Skip to content

patina-v8.1.0

Choose a tag to compare

View all tags
@github-actions github-actions released this 18 Sep 17:58
· 745 commits to refs/heads/main since this release
patina-v8.1.0
572ca24
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

What's Changed

  • Update FV protocol implementations to make business logic more rusty @joschock (#729)
    Change Details
      ## Description

    Refactors fv.rs routines to split FFI APIs for the FV protocols from the FV business logic so that the business logic can use more idiomatic rust. Closes #715.

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    Unit tests for FV module continue to pass.

    Integration Instructions

    N/A.



  • .github/ci-workflow.yml: Run CI on feature and major branches @makubacki (#735)
    Change Details
      ## Description

    Run CI on "feature" branches and the "major" branch so changes are compatible with the requirements to merge to main.

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    • PR into a "feature/test" branch and a "major" branch on fork and confirm the workflows start

    Integration Instructions

    • N/A

  • [REBASE \& FF] dxe\_core: Introduce 32 bit memory prioritization option \& resolve allocator issues @cfernald (#732)
    Change Details
      ## Description

    This pull request adds functionality for prioritize allocating 32-bit address for platforms that need to support buggy external code. This change includes a few incidental changes details below for consistency and fixes related to this functionality.

    Introduce a prioritize_32_memory option for the core & GCD

    Some platforms may encounter issues using the Patina default top-down allocation strategy as it will prioritize high memory and some components may not properly handle 64 bit addresses. This could be because of fixed memory sizes or use of 32 bit variables to store addresses.

    This new routine will attempt tp priorritize 32 bit memory when used. This specifically applies to top down allocations as bottom up will inherently prioritize low memory and address allocations are not applicable.

    Switch TopDown to use max address for PI spec

    The PI specified allocation type for TopDown specifies a max address and not a min address for both memory and IO in section 7 as "EfiGcdAllocateMaxAddressSearchTopDown". The correct definition was used, but converted into an inverse type. This commit resolves this inconsistency and converts all top down searches to use a max instead of a min.

    Change ALLOCATE_MAX_ADDRESS to still use TopDown allocations

    For consistency this commit changes MAX_ADDRESS allocations to also use TopDown but with the max address set. This ensures that low memory is reserved for when it is actually needed and prevents fragmenting memory.

    Additionally, this change removed a condition in the allocate_pages_should_allocate_pages test as it incorrectly assumes MAX_ADDRESS follows a bottom up scheme.

    CLOSES #191

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    • Q35 OS Boot
    • SBSA OS Boot
    • Intel platform OS Boot.

    Integration Instructions

    N/A



  • Set RUSTC\_BOOTSTRAP for rust analyzer tests @cfernald (#730)
    Change Details
      ## Description

    Bootstrap is not generally needed anymore, but to use the "Run test" or "Debug" options that are available through rust analyzer, it needs to be set for that environment.

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    N/A

    Integration Instructions

    N/A



  • chore: Release @Javagedes (#728)
    Change Details
      ## Description
    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    N/A

    Integration Instructions

    N/A



🚀 Features & ✨ Enhancements

  • ParamFunction: Support Additional Self Types @Javagedes (#745)
    Change Details
      ## Description

    This commit enhances the blanket ParamFunction trait implementation to support functions that take &Self::In and &mut Self::In in addition to Self::In.

    What this translates to is that struct / enum components using the #[derive(IntoComponent)] macro can now define entry point functions that take &self or &mut self as their first parameter (in addition to self). This enhancement paves the way for future support of multi-callable dependency injectable types, including but not limited to components.

    Additionally, this commit introduces a custom error message when a function is being used as a ParamFunction but does not meet the function signature requirements. This should help users quickly identify and rectify issues with their function signatures.

    Improves the error message when the function you declare as your entry point does not meet the requirements of being a ParamFunction

    image
    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    CI continues to pass. Existing components continue to execute as expected.

    Integration Instructions

    N/A - But users may now have functions with the interface &self or &mut self



🐛 Bug Fixes

  • dxe\_core: Move runtime event handling into runtime module @cfernald (#738)
    Change Details
      ## Description

    This commit moves the handling of runtime events into the runtime.rs module and allocates them out of runtime memory. Runtime events, by definition and standard, are intended to be allocated in runtime memory and handled by a runtime driver. Since Patina handles the set_virtual_address_map itself, it must try to isolate that logic as much as possible from leveraging boot services resources.

    This commit specifically resolves an issue on Windows where the set virtual address map is not called from the full firmware context. In this seperate context, the runtime services memory is mapped, but newer boot services data allocations may not be mapped. Moving the allocations to runtime memory works around potential access violations when traversing the tree that contains events that may have been allocated from a later heap.

    CLOSES #337

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    Boot to Windows on Q35

    Integration Instructions

    N/A



📖 Documentation Updates

  • RFC Amendment: Add feature branch merge criteria to release process @makubacki (#748)
    Change Details
      ## Description

    Resolves #743

    Status: FCP

    Adds a section that describes the criteria for work on a feature branch to be merged to the main branch.

    The goal is to provide clarity so feature owners can plan how to organize their work to support the requirements and the process of adding new features is consistent over time for Patina consumers.

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    • cargo make all

    Integration Instructions

    • N/A

  • Add Patina Rust version update policy @makubacki (#742)
    Change Details
      ## Description

    • Adds Rust version update guidance to README.md.

    • Adds a new workflow to help ensure Rust version update guidance is not forgotten when the Rust version is updated.

    • Impacts functionality?

    • Impacts security?

    • Breaking change?

    • Includes tests?

    • Includes documentation?

    How This Was Tested

    • Run workflow on a fork checking:
    • Workflow does not run when rust-toolchain.toml is not modified
    • Workflow runs when rust-toolchain.toml is modified
    • Comment is left on the PR when expected
    • Old and New versions in the PR are correct per the PR changes
    • A comment is not left if one was already left on the PR

    Example

    An example of the comment posted when the toolchain version is modified:

    image

    Integration Instructions

    • N/A

  • README.md: Add high-level roadmap @makubacki (#733)
    Change Details
      ## Description

    Provide a brief roadmap of Patina's upcoming priorities.

    This is in the readme as it is intended to be easy-to-find on a user's first visit to the project.

    • Impacts functionality?
    • Impacts security?
    • Breaking change?
    • Includes tests?
    • Includes documentation?

    How This Was Tested

    • cargo make all

    Integration Instructions

    • N/A

Full Changelog: patina-v8.0.0...v8.1.0

Contributors

cfernald, makubacki, and 2 other contributors
Assets 2
Loading