Windows installer: find the Wintun licence at the archive root

Co-Authored-By: Claude <noreply@anthropic.com>

.cargo/config.toml

@@ -0,0 +1,5 @@
+# Windows: link the C runtime statically, so the agent and the tray start on a
+# machine that has no Visual C++ redistributable installed. The installer ships
+# nothing but our own files and Wintun.
+[target.x86_64-pc-windows-msvc]
+rustflags = ["-C", "target-feature=+crt-static"]

.gitea/workflows/release.yml

@@ -190,6 +190,29 @@ jobs:
             tar -C dist/pkg -czf "${archive}" "${{ matrix.asset }}"
           fi
 
+      # The Windows installer: the agent as a service, Wintun, the tray and the
+      # PATH entry in one setup program. Wintun is the signed library from
+      # wintun.net, pinned by hash so a changed download fails the build.
+      - name: Build the Windows installer
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          $ErrorActionPreference = 'Stop'
+          $version = $env:GITHUB_REF_NAME.TrimStart('v')
+          Invoke-WebRequest -Uri https://www.wintun.net/builds/wintun-0.14.1.zip -OutFile wintun.zip
+          $hash = (Get-FileHash wintun.zip -Algorithm SHA256).Hash
+          if ($hash -ne '07C256185D6EE3652E09FA55C0B673E2624B565E02C4B9091C79CA7D2F24EF51') {
+            throw "wintun.zip has an unexpected hash: $hash"
+          }
+          Expand-Archive wintun.zip -DestinationPath wintun-dl
+          choco install innosetup -y --no-progress
+          $iscc = Join-Path ${env:ProgramFiles(x86)} 'Inno Setup 6\ISCC.exe'
+          $stage = Join-Path $PWD 'dist/pkg/${{ matrix.asset }}'
+          & $iscc "/DAppVersion=$version" "/DSourceDir=$stage" `
+            "/DWintunDir=$(Join-Path $PWD 'wintun-dl/wintun/bin/amd64')" `
+            "/O$PWD" dist/windows/tsunagi.iss
+          if ($LASTEXITCODE -ne 0) { throw "ISCC failed with $LASTEXITCODE" }
+
       # A dependency-free .deb for the x86_64 and aarch64 Linux builds: it drops
       # the static binary, the systemd service and the polkit rule. Built with
       # dpkg-deb (present on the runner); see dist/deb/build-deb.sh.
@@ -251,6 +274,7 @@ jobs:
             ${{ matrix.asset }}-*.*
             *.deb
             *.pkg.tar.zst
+            tsunagi-setup-*.exe
           if-no-files-found: error
 
   publish:
@@ -322,5 +346,5 @@ jobs:
               cat up.json >&2 || true
               rc=1
             fi
-          done < <(find artifacts -type f \( -name '*.tar.gz' -o -name '*.zip' -o -name '*.deb' -o -name '*.pkg.tar.zst' \) -print0)
+          done < <(find artifacts -type f \( -name '*.tar.gz' -o -name '*.zip' -o -name '*.deb' -o -name '*.pkg.tar.zst' -o -name 'tsunagi-setup-*.exe' \) -print0)
           exit "$rc"

.github/workflows/release.yml

@@ -192,6 +192,29 @@ jobs:
             tar -C dist/pkg -czf "${archive}" "${{ matrix.asset }}"
           fi
 
+      # The Windows installer: the agent as a service, Wintun, the tray and the
+      # PATH entry in one setup program. Wintun is the signed library from
+      # wintun.net, pinned by hash so a changed download fails the build.
+      - name: Build the Windows installer
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          $ErrorActionPreference = 'Stop'
+          $version = $env:GITHUB_REF_NAME.TrimStart('v')
+          Invoke-WebRequest -Uri https://www.wintun.net/builds/wintun-0.14.1.zip -OutFile wintun.zip
+          $hash = (Get-FileHash wintun.zip -Algorithm SHA256).Hash
+          if ($hash -ne '07C256185D6EE3652E09FA55C0B673E2624B565E02C4B9091C79CA7D2F24EF51') {
+            throw "wintun.zip has an unexpected hash: $hash"
+          }
+          Expand-Archive wintun.zip -DestinationPath wintun-dl
+          choco install innosetup -y --no-progress
+          $iscc = Join-Path ${env:ProgramFiles(x86)} 'Inno Setup 6\ISCC.exe'
+          $stage = Join-Path $PWD 'dist/pkg/${{ matrix.asset }}'
+          & $iscc "/DAppVersion=$version" "/DSourceDir=$stage" `
+            "/DWintunDir=$(Join-Path $PWD 'wintun-dl/wintun/bin/amd64')" `
+            "/O$PWD" dist/windows/tsunagi.iss
+          if ($LASTEXITCODE -ne 0) { throw "ISCC failed with $LASTEXITCODE" }
+
       # A dependency-free .deb for the x86_64 and aarch64 Linux builds: it drops
       # the static binary, the systemd service and the polkit rule. Built with
       # dpkg-deb (present on the runner); see dist/deb/build-deb.sh.
@@ -250,6 +273,7 @@ jobs:
             ${{ matrix.asset }}-*.*
             *.deb
             *.pkg.tar.zst
+            tsunagi-setup-*.exe
           if-no-files-found: error
 
   publish:
@@ -278,7 +302,7 @@ jobs:
 
           # download-artifact nests each artifact under its own directory, so
           # the archives are artifacts/<asset>/<file>.
-          mapfile -t files < <(find artifacts -type f \( -name '*.tar.gz' -o -name '*.zip' -o -name '*.deb' -o -name '*.pkg.tar.zst' \))
+          mapfile -t files < <(find artifacts -type f \( -name '*.tar.gz' -o -name '*.zip' -o -name '*.deb' -o -name '*.pkg.tar.zst' -o -name 'tsunagi-setup-*.exe' \))
           printf 'asset: %s\n' "${files[@]}"
 
           prerelease=()

Cargo.lock

@@ -6250,6 +6250,7 @@ dependencies = [
  "tsunagi-tcp-tls",
  "tsunagi-wg-quic",
  "tsunagi-wireguard",
+ "windows-service",
 ]
 
 [[package]]
@@ -7135,6 +7136,17 @@ dependencies = [
  "windows-link",
 ]
 
+[[package]]
+name = "windows-service"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "857224b3b211c6f3616921f081ee54721ee3ad2ace2fac6a6337e032f7b4dcf2"
+dependencies = [
+ "bitflags 2.13.2",
+ "widestring",
+ "windows-sys 0.61.2",
+]
+
 [[package]]
 name = "windows-strings"
 version = "0.1.0"

crates/tsunagi-core/src/agent/mod.rs

@@ -800,6 +800,11 @@ impl Agent {
         // itself stays, because the agent owns it and other networks may
         // still be using it.
         self.inner.routes.remove_network(network_id);
+        if let Some(interface) = self.inner.interface.get() {
+            interface
+                .set_underlay(network_id, Vec::new(), Vec::new())
+                .await;
+        }
         if let Some(interface) = self.inner.interface.get()
             && let Err(err) = interface.sync_addresses().await
         {

crates/tsunagi-core/src/agent/network.rs

@@ -712,6 +712,34 @@ impl Runtime {
         self.start_dials();
         // Last, so it carries what this round learned.
         self.introduce_peers();
+        self.refresh_underlay().await;
+    }
+
+    /// Hands the interface the underlay destinations this network depends
+    /// on: the verified direct IPv4 paths of every control connection, and
+    /// the relays in use. Only hosts that cannot exempt the agent from an
+    /// exit node's default route read it.
+    async fn refresh_underlay(&self) {
+        let Some(interface) = self.params.interface.clone() else {
+            return;
+        };
+        let mut addresses = Vec::new();
+        let mut hosts = Vec::new();
+        for session in self.sessions.values() {
+            for path in snapshot_connection(&session.conn).paths {
+                match path.remote {
+                    PathAddr::Ip(std::net::SocketAddr::V4(socket)) => addresses.push(*socket.ip()),
+                    PathAddr::Relay(url) => hosts.extend(relay_host(&url)),
+                    _ => {}
+                }
+            }
+        }
+        for url in self.params.adapter.snapshot().relay_urls {
+            hosts.extend(relay_host(&url));
+        }
+        interface
+            .set_underlay(self.network_id, addresses, hosts)
+            .await;
     }
 
     fn add_candidate(&mut self, candidate: Candidate) {
@@ -2448,3 +2476,48 @@ mod tests {
         assert_eq!(chosen_restored.copied(), Some("wg"));
     }
 }
+
+/// The host name of a relay URL such as `https://relay.example:443/`.
+fn relay_host(url: &str) -> Option<String> {
+    let rest = url.split_once("://").map_or(url, |(_, rest)| rest);
+    let authority = rest.split(['/', '?', '#']).next()?;
+    let host = authority.rsplit_once('@').map_or(authority, |(_, h)| h);
+    let host = if host.starts_with('[') {
+        host.split_once(']')?.0.trim_start_matches('[')
+    } else {
+        host.split(':').next()?
+    };
+    (!host.is_empty()
+        && host.len() <= 253
+        && host
+            .chars()
+            .all(|c| c.is_ascii_alphanumeric() || c == '.' || c == '-' || c == ':'))
+    .then(|| host.to_ascii_lowercase())
+}
+
+#[cfg(test)]
+mod relay_host_tests {
+    use super::relay_host;
+
+    #[test]
+    fn the_host_is_taken_out_of_a_relay_url() {
+        assert_eq!(
+            relay_host("https://relay.example.com./").as_deref(),
+            Some("relay.example.com.")
+        );
+        assert_eq!(
+            relay_host("https://Euw1-1.relay.iroh.network:443/x").as_deref(),
+            Some("euw1-1.relay.iroh.network")
+        );
+        assert_eq!(
+            relay_host("http://192.0.2.7:3340").as_deref(),
+            Some("192.0.2.7")
+        );
+        assert_eq!(
+            relay_host("https://[2001:db8::1]:443/").as_deref(),
+            Some("2001:db8::1")
+        );
+        assert_eq!(relay_host("https:///").as_deref(), None);
+        assert_eq!(relay_host("https://bad host/").as_deref(), None);
+    }
+}

crates/tsunagi-core/src/overlay/hostrules/exit.rs

@@ -39,6 +39,13 @@ pub struct ExitHostPlan {
     /// reads it, to keep this agent's own traffic *to* the overlay inside it
     /// where the host has no per-user routing (macOS).
     pub overlay: Vec<Ipv4Range>,
+    /// Underlay addresses this agent talks to directly and must keep reaching
+    /// outside the overlay's default route. Only filled while `client` is set
+    /// and only for hosts that report [`ExitHostRules::needs_bypass`].
+    pub bypass: Vec<std::net::Ipv4Addr>,
+    /// Host names (relay servers) that must stay reachable the same way; the
+    /// host rules resolve them.
+    pub bypass_hosts: Vec<String>,
 }
 
 impl ExitHostPlan {
@@ -76,6 +83,13 @@ pub trait ExitHostRules: Send + Sync + std::fmt::Debug + 'static {
     /// when nothing is there: it runs on the shutdown path, and once at
     /// startup to take away what a crashed run left.
     fn clear<'a>(&'a self, interface: &'a str) -> BoxFuture<'a, ()>;
+
+    /// Whether using an exit node on this host needs the underlay addresses
+    /// the agent talks to, because the host cannot exempt the agent's own
+    /// traffic from the default route by user id (Windows). The default is no.
+    fn needs_bypass(&self) -> bool {
+        false
+    }
 }
 
 /// The tag every firewall object of one offered range carries.
@@ -103,6 +117,7 @@ pub struct MockExitRules {
     calls: Arc<Mutex<Vec<String>>>,
     forwarding: Arc<Mutex<Option<bool>>>,
     failure: Arc<Mutex<Option<String>>>,
+    needs_bypass: Arc<Mutex<bool>>,
 }
 
 impl MockExitRules {
@@ -118,6 +133,11 @@ impl MockExitRules {
         *lock(&self.forwarding) = forwarding;
     }
 
+    /// Whether this host asks for the underlay addresses, like Windows.
+    pub fn set_needs_bypass(&self, needs: bool) {
+        *lock(&self.needs_bypass) = needs;
+    }
+
     /// Makes every step fail with this reason from now on.
     pub fn fail_with(&self, reason: Option<String>) {
         *lock(&self.failure) = reason;
@@ -142,6 +162,10 @@ fn lock<T>(mutex: &Mutex<T>) -> std::sync::MutexGuard<'_, T> {
 }
 
 impl ExitHostRules for MockExitRules {
+    fn needs_bypass(&self) -> bool {
+        *lock(&self.needs_bypass)
+    }
+
     fn apply<'a>(&'a self, plan: &'a ExitHostPlan) -> BoxFuture<'a, ExitHostReport> {
         Box::pin(async move {
             lock(&self.calls).push(format!("apply:{}", plan.interface));
@@ -209,6 +233,7 @@ mod tests {
             offer: vec![range()],
             client: true,
             overlay: vec![range()],
+            ..Default::default()
         };
         let report = rules.apply(&plan).await;
         assert_eq!(report.offer[0].1, RuleOutcome::Applied);
@@ -224,6 +249,7 @@ mod tests {
                 offer: vec![range()],
                 client: false,
                 overlay: Vec::new(),
+                ..Default::default()
             })
             .await;
         assert_eq!(report.forwarding, Some(false));

crates/tsunagi-core/src/overlay/hostrules/macos_exit.rs

@@ -623,6 +623,7 @@ destination: default\n\
             offer: if offer { vec![range()] } else { Vec::new() },
             client,
             overlay: vec![range()],
+            ..Default::default()
         }
     }
 

crates/tsunagi-core/src/overlay/hostrules/windows.rs

@@ -162,12 +162,36 @@ impl WindowsHostRules {
     }
 }
 
+/// The IPv4 addresses a host name resolves to; empty when it does not.
+fn resolve_host(host: &str) -> Vec<std::net::Ipv4Addr> {
+    use std::net::{IpAddr, ToSocketAddrs};
+    (host, 443)
+        .to_socket_addrs()
+        .map(|addrs| {
+            addrs
+                .filter_map(|addr| match addr.ip() {
+                    IpAddr::V4(ip) => Some(ip),
+                    IpAddr::V6(_) => None,
+                })
+                .collect()
+        })
+        .unwrap_or_default()
+}
+
 impl ExitHostRules for WindowsHostRules {
+    fn needs_bypass(&self) -> bool {
+        true
+    }
+
     fn apply<'a>(&'a self, plan: &'a ExitHostPlan) -> BoxFuture<'a, ExitHostReport> {
         Box::pin(async move {
             let job = plan.clone();
             tokio::task::spawn_blocking(move || {
-                windows_exit::apply_plan(&|script| run_powershell_output(script), &job)
+                windows_exit::apply_plan(
+                    &|script| run_powershell_output(script),
+                    &resolve_host,
+                    &job,
+                )
             })
             .await
             .unwrap_or_else(|err| {

crates/tsunagi-core/src/overlay/hostrules/windows_exit.rs

@@ -23,24 +23,46 @@
 //!
 //! # Using
 //!
-//! Not available. Windows has no per-user or per-process routing, so the
+//! Windows has no per-user or per-process routing, so the Linux scheme (an
-//! agent's own QUIC, relay and DHT traffic cannot be kept out of a default
+//! exemption for the agent's user) does not carry over. Instead the agent's
-//! route through the overlay: it would loop into the tunnel it carries. Rather
+//! own underlay traffic is exempted by *destination*:
-//! than install something that stops the agent working, the using side
+//!
-//! reports why it is not in place and changes nothing.
+//! ```text
+//! New-NetRoute 0.0.0.0/1   on the overlay interface, metric 1
+//! New-NetRoute 128.0.0.0/1 on the overlay interface, metric 1
+//! New-NetRoute <ip>/32 via the physical default gateway, metric 7373
+//! ```
+//!
+//! The halves out-rank the default route without replacing it, so every more
+//! specific route (LAN, another VPN) keeps working. A `/32` per address the
+//! agent talks to directly (the verified direct paths of its connections, and
+//! the relays it uses, resolved when the rules are applied) keeps that traffic
+//! on the physical network instead of looping into the tunnel it carries. The
+//! metric 7373 marks those routes as ours, which is how stale ones are found
+//! and removed. Everything is written to the active store, so a reboot clears
+//! it. IPv6 is blackholed with `::/1` and `8000::/1` on the overlay interface,
+//! which has no IPv6: applications fall back to IPv4 after a delay, where
+//! Linux and macOS answer at once. The bypass follows the agent's paths as
+//! they change; a path that appears after the last apply is reached through
+//! the tunnel until the next one.
+//!
+//! The bypass routes are written before the halves, and removed after them.
 
 use super::RuleOutcome;
+use std::net::Ipv4Addr;
+
 use super::exit::{ExitHostPlan, ExitHostReport, exit_tag, exit_tag_prefix};
 use crate::state::Ipv4Range;
 
-/// Why the using side is not installed.
+/// The route metric that marks a bypass route as this agent's.
-pub(super) const CLIENT_UNAVAILABLE: &str = "using an exit node is not available on Windows yet: \
+const BYPASS_METRIC: u32 = 7373;
-     Windows cannot keep the agent's own traffic out of a default route through the overlay, \
-     so it would loop";
 
 /// Runs a PowerShell script and returns what it printed.
 pub(super) type RunScript<'a> = &'a dyn Fn(&str) -> Result<String, String>;
 
+/// Resolves a host name to its IPv4 addresses.
+pub(super) type Resolve<'a> = &'a dyn Fn(&str) -> Vec<Ipv4Addr>;
+
 /// A PowerShell single-quoted literal, with any embedded quote doubled.
 fn ps_literal(value: &str) -> String {
     format!("'{}'", value.replace('\'', "''"))
@@ -86,6 +108,125 @@ fn forwarding_script(interface: &str) -> String {
     )
 }
 
+/// Prints `<interface index> <next hop>` of the best default route that is
+/// not through the overlay interface.
+fn gateway_script(interface: &str) -> String {
+    let alias = ps_literal(interface);
+    format!(
+        "$ErrorActionPreference = 'Stop'\n\
+         $alias = {alias}\n\
+         $best = Get-NetRoute -DestinationPrefix '0.0.0.0/0' -AddressFamily IPv4 \
+         -PolicyStore ActiveStore | Where-Object {{ $_.InterfaceAlias -ne $alias -and \
+         $_.NextHop -ne '0.0.0.0' }} | Sort-Object {{ $_.RouteMetric + \
+         (Get-NetIPInterface -InterfaceIndex $_.InterfaceIndex -AddressFamily IPv4).InterfaceMetric }} | \
+         Select-Object -First 1\n\
+         if (-not $best) {{ throw 'the host has no default gateway outside the overlay' }}\n\
+         \"$($best.InterfaceIndex) $($best.NextHop)\"\n"
+    )
+}
+
+/// The default gateway as the script above printed it.
+fn parse_gateway(text: &str) -> Option<(u32, Ipv4Addr)> {
+    let mut words = text.split_whitespace();
+    let index = words.next()?.parse().ok()?;
+    let hop = words.next()?.parse().ok()?;
+    words.next().is_none().then_some((index, hop))
+}
+
+/// Replaces the bypass routes with one per address, through the gateway.
+fn bypass_script(index: u32, gateway: Ipv4Addr, addresses: &[Ipv4Addr]) -> String {
+    let list: Vec<String> = addresses
+        .iter()
+        .map(|address| ps_literal(&address.to_string()))
+        .collect();
+    format!(
+        "$ErrorActionPreference = 'Stop'\n\
+         Get-NetRoute -AddressFamily IPv4 -PolicyStore ActiveStore -ErrorAction SilentlyContinue | \
+         Where-Object {{ $_.RouteMetric -eq {BYPASS_METRIC} -and $_.DestinationPrefix.EndsWith('/32') }} | \
+         Remove-NetRoute -Confirm:$false -ErrorAction SilentlyContinue\n\
+         foreach ($ip in @({})) {{\n\
+         New-NetRoute -DestinationPrefix \"$ip/32\" -InterfaceIndex {index} -NextHop '{gateway}' \
+         -RouteMetric {BYPASS_METRIC} -PolicyStore ActiveStore -ErrorAction SilentlyContinue | Out-Null\n\
+         }}\n",
+        list.join(", ")
+    )
+}
+
+/// Sends everything else through the overlay interface.
+fn halves_script(interface: &str) -> String {
+    let alias = ps_literal(interface);
+    format!(
+        "$ErrorActionPreference = 'Stop'\n\
+         $alias = {alias}\n\
+         foreach ($p in '0.0.0.0/1', '128.0.0.0/1') {{\n\
+         Get-NetRoute -DestinationPrefix $p -InterfaceAlias $alias -PolicyStore ActiveStore \
+         -ErrorAction SilentlyContinue | Remove-NetRoute -Confirm:$false -ErrorAction SilentlyContinue\n\
+         New-NetRoute -DestinationPrefix $p -InterfaceAlias $alias -NextHop '0.0.0.0' \
+         -RouteMetric 1 -PolicyStore ActiveStore | Out-Null\n\
+         }}\n\
+         foreach ($p in '::/1', '8000::/1') {{\n\
+         Get-NetRoute -DestinationPrefix $p -InterfaceAlias $alias -PolicyStore ActiveStore \
+         -ErrorAction SilentlyContinue | Remove-NetRoute -Confirm:$false -ErrorAction SilentlyContinue\n\
+         New-NetRoute -DestinationPrefix $p -InterfaceAlias $alias -NextHop '::' \
+         -RouteMetric 1 -PolicyStore ActiveStore -ErrorAction SilentlyContinue | Out-Null\n\
+         }}\n"
+    )
+}
+
+/// Removes the halves and every bypass route. What is not there is the state
+/// wanted.
+fn remove_client_script(interface: &str) -> String {
+    let alias = ps_literal(interface);
+    format!(
+        "$ErrorActionPreference = 'SilentlyContinue'\n\
+         $alias = {alias}\n\
+         foreach ($p in '0.0.0.0/1', '128.0.0.0/1', '::/1', '8000::/1') {{\n\
+         Get-NetRoute -DestinationPrefix $p -InterfaceAlias $alias -PolicyStore ActiveStore | \
+         Remove-NetRoute -Confirm:$false\n\
+         }}\n\
+         Get-NetRoute -AddressFamily IPv4 -PolicyStore ActiveStore | \
+         Where-Object {{ $_.RouteMetric -eq {BYPASS_METRIC} -and $_.DestinationPrefix.EndsWith('/32') }} | \
+         Remove-NetRoute -Confirm:$false\n\
+         exit 0\n"
+    )
+}
+
+/// Every address to keep direct: the ones given and what the names resolve
+/// to, sorted and without duplicates.
+fn bypass_addresses(
+    resolve: Resolve<'_>,
+    addresses: &[Ipv4Addr],
+    hosts: &[String],
+) -> Vec<Ipv4Addr> {
+    let mut all: std::collections::BTreeSet<Ipv4Addr> = addresses.iter().copied().collect();
+    for host in hosts {
+        all.extend(resolve(host));
+    }
+    all.into_iter()
+        .filter(|a| !a.is_loopback() && !a.is_unspecified() && !a.is_multicast())
+        .collect()
+}
+
+fn apply_client(
+    run: RunScript<'_>,
+    resolve: Resolve<'_>,
+    plan: &ExitHostPlan,
+) -> Result<(), String> {
+    let interface = plan.interface.as_str();
+    let gateway = run(&gateway_script(interface))?;
+    let (index, hop) = parse_gateway(&gateway)
+        .ok_or_else(|| format!("cannot read the default gateway from `{}`", gateway.trim()))?;
+    let addresses = bypass_addresses(resolve, &plan.bypass, &plan.bypass_hosts);
+    run(&bypass_script(index, hop, &addresses))?;
+    if let Err(why) = run(&halves_script(interface)) {
+        // Do not leave a half-installed set: bypass routes alone are harmless
+        // but the next apply starts clean.
+        let _ = run(&remove_client_script(interface));
+        return Err(why);
+    }
+    Ok(())
+}
+
 fn parse_forwarding(text: &str) -> Option<bool> {
     match text.trim().to_ascii_lowercase().as_str() {
         "enabled" | "1" => Some(true),
@@ -95,7 +236,11 @@ fn parse_forwarding(text: &str) -> Option<bool> {
 }
 
 /// Makes the host match the plan.
-pub(super) fn apply_plan(run: RunScript<'_>, plan: &ExitHostPlan) -> ExitHostReport {
+pub(super) fn apply_plan(
+    run: RunScript<'_>,
+    resolve: Resolve<'_>,
+    plan: &ExitHostPlan,
+) -> ExitHostReport {
     let interface = plan.interface.as_str();
     // What an earlier plan left and this one no longer wants.
     let pruned = run(&prune_script(interface, &plan.offer));
@@ -113,6 +258,16 @@ pub(super) fn apply_plan(run: RunScript<'_>, plan: &ExitHostPlan) -> ExitHostRep
             (*range, outcome)
         })
         .collect();
+    let client = if plan.client {
+        Some(match apply_client(run, resolve, plan) {
+            Ok(()) => RuleOutcome::Applied,
+            Err(why) => RuleOutcome::Failed(why),
+        })
+    } else {
+        // Whatever an earlier plan left must not stay in force.
+        let _ = run(&remove_client_script(interface));
+        None
+    };
     let forwarding = if plan.offer.is_empty() {
         None
     } else {
@@ -123,9 +278,7 @@ pub(super) fn apply_plan(run: RunScript<'_>, plan: &ExitHostPlan) -> ExitHostRep
     ExitHostReport {
         offer,
         forwarding,
-        client: plan
+        client,
-            .client
-            .then(|| RuleOutcome::Failed(CLIENT_UNAVAILABLE.to_string())),
     }
 }
 
@@ -134,6 +287,9 @@ pub(super) fn clear_all(run: RunScript<'_>, interface: &str) {
     if let Err(why) = run(&prune_script(interface, &[])) {
         tracing::warn!(interface, "cannot remove the exit node NAT: {why}");
     }
+    if let Err(why) = run(&remove_client_script(interface)) {
+        tracing::warn!(interface, "cannot remove the exit node routes: {why}");
+    }
 }
 
 #[cfg(test)]
@@ -154,6 +310,7 @@ mod tests {
             offer: vec![range()],
             client,
             overlay: vec![range()],
+            ..Default::default()
         }
     }
 
@@ -210,14 +367,16 @@ mod tests {
                 String::new()
             })
         };
-        let report = apply_plan(&run, &plan(false));
+        let report = apply_plan(&run, &no_names, &plan(false));
         assert_eq!(report.offer, vec![(range(), RuleOutcome::Applied)]);
         assert_eq!(report.forwarding, Some(false));
         assert_eq!(report.client, None);
         let calls = calls.borrow();
         assert!(calls[0].contains("StartsWith"), "{calls:?}");
         assert!(calls[1].contains("New-NetNat"), "{calls:?}");
-        assert!(calls[2].contains("Get-NetIPInterface"), "{calls:?}");
+        // Without a client, the routes of an earlier plan are taken away.
+        assert!(calls[2].contains("RouteMetric -eq 7373"), "{calls:?}");
+        assert!(calls[3].contains("Get-NetIPInterface"), "{calls:?}");
     }
 
     #[test]
@@ -229,39 +388,216 @@ mod tests {
                 Ok(String::new())
             }
         };
-        let report = apply_plan(&run, &plan(false));
+        let report = apply_plan(&run, &no_names, &plan(false));
         let RuleOutcome::Failed(why) = &report.offer[0].1 else {
             panic!("{report:?}");
         };
         assert!(why.contains("already exists"), "{why}");
     }
 
+    fn no_names(_: &str) -> Vec<Ipv4Addr> {
+        Vec::new()
+    }
+
+    fn client_plan() -> ExitHostPlan {
+        ExitHostPlan {
+            interface: "tsun0".into(),
+            offer: Vec::new(),
+            client: true,
+            overlay: vec![range()],
+            bypass: vec![Ipv4Addr::new(138, 201, 61, 182)],
+            bypass_hosts: vec!["relay.example".into()],
+        }
+    }
+
     #[test]
-    fn using_an_exit_node_is_refused_and_changes_nothing() {
+    fn the_gateway_is_read_from_the_script_output() {
-        let calls = RefCell::new(0usize);
+        assert_eq!(
-        let run = |script: &str| -> Result<String, String> {
+            parse_gateway("12 192.168.1.1\r\n"),
-            *calls.borrow_mut() += 1;
+            Some((12, Ipv4Addr::new(192, 168, 1, 1)))
-            // No script may touch routes.
+        );
-            assert!(!script.contains("Route"), "{script}");
+        assert_eq!(parse_gateway(""), None);
-            Ok(String::new())
+        assert_eq!(parse_gateway("12"), None);
+        assert_eq!(parse_gateway("12 gw"), None);
+        assert_eq!(parse_gateway("12 1.1.1.1 extra"), None);
+        let script = gateway_script("tsun0");
+        assert!(script.contains("-ne $alias"));
+        assert!(script.contains("-ne '0.0.0.0'"));
+    }
+
+    #[test]
+    fn bypass_routes_are_marked_host_routes_through_the_physical_gateway() {
+        let script = bypass_script(
+            12,
+            Ipv4Addr::new(192, 168, 1, 1),
+            &[Ipv4Addr::new(1, 2, 3, 4), Ipv4Addr::new(5, 6, 7, 8)],
+        );
+        assert!(script.contains("@('1.2.3.4', '5.6.7.8')"), "{script}");
+        assert!(script.contains("-InterfaceIndex 12 -NextHop '192.168.1.1'"));
+        assert!(script.contains("-RouteMetric 7373"));
+        assert!(script.contains("-PolicyStore ActiveStore"));
+        // Stale marked routes go first, and only /32 ones with our metric.
+        assert!(script.find("Remove-NetRoute").unwrap() < script.find("New-NetRoute").unwrap());
+        assert!(script.contains("RouteMetric -eq 7373"));
+        assert!(script.contains("EndsWith('/32')"));
+    }
+
+    #[test]
+    fn the_halves_go_through_the_overlay_and_ipv6_is_blackholed() {
+        let script = halves_script("tsun0");
+        assert!(script.contains("'0.0.0.0/1', '128.0.0.0/1'"));
+        assert!(script.contains("-InterfaceAlias $alias -NextHop '0.0.0.0'"));
+        assert!(script.contains("'::/1', '8000::/1'"));
+        assert!(script.contains("-PolicyStore ActiveStore"));
+        assert!(
+            !script.contains("0.0.0.0/0"),
+            "the default route is left alone"
+        );
+        assert!(halves_script("x'y").contains("'x''y'"));
+    }
+
+    #[test]
+    fn addresses_are_resolved_merged_and_cleaned() {
+        let resolve = |host: &str| -> Vec<Ipv4Addr> {
+            match host {
+                "relay.example" => vec![Ipv4Addr::new(9, 9, 9, 9), Ipv4Addr::new(1, 2, 3, 4)],
+                _ => Vec::new(),
+            }
         };
-        let report = apply_plan(&run, &plan(true));
+        let all = bypass_addresses(
+            &resolve,
+            &[
+                Ipv4Addr::new(1, 2, 3, 4),
+                Ipv4Addr::LOCALHOST,
+                Ipv4Addr::UNSPECIFIED,
+            ],
+            &["relay.example".into(), "gone.example".into()],
+        );
+        assert_eq!(
+            all,
+            vec![Ipv4Addr::new(1, 2, 3, 4), Ipv4Addr::new(9, 9, 9, 9)]
+        );
+    }
+
+    #[test]
+    fn the_bypass_is_written_before_the_halves_that_need_it() {
+        let calls = RefCell::new(Vec::new());
+        let run = |script: &str| -> Result<String, String> {
+            calls.borrow_mut().push(script.to_string());
+            Ok(if script.contains("Select-Object -First 1") {
+                "12 192.168.1.1\n".to_string()
+            } else {
+                String::new()
+            })
+        };
+        let resolve = |_: &str| vec![Ipv4Addr::new(9, 9, 9, 9)];
+        let report = apply_plan(&run, &resolve, &client_plan());
+        assert_eq!(report.client, Some(RuleOutcome::Applied));
+        let calls = calls.borrow();
+        let at = |needle: &str| calls.iter().position(|c| c.contains(needle)).unwrap();
+        assert!(at("Select-Object -First 1") < at("-RouteMetric 7373"));
+        assert!(at("-RouteMetric 7373") < at("'0.0.0.0/1', '128.0.0.0/1'"));
+        let bypass = &calls[at("-RouteMetric 7373")];
+        assert!(
+            bypass.contains("'9.9.9.9'") && bypass.contains("'138.201.61.182'"),
+            "{bypass}"
+        );
+    }
+
+    #[test]
+    fn nothing_goes_through_the_overlay_when_the_bypass_cannot_be_written() {
+        let run = |script: &str| -> Result<String, String> {
+            if script.contains("Select-Object -First 1") {
+                Ok("12 192.168.1.1".to_string())
+            } else if script.contains("-RouteMetric 7373") && script.contains("foreach ($ip") {
+                Err("Access is denied".to_string())
+            } else {
+                assert!(!script.contains("-NextHop '0.0.0.0'"), "{script}");
+                Ok(String::new())
+            }
+        };
+        let report = apply_plan(&run, &no_names, &client_plan());
+        assert_eq!(
+            report.client,
+            Some(RuleOutcome::Failed("Access is denied".into()))
+        );
+    }
+
+    #[test]
+    fn a_missing_gateway_is_a_failure_that_installs_no_route() {
+        let run = |script: &str| -> Result<String, String> {
+            assert!(!script.contains("New-NetRoute"), "{script}");
+            if script.contains("Select-Object -First 1") {
+                Err("the host has no default gateway outside the overlay".into())
+            } else {
+                Ok(String::new())
+            }
+        };
+        let report = apply_plan(&run, &no_names, &client_plan());
         let Some(RuleOutcome::Failed(why)) = report.client else {
             panic!("{report:?}");
         };
-        assert!(why.contains("not available on Windows"), "{why}");
+        assert!(why.contains("no default gateway"), "{why}");
-        assert!(why.contains("loop"), "{why}");
+    }
+
+    #[test]
+    fn a_failed_halves_step_removes_what_was_written() {
+        let calls = RefCell::new(Vec::new());
+        let run = |script: &str| -> Result<String, String> {
+            calls.borrow_mut().push(script.to_string());
+            if script.contains("Select-Object -First 1") {
+                Ok("12 192.168.1.1".to_string())
+            } else if script.contains("'128.0.0.0/1'") && script.contains("New-NetRoute") {
+                Err("boom".to_string())
+            } else {
+                Ok(String::new())
+            }
+        };
+        let report = apply_plan(&run, &no_names, &client_plan());
+        assert!(matches!(report.client, Some(RuleOutcome::Failed(_))));
+        assert!(calls.borrow().last().unwrap().contains("Remove-NetRoute"));
+    }
+
+    #[test]
+    fn dropping_the_client_removes_its_routes_and_clearing_does_too() {
+        let calls = RefCell::new(Vec::new());
+        let run = |script: &str| -> Result<String, String> {
+            calls.borrow_mut().push(script.to_string());
+            Ok(String::new())
+        };
+        let report = apply_plan(&run, &no_names, &plan(false));
+        assert_eq!(report.client, None);
+        assert!(
+            calls
+                .borrow()
+                .iter()
+                .any(|c| c.contains("RouteMetric -eq 7373"))
+        );
+        let script = remove_client_script("tsun0");
+        assert!(script.contains("'0.0.0.0/1', '128.0.0.0/1', '::/1', '8000::/1'"));
+        assert!(script.contains("exit 0"));
+        // Only routes on our interface, or marked ours, are touched.
+        assert!(script.contains("-InterfaceAlias $alias"));
+        calls.borrow_mut().clear();
+        clear_all(&run, "tsun0");
+        assert!(
+            calls
+                .borrow()
+                .iter()
+                .any(|c| c.contains("RouteMetric -eq 7373"))
+        );
     }
 
     #[test]
     fn clearing_prunes_everything_for_the_interface() {
-        let seen = RefCell::new(String::new());
+        let seen = RefCell::new(Vec::<String>::new());
         let run = |script: &str| -> Result<String, String> {
-            *seen.borrow_mut() = script.to_string();
+            seen.borrow_mut().push(script.to_string());
             Ok(String::new())
         };
         clear_all(&run, "tsun0");
-        assert!(seen.borrow().contains("$keep = @()"));
+        let seen = seen.borrow();
-        assert!(seen.borrow().contains("'tsunagi-exit:tsun0:'"));
+        assert!(seen[0].contains("$keep = @()"));
+        assert!(seen[0].contains("'tsunagi-exit:tsun0:'"));
     }
 }

crates/tsunagi-core/src/overlay/interface.rs

@@ -164,6 +164,17 @@ pub struct Interface {
     rules_report: std::sync::Mutex<Option<BroadcastRulesReport>>,
     /// What the exit-node rules were last asked for, and what came of it.
     exit_state: std::sync::Mutex<ExitState>,
+    /// Where each network's runtime reaches its peers and relays outside the
+    /// overlay; read only by hosts that cannot exempt the agent from the
+    /// exit node's default route.
+    underlay: std::sync::Mutex<std::collections::HashMap<NetworkId, Underlay>>,
+}
+
+/// Underlay destinations one network's runtime currently depends on.
+#[derive(Debug, Default, Clone, PartialEq, Eq)]
+struct Underlay {
+    addresses: std::collections::BTreeSet<std::net::Ipv4Addr>,
+    hosts: std::collections::BTreeSet<String>,
 }
 
 /// The exit-node host rules as this interface last left them.
@@ -275,6 +286,7 @@ impl Interface {
             host_rules,
             rules_report: std::sync::Mutex::new(None),
             exit_state: std::sync::Mutex::new(ExitState::default()),
+            underlay: std::sync::Mutex::new(std::collections::HashMap::new()),
         })
     }
 
@@ -350,11 +362,20 @@ impl Interface {
             .collect();
         ranges.sort_by_key(|range| (range.base, range.prefix_len));
         ranges.dedup();
+        let client = self.routes.exit_via().is_some();
+        let overlay = self.routes.overlay_ranges();
+        let (bypass, bypass_hosts) = if client && exit.needs_bypass() {
+            self.underlay_targets(&overlay)
+        } else {
+            (Vec::new(), Vec::new())
+        };
         let wanted = ExitHostPlan {
             interface: self.name.clone(),
             offer: ranges,
-            client: self.routes.exit_via().is_some(),
+            client,
-            overlay: self.routes.overlay_ranges(),
+            overlay,
+            bypass,
+            bypass_hosts,
         };
 
         let first = !std::mem::replace(&mut self.exit_state().cleaned, true);
@@ -426,6 +447,68 @@ impl Interface {
         state.report = Some(report);
     }
 
+    /// Tells the interface where a network's runtime reaches peers and
+    /// relays outside the overlay. When the host needs those kept out of the
+    /// exit node's default route and the set changed, the rules follow.
+    pub async fn set_underlay(
+        &self,
+        network: NetworkId,
+        addresses: impl IntoIterator<Item = std::net::Ipv4Addr>,
+        hosts: impl IntoIterator<Item = String>,
+    ) {
+        let new = Underlay {
+            addresses: addresses.into_iter().collect(),
+            hosts: hosts.into_iter().collect(),
+        };
+        let changed = {
+            let mut map = match self.underlay.lock() {
+                Ok(guard) => guard,
+                Err(poisoned) => poisoned.into_inner(),
+            };
+            if new == Underlay::default() {
+                map.remove(&network).is_some()
+            } else {
+                map.insert(network, new.clone()) != Some(new)
+            }
+        };
+        let needed = changed
+            && self.routes.exit_via().is_some()
+            && self
+                .host_rules
+                .as_ref()
+                .and_then(|rules| rules.exit_rules())
+                .is_some_and(|exit| exit.needs_bypass());
+        if needed {
+            self.sync_exit_rules().await;
+        }
+    }
+
+    /// Every underlay destination worth keeping direct, minus anything that
+    /// is the overlay itself, loopback or link-local.
+    fn underlay_targets(
+        &self,
+        overlay: &[crate::state::Ipv4Range],
+    ) -> (Vec<std::net::Ipv4Addr>, Vec<String>) {
+        let map = match self.underlay.lock() {
+            Ok(guard) => guard,
+            Err(poisoned) => poisoned.into_inner(),
+        };
+        let mut addresses = std::collections::BTreeSet::new();
+        let mut hosts = std::collections::BTreeSet::new();
+        for underlay in map.values() {
+            addresses.extend(underlay.addresses.iter().copied().filter(|address| {
+                !address.is_loopback()
+                    && !address.is_link_local()
+                    && !address.is_multicast()
+                    && !address.is_unspecified()
+                    && !address.is_broadcast()
+                    && !overlay.iter().any(|range| range.contains(*address))
+            }));
+            hosts.extend(underlay.hosts.iter().cloned());
+        }
+        (addresses.into_iter().collect(), hosts.into_iter().collect())
+    }
+
     fn exit_state(&self) -> std::sync::MutexGuard<'_, ExitState> {
         match self.exit_state.lock() {
             Ok(guard) => guard,
@@ -1343,6 +1426,65 @@ mod tests {
         assert_eq!(interface.counters().received, 1);
     }
 
+    #[tokio::test]
+    async fn the_underlay_reaches_the_rules_only_for_hosts_that_need_it_and_only_while_using_one() {
+        let rules = MockHostRules::new();
+        let (interface, routes, id) = managed_interface(&rules, false).await;
+        let relay = || vec!["relay.example".to_string()];
+        let direct = || {
+            vec![
+                Ipv4Addr::new(138, 201, 61, 182),
+                Ipv4Addr::new(10, 13, 37, 9), // the overlay itself
+                Ipv4Addr::LOCALHOST,
+                Ipv4Addr::new(169, 254, 1, 1),
+            ]
+        };
+
+        // A host that does not ask never gets them, so Linux and macOS do not
+        // re-apply their rules whenever a path moves.
+        routes.set_exit(id, using(peer(2)));
+        interface.set_underlay(id, direct(), relay()).await;
+        interface.sync_exit_rules().await;
+        let plan = rules.exit().installed().unwrap();
+        assert!(plan.client);
+        assert!(plan.bypass.is_empty() && plan.bypass_hosts.is_empty());
+
+        // A host that does gets the usable ones, and follows changes.
+        rules.exit().set_needs_bypass(true);
+        interface.set_underlay(id, direct(), relay()).await;
+        interface.sync_exit_rules().await;
+        let plan = rules.exit().installed().unwrap();
+        assert_eq!(plan.bypass, vec![Ipv4Addr::new(138, 201, 61, 182)]);
+        assert_eq!(plan.bypass_hosts, relay());
+        let applies = || {
+            rules
+                .exit()
+                .calls()
+                .iter()
+                .filter(|c| c.starts_with("apply"))
+                .count()
+        };
+        let before = applies();
+        interface.set_underlay(id, direct(), relay()).await;
+        assert_eq!(applies(), before, "an unchanged set is not applied again");
+        interface
+            .set_underlay(id, vec![Ipv4Addr::new(5, 6, 7, 8)], relay())
+            .await;
+        assert_eq!(applies(), before + 1);
+        assert_eq!(
+            rules.exit().installed().unwrap().bypass,
+            vec![Ipv4Addr::new(5, 6, 7, 8)]
+        );
+
+        // Not using an exit node: nothing to keep direct.
+        routes.set_exit(id, Default::default());
+        interface.sync_exit_rules().await;
+        interface
+            .set_underlay(id, vec![Ipv4Addr::new(1, 1, 1, 1)], relay())
+            .await;
+        assert!(rules.exit().installed().is_none());
+    }
+
     #[tokio::test]
     async fn the_exit_rules_follow_the_routing_table_and_leave_with_the_interface() {
         let rules = MockHostRules::new();

crates/tsunagi/Cargo.toml

@@ -34,6 +34,9 @@ netwatch = "0.19.3"
 [target.'cfg(unix)'.dependencies]
 nix = { version = "0.31", features = ["user"] }
 
+[target."cfg(windows)".dependencies]
+windows-service = "0.8"
+
 [dev-dependencies]
 tokio = { version = "1.53", features = ["rt", "rt-multi-thread", "sync", "time", "macros", "process"] }
 tempfile.workspace = true

crates/tsunagi/src/main.rs

@@ -11,6 +11,9 @@ use std::path::PathBuf;
 use std::sync::Arc;
 use std::time::Duration;
 
+#[cfg(windows)]
+mod winservice;
+
 use clap::{Args, Parser, Subcommand, ValueEnum};
 use tsunagi::agent::Event;
 use tsunagi::config::{AgentConfig, StoragePaths, TransportPolicy};
@@ -84,6 +87,10 @@ enum Command {
     ExitNode(ExitNodeArgs),
     /// Removes everything this device has stored and starts over.
     Wipe(WipeArgs),
+    /// Runs the agent under the Windows service manager. Registered by the
+    /// installer, not typed.
+    #[command(hide = true)]
+    Service(Box<UpArgs>),
 }
 
 #[derive(Debug, Args)]
@@ -757,12 +764,26 @@ fn main() -> std::process::ExitCode {
 
     let is_debug = log_filter.contains("debug") || log_filter.contains("trace");
 
-    tracing_subscriber::fmt()
+    let builder = tracing_subscriber::fmt()
         .with_env_filter(tracing_subscriber::EnvFilter::new(&log_filter))
         .with_target(is_debug)
-        .compact()
+        .compact();
-        .with_writer(std::io::stderr)
+    // A service has no console: it writes to a file.
-        .init();
+    match matches!(cli.command, Command::Service(_))
+        .then(open_service_log)
+        .flatten()
+    {
+        Some(file) => builder
+            .with_ansi(false)
+            .with_writer(std::sync::Mutex::new(file))
+            .init(),
+        None => builder.with_writer(std::io::stderr).init(),
+    }
+
+    #[cfg(windows)]
+    if let Command::Service(args) = cli.command {
+        return run_service(*args);
+    }
 
     // The library never starts a runtime; this binary owns it.
     let runtime = match tokio::runtime::Builder::new_multi_thread()
@@ -785,6 +806,89 @@ fn main() -> std::process::ExitCode {
     }
 }
 
+/// The log file of the Windows service, `%ProgramData%\Tsunagi\logs\tsng.log`.
+///
+/// The previous one is kept as `tsng.log.old` once it passes a few megabytes,
+/// so a service that logs for months does not fill the disk. `None` anywhere
+/// else, and when the file cannot be opened.
+fn open_service_log() -> Option<std::fs::File> {
+    if !cfg!(windows) {
+        return None;
+    }
+    let root = std::env::var_os("ProgramData")?;
+    open_log_in(&PathBuf::from(root).join("Tsunagi").join("logs"))
+}
+
+/// Opens `tsng.log` in a directory for appending, moving a large one aside.
+fn open_log_in(dir: &std::path::Path) -> Option<std::fs::File> {
+    std::fs::create_dir_all(dir).ok()?;
+    let log = dir.join("tsng.log");
+    if std::fs::metadata(&log).is_ok_and(|meta| meta.len() > SERVICE_LOG_LIMIT) {
+        let _ = std::fs::rename(&log, dir.join("tsng.log.old"));
+    }
+    std::fs::OpenOptions::new()
+        .create(true)
+        .append(true)
+        .open(log)
+        .ok()
+}
+
+/// Past this size the service log is moved aside, keeping one older file.
+const SERVICE_LOG_LIMIT: u64 = 8 * 1024 * 1024;
+
+/// The directories and control socket of the Windows service, unless the
+/// service command line names its own.
+///
+/// A service runs as the system account, whose profile is the wrong home for
+/// a device's identity, so the state is kept under `%ProgramData%\Tsunagi`,
+/// where the installer locks it to administrators and the system. The control
+/// socket stays at the system-wide default every client already looks at,
+/// which is what lets a tray and an unelevated `tsng status` find the service.
+#[cfg_attr(not(windows), allow(dead_code))]
+fn service_defaults(mut args: UpArgs, program_data: &std::path::Path) -> UpArgs {
+    let root = program_data.join("Tsunagi");
+    args.paths
+        .state_dir
+        .get_or_insert_with(|| root.join("data"));
+    args.paths
+        .cache_dir
+        .get_or_insert_with(|| root.join("cache"));
+    args.control_socket
+        .get_or_insert_with(tsunagi::ipc::default_control_socket_path);
+    args
+}
+
+/// Hands the process to the Windows service manager, with `up` as the body.
+#[cfg(windows)]
+fn run_service(args: UpArgs) -> std::process::ExitCode {
+    let program_data = std::env::var_os("ProgramData")
+        .map(PathBuf::from)
+        .filter(|path| path.is_absolute())
+        .unwrap_or_else(|| PathBuf::from(r"C:\ProgramData"));
+    let args = service_defaults(args, &program_data);
+    let args = std::sync::Mutex::new(Some(args));
+    let body = move || -> Result<(), String> {
+        let Some(args) = args.lock().ok().and_then(|mut slot| slot.take()) else {
+            return Err("the service was started twice".to_string());
+        };
+        let runtime = tokio::runtime::Builder::new_multi_thread()
+            .enable_all()
+            .build()
+            .map_err(|err| format!("cannot start the async runtime: {err}"))?;
+        runtime.block_on(up(args)).map_err(|err| err.to_string())
+    };
+    match winservice::run(Box::new(body)) {
+        Ok(()) => std::process::ExitCode::SUCCESS,
+        Err(err) => {
+            eprintln!(
+                "error: cannot run as a service ({err}); `tsng service` is started by the \
+                 Windows service manager, use `tsng up` to run the agent yourself"
+            );
+            std::process::ExitCode::FAILURE
+        }
+    }
+}
+
 async fn run(command: Command) -> Result<(), Box<dyn std::error::Error>> {
     match command {
         Command::Id(args) => id(args).await,
@@ -796,6 +900,11 @@ async fn run(command: Command) -> Result<(), Box<dyn std::error::Error>> {
         Command::Dns(args) => dns_command(args).await,
         Command::ExitNode(args) => exit_node_command(args).await,
         Command::Wipe(args) => wipe(args).await,
+        // Taken in `main`, before there is a runtime: the service manager
+        // wants the process's thread, not an async task.
+        Command::Service(_) => Err("`tsng service` is started by the Windows service \
+                                    manager; use `tsng up` to run the agent yourself"
+            .into()),
     }
 }
 
@@ -4900,7 +5009,14 @@ async fn stop_signal() -> &'static str {
             _ = terminate.recv() => "terminated",
         }
     }
-    #[cfg(not(unix))]
+    #[cfg(windows)]
+    {
+        tokio::select! {
+            _ = tokio::signal::ctrl_c() => "interrupted",
+            _ = winservice::stop_requested() => "stopped by the service manager",
+        }
+    }
+    #[cfg(not(any(unix, windows)))]
     {
         let _ = tokio::signal::ctrl_c().await;
         "interrupted"
@@ -5943,3 +6059,93 @@ mod network_context_tests {
         assert_eq!(standing, NetworkStanding::Stopped);
     }
 }
+
+#[cfg(test)]
+mod service_log_tests {
+    #![allow(clippy::unwrap_used, clippy::expect_used, clippy::panic)]
+
+    use std::io::Write;
+
+    use super::*;
+
+    #[test]
+    fn the_log_appends_and_a_large_one_is_moved_aside() {
+        let dir = tempfile::tempdir().unwrap();
+        let logs = dir.path().join("logs");
+        let mut file = open_log_in(&logs).unwrap();
+        writeln!(file, "one").unwrap();
+        drop(file);
+        let mut file = open_log_in(&logs).unwrap();
+        writeln!(file, "two").unwrap();
+        drop(file);
+        assert_eq!(
+            std::fs::read_to_string(logs.join("tsng.log")).unwrap(),
+            "one\ntwo\n"
+        );
+
+        let big = std::fs::File::options()
+            .append(true)
+            .open(logs.join("tsng.log"))
+            .unwrap();
+        big.set_len(SERVICE_LOG_LIMIT + 1).unwrap();
+        drop(big);
+        let file = open_log_in(&logs).unwrap();
+        assert_eq!(file.metadata().unwrap().len(), 0);
+        assert!(logs.join("tsng.log.old").exists());
+    }
+
+    #[test]
+    fn a_log_that_cannot_be_opened_is_none() {
+        let dir = tempfile::tempdir().unwrap();
+        let file = dir.path().join("file");
+        std::fs::write(&file, b"x").unwrap();
+        // A directory cannot be made under a regular file.
+        assert!(open_log_in(&file.join("logs")).is_none());
+    }
+
+    #[test]
+    fn the_service_keeps_its_state_under_program_data_unless_told_otherwise() {
+        use clap::Parser;
+        let parse = |extra: &[&str]| {
+            let mut words = vec!["tsng", "service"];
+            words.extend_from_slice(extra);
+            match Cli::try_parse_from(words).unwrap().command {
+                Command::Service(args) => *args,
+                other => panic!("{other:?}"),
+            }
+        };
+        let root = std::path::Path::new("/pd");
+        let args = service_defaults(parse(&[]), root);
+        assert_eq!(
+            args.paths.state_dir,
+            Some(root.join("Tsunagi").join("data"))
+        );
+        assert_eq!(
+            args.paths.cache_dir,
+            Some(root.join("Tsunagi").join("cache"))
+        );
+        assert_eq!(
+            args.control_socket,
+            Some(tsunagi::ipc::default_control_socket_path())
+        );
+
+        let args = service_defaults(
+            parse(&["--state-dir", "/mine", "--control-socket", "/mine/sock"]),
+            root,
+        );
+        assert_eq!(args.paths.state_dir, Some(PathBuf::from("/mine")));
+        assert_eq!(args.control_socket, Some(PathBuf::from("/mine/sock")));
+    }
+
+    #[test]
+    fn the_service_command_is_refused_off_the_service_manager() {
+        use clap::Parser;
+        let cli = Cli::try_parse_from(["tsng", "service"]).unwrap();
+        assert!(matches!(cli.command, Command::Service(_)));
+        let runtime = tokio::runtime::Builder::new_current_thread()
+            .build()
+            .unwrap();
+        let err = runtime.block_on(run(cli.command)).unwrap_err();
+        assert!(err.to_string().contains("service manager"), "{err}");
+    }
+}

crates/tsunagi/src/winservice.rs

@@ -0,0 +1,121 @@
+//! The agent as a Windows service.
+//!
+//! `tsng service` is what the installer registers with the service control
+//! manager (`sc create`); nobody types it. The manager starts the process,
+//! which must call into the dispatcher quickly and then report its state, so
+//! this is the only place the Windows service API is touched. Everything the
+//! agent does is the same `up` that runs in a console; the only differences
+//! are that it stops when the manager says so rather than on Ctrl-C, and that
+//! it logs to a file because a service has no console.
+//!
+//! The one `unsafe` here is not ours: `define_windows_service!` expands to the
+//! `extern "system"` entry point the manager calls, which has to be unsafe.
+
+#![allow(unsafe_code)]
+
+use std::ffi::OsString;
+use std::sync::OnceLock;
+use std::time::Duration;
+
+use windows_service::service::{
+    ServiceControl, ServiceControlAccept, ServiceExitCode, ServiceState, ServiceStatus, ServiceType,
+};
+use windows_service::service_control_handler::{
+    self, ServiceControlHandlerResult, ServiceStatusHandle,
+};
+use windows_service::{define_windows_service, service_dispatcher};
+
+/// The name the installer registers the service under.
+pub const SERVICE_NAME: &str = "Tsunagi";
+
+/// How long the manager waits for a stop before it gives up on us. The agent's
+/// own shutdown is bounded and shorter.
+const STOP_WAIT: Duration = Duration::from_secs(25);
+
+/// Set when the manager asks the service to stop.
+static STOP: tokio::sync::Notify = tokio::sync::Notify::const_new();
+
+/// What `up` does, handed over by `main` because the manager's entry point
+/// cannot take arguments.
+type Body = Box<dyn Fn() -> Result<(), String> + Send + Sync>;
+static BODY: OnceLock<Body> = OnceLock::new();
+
+/// Where the manager's status handle is kept, so the stop handler can report.
+static STATUS: OnceLock<ServiceStatusHandle> = OnceLock::new();
+
+/// Resolves when the manager has asked the service to stop.
+pub async fn stop_requested() {
+    STOP.notified().await;
+}
+
+define_windows_service!(ffi_service_main, service_main);
+
+/// Hands the process to the service control manager and returns when the
+/// service has stopped. Fails at once when not started by the manager, which
+/// is how running `tsng service` in a console is reported.
+pub fn run(body: Body) -> Result<(), windows_service::Error> {
+    let _ = BODY.set(body);
+    service_dispatcher::start(SERVICE_NAME, ffi_service_main)
+}
+
+fn service_main(_arguments: Vec<OsString>) {
+    let result = serve();
+    let code = match &result {
+        Ok(()) => ServiceExitCode::Win32(0),
+        Err(err) => {
+            tracing::error!("the service failed: {err}");
+            // A non-zero exit lets the recovery actions restart it.
+            ServiceExitCode::ServiceSpecific(1)
+        }
+    };
+    if let Some(handle) = STATUS.get() {
+        let _ = handle.set_service_status(status(ServiceState::Stopped, code, Duration::ZERO));
+    }
+}
+
+fn serve() -> Result<(), String> {
+    let handle = service_control_handler::register(SERVICE_NAME, |control| match control {
+        ServiceControl::Stop | ServiceControl::Shutdown => {
+            if let Some(handle) = STATUS.get() {
+                let _ = handle.set_service_status(status(
+                    ServiceState::StopPending,
+                    ServiceExitCode::Win32(0),
+                    STOP_WAIT,
+                ));
+            }
+            STOP.notify_one();
+            ServiceControlHandlerResult::NoError
+        }
+        ServiceControl::Interrogate => ServiceControlHandlerResult::NoError,
+        _ => ServiceControlHandlerResult::NotImplemented,
+    })
+    .map_err(|err| format!("cannot register with the service control manager: {err}"))?;
+    let _ = STATUS.set(handle);
+    handle
+        .set_service_status(status(
+            ServiceState::Running,
+            ServiceExitCode::Win32(0),
+            Duration::ZERO,
+        ))
+        .map_err(|err| format!("cannot report the service as running: {err}"))?;
+    match BODY.get() {
+        Some(body) => body(),
+        None => Err("the service has nothing to run".to_string()),
+    }
+}
+
+fn status(state: ServiceState, exit_code: ServiceExitCode, wait_hint: Duration) -> ServiceStatus {
+    ServiceStatus {
+        service_type: ServiceType::OWN_PROCESS,
+        current_state: state,
+        controls_accepted: if state == ServiceState::Running {
+            ServiceControlAccept::STOP | ServiceControlAccept::SHUTDOWN
+        } else {
+            ServiceControlAccept::empty()
+        },
+        exit_code,
+        checkpoint: 0,
+        wait_hint,
+        process_id: None,
+    }
+}

dist/windows/README.md

@@ -1,5 +1,34 @@
 # Running tsunagi on Windows
 
+## The installer (recommended)
+
+`tsunagi-setup-<version>-x86_64.exe` does everything in one go. Run it as an
+administrator (it asks):
+
+- installs `tsng.exe`, the tray app and **Wintun** to `C:\Program Files\Tsunagi`;
+- registers and starts the **Tsunagi** Windows service (it starts at boot and
+  restarts after a crash), running as the system account;
+- adds `tsng` to the PATH and an inbound Windows Firewall rule for the agent
+  (both are tasks you can untick);
+- starts the tray icon at sign-in.
+
+Where things live: the device's identity and networks in
+`%ProgramData%\Tsunagi\data` (readable only by administrators and the system),
+the log in `%ProgramData%\Tsunagi\logs\tsng.log`. The service keeps its own
+state: networks joined by an agent you started by hand (`tsng up`, which keeps
+its data in your user profile) are not carried over; join them again with
+`tsng join`. The installer stops such an agent, because it cannot run beside the
+service.
+
+`tsng status`, `tsng join` and the tray work from an ordinary, non-elevated
+terminal. Manage the service with the Services app or `sc stop Tsunagi` /
+`sc start Tsunagi`. Silent install: `/VERYSILENT /SUPPRESSMSGBOXES`; choose
+tasks with `/TASKS="!autostart,addtopath,firewall"`. Uninstalling removes the
+service, the firewall rule and the PATH entry, and asks before deleting the
+data.
+
+## By hand (the zip)
+
 This archive holds `tsng.exe`. The overlay interface is a Wintun adapter, so
 the agent needs two things to carry real traffic:
 

dist/windows/tsunagi.iss

@@ -0,0 +1,242 @@
+; Inno Setup script for the Tsunagi Windows installer.
+;
+; One installer that leaves a working machine:
+;   * the agent (tsng.exe), as a Windows service that starts at boot
+;   * Wintun, the driver library behind the overlay interface
+;   * the tray application, started at sign-in
+;   * `tsng` on the PATH, and an inbound Defender Firewall rule for the agent
+;
+; Built in CI (see .github/workflows/release.yml); locally:
+;   ISCC.exe /DAppVersion=0.1.0 /DSourceDir=<dir with tsng.exe and tsunagi-tray.exe>
+;            /DWintunDir=<wintun\bin\amd64> /O<output dir> dist\windows\tsunagi.iss
+; Silent install: tsunagi-setup-<version>-x86_64.exe /VERYSILENT /SUPPRESSMSGBOXES
+; Pick tasks with /TASKS="!autostart,addtopath,firewall".
+
+#ifndef AppVersion
+  #define AppVersion "0.0.0"
+#endif
+#ifndef SourceDir
+  #error SourceDir must point at the directory holding tsng.exe and tsunagi-tray.exe
+#endif
+#ifndef WintunDir
+  #error WintunDir must point at the directory holding wintun.dll
+#endif
+
+#define ServiceName "Tsunagi"
+
+[Setup]
+AppId={{F635E5DD-250D-499F-B0B3-AFC7CC64A15D}
+AppName=Tsunagi
+AppVersion={#AppVersion}
+AppPublisher=Tsunagi
+DefaultDirName={autopf}\Tsunagi
+DefaultGroupName=Tsunagi
+DisableProgramGroupPage=yes
+PrivilegesRequired=admin
+ArchitecturesAllowed=x64compatible
+ArchitecturesInstallIn64BitMode=x64compatible
+OutputBaseFilename=tsunagi-setup-{#AppVersion}-x86_64
+Compression=lzma2
+SolidCompression=yes
+WizardStyle=modern
+ChangesEnvironment=yes
+UninstallDisplayName=Tsunagi
+; The service and the tray are stopped by hand below, so the installer does
+; not need to ask Windows to close applications.
+CloseApplications=no
+
+[Tasks]
+Name: "autostart"; Description: "Start the tray icon when I sign in"; GroupDescription: "Tsunagi:"
+Name: "addtopath"; Description: "Add tsng to the PATH"; GroupDescription: "Tsunagi:"
+Name: "firewall"; Description: "Allow the agent through Windows Firewall (direct connections between devices)"; GroupDescription: "Tsunagi:"
+
+[Files]
+Source: "{#SourceDir}\tsng.exe"; DestDir: "{app}"; Flags: ignoreversion
+Source: "{#SourceDir}\tsunagi-tray.exe"; DestDir: "{app}"; Flags: ignoreversion
+; Wintun is the unmodified, signed library from wintun.net; its licence allows
+; shipping it with an application and asks for the licence text beside it.
+; In the archive the text sits at its root, two levels above bin\amd64.
+Source: "{#WintunDir}\wintun.dll"; DestDir: "{app}"; Flags: ignoreversion
+Source: "{#WintunDir}\..\..\LICENSE.txt"; DestDir: "{app}\licenses"; DestName: "wintun-LICENSE.txt"; Flags: ignoreversion
+
+[Dirs]
+; Identity and network secrets live here; locked down below.
+Name: "{commonappdata}\Tsunagi\data"
+Name: "{commonappdata}\Tsunagi\cache"
+Name: "{commonappdata}\Tsunagi\logs"
+
+[Icons]
+Name: "{group}\Tsunagi"; Filename: "{app}\tsunagi-tray.exe"
+Name: "{commonstartup}\Tsunagi"; Filename: "{app}\tsunagi-tray.exe"; Tasks: autostart
+
+[Registry]
+Root: HKLM; Subkey: "SYSTEM\CurrentControlSet\Control\Session Manager\Environment"; \
+    ValueType: expandsz; ValueName: "Path"; ValueData: "{olddata};{app}"; \
+    Check: NeedsAddPath('{app}'); Tasks: addtopath
+
+[Run]
+; Only the system account and administrators may read the state: it holds the
+; device's key and every network's secret. SIDs, so it works in any language.
+Filename: "{sys}\icacls.exe"; \
+    Parameters: """{commonappdata}\Tsunagi\data"" /inheritance:r /grant:r *S-1-5-18:(OI)(CI)F *S-1-5-32-544:(OI)(CI)F"; \
+    Flags: runhidden; StatusMsg: "Protecting the agent's data..."
+Filename: "{sys}\icacls.exe"; \
+    Parameters: """{commonappdata}\Tsunagi\cache"" /inheritance:r /grant:r *S-1-5-18:(OI)(CI)F *S-1-5-32-544:(OI)(CI)F"; \
+    Flags: runhidden
+Filename: "{sys}\icacls.exe"; \
+    Parameters: """{commonappdata}\Tsunagi\logs"" /inheritance:r /grant:r *S-1-5-18:(OI)(CI)F *S-1-5-32-544:(OI)(CI)F"; \
+    Flags: runhidden
+; The tray runs as the person who installed, not elevated, like it does at sign-in.
+Filename: "{app}\tsunagi-tray.exe"; Description: "Start the Tsunagi tray"; \
+    Flags: nowait postinstall skipifsilent runasoriginaluser
+
+[UninstallRun]
+Filename: "{sys}\netsh.exe"; Parameters: "advfirewall firewall delete rule name=""Tsunagi"""; \
+    Flags: runhidden; RunOnceId: "RemoveFirewallRule"
+
+[Code]
+const
+  ServiceName = '{#ServiceName}';
+  EnvKey = 'SYSTEM\CurrentControlSet\Control\Session Manager\Environment';
+
+function Sc(const Params: String): Integer;
+var
+  Code: Integer;
+begin
+  if Exec(ExpandConstant('{sys}\sc.exe'), Params, '', SW_HIDE, ewWaitUntilTerminated, Code) then
+    Result := Code
+  else
+    Result := -1;
+end;
+
+function ServiceExists: Boolean;
+begin
+  // 1060 is ERROR_SERVICE_DOES_NOT_EXIST.
+  Result := Sc('query ' + ServiceName) <> 1060;
+end;
+
+procedure StopService;
+var
+  Code: Integer;
+begin
+  // `net stop` waits until the service has stopped, or says it was not running.
+  Exec(ExpandConstant('{sys}\net.exe'), 'stop ' + ServiceName, '', SW_HIDE, ewWaitUntilTerminated, Code);
+end;
+
+procedure StopTray;
+var
+  Code: Integer;
+begin
+  Exec(ExpandConstant('{sys}\taskkill.exe'), '/F /IM tsunagi-tray.exe', '', SW_HIDE, ewWaitUntilTerminated, Code);
+end;
+
+// An agent started by hand (`tsng up`) holds the control pipe and the Wintun
+// adapter, and the file about to be replaced: it cannot run beside the service.
+procedure StopStrayAgents;
+var
+  Code: Integer;
+begin
+  Exec(ExpandConstant('{sys}\taskkill.exe'), '/F /IM tsng.exe', '', SW_HIDE, ewWaitUntilTerminated, Code);
+end;
+
+function NeedsAddPath(Param: String): Boolean;
+var
+  Current: String;
+begin
+  if not RegQueryStringValue(HKLM, EnvKey, 'Path', Current) then
+  begin
+    Result := True;
+    exit;
+  end;
+  Result := Pos(';' + Uppercase(ExpandConstant(Param)) + ';', ';' + Uppercase(Current) + ';') = 0;
+end;
+
+procedure RemoveFromPath;
+var
+  Current, Entry: String;
+  At: Integer;
+begin
+  if not RegQueryStringValue(HKLM, EnvKey, 'Path', Current) then
+    exit;
+  Entry := ';' + ExpandConstant('{app}');
+  At := Pos(Uppercase(Entry), Uppercase(Current));
+  if At > 0 then
+  begin
+    Delete(Current, At, Length(Entry));
+    RegWriteExpandStringValue(HKLM, EnvKey, 'Path', Current);
+  end;
+end;
+
+// Stops what is about to be replaced: the files are in use while it runs.
+function PrepareToInstall(var NeedsRestart: Boolean): String;
+begin
+  StopTray;
+  if ServiceExists then
+    StopService;
+  StopStrayAgents;
+  Result := '';
+end;
+
+procedure InstallService;
+var
+  App: String;
+begin
+  App := ExpandConstant('{app}\tsng.exe');
+  // The command line is `tsng service`: it keeps its state under
+  // %ProgramData%\Tsunagi and serves the system-wide control socket.
+  if ServiceExists then
+    Sc('config ' + ServiceName + ' binPath= "\"' + App + '\" service" start= auto obj= LocalSystem')
+  else
+    Sc('create ' + ServiceName + ' binPath= "\"' + App + '\" service" start= auto obj= LocalSystem DisplayName= "Tsunagi agent"');
+  Sc('description ' + ServiceName + ' "Tsunagi: private mesh network agent"');
+  // Restart after a crash, backing off, and forget the failures after a day.
+  Sc('failure ' + ServiceName + ' reset= 86400 actions= restart/5000/restart/5000/restart/30000');
+  Sc('start ' + ServiceName);
+end;
+
+procedure InstallFirewallRule;
+var
+  Code: Integer;
+  Netsh: String;
+begin
+  Netsh := ExpandConstant('{sys}\netsh.exe');
+  Exec(Netsh, 'advfirewall firewall delete rule name="Tsunagi"', '', SW_HIDE, ewWaitUntilTerminated, Code);
+  if WizardIsTaskSelected('firewall') then
+    Exec(Netsh,
+      'advfirewall firewall add rule name="Tsunagi" dir=in action=allow enable=yes profile=any program="' +
+      ExpandConstant('{app}\tsng.exe') + '"',
+      '', SW_HIDE, ewWaitUntilTerminated, Code);
+end;
+
+procedure CurStepChanged(CurStep: TSetupStep);
+begin
+  if CurStep = ssPostInstall then
+  begin
+    InstallFirewallRule;
+    InstallService;
+  end;
+end;
+
+procedure CurUninstallStepChanged(CurUninstallStep: TUninstallStep);
+begin
+  if CurUninstallStep = usUninstall then
+  begin
+    StopTray;
+    if ServiceExists then
+    begin
+      StopService;
+      Sc('delete ' + ServiceName);
+    end;
+    StopStrayAgents;
+    RemoveFromPath;
+  end;
+  if CurUninstallStep = usPostUninstall then
+  begin
+    // The data is the device's identity and its networks: kept unless asked,
+    // and never removed by a silent uninstall.
+    if (not UninstallSilent) and
+       (MsgBox('Also delete this device''s identity, networks and logs?' + #13#10 +
+               'Choose No to keep them for a later reinstall.', mbConfirmation, MB_YESNO or MB_DEFBUTTON2) = IDYES) then
+      DelTree(ExpandConstant('{commonappdata}\Tsunagi'), True, True, True);
+  end;
+end;

docs/exit-node.md

@@ -82,18 +82,25 @@ is built from `pf` and `route`. The agent runs as root there.
 
 ## Windows
 
-Offering only. Each offered range gets a WinNAT object
+- **Offering:** each offered range gets a WinNAT object
-`tsunagi-exit:<interface>:<range>` (`New-NetNat`), removed with the offer, the
+  `tsunagi-exit:<interface>:<range>` (`New-NetNat`), removed with the offer, the
-network and the agent. Forwarding is read for the overlay interface and never
+  network and the agent. Forwarding is read for the overlay interface and never
-changed; Windows forwards per interface, so enable it on the ones involved
+  changed; Windows forwards per interface, so enable it on the ones involved
-(`Get-NetIPInterface | Set-NetIPInterface -Forwarding Enabled`, administrator
+  (`Get-NetIPInterface | Set-NetIPInterface -Forwarding Enabled`, administrator
-PowerShell). Some Windows versions allow one NAT per host; a second range then
+  PowerShell). Some Windows versions allow one NAT per host; a second range then
-reports its own failure.
+  reports its own failure.
-
+- **Using:** Windows has no per-user routing, so the agent's own traffic is kept
-Using one is **not available**: Windows has no per-user routing, so the agent's
+  off the tunnel *by destination*. `0.0.0.0/1` and `128.0.0.0/1` go through the
-own traffic cannot be kept out of a default route through the overlay and would
+  overlay interface; a `/32` route with metric 7373 (the marker for "ours")
-loop. `tsng exit-node <name>` reports that and installs nothing. This was
+  through the physical default gateway is added for every address the agent
-written without a Windows host to run it on.
+  talks to directly: verified direct paths of its connections, plus the relays
+  it uses, resolved when the rules are applied. The set follows path changes. A
+  path that appears after the last apply goes through the tunnel until the next
+  one. IPv6 is blackholed with `::/1` and `8000::/1` on the overlay interface,
+  so applications fall back to IPv4 after a delay rather than at once.
+  Everything is in the active store: a reboot clears it.
+- Not verified on a live host when this was written. To inspect:
+  `Get-NetRoute -PolicyStore ActiveStore | ? { $_.RouteMetric -eq 7373 -or $_.InterfaceAlias -like 'tsun*' }`.
 
 ## What it does not do