Fix: main thread timeout early than work thread
This commit is contained in:
@@ -25,8 +25,11 @@ This document tracks the current DNS leak detector implementation against the de
|
||||
- `dns leak watch --iface-diag` (diagnostics for capture-capable interfaces).
|
||||
- Interface selection:
|
||||
- per-interface open timeout to avoid capture hangs
|
||||
- stable default pick (up, non-loopback, named ethernet/wlan) before fallback scan
|
||||
- ordered scan prefers non-loopback + named ethernet/wlan and interfaces with IPs
|
||||
- verbose logging of interface selection attempts (use `-v` / `-vv`)
|
||||
- overall watch timeout accounts for worst-case interface scan time
|
||||
- Capture loop:
|
||||
- receiver runs in a worker thread; main loop polls with a short timeout to avoid blocking
|
||||
|
||||
## Partially implemented
|
||||
- Route/interface classification: heuristic only (loopback/tunnel/physical by iface name).
|
||||
|
||||
Reference in New Issue
Block a user