From: Brijesh Singh <brijesh.singh@amd.com>
To: linux-kernel@vger.kernel.org, x86@kernel.org, kvm@vger.kernel.org
Cc: Thomas Gleixner <tglx@linutronix.de>, Borislav Petkov <bp@suse.de>,
        Joerg Roedel <joro@8bytes.org>, "Michael S . Tsirkin" <mst@redhat.com>,
        Paolo Bonzini <pbonzini@redhat.com>,
        =?UTF-8?q?=5C=22Radim=20Kr=C4=8Dm=C3=A1=C5=99=5C=22?= 
        <rkrcmar@redhat.com>,
        Tom Lendacky <thomas.lendacky@amd.com>,
        Brijesh Singh <brijesh.singh@amd.com>
Subject: [RFC Part2 PATCH v3 26/26] KVM: X86: Restart the guest when insn_len is zero and SEV is enabled
Date: Mon, 24 Jul 2017 15:03:03 -0500
Message-Id: <20170724200303.12197-27-brijesh.singh@amd.com>
In-Reply-To: <20170724200303.12197-1-brijesh.singh@amd.com>
References: <20170724200303.12197-1-brijesh.singh@amd.com>
MIME-Version: 1.0
Content-Type: text/plain
SpamDiagnosticOutput: 1:99
SpamDiagnosticMetadata: NSPM
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 24 Jul 2017 20:03:53.1191 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BY2PR12MB0145
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 1901
Lines: 48

On AMD platform, under certain conditions insn_len may be zero on #NPF.
This can happen if guest gets a page-fault on data access, but HW table
walker is not able to read the instruction page (e.g instuction 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.

I have encountered this issue when running kernbench inside the guest.

Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
---
 arch/x86/kvm/mmu.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index ccb70b8..be41ad0 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -4850,6 +4850,23 @@ int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gva_t cr2, u64 error_code,
 	if (mmio_info_in_cache(vcpu, cr2, direct))
 		emulation_type = 0;
 emulate:
+	/*
+	 * On AMD platform, under certain conditions insn_len may be zero on #NPF.
+	 * This can happen if guest gets a page-fault on data access, but HW table
+	 * walker is not able to read the instruction page (e.g instuction page
+	 * is not present).
+	 *
+	 * Typically, when insn_len is zero, x86_emulate_instruction() walks the
+	 * guest page table and fetches the instruction bytes. When SEV is active,
+	 * the guest memory is encrypted with guest key hence we will not able to
+	 * fetch the instruction bytes. In those cases we simply restart the guest.
+	 */
+	if (unlikely(!insn_len)) {
+		if (kvm_x86_ops->memory_encryption_enabled &&
+			kvm_x86_ops->memory_encryption_enabled(vcpu))
+			return 1;
+	}
+
 	er = x86_emulate_instruction(vcpu, cr2, emulation_type, insn, insn_len);
 
 	switch (er) {
-- 
2.9.4