Added function get_enabled to export vring enable status.

[WeiChungHsu]

Summary of the PR

Added function get_enabled to export vring enable status.

Please summarize here why the changes in this PR are needed.
I am working on snapshot/restore feature for vhost-device-vsock crate. And during snapshot, system sends "FrontendReq::SET_VRING_ENABLE, enable false" to vhost-device-vsock device. Then the vring was disabled. However, the vring tx/rx queue still execute iter() which will return error and then system (vring) hang because vring has been disabled.

My proposed solution to solve this snapshot/restore hang issue is adding vring enable status checking before access/iter() the vring. Therefore, we need this get_enabled function to check if vring is enabled or disabled. I believe in the future, other vhost device may have similar hang issue when it executes snapshot.

Requirements

Before submitting your PR, please make sure you addressed the following
requirements:

• All commits in this PR have Signed-Off-By trailers (with
  git commit -s), and the commit message has max 60 characters for the
  summary and max 75 characters for each description line.
• All added/changed functionality has a corresponding unit/integration
  test.
• All added/changed public-facing functionality has entries in the "Upcoming
  Release" section of CHANGELOG.md (if no such section exists, please create one).
• Any newly added unsafe code is properly documented.

[Ablu]

Do you have an example how the corresponding change looks on the vhost-device-vsock side?

[WeiChungHsu]

Do you have an example how the corresponding change looks on the vhost-device-vsock side?

When the Vring is disabled (and queue set to not ready), the vhu_vsock_thread.rs still try to iterate and read RX queue, therefore, it hang at this line:
https://github.com/rust-vmm/vhost-device/blob/main/vhost-device-vsock/src/vhu_vsock_thread.rs#L591-L594

Furthermore, it hang because this Queue iter() operation return error at this line:
https://github.com/rust-vmm/vm-virtio/blob/main/virtio-queue/src/queue.rs#L592-L597

My solution in vhu_vsock_thread.rs:
let mut vring_mut = vring.get_mut();

if !vring_mut.get_enabled() {
    return Ok(true);
}

    let queue = vring_mut.get_queue_mut();

    while let Some(mut avail_desc) = queue
        .iter(atomic_mem.memory())
        .map_err(|_| Error::IterateQueue)?
        .next()

[Ablu]

Furthermore, it hang because this Queue iter() operation return error at this line:
https://github.com/rust-vmm/vm-virtio/blob/main/virtio-queue/src/queue.rs#L592-L597

Hm. But shouldn't it error out in that case versus hanging? I do not fully understand where the hang would happen.

Ideally, I think it would be great if we can avoid easy footguns by designing the API in a way that avoids them altogether. I assume chances are that people will forget this check in other implementations too?

[stefano-garzarella]

@WeiChungHsu can you fix CI issues and rebase this?

[WeiChungHsu]

Do you have an example how the corresponding change looks on the vhost-device-vsock side?

Hi,

Couple days ago, I created a PR for vhost-device-vsock side. It used this new vring function to check if vring is enabled/disabled.
rust-vmm/vhost-device#827

Sorry for my late reply.

[WeiChungHsu]

@WeiChungHsu can you fix CI issues and rebase this?

Yes, I am working on it now.

[stefano-garzarella]

@WeiChungHsu please avoid merging from main, instead rebase this PR on current main.