runtime/wasm: add WASI single-worker scheduler (based on #2192, #2208)

[cpunion]

Depends on #2192 and #2208.

Part of #2152.

Problem

GOOS=wasip1 GOARCH=wasm llgo build can produce a module after #2192, but goroutines still select the pthread runtime. The resulting executable imports shared host memory and WASI pthread facilities, so ordinary single-worker Wasmtime execution cannot run Go scheduler behavior.

Implementation

• Make WASI pthread scheduling an explicit LLGO_WASI_THREADS=1 compatibility mode.
• Add a raw wasm32 execution-context backend and a one-M/one-P FIFO scheduler for default WASI Preview1 builds.
• Run package initialization and main.main as the first schedulable task while keeping the host entry on the system stack.
• Declare Asyncify as a typed target capability; executable linking writes a temporary core module, runs pinned Binaryen instrumentation, and atomically publishes the final module. Archives and shared libraries are unchanged.
• Compile LLVM 19 SjLj with legacy wasm EH, apply Asyncify, then translate the final module to standardized exnref EH.
• Keep the existing JS wasm scheduler behavior behind the same internal context boundary.
• Depend on ssa/wasm: use static defer continuation dispatch #2208 for static wasm defer continuation dispatch; without it, LLVM 19 crashes while selecting runtime.EnsureLocalInitializer after compiler/runtime: add //llgo:tls and //llgo:gls package variables #2079.

The implementation is intentionally limited to the scheduler/context foundation. Channel blocking, timers and host async integration, wasm GC integration, root/safepoint work, and optional WASI Preview2 support remain follow-up work tracked by #2152.

Validation

Bounded locally with GOMAXPROCS=2, -p=1, and a 15 GiB ceiling:

• macOS: J32, J64, and P1 scheduler fixtures pass, including deadlock behavior and 5,000 serial goroutine lifecycles.
• macOS: P1 passes -O0, -O3, and ThinLTO builds; the final module validates and has no default pthread/shared-memory imports.
• macOS: explicit LLGO_WASI_THREADS=1 still builds the historical pthread/shared-memory artifact.
• macOS: executable and archive post-link guards, the external-module _demo/c/helloc P1 build/execution path, Go 1.26.5 tests, and Go 1.24.2 host tests pass.
• Ubuntu 24.04 amd64 container: focused build/crosscompile/runtime tests pass; P1 builds, validates, runs under Wasmtime 39.0.1, and reports the expected deadlock.
• Full LTO still hits the existing LLVM 19 SelectionDAG crash in runtime.printany; the same failure reproduces on the unmodified runtime/wasm: add single-worker Asyncify scheduler #2192 base.
• The updated compiler/runtime: add //llgo:tls and //llgo:gls package variables #2079/ssa/wasm: use static defer continuation dispatch #2208 integration passes the full GitHub Ubuntu/macOS matrix and Codecov; the reproduced LLVM 19 WASI build, validation, and execution also pass locally.

Independent diff over #2192 and #2208: 26 files, +1,199/-35. This includes the focused post-link lifecycle coverage added after the initial review.

[codecov]

Codecov Report

❌ Patch coverage is 96.93252% with 5 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
internal/build/wasm_postlink.go	95.23%	2 Missing and 1 partial ⚠️
internal/build/build.go	85.71%	1 Missing and 1 partial ⚠️

📢 Thoughts on this report? Let us know!