 ` [PATCH 05/14] KVM: x86: Emulator fixes for eip canonical checks on near branches Paolo Bonzini
2014-10-24 17:53   ` Andy Lutomirski
2014-10-25 19:57     ` Nadav Amit
2014-10-25 23:51       ` Andy Lutomirski
2014-10-24 15:07 ` [PATCH 06/14] KVM: x86: Handle errors when RIP is set during far jumps Paolo Bonzini
2014-10-24 15:07 ` [PATCH 07/14] kvm: vmx: handle invvpid vm exit gracefully Paolo Bonzini
2014-10-24 15:07 ` [PATCH 08/14] kvm: x86: don't kill guest on unknown exit reason Paolo Bonzini
2014-10-24 17:57   ` Andy Lutomirski [this message]
2014-10-24 21:54     ` Paolo Bonzini
2014-10-24 22:26       ` Andy Lutomirski
2014-10-24 15:07 ` [PATCH 09/14] KVM: x86: Decoding guest instructions which cross page boundary may fail Paolo Bonzini
2014-10-24 15:07 ` [PATCH 10/14] KVM: emulate: avoid accessing NULL ctxt->memopp Paolo Bonzini
2014-10-24 15:07 ` [PATCH 11/14] KVM: x86: Emulator does not decode clflush well Paolo Bonzini
2014-10-24 15:07 ` [PATCH 12/14] KVM: x86: PREFETCH and HINT_NOP should have SrcMem flag Paolo Bonzini
2014-10-24 15:07 ` [PATCH 13/14] kvm: fix excessive pages un-pinning in kvm_iommu_map error path Paolo Bonzini
2014-10-24 15:58   ` Quentin Casasnovas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=544A9320.6010102@amacapital.net \
    --to=luto@amacapital.net \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).