fix: llvm 21.1.4 build

[16bit-ykiko]

Summary by CodeRabbit

• New Features

  • Added configurable build mode option for LLVM (debug, release, releasedbg)
  • Enabled Address Sanitizer in debug builds
  • Support for git-based versioning

• Improvements

  • Platform-specific optimizations: lld linker on Linux, arm64 support on macOS, clang-cl on Windows
  • Enhanced clang and clang-tools-extra inclusion in builds
  • Broadened LLVM target builds to include all targets to improve cross-compiler support and builtin type generation

[coderabbitai]

Walkthrough

Updated LLVM build configuration in xmake.lua: changed LLVM_TARGETS_TO_BUILD from Native to all within the on_install CMake configuration to broaden built targets for cross-compilation and builtin type generation.

Changes

Cohort / File(s)	Summary
LLVM build flag update xmake.lua	Changed LLVM_TARGETS_TO_BUILD value from Native to all inside the on_install CMake configuration (adjusts targets built by LLVM to include all architectures).

Sequence Diagram(s)

(Skipped — change is a single CMake flag adjustment and does not alter control flow.)

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

• Quick verification of CMake flag placement and that downstream packaging/CI expectations handle building all targets.
• Confirm no unintended build-time or size regressions in CI/tooling that assume Native.

Suggested reviewers

• aurora0x27

Poem

🐰
I hopped through flags both small and tall,
Swapped "Native" for "all" to welcome all.
Cross-compilers cheer, builtin types align,
A rabbit's small tweak — now many targets shine.

Pre-merge checks and finishing touches

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	You can run @coderabbitai generate docstrings to improve docstring coverage.
Title Check	❓ Inconclusive	The title "fix: llvm 21.1.4 build" refers to a real aspect of the change (LLVM build configuration) but lacks specificity about what the fix entails. While the changeset does fix an LLVM build issue by changing LLVM_TARGETS_TO_BUILD from "Native" to "all" to support cross-compilers, the title doesn't clearly convey this main change. A teammate scanning history would see the title and understand only that there is a build-related fix, without understanding the specific nature of the change or its impact.	Consider revising the title to be more specific about the actual change. For example: "build all llvm targets instead of native targets" or "enable all llvm build targets for cross-compiler support" would more clearly communicate the primary change to reviewers scanning the project history.

✅ Passed checks (1 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch fix-21.1.4

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between de49d1f and 438aae1.

📒 Files selected for processing (1)

• xmake.lua (1 hunks)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (5)

• GitHub Check: build (releasedbg, n)
• GitHub Check: build (debug, n)
• GitHub Check: build (releasedbg, n)
• GitHub Check: build (debug, n)
• GitHub Check: build (releasedbg, MD, n)

🔇 Additional comments (1)

xmake.lua (1)

97-99: Good explanation for the change.

The comment clearly explains the rationale for building all targets. However, consider adding a reference to the specific issue or build failure this fixes (e.g., issue number or brief description of the builtin type generation problem).