On Tue, 2024-07-30 at 07:13 -0700, Randy Dunlap wrote:

 	 
External email : Please do not click links or open attachments until
you have verified the sender or the content.
 Hi--

On 7/30/24 1:24 AM, Liju-clr Chen wrote:

quoted

From: Yingshiuan Pan <redacted>

GenieZone is MediaTek proprietary hypervisor solution, and it is

running

quoted

in EL2 stand alone as a type-I hypervisor. It is a pure EL2
implementation which implies it does not rely any specific host VM,

and

quoted

this behavior improves GenieZone's security as it limits its

interface.

quoted

Signed-off-by: Yingshiuan Pan <redacted>
Co-developed-by: Yi-De Wu <redacted>
Signed-off-by: Yi-De Wu <redacted>
Signed-off-by: Liju Chen <redacted>
---
 Documentation/virt/geniezone/introduction.rst | 87

+++++++++++++++++++

quoted

 Documentation/virt/index.rst                  |  1 +
 MAINTAINERS                                   |  6 ++
 3 files changed, 94 insertions(+)
 create mode 100644 Documentation/virt/geniezone/introduction.rst

diff --git a/Documentation/virt/geniezone/introduction.rst

b/Documentation/virt/geniezone/introduction.rst

quoted

new file mode 100644
index 000000000000..f280476228b3

--- /dev/null
+++ b/Documentation/virt/geniezone/introduction.rst

@@ -0,0 +1,87 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+======================
+GenieZone Introduction
+======================
+
+Overview
+========
+GenieZone hypervisor (gzvm) is a type-1 hypervisor that supports

various virtual

quoted

+machine types and provides security features such as TEE-like

scenarios and

quoted

+secure boot. It can create guest VMs for security use cases and

has

quoted

+virtualization capabilities for both platform and interrupt.

Although the

quoted

+hypervisor can be booted independently, it requires the assistance

of GenieZone

quoted

+hypervisor kernel driver(also named gzvm) to leverage the ability

of Linux

                     driver (also

quoted

+kernel for vCPU scheduling, memory management, inter-VM

communication and virtio

quoted

+backend support.
+
+Supported Architecture
+======================
+GenieZone now only supports MediaTek ARM64 SoC.
+
+Features
+========
+
+- vCPU Management
+
+  VM manager aims to provide vCPUs on the basis of time sharing on

physical

quoted

+  CPUs. It requires Linux kernel in host VM for vCPU scheduling

and VM power

quoted

+  management.
+
+- Memory Management
+
+  Direct use of physical memory from VMs is forbidden and designed

to be

quoted

+  dictated to the privilege models managed by GenieZone hypervisor

for security

quoted

+  reason. With the help of gzvm module, the hypervisor would be

able to manipulate

Is this change acceptable?:

             With the help of the gzvm module, the hypervisor is able
to manipulate

Hi Randy,
Sure, thank you for your comment.
Will update in next version.

quoted

+  memory as objects.
+
+- Virtual Platform
+
+  We manage to emulate a virtual mobile platform for guest OS

running on guest

     s/We manage to emulate/The gzvm hypervisor emulates/

or something like that...

quoted

+  VM. The platform supports various architecture-defined devices,

such as

quoted

+  virtual arch timer, GIC, MMIO, PSCI, and exception

watching...etc.

quoted

+
+- Inter-VM Communication
+
+  Communication among guest VMs was provided mainly on RPC. More

communication

                                   is provided

quoted

+  mechanisms were to be provided in the future based on VirtIO-

vsock.

                are to be provided
or
                will be provided

quoted

+
+- Device Virtualization
+
+  The solution is provided using the well-known VirtIO. The gzvm

module would

                                                           The gzvm
module

quoted

+  redirect MMIO traps back to VMM where the virtual devices are

mostly emulated.

     redirects

quoted

+  Ioeventfd is implemented using eventfd for signaling host VM

that some IO

quoted

+  events in guest VMs need to be processed.
+
+- Interrupt virtualization
+
+  All Interrupts during some guest VMs running would be handled by

GenieZone

         interrupts                               are handled

quoted

+  hypervisor with the help of gzvm module, both virtual and

physical ones.

quoted

+  In case there's no guest VM running out there, physical

interrupts would be

                     no guest VM running, physical interrupts are

quoted

+  handled by host VM directly for performance reason. Irqfd is

also implemented

quoted

+  using eventfd for accepting vIRQ requests in gzvm module.
+
+Platform architecture component
+===============================
+
+- vm
+
+  The vm component is responsible for setting up the capability

and memory

quoted

+  management for the protected VMs. The capability is mainly about

the lifecycle

quoted

+  control and boot context initialization. And the memory

management is highly

quoted

+  integrated with ARM 2-stage translation tables to convert VA to

IPA to PA

quoted

+  under proper security measures required by protected VMs.
+
+- vcpu
+
+  The vcpu component is the core of virtualizing aarch64 physical

CPU runnable,

The ending "runnable" doesn't seem to fit here - or I just can't
parse that.

quoted

+  and it controls the vCPU lifecycle including creating, running

and destroying.

quoted

+  With self-defined exit handler, the vm component would be able

to act

                                     the vm component is able to act

quoted

+  accordingly before terminated.

                 before termination.
or
                 before being terminated.
or
                 before exit.

quoted

+
+- vgic
+
+  The vgic component exposes control interfaces to Linux kernel

via irqchip, and

quoted

+  we intend to support all SPI, PPI, and SGI. When it comes to

virtual

quoted

+  interrupts, the GenieZone hypervisor would write to list

registers and trigger

                               hypervisor writes to list registers
and triggers

quoted

+  vIRQ injection in guest VMs via GIC.

HTH.
-- 
~Randy

BR
Liju