wchain: from "app is hung" to real wait edges

Date: February 05, 2026

wchain gives you thread-level wait-chain topology quickly, which is the missing middle between "it froze" and opening a full debugger session.

Immediate Snapshot

_test\wchain.exe --self --all-threads --include-running

PID 1234 (C:\tools\_test\wchain.exe)
Filter: all threads, include running threads

TID 32680 [GUI] [running] nodes=1 cycle=no
TID 27704 [non-GUI] [running] nodes=1 cycle=no
Summary: total=4 gui=1 inspected=4 displayed=4 blocked=0 failures=0

Useful as a quick sanity pass before deciding whether to attach debugger or gather dumps.

Workflow 1: UI-Hang Triage

_test\wchain.exe --pid <targetPid>

Default behavior is already tuned for GUI blocked-thread focus, so you get a concise first read instead of a wall of thread noise.

Workflow 2: Full Topology During Incident

_test\wchain.exe --pid <targetPid> --all-threads --include-running --verbose

Use this mode when you need a broad process state report and API detail for inaccessible threads.

Guardrails

WCT gives chain topology, not full scheduler truth. Treat this as high-value triage context and pair with traces/dumps for final diagnosis.