Merge pull request 'Enable netstack to hopefully fix DNS' (#31) from fix/add-gvisor-netstack into main

Reviewed-on: #31

Dockerfile

@@ -70,49 +70,6 @@ WORKDIR /src/tailscale
 # disables the filter at runtime for debugging — no rebuild needed.
 COPY patches/stderr_verbosity_filter.go cmd/tailscaled/
 
-# Patch net/tstun/wrap.go: fix panic("unreachable") in invertGSOChecksum for
-# ts_omit_netstack builds.
-#
-# invertGSOChecksum is a gVisor/GSO helper that inverts a transport-layer
-# checksum before/after SNAT when gVisor hands us a segment with a partial
-# checksum (NeedsCsum=true). It is only meaningful when netstack (gVisor) is
-# compiled in (HasNetstack=true).
-#
-# The function correctly guards its body with:
-#   if !buildfeatures.HasNetstack { panic("unreachable") }
-#
-# When built with ts_omit_netstack, HasNetstack is a const false, so that guard
-# evaluates to `if true { panic(...) }` — the function always panics.
-#
-# The problem: invertGSOChecksum is called unconditionally from injectedRead()
-# (twice, around pc.snat()), even for the res.data path where res.packet==nil
-# and gso is a zero-value netstack_GSO (NeedsCsum=false). The HasNetstack
-# guard in the res.packet branch does NOT protect these calls.
-#
-# As a result, any code path that injects an outbound packet via InjectOutbound()
-# — which happens when enabling exit-node use (Tailscale sends TSMP messages
-# and synthesizes packets through the TUN injection path) — hits injectedRead
-# with res.data!=nil, calls invertGSOChecksum, and crashes with:
-#   panic: unreachable
-#   tailscale.com/net/tstun.invertGSOChecksum(...)
-#   tailscale.com/net/tstun.(*Wrapper).injectedRead(...)  wrap.go:1077
-#
-# Fix: replace the `panic("unreachable")` with a `return` in invertGSOChecksum.
-# When HasNetstack=false (ts_omit_netstack), a zero-value netstack_GSO always
-# has NeedsCsum=false, so the function is correctly a no-op anyway. This matches
-# what the function would do if the rest of its body ran: NeedsCsum=false → return.
-#
-# The sed expression targets the function precisely: it matches the three-line
-# sequence that opens invertGSOChecksum's HasNetstack guard, and replaces only
-# the panic line with return. The pattern is stable across minor reformats
-# because it anchors on the literal function comment and the specific panic string.
-#
-# See tailscale/tailscale issue for context (no upstream fix as of v1.98.5):
-# panic happens when using exit-node via a ts_omit_netstack build.
-RUN sed -i \
-    -e '/func invertGSOChecksum/,/^}/ s/\t\tpanic("unreachable")/\t\treturn/' \
-    net/tstun/wrap.go
-
 # Build a minimal combined binary (tailscale CLI + tailscaled daemon in one file).
 #
 # Tag strategy — ALLOWLIST, not blocklist:
@@ -148,6 +105,34 @@ RUN sed -i \
 #                         waiting for completion") WITHOUT printing the auth URL
 #                         or confirming success. Including it makes interactive
 #                         'up' behave normally (blocks, prints login URL).
+#   netstack            — gVisor userspace network stack. Counter-intuitively
+#                         REQUIRED even though the router uses a real kernel TUN
+#                         (NOT --tun=userspace-networking). In v1.98.5 the
+#                         100.100.100.100:53 MagicDNS listener is served ONLY by
+#                         netstack's handleLocalPackets, installed via
+#                         PreFilterPacketOutboundToWireGuardNetstackIntercept.
+#                         The non-netstack "engine" interceptor that the wrap.go
+#                         comments claim handles quad-100 "if netstack is not
+#                         installed" does NOT actually do so on Linux (its body
+#                         only reflects loopback on darwin/ios/plan9, else
+#                         Accept). So with ts_omit_netstack, NOTHING absorbs
+#                         packets to 100.100.100.100: queries fall through to
+#                         WireGuard, no peer owns that IP, and even tailnet-name
+#                         resolution (and 'ping host.tailnet.ts.net') times out.
+#                         The 'dns' tag links the resolver but nothing routes
+#                         packets to it without netstack — the two tags are
+#                         independent (dns has no Dep on netstack). Omitting
+#                         netstack ALSO triggered a panic("unreachable") in
+#                         net/tstun.invertGSOChecksum on the exit-node inject
+#                         path (HasNetstack=const false made the guard always
+#                         panic); enabling netstack makes that guard dead code,
+#                         fixing the crash as a side effect. Cost (arm64, vs a
+#                         netstack-omitted build): ~+0.5 MB extracted on flash
+#                         and ~+2.3 MB resident RAM after UPX decompression —
+#                         measured, acceptable for a 16 MB-flash router.
+#   gro                 — Generic Receive Offload (perf). Depends on netstack;
+#                         pulled in with it. Small, and improves throughput on
+#                         the netstack DNS/inject path.
 #
 # Everything else remains omitted, including (rationale):
 #   clientupdate  — DELIBERATELY removed. The built-in updater would download
@@ -172,9 +157,11 @@ RUN sed -i \
 #                   which is exactly the flash wear we want to avoid.
 #   logtail       — no persistent log writes to flash; also pass
 #                   --no-logs-no-support at runtime
-#   netstack+gro  — userspace networking; router uses kernel TUN
 #   ssh           — not needed; access via MikroTik SSH + tailscale CLI
 #   all GUI/desktop/cloud/k8s features — irrelevant for a headless router
+#
+# NOTE: netstack/gro are NOT in this omit list — see the opted-in section above
+# for why MagicDNS quad-100 serving structurally requires them in v1.98.5.
 
 RUN mkdir -p /out && \
     ALL_OMIT=$(GOOS= GOARCH= go run ./cmd/featuretags --min --add=osrouter) && \
@@ -191,6 +178,8 @@ RUN mkdir -p /out && \
           -e 's/ts_omit_iptables,\{0,1\}//g' \
           -e 's/ts_omit_unixsocketidentity,\{0,1\}//g' \
           -e 's/ts_omit_ipnbus,\{0,1\}//g' \
+          -e 's/ts_omit_netstack,\{0,1\}//g' \
+          -e 's/ts_omit_gro,\{0,1\}//g' \
           -e 's/,$//' \
     ) && \
     echo "Build tags: ${TAGS}" && \

docs/DESIGN.md

@@ -15,22 +15,26 @@ Measured flattened rootfs for the arm64 image:
 
 | Component | On-disk size |
 |---|---|
-| `tailscale.combined` (UPX-compressed) | ~2.98 MB |
+| `tailscale.combined` (UPX-compressed) | ~3.47 MB |
 | custom static busybox (UPX, ~100 applets) | ~218 kB |
 | CA certificates | ~213 kB |
-| **Total extracted rootfs** | **~3.4 MB** |
+| **Total extracted rootfs** | **~3.9 MB** |
 
-(The compressed image / transfer tarball is ~3.3–4.3 MB depending on arch.)
+The `tailscale.combined` figure includes `netstack` (gVisor), which adds
+~0.5 MB on disk over a netstack-omitted build — a deliberate inclusion, see
+[Why netstack is required (even with a kernel TUN)](#why-netstack-is-required-even-with-a-kernel-tun).
+
+(The compressed image / transfer tarball is ~3.8–4.3 MB depending on arch.)
 
 | Arch | Image (compressed) |
 |---|---|
-| amd64  | ~4.2 MB |
-| arm64  | ~3.5 MB |
-| arm/v7 | ~3.5 MB |
+| amd64  | ~4.3 MB |
+| arm64  | ~4.0 MB |
+| arm/v7 | ~4.0 MB |
 
-On a deployed RouterOS device the container consumes **~3.7 MiB of flash**
+On a deployed RouterOS device the container consumes **~4.2 MiB of flash**
 (measured by `free-hdd-space` delta). Note that `du` *inside* the container
-reports roughly double that (~7 MB) — that is RouterOS block-allocation
+reports roughly double that (~8 MB) — that is RouterOS block-allocation
 rounding, **not** real usage or duplication; see
 [Avoiding overlayfs layer duplication](#avoiding-overlayfs-layer-duplication)
 for how to measure correctly.
@@ -118,13 +122,13 @@ delta**, not `du`:
 /system/resource/print     # note free-hdd-space before and after adding the container
 ```
 
-The container should consume **~3.7 MiB** of flash (e.g. 94.6 → 90.9 MiB free).
+The container should consume **~4.2 MiB** of flash (e.g. 94.6 → 90.4 MiB free).
 
 Do **not** trust `du` inside the container for this. Busybox `du` reports
-*allocated blocks*, and RouterOS's container store rounds a ~3 MB file up to
-~6 MB of blocks — so `du -sx /` reports ~7 MB even though real flash use is
-~3.7 MB. `ls -la /usr/local/bin` confirms the binary's true content size
-(~3.1 MB) and that it is a single file with two symlinks (no duplication).
+*allocated blocks*, and RouterOS's container store rounds the ~3.5 MB binary up
+to ~7 MB of blocks — so `du -sx /` reports ~8 MB even though real flash use is
+~4.2 MB. `ls -la /usr/local/bin` confirms the binary's true content size
+(~3.5 MB) and that it is a single file with two symlinks (no duplication).
 The image itself carries the binary in exactly one layer (verified at the blob
 level); the inflation is purely the filesystem's block accounting.
 
@@ -149,7 +153,8 @@ that's a separate build, not just a `--platform` change.
 | `advertise-routes` | Expose LAN subnets to the tailnet |
 | `use-exit-node` | Route the router's own traffic via a remote exit node |
 | `accept-routes` | Receive subnet routes from other tailnet nodes |
-| DNS / MagicDNS | Resolve `*.ts.net` names |
+| DNS / MagicDNS | Resolve `*.ts.net` names (resolver + resolv.conf manager). **Note:** serving `100.100.100.100` also requires `netstack` — see [Why netstack is required (even with a kernel TUN)](#why-netstack-is-required-even-with-a-kernel-tun) |
+| `netstack` + `gro` | gVisor userspace stack. Counter-intuitively **required** to serve MagicDNS on `100.100.100.100`, even though the router uses a real kernel TUN — see [Why netstack is required (even with a kernel TUN)](#why-netstack-is-required-even-with-a-kernel-tun) |
 | portmapper (NAT-PMP/PCP/UPnP) | Punch through upstream NAT |
 | listenrawdisco | Raw socket disco for better NAT traversal |
 | health | Powers `tailscale status` output |
@@ -166,7 +171,6 @@ that's a separate build, not just a `--platform` change.
 | `cachenetmap` | **Deliberately removed** — see [Why netmap disk-caching is removed](#why-netmap-disk-caching-is-removed) |
 | `logtail` | Would attempt persistent log writes; wear flash. Removing it also removes stderr verbosity filtering — restored by an injected filter, see [Log verbosity filtering](#log-verbosity-filtering) |
 | `netlog` | Network flow logging; separate concern |
-| `netstack` + `gro` | Userspace/gVisor networking; router uses kernel TUN |
 | `ssh` | Access via MikroTik SSH + `tailscale` CLI instead |
 | `linuxdnsfight` | inotify on `/etc/resolv.conf`; no systemd in container |
 | `networkmanager` / `resolved` / `dbus` / `sdnotify` | No systemd stack in container |
@@ -226,6 +230,84 @@ the in-memory resilience (the common case) while eliminating per-netmap flash
 writes. Only `tailscaled.state` (written on auth / key rotation) ever touches
 flash.
 
+### Why netstack is required (even with a kernel TUN)
+
+This is the least obvious inclusion in the build, so it is documented in full.
+
+`netstack` is Tailscale's embedded **gVisor userspace TCP/IP stack**. The
+natural assumption — and what earlier versions of this build acted on — is that
+a router which owns a **real kernel TUN device** (it is *not* run with
+`--tun=userspace-networking`) has no use for a userspace stack, so `netstack`
+(and its dependent `gro`) can be omitted to save space. That assumption is
+**wrong for one specific, important path: MagicDNS.**
+
+**MagicDNS on `100.100.100.100` is served only by netstack.** In Tailscale
+v1.98.5 the in-process listener for the Tailscale service IP
+(`100.100.100.100:53`, UDP) is installed exclusively by netstack's
+`handleLocalPackets`, wired into the TUN wrapper as
+`PreFilterPacketOutboundToWireGuardNetstackIntercept`
+(`wgengine/netstack/netstack.go`). When a packet leaves the host toward
+`100.100.100.100`, this hook absorbs it into the gVisor stack, whose UDP-53
+acceptor runs the MagicDNS resolver.
+
+**The "engine fallback" does not actually exist.** The TUN wrapper consults a
+second hook, `PreFilterPacketOutboundToWireGuardEngineIntercept`, and a comment
+in `net/tstun/wrap.go` claims it "primarily handles quad-100 if netstack is not
+installed." In v1.98.5 that comment is **false on Linux**: the engine
+`handleLocalPackets` (`wgengine/userspace.go`) only reflects loopback on
+darwin/ios/plan9 and otherwise returns `Accept` — it never touches
+`100.100.100.100`. So with `ts_omit_netstack` there is **no** code that absorbs
+quad-100 packets at all.
+
+**`dns` and `netstack` are independent tags.** The `dns` feature (which this
+build opts in) links the resolver and the `/etc/resolv.conf` manager, but it has
+no dependency on `netstack` and does **not** install any quad-100 transport.
+The net result of `dns` on + `netstack` off is a resolver that is correctly
+wired up but that **never receives any packets** — the worst kind of silent
+breakage. Symptoms observed on the device:
+
+- `/etc/resolv.conf` correctly points at `100.100.100.100` (the manager works),
+- but `dig anything @100.100.100.100` from inside the container **times out**
+  ("no servers could be reached"),
+- and even tailnet-internal names fail: `ping host.<tailnet>.ts.net` →
+  `bad address` (a name that needs **no** upstream forwarding still can't
+  resolve, proving the listener itself is dead, not an upstream-resolver issue),
+- while `ping 1.1.1.1` (a raw IP needing no DNS) works fine over the kernel data
+  path — confirming forwarding/exit-node connectivity is unaffected and isolating
+  the fault to DNS serving.
+
+**It also fixed a crash.** Omitting `netstack` set `buildfeatures.HasNetstack`
+to a compile-time `false`, which turned the guard in
+`net/tstun.invertGSOChecksum` (`if !HasNetstack { panic("unreachable") }`) into
+an always-panic. That function is called on the packet-injection path used when
+enabling exit-node mode, producing `panic: unreachable` and a daemon restart
+loop. Enabling `netstack` makes `HasNetstack` a const `true`, so the guard
+becomes dead code and the crash disappears as a side effect — fixed at the root
+cause rather than patched around.
+
+**Cost.** Measured on arm64, a netstack-enabled build versus a netstack-omitted
+one:
+
+| Metric | netstack omitted | netstack enabled | Delta |
+|---|---|---|---|
+| Extracted rootfs (flash) | ~3.42 MB | ~3.91 MB | **+0.49 MB** |
+| `tailscale.combined` on disk (UPX) | ~2.99 MB | ~3.47 MB | +0.48 MB |
+| Resident RAM after UPX decompress | ~12.25 MB | ~14.56 MB | **+2.31 MB** |
+
+The flash cost (~0.5 MB) is negligible on a 16 MB-class device. The RAM cost
+(~2.3 MB resident) is the real consideration on low-memory models, but is
+acceptable given that without it MagicDNS is entirely non-functional. The
+trade is: **half a megabyte of flash to make MagicDNS work at all.** `gro`
+(Generic Receive Offload) depends on `netstack` and is pulled in alongside it;
+it is small and improves throughput on the netstack path.
+
+**Caveat for future Tailscale bumps.** This coupling (quad-100 serving living
+only in netstack) is an upstream implementation detail, not a stable contract.
+If a future release adds a genuine non-netstack quad-100 path — or the daemon
+itself is refactored — re-test whether `netstack` can be dropped again. The
+canary is simple: from inside the container, `dig google.com @100.100.100.100`
+must return answers and `ping <host>.<tailnet>.ts.net` must resolve.
+
 ### Log verbosity filtering
 
 Upstream `tailscaled` embeds verbosity tags (`[v1]`, `[v2]`, …) inside its log