linux-kernel - Re: [PATCH 1/4] nvmx: use warn_on for buggy cases when emulating invept/invvpid

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 21 Jul 2016 11:13:37 +0200
From:	Paolo Bonzini <pbonzini@...hat.com>
To:	Bandan Das <bsd@...hat.com>, kvm@...r.kernel.org
Cc:	rkrcmar@...hat.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/4] nvmx: use warn_on for buggy cases when emulating
 invept/invvpid



On 21/07/2016 00:25, Bandan Das wrote:
> If L1 hypervisor decides to try out something weird, alert the
> host but only less aggressively. Also, remove the comment
> regarding nested vpid support since it is no longer valid.
> 
> Signed-off-by: Bandan Das <bsd@...hat.com>
> ---
>  arch/x86/kvm/vmx.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
> index 64a79f2..9fd0681 100644
> --- a/arch/x86/kvm/vmx.c
> +++ b/arch/x86/kvm/vmx.c
> @@ -2854,7 +2854,6 @@ static int vmx_get_vmx_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 *pdata)
>  			vmx->nested.nested_vmx_secondary_ctls_high);
>  		break;
>  	case MSR_IA32_VMX_EPT_VPID_CAP:
> -		/* Currently, no nested vpid support */

This is okay.

>  		*pdata = vmx->nested.nested_vmx_ept_caps |
>  			((u64)vmx->nested.nested_vmx_vpid_caps << 32);
>  		break;
> @@ -7462,7 +7461,7 @@ static int handle_invept(struct kvm_vcpu *vcpu)
>  		break;
>  	default:
>  		/* Trap single context invalidation invept calls */
> -		BUG_ON(1);
> +		WARN_ON(1);
>  		break;
>  	}
>  
> @@ -7525,7 +7524,7 @@ static int handle_invvpid(struct kvm_vcpu *vcpu)
>  		break;
>  	default:
>  		/* Trap individual address invalidation invvpid calls */
> -		BUG_ON(1);
> +		WARN_ON(1);
>  		break;
>  	}
>  
> 

These are BUGs because they are checked above:

        if (!(types & (1UL << type))) {
                nested_vmx_failValid(vcpu,
                                VMXERR_INVALID_OPERAND_TO_INVEPT_INVVPID);
                skip_emulated_instruction(vcpu);
                return 1;
        }

Guest-triggerable WARNs are only just a little better than
guest-triggerable BUGs.  Guest-triggerable messages should be
rate-limited printk.

I don't object to the change, but the commit message should be
modified (or the change dropped).

Paolo