fix IovDeque for non 4K pages

[ShadowCurse]

Changes

The L const generic was determining the maximum number of iov
elements in the IovDeque. This cases the issue when the host kernel
uses pages which can contain more entries than L. For example usual
4K pages can contain 256 iovs while 16K pages can contain 1024 iovs.
Current implementation on 16K (and any other bigger than 4K page size)
will continue wrap IovDeque when it reaches 256'th element. This
breaks the implementation since elements written past 256'th index will
not be 'duplicated' at the beginning of the queue.

Curren implementation expects this behavior:

 page 1 page 2
|ABCD|#|ABCD|
      ^ will wrap here

With big page sizes current impl will:

 page 1              page2
|ABCD|EFGD________|#|ABCDEFGD________|
     ^ sill wrap here
                   ^ but should wrap here

The solution is to calculate the maximum capacity the IovDeque can
hold, and use it for wrapping purposes. This capacity is allowed to be
bigger than L. The actual used number of entries in the queue will
still be guarded by the L parameter used in the is_full method.

Reason

Fixes #5217

License Acceptance

By submitting this pull request, I confirm that my contribution is made under
the terms of the Apache 2.0 license. For more information on following Developer
Certificate of Origin and signing off your commits, please check
CONTRIBUTING.md.

PR Checklist

• I have read and understand CONTRIBUTING.md.
• I have run tools/devtool checkstyle to verify that the PR passes the
  automated style checks.
• I have described what is done in these changes, why they are needed, and
  how they are solving the problem in a clear and encompassing way.
• I have updated any relevant documentation (both in code and in the docs)
  in the PR.
• I have mentioned all user-facing changes in CHANGELOG.md.
• If a specific issue led to this PR, this PR closes the issue.
• When making API changes, I have followed the
  Runbook for Firecracker API changes.
• I have tested all new and changed functionalities in unit tests and/or
  integration tests.
• I have linked an issue to every new TODO.

---

• This functionality cannot be added in rust-vmm.

[codecov]

Codecov Report

Attention: Patch coverage is 87.50000% with 1 line in your changes missing coverage. Please review.

Project coverage is 82.93%. Comparing base (331ffec) to head (9b52af9).

Files with missing lines	Patch %	Lines
src/vmm/src/devices/virtio/iov_deque.rs	87.50%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5222      +/-   ##
==========================================
+ Coverage   82.88%   82.93%   +0.05%     
==========================================
  Files         250      250              
  Lines       26936    26942       +6     
==========================================
+ Hits        22325    22344      +19     
+ Misses       4611     4598      -13     

Flag	Coverage Δ
5.10-c5n.metal	83.37% <87.50%> (+<0.01%)	⬆️
5.10-m5n.metal	83.37% <87.50%> (+<0.01%)	⬆️
5.10-m6a.metal	82.58% <87.50%> (-0.01%)	⬇️
5.10-m6g.metal	79.20% <87.50%> (+<0.01%)	⬆️
5.10-m6i.metal	83.36% <87.50%> (-0.01%)	⬇️
5.10-m7a.metal-48xl	82.57% <87.50%> (?)
5.10-m7g.metal	79.20% <87.50%> (+<0.01%)	⬆️
5.10-m7i.metal-24xl	83.32% <87.50%> (?)
5.10-m7i.metal-48xl	83.33% <87.50%> (?)
5.10-m8g.metal-24xl	79.19% <87.50%> (?)
5.10-m8g.metal-48xl	79.19% <87.50%> (?)
6.1-c5n.metal	83.42% <87.50%> (+<0.01%)	⬆️
6.1-m5n.metal	83.41% <87.50%> (-0.01%)	⬇️
6.1-m6a.metal	82.64% <87.50%> (+<0.01%)	⬆️
6.1-m6g.metal	79.20% <87.50%> (+<0.01%)	⬆️
6.1-m6i.metal	83.40% <87.50%> (-0.01%)	⬇️
6.1-m7a.metal-48xl	82.62% <87.50%> (?)
6.1-m7g.metal	79.20% <87.50%> (+<0.01%)	⬆️
6.1-m7i.metal-24xl	83.42% <87.50%> (?)
6.1-m7i.metal-48xl	83.43% <87.50%> (?)
6.1-m8g.metal-24xl	79.19% <87.50%> (?)
6.1-m8g.metal-48xl	79.19% <87.50%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Manciukic]

could you to add a unit test to verify the behaviour when L < capacity? For example, using L=64 and testing a scenario where it was failing before this change

[louwers]

Confirmed that this fix seems to solve the problem I reported.

The issue is no longer reproducible.