Got "overflow when subtracting duration from instant" when "adapter.get_config()"

[Effect-Wei]

Hello, I got thread panicked when running adapter.get_config()
I'm using Windows 11 and rustc 1.77.1 (7cf61ebde 2024-03-27)
Maybe similar to this issue?

Here's the stacktrace:

thread '<unnamed>' panicked at library\std\src\time.rs:433:33:
overflow when subtracting duration from instant
stack backtrace:
   0: std::panicking::begin_panic_handler
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\std\src\panicking.rs:647
   1: core::panicking::panic_fmt
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\core\src\panicking.rs:72
   2: core::panicking::panic_display
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\core\src\panicking.rs:196
   3: core::panicking::panic_str
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\core\src\panicking.rs:171
   4: core::option::expect_failed
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\core\src\option.rs:1988
   5: core::option::Option::expect
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\core\src\option.rs:894
   6: std::time::impl$3::sub
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97/library\std\src\time.rs:434
   7: wireguard_nt::adapter::Adapter::get_config
             at C:\Users\name\.cargo\registry\src\mirrors.tuna.tsinghua.edu.cn-2eab394af869c8a2\wireguard-nt-0.3.0\src\adapter.rs:576
   8: my_code::wg_connect::closure$0
             at .\src\main.rs:191
   9: core::hint::black_box
             at /rustc/7cf61ebde7b22796c69757901dd346d0fe70bd97\library\core\src\hint.rs:334
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

Related part of my code:

    let _thread = std::thread::spawn(move || {
        std::thread::sleep(Duration::from_secs(1));
        'outer: loop {
            for _ in 0..5 {
                if stop.load(Ordering::Relaxed) {
                    drop(adapter);
                    break 'outer;
                }
                std::thread::sleep(Duration::from_millis(200));
            }
            let stats = adapter.get_config(); //*** Line 191 ***
            for peer in stats.peers {
                //let handshake_age = Instant::now().saturating_duration_since(peer.last_handshake);
                let handshake_age = Duration::from_secs(0);
                window
                    .emit(
                        "wg-statistics",
                        WgStatistics {
                            up: peer.tx_bytes,
                            down: peer.rx_bytes,
                            handshake_age: handshake_age.as_secs(),
                        },
                    )
                    .unwrap();
                println!(
                    "  {:?}, up: {}, down: {}, handshake: {:.1}s ago",
                    peer.allowed_ips,
                    peer.tx_bytes,
                    peer.rx_bytes,
                    handshake_age.as_secs()
                );
            }
        }
    });

[TroyNeubauer]

Good catch, thanks for filing!

Using SystemTime and checked_sub should fix this, will write a PR shortly.

I'm curious what your time setting is? Assuming its normal (April 2024?)

[Effect-Wei]

Is's April 2, 2024 23:04 (UTC+8) on my computer and server. This is the right time.

I also did some experiments and found that this may be because I adapter.get_config() before the first handshake. For example:

connect();
sleep(10s);
adapter.get_config()
print(handshakeAge);

It always give Last handshake: 5s ago instead of 10s and no more panic.

Hope this information helps.

(And sorry for my bad English)

[TroyNeubauer]

I have a possible fix out in #12. I was unable to replicate the panic, so could you test the fix-handshake-time-panic branch for me?

Also I changed WireguardPeer::last_handshake to be a SystemTime to better match the C code, which will be a breaking change

[Effect-Wei]

(Sorry for replying so late)

This fix works! The same code (with necessary changes) will no longer panic.

Now it gives:

handshake: 1712415051s ago    // about 54 years ago, that's 1970
handshake: 1712415052s ago
handshake: 1712415053s ago
handshake: 1712415054s ago
handshake: 1712415055s ago
handshake: 0s ago             // first handshake
handshake: 1s ago
...

I guess I can change my code to show "N/As ago" or whatever. It doesn't matter.

Thanks for your efforts!

[TroyNeubauer]

No worries, I'll change last_handshake to be an Option<SystemTime> to make this more clear and upload a new version soon!

[TroyNeubauer]

Just released v0.4.0.

Btw your english is good!