linux-kernel - Re: [Part2 PATCH v5 31/31] KVM: X86: Restart the guest when insn

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 for Android: free password hash cracker in your pocket

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

Message-ID: <639d43f6-c403-4fff-7152-9814d9c5b9e6@amd.com>
Date:   Wed, 18 Oct 2017 07:19:59 -0500
From:   Brijesh Singh <brijesh.singh@....com>
To:     Paolo Bonzini <pbonzini@...hat.com>, x86@...nel.org,
        kvm@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:     brijesh.singh@....com, Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H. Peter Anvin" <hpa@...or.com>,
        Radim Krčmář <rkrcmar@...hat.com>,
        Joerg Roedel <joro@...tes.org>, Borislav Petkov <bp@...e.de>,
        Tom Lendacky <thomas.lendacky@....com>
Subject: Re: [Part2 PATCH v5 31/31] KVM: X86: Restart the guest when insn_len
 is zero and SEV is enabled

Hi Paolo,


On 10/18/17 4:26 AM, Paolo Bonzini wrote:
> On 04/10/2017 15:17, Brijesh Singh wrote:
>> +	/*
>> +	 * On AMD platforms, under certain conditions insn_len may be zero on #NPF.
>> +	 * This can happen if a guest gets a page-fault on data access but the HW
>> +	 * table walker is not able to read the instruction page (e.g instruction
>> +	 * page is not present in memory).
>> +	 *
>> +	 * Typically, when insn_len is zero, x86_emulate_instruction() walks the
>> +	 * guest page table and fetches the instruction bytes from guest memory.
>> +	 * When SEV is enabled, the guest memory is encrypted with guest-specific
>> +	 * key hence hypervisor will not able to fetch the instruction bytes.
>> +	 * In those cases we simply restart the guest.
>> +	 */
>> +	if (unlikely(!insn_len) &&
>> +	    kvm_x86_ops->mem_enc_enabled &&
>> +	    kvm_x86_ops->mem_enc_enabled(vcpu))
>> +		return 1;
>> +
> Is it needed to test mem_enc_enabled?  Could it instead test for the
> availability of decode assists?

We can use X86_FEATURE_DECODEASSIST but since mmu.c is common file hence
I was not sure if its safe for the VMX.  Is it okay to use decode assist
feature bit in mm.c ?

>
> Thanks,
>
> Paolo