Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Zenghui Yu <yuzenghui@huawei.com>
To:     <kvmarm@lists.cs.columbia.edu>, <maz@kernel.org>
CC:     <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>, <eric.auger@redhat.com>,
        <james.morse@arm.com>, <julien.thierry.kdev@gmail.com>,
        <suzuki.poulose@arm.com>, <wanghaibin.wang@huawei.com>,
        Zenghui Yu <yuzenghui@huawei.com>
Subject: [PATCH 2/2] KVM: arm64: vgic: Forbid invalid userspace Distributor accesses
Date:   Fri, 13 Nov 2020 22:28:01 +0800
Message-ID: <20201113142801.1659-3-yuzenghui@huawei.com>
In-Reply-To: <20201113142801.1659-1-yuzenghui@huawei.com>
References: <20201113142801.1659-1-yuzenghui@huawei.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7BIT
Content-Type:   text/plain; charset=US-ASCII
Precedence: bulk

Accessing registers in the Distributor before setting its base address
should be treated as an invalid userspece behaviour. But KVM implicitly
allows it as we handle the access anyway, regardless of whether the base
address is set or not.

Fix this issue by informing userspace what had gone wrong (-ENXIO).

Signed-off-by: Zenghui Yu <yuzenghui@huawei.com>
---
 arch/arm64/kvm/vgic/vgic-mmio-v3.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/arm64/kvm/vgic/vgic-mmio-v3.c b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
index 30e370585a27..6a9e5eb311f0 100644
--- a/arch/arm64/kvm/vgic/vgic-mmio-v3.c
+++ b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
@@ -1029,11 +1029,15 @@ void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg, bool allow_group1)
 int vgic_v3_dist_uaccess(struct kvm_vcpu *vcpu, bool is_write,
 			 int offset, u32 *val)
 {
+	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
 	struct vgic_io_device dev = {
 		.regions = vgic_v3_dist_registers,
 		.nr_regions = ARRAY_SIZE(vgic_v3_dist_registers),
 	};
 
+	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base))
+		return -ENXIO;
+
 	return vgic_uaccess(vcpu, &dev, is_write, offset, val);
 }
 
-- 
2.19.1