[Based on #2159] debug: keep optimization and SSA rewrites DWARF-safe

[cpunion]

Based on #2159.

Problem

DWARF must not change executable semantics. Enabling the normal optimization pipeline for debug builds exposed four independent places where debug metadata affected code generation:

• linked builds skipped the configured LLVM pass pipeline whenever DWARF was emitted
• conditional defer lowering created inlinable runtime calls without !dbg, so full LTO verification aborted
• ssa.DebugRef users blocked LLGo's return-load and single-case-select order repairs
• ssa.DebugRef users blocked otherwise valid static global slice folding

Running full LTO with DWARF also exposes aggregate Go string values to the reflect MethodByName plugin. The prerequisite #2159 handles that plugin boundary independently and preserves both aggregate and legacy split-string analysis. This PR contains no LTO plugin implementation changes.

Changes

• run the normal LLVM pass pipeline for linked modes with or without DWARF; keep ModeGen as raw generated IR
• derive the compile unit's optimized marker from the actual pass decision and optimization level
• inherit the source builder's location when defer lowering splits the entry block
• move ssa.DebugRef users together with repaired return loads and select assignment dependencies, while still rejecting executable early uses
• ignore DebugRef pseudo-users when proving a static slice initializer

This consolidates the changes previously proposed in #2144 and #2146. It does not change the default -w policy, Python-specific debugging, pclntab policy, C ABI variable homes, LLDB language behavior, or non-debug optimization policy.

Fixes #2118.
Fixes #2121.
Fixes #2122.

Verification

• macOS, Go 1.26.5 / LLVM 19: full ssa and internal/build package tests
• macOS: DWARF plus full-LTO runtime coverage and return/select semantic acceptance
• Ubuntu arm64: DWARF plus LTO-plugin runtime coverage and four symbol-pruning cases
• Ubuntu arm64: focused cl, ssa, internal/build, and test/go coverage run (61.5% aggregate package coverage)
• changed decision, debug-location, value-ordering, and defer-init helpers are exercised directly

Ubuntu validation was limited to 2 CPUs, 15 GiB memory, and 512 PIDs.

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[codecov]

Codecov Report

❌ Patch coverage is 93.75000% with 1 line in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
cl/static_init.go	87.50%	0 Missing and 1 partial ⚠️

📢 Thoughts on this report? Let us know!

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.