Skip to content

Completed FileX Windows simulator regression support - #100

Open
fdesbiens wants to merge 8 commits into
eclipse-threadx:devfrom
fdesbiens:windows-sim-ports-refresh-2026-09-01
Open

fdesbiens wants to merge 8 commits into
eclipse-threadx:devfrom
fdesbiens:windows-sim-ports-refresh-2026-09-01

Conversation

@fdesbiens

Copy link
Copy Markdown
Contributor

Summary

  • add and stabilize the FileX Win32 MSVC simulator port
  • provide CMake/Ninja PowerShell build and test support for Win32 and Win64
  • preserve pointer width in the Windows port and RAM-disk regression helpers
  • make regression test targets safe under the legacy MSVC path limit while retaining descriptive executable and CTest names
  • fix Windows-specific directory and file regression assumptions

This refresh incorporates all useful work from the historical windows-sim-ports-completion branch; all four of that branch's patches are patch-equivalent here, so no legacy change remains to salvage.

Depends on eclipse-threadx/threadx#736.

Validation

  • Win32: 1,224/1,224 tests passed across all nine configurations
  • Win64: 1,224/1,224 tests passed across all nine configurations
  • Total: 2,448/2,448 tests passed

All builds used CMake and Ninja with MSVC and the ThreadX branch from eclipse-threadx/threadx#736. Tests ran serially with failed-test retry enabled, including the long fault-tolerance cases.

Notes

  • No external dependency was added.
  • Existing MSVC warnings in upstream regression sources remain unchanged.

The Win32 VS 2019 port has an include directory but no CMakeLists.txt, which prevents FileX from being consumed as a CMake dependency. Wire the header-only port into the build.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Removed a stray copyright line from the version initializer, corrected the copyright year, and aligned the Win32 version string with the target release.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Regenerated the Win32 port header with regression hooks, aligned MSVC large-buffer declarations, enabled large-address-aware Win32 test executables, and corrected the default ThreadX path.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Guarded fault-tolerant standalone buffers so non-fault-tolerant MSVC configurations do not declare zero-length arrays.

Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
Assisted-by: Codex (gpt-5.6-sol) <codex@openai.com>
@rahmanih

Copy link
Copy Markdown
Contributor

LGTM 👍

The branch added the older per-edit disclosure, which names the product and the
model version. That form accumulates: a second agent, or the same agent on a
newer model, appends another line rather than recognising the one already there,
and the spelling drifts with every release.

All 10 lines become the fixed text, written with the comment character each file
uses -- `//` in C sources and headers, `#` in CMakeLists and PowerShell:

    Portions of this file were generated with AI assistance.

Attribution for the change itself stays on the commit, in the Assisted-by
trailer. The AI Disclosure headers on the files this branch creates are
unchanged, and keep naming their product and model: a file is created once, so
that record cannot grow.

Comment-only, one line replaced per file, no file left with zero or more than
one. The new-file AI Disclosure headers are untouched.

Assisted-by: Claude Code (Opus 5) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants