Virtio blk #36

DINGBROK423 · 2025-12-23T09:53:40Z

为接入虚拟块设备做了相关适配

Copilot

Pull request overview

This PR adds support for virtual block devices (Virtio-blk) to the hypervisor by integrating MMIO-based virtio block device configuration and enhancing memory management capabilities.

Wraps the VM's address space in Arc<Mutex<>> to enable safe concurrent access from virtio device handlers
Implements memory region mapping logic with support for identical mapping, allocated mapping, and reserved mapping types
Adds guest memory access closures (read/write operations and IRQ injection) for virtio device communication

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 11 comments.

File	Description
src/vm.rs	Core changes: wraps address_space in Arc<Mutex<>>, adds memory region mapping logic, implements guest memory access closures for virtio devices, updates all address_space access sites with lock() calls
src/config.rs	Adds VirtioBlkMmioDeviceConfig support, exposes memory_regions field and related methods, adds get_vcpu_affinities_pcpu_ids method
Cargo.toml	Updates dependencies to personal fork branches for axdevice and axvmconfig to support virtio-blk features

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-12-23T09:56:58Z

src/vm.rs

+                match addr_space.translated_byte_buffer(addr, data.len()) {
+                    Some(mut buffer) => {
+                        let mut copied_bytes = 0;
+                        for (_i, chunk) in buffer.iter_mut().enumerate() {


The variable _i in the iterator enumerate is unused (indicated by the underscore prefix). Consider removing the enumerate entirely and just using iter_mut() directly since the index is not needed.

Suggested change

for (_i, chunk) in buffer.iter_mut().enumerate() {

for chunk in buffer.iter_mut() {

Copilot · 2025-12-23T09:56:59Z

src/vm.rs

+        let vcpu_id_pcpu_sets = config.phys_cpu_ls.get_vcpu_affinities_pcpu_ids();
+


The variable vcpu_id_pcpu_sets is declared but never used in this function. Consider removing this unused variable declaration to avoid confusion and improve code cleanliness.

Suggested change

let vcpu_id_pcpu_sets = config.phys_cpu_ls.get_vcpu_affinities_pcpu_ids();

Copilot · 2025-12-23T09:56:59Z

src/config.rs

    pub fn interrupt_mode(&self) -> VMInterruptMode {
        self.interrupt_mode
    }
+


The newly added public method get_vcpu_affinities_pcpu_ids() lacks documentation comments. Public API methods should have doc comments explaining their purpose, return value format, and usage to improve code maintainability and API usability.

Suggested change

/// Returns the vCPU-to-pCPU affinity information for this VM.

///

/// Each element in the returned vector is a tuple of

/// `(vcpu_index, pcpu_id, pcpu_set_index)`:

/// - `vcpu_index`: the index of the vCPU within this VM.

/// - `pcpu_id`: an optional physical CPU identifier; `None` means the vCPU

/// is not pinned to a specific physical CPU.

/// - `pcpu_set_index`: the index into the configured physical CPU set that

/// this vCPU is associated with.

///

/// This is a convenience wrapper around the underlying physical CPU list,

/// intended for callers that need to inspect or configure vCPU pinning.

Copilot · 2025-12-23T09:56:59Z

src/vm.rs

+                        Ok(_) => {},
+                        Err(e) => {
+                            if e != AxErrorKind::AlreadyExists {
+                                return Err(e.into());
+                            }
+                            warn!("Memory region [{:#x}] already mapped, skipping...", mem_region.gpa);
+                        }


This error handling pattern is duplicated from lines 240-244. The code for handling the AlreadyExists case is repeated for both MapIdentical and MapAlloc. Consider extracting this into a helper function to reduce code duplication and improve maintainability.

Copilot · 2025-12-23T09:56:59Z

src/vm.rs

+        let address_space = Arc::new(Mutex::new(
+            AddrSpace::new_empty(GuestPhysAddr::from(VM_ASPACE_BASE), VM_ASPACE_SIZE)?
+        ));


Adding Arc<Mutex<>> around the address_space introduces potential for lock contention and complexity. While this enables shared access needed for the guest memory access closures, consider whether a more fine-grained locking strategy or a different design pattern (such as channels or read-write locks) might better serve the concurrent access patterns. The current approach requires acquiring the lock for every address space operation, which could become a bottleneck.

Copilot · 2025-12-23T09:57:00Z

src/vm.rs

+                    ) {
+                        Ok(_) => {},
+                        Err(e) => {
+                            if e != AxErrorKind::AlreadyExists {


The error comparison if e != AxErrorKind::AlreadyExists is checking an error against an error kind. This is a type mismatch - e appears to be of type AxError based on the context (result of map_alloc), but it's being compared to AxErrorKind. You should extract the error kind from the error before comparison, for example using a method like .kind() if available, or match on the error's kind field.

Copilot · 2025-12-23T09:57:00Z

src/vm.rs

+                        }
+                    }
+
+


There are two consecutive blank lines here. Consider removing one to maintain consistent spacing throughout the codebase.

Suggested change

Copilot · 2025-12-23T09:57:00Z

Cargo.toml

+axdevice = { git = "https://github.com/DINGBROK423/axdevice.git", branch = "virt-blk", default-features = false }
 axdevice_base = "0.1"
-axvmconfig = { version = "0.1", default-features = false }
+axvmconfig = { git = "https://github.com/DINGBROK423/axvmconfig.git", branch = "merge_add_virtio_mmio", default-features = false }


The dependencies are pointing to personal GitHub forks (DINGBROK423) instead of the official arceos-hypervisor organization repositories. For production use, these should point to stable, official releases or the main organization repositories to ensure maintenance, stability, and trustworthiness of the dependencies.

Copilot · 2025-12-23T09:57:01Z

Cargo.toml


 [patch.crates-io]
-axvmconfig = { git = "https://github.com/arceos-hypervisor/axvmconfig.git", branch = "next" }
+axvmconfig = { git = "https://github.com/DINGBROK423/axvmconfig.git", branch = "merge_add_virtio_mmio", default-features = false }


This line appears to be redundant - it patches axvmconfig to the same repository and branch that's already specified in the dependencies section at line 30. Consider removing this patch entry to avoid confusion and potential version conflicts.

Suggested change

axvmconfig = { git = "https://github.com/DINGBROK423/axvmconfig.git", branch = "merge_add_virtio_mmio", default-features = false }

Copilot · 2025-12-23T09:57:01Z

src/config.rs

+    /// Returns configurations related to VM memory regions.
+    pub fn memory_regions(&self) -> &Vec<VmMemConfig> {
+        &self.memory_regions
+    }

-    // /// Adds a new memory region to the VM configuration.
-    // pub fn add_memory_region(&mut self, region: VmMemConfig) {
-    //     self.memory_regions.push(region);
-    // }
+    /// Adds a new memory region to the VM configuration.
+    pub fn add_memory_region(&mut self, region: VmMemConfig) {
+        self.memory_regions.push(region);
+    }


The newly added public methods memory_regions(), add_memory_region(), and virtio_blk_mmio() lack documentation comments. Public API methods should have doc comments explaining their purpose, parameters, and return values to improve code maintainability and API usability.

DINGBROK423 added 4 commits December 4, 2025 20:00

WIP: Add Virtio-blk support

65acdab

Merge upstream/next into virtio-blk

11f81a5

Fix compilation errors in vm.rs

a28c7eb

Integrated virtual block device

d432418

Copilot AI review requested due to automatic review settings December 23, 2025 09:53

Copilot started reviewing on behalf of DINGBROK423 December 23, 2025 09:54 View session

Copilot AI reviewed Dec 23, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Virtio blk #36

Virtio blk #36

Uh oh!

DINGBROK423 commented Dec 23, 2025

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Copilot AI Dec 23, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

	for (_i, chunk) in buffer.iter_mut().enumerate() {
	for chunk in buffer.iter_mut() {

		let vcpu_id_pcpu_sets = config.phys_cpu_ls.get_vcpu_affinities_pcpu_ids();

+    /// Returns the vCPU-to-pCPU affinity information for this VM.
+    ///
+    /// Each element in the returned vector is a tuple of
+    /// `(vcpu_index, pcpu_id, pcpu_set_index)`:
+    /// - `vcpu_index`: the index of the vCPU within this VM.
+    /// - `pcpu_id`: an optional physical CPU identifier; `None` means the vCPU
+    ///   is not pinned to a specific physical CPU.
+    /// - `pcpu_set_index`: the index into the configured physical CPU set that
+    ///   this vCPU is associated with.
+    ///
+    /// This is a convenience wrapper around the underlying physical CPU list,
+    /// intended for callers that need to inspect or configure vCPU pinning.

Virtio blk #36

Are you sure you want to change the base?

Virtio blk #36

Uh oh!

Conversation

DINGBROK423 commented Dec 23, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 23, 2025

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant