* [client] Stop dumping the whole device to clear one peer endpoint
Clearing a peer's endpoint has to remove and re-add the peer, because neither the
netlink API nor the wireguard-go UAPI can clear an endpoint in place. To keep the
peer's allowed IPs across that dance, RemoveEndpointAddress read them back from the
device: a full wgctrl.Device() dump on the kernel path, a full IpcGet plus text parse
on the userspace one. Both cost a round trip proportional to the entire network map,
both run under the interface lock, and both run on every relay and ICE transition.
On a routing peer with ~15700 peers that is megabytes of netlink traffic per
transition, at a measured 713 transitions per minute, with every other configuration
operation queued behind it. RemoveAllowedIP paid the same price for the same reason.
The allowed IPs cannot come from the caller: peer.Conn knows the peer's own overlay
addresses, while the routed prefixes are attached separately by the route manager's
refcounter, so a caller-supplied set would silently drop every route behind the peer.
The configurer is the only writer of its device's peer set, so it can keep an
authoritative mirror of what it configured and answer from memory instead. The mirror
is fed by every operation that changes a peer's allowed IPs and reset by a device
reconfiguration that replaces the peer set. A peer the mirror has not seen, which is
what an out-of-band reconfiguration leaves behind, still falls back to reading the
device and seeds the mirror from it.
Prefixes are unmapped on the way in, so a v4-mapped address compares equal to the
plain v4 prefix for the same network rather than registering as a second entry.
Measured on a userspace device, allocations to clear one endpoint:
peers 64 256 1024 4096
before 1452 - 21617 -
after 91 91 91 91
* [client] Keep update-only allowed IP adds out of the peer mirror
AddAllowedIP configures the device with update_only, which is a silent no-op when
the peer does not exist, so its success says nothing about whether the device took
the prefix. Recording it unconditionally let the mirror hold a peer the device had
dropped, and RemoveEndpointAddress re-adds a peer without update_only: clearing the
endpoint of such a peer recreated it, carrying allowed IPs the device never held.
Allowed IPs are unique per device, so the recreated peer takes those prefixes away
from the peer that legitimately holds them.
This is not a theoretical window. Under lazy connections a routing peer's device
entry is torn down and re-created on the idle transition, and a routed prefix
re-added during that window is lost exactly because of update_only (#6863).
Allowed IP adds now merge only onto a peer the store already knows, which mirrors
the device: the operations that can create a peer record it, the update-only ones
do not. A peer missing from the store still falls back to reading the device.
* [client] Hand a prefix over to its new owner in the peer mirror
An allowed IP belongs to exactly one peer: configuring a prefix on a peer takes it
away from whichever peer held it before, and the configurer leaves that handover to
the device rather than removing the prefix from the previous holder itself, which is
what UpdatePeer's "wg will handle duplicated peer IP" refers to. The mirror recorded
the prefix on the new peer while leaving it listed under the old one, so clearing the
old peer's endpoint rewrote its allowed IPs from that stale list and took the prefix
back from the peer that now owns it. Traffic for the routed prefix then went to the
wrong peer. Reading the device before each write used to rule this out.
The store now tracks the owner of each prefix and performs the same handover, so
rewriting one peer's list cannot reclaim a prefix another peer holds.
Prefixes are also masked on the way in. A device stores them masked, so a caller
passing host bits would otherwise fail to match what a device fallback seeded and
could never remove that prefix by value. Conversion back from the device now keys
the v4-mapped decision on the mask width as well, so a genuine v6 prefix inside the
mapped range stays v6 instead of being dropped as an invalid v4 prefix.
* [client] Keep a mapped v6 prefix below /96 out of the v4 form
normalizePrefix unmapped any v4-mapped address before masking it, keeping the
original prefix length. For a genuine v6 prefix inside the mapped range, such as
::ffff:0:0/64, that pairs a v4 address with a v6 sized mask: netip.PrefixFrom
returns an invalid prefix and Masked turns it into the zero prefix. The store then
held a prefix whose Bits is -1, which cannot reproduce the allowed IP the device
was given, so re-adding the peer after an endpoint removal could fail once the
peer had already been removed.
Masking now comes first, and it also decides the address family: only a prefix at
least 96 bits long keeps the mapped marker through the mask, so anything shorter
inside that range is v6 and stays v6.
* [client] Record a peer created by a preshared key write
Setting a preshared key without updateOnly creates the peer when it is absent, and
Rosenpass applies a peer's first key exactly that way, since applyKeyLocked passes
the peer's initialized flag. The store ignored that operation, so the peer could
exist on the device while the store treated it as unknown.
An update-only allowed IP add on such a peer then succeeded on the device, which
moved the prefix away from its previous holder, while the store skipped the peer
and left the previous holder still claiming it. Clearing that holder's endpoint
rewrote it from the stale claim and took the prefix back, leaving the peer that
owns the route with nothing.
Every device operation that can create a peer now records it, which is the same
rule the update-only operations already follow from the other side.
* [client] Match a peer on the parsed key instead of its base64 form
getPeer scanned the device comparing Key.String to the caller's key. wgtypes.Key
is a 32 byte array, so it compares directly, while String base64 encodes it into a
fresh allocation on every iteration. The scan therefore allocated once per peer on
the device to find a single peer, and on a large network that is tens of thousands
of allocations per lookup.
The key is parsed once up front and the arrays are compared. Behaviour is
unchanged: the callers already parse the same key before reaching here, so the new
parse error is unreachable in practice and only guards the helper on its own.
* [client] Normalize prefixes on their way to the device
Prefixes were normalized when recorded but not when written, so a caller's raw prefix
reached the device while a different form was kept for it. The conversion is also where
a mapped prefix goes wrong: net.IPNet prints a v4-mapped address as v4 but takes the
length from its 16 byte mask, so ::ffff:10.1.2.3/64 is handed to a userspace device as
10.1.2.3/0 — an allowed IP matching every v4 address, on a peer that was meant to carry
one /64.
prefixesToIPNets now normalizes, and the two hand-built conversions in AddAllowedIP go
through it, so there is a single place where a prefix is turned into something a device
is given and it cannot disagree with what is recorded for it.
* [client] Parse the endpoint before configuring the peer
The userspace UpdatePeer parsed the endpoint address after the device had already been
configured, and returned on a parse failure. The device was then left holding a peer
that neither the activity recorder nor the allowed IP store had been told about, so the
peer was invisible to the wake path and the prefix handover for its allowed IPs never
happened, leaving the previous holder still claiming them.
The parse now happens before anything is written, so the only failure left after the
device is touched is one the caller cannot cause.
* [client] Keep the record when a peer removal fails
The two configurers disagreed: the kernel one dropped its record only once the device
had accepted the removal, the userspace one dropped it either way. Removing a peer is a
single device write, so a failure leaves the peer exactly as it was, with the allowed IPs
the record still describes. Dropping it there asserts nothing useful and only sends the
next caller to read the whole device back for an answer it already had.
The userspace one now follows the kernel and returns early on failure.
* [client] Write down what the allowed IP store does not guarantee
Two properties were relied on without being stated. The store's lock covers its map and
not the device write beside it, so consistency between the two rests on callers being
serialized, which WGIface does with its mutex; anyone removing that would have no way to
learn it mattered. And the fallback to the device only covers a peer the store has never
seen, so a peer first recorded from empty while the device already held prefixes keeps
only what was recorded, and the next endpoint removal drops the rest.
* [client] Key the allowed IP store on the parsed peer key
The store keyed on the textual key, so a lookup compared 44 byte strings while the
callers all held the parsed key already and the configurer had to carry both forms.
wgtypes.Key is a 32 byte array and compares directly, which is what getPeer was changed
to do for the same reason.
The store and its helpers now take wgtypes.Key, the callers pass the key they parsed on
entry, and the textual form survives only where something outside speaks it: parseStatus
reports peers that way, so the userspace fallback converts once for its scan.
* [client] Document the configurer methods the store changed
The exported configurer methods now carry what the allowed IP store made true of them:
when the mirror is reset, that a peer update merges its prefixes and takes them from
their previous owner, that an update-only add on an absent peer does nothing, and what
each side does with its record when a device write fails — where the two configurers
differ, since the userspace one reports a prefix it does not have and the kernel one
treats it as a no-op. mergeLocked states the lock its callers must already hold.
Docstrings that only restated the name of a test are left out; the tests explain the
scenario they set up in the body, where the explanation belongs.
* [client] Export the only-owner-writable path check from elevate
Pure refactor, no behavior change: the existing checkOnlyOwnerWritable gets a
thin exported wrapper so callers outside the elevation path can reuse it. No
call site changes here.
* [client] Validate the saved service parameters before applying them
The install reads <stateDir>/service.json and applies it to the service it then
registers: its arguments, its config path and its environment. The restricted
ACL that saveServiceParams puts on the state directory is applied when the file
is written, which is not necessarily before the file is first read, so the
install now checks the file rather than assuming it.
A file whose ownership or permissions are not the ones saveServiceParams
produces is treated as absent, and the install proceeds with its defaults. The
check covers the directories above the file as well, so what is checked is what
is read.
* [client] Restrict which environment variables the service is registered with
--service-env, and the service.json it persists to, accepted any name. A small
set of them decides how a process resolves the executables and libraries it
loads, and the daemon needs none of those: it now refuses them when they are
passed explicitly, and drops them with a warning when they come back from a
service.json written by an older version, so an upgrade does not fail over a
variable nobody needs.
* [client] Resolve netsh by absolute path
The lookup consulted PATH first and fell back to System32, in both the copy the
userspace firewall uses and the one that tears the interface down. It now asks
Windows for the system directory, so the resolution no longer depends on the
environment the service happens to be started with.
* [client] Move the System32 lookup into a package both callers share
Pure refactor, no behavior change: client/iface and client/firewall/uspfilter
carried a copy each of the same function, and neither imports the other, so the
body moves to client/internal/wincmd — alongside winregistry, which is where
the client's other Windows-only helper already lives. Both call sites now read
wincmd.System32("netsh").
* [client] Cover the System32 lookup with a test
Asserts what the previous commits changed: the lookup is absolute, and neither
PATH nor %SystemRoot% moves it.
* [client] Refuse the loader environment families by prefix
Review follow-up on the previous commit:
- LD_* and DYLD_* are now refused whole rather than name by name. Their members
differ per platform and libc and grow with new OS releases, so a list of them
is out of date as soon as it is written — DYLD_FALLBACK_LIBRARY_PATH and
DYLD_FALLBACK_FRAMEWORK_PATH were already missing from it.
- The names are folded to upper case only on Windows, where a variable is the
same one however it is spelled. Elsewhere the environment is case-sensitive,
so Path and PATH are two variables and only the exact spelling is the one that
is read; the fold refused the wrong one.
- TEMP and TMP stay in the denylist, but the rationale and the message now say
what they actually decide: where the service writes, not what it loads.
Android 16+ local network protection derives the blocked prefixes from the interface address prefix. A /16 address turns the whole overlay into a local network, so apps without ACCESS_LOCAL_NETWORK cannot reach any peer. Pass the address as /32 and /128 and add the overlay networks to the route list that the Android side turns into VPN routes, on both the initial create and the renew path.
* [client] Add tests for the ICEBind open and close races
Running many embedded clients in one process intermittently wedges interface
creation. A goroutine dump taken from 50 clients shows ten of them parked for
seven minutes in Device.IpcSet, in closeBindLocked waiting on
device.net.stopping.Wait, holding device.net while every other device
goroutine queues behind it on Device.Up.
Open writes s.closed and Close reads it with no synchronisation, and Close
also closes s.closedChan without the mutex that Open swaps it under. Two
Closes can both pass the check and close the same channel, and a Close racing
an Open can mark the bind closed while a live channel and live receive
functions remain, after which every later Close takes its early return and
runs neither close(closedChan) nor StdNetBind.Close. The receive functions
never stop, so stopping.Wait never returns.
These tests do not fix that. The first pins the contract closeBindLocked
depends on and passes today. The other two fail under -race, reporting the
races at the three sites above, and pass again once closed and closedChan are
guarded consistently.
* [client] Release parked receivers so reopening a bind cannot stall
receiveRelayed held closedChanMu for the whole of its blocking select, so a
parked receiver kept the read lock indefinitely and Open could never take the
write lock it needs to install a fresh closedChan. wireguard-go reaches Open
from Device.IpcSet and Device.Up with device.net held, so the stall took the
device lock with it: interface creation never finished, every other device
goroutine queued behind Device.Up, and Engine.Start never returned.
Callers now copy the channel under a short read lock and select on the copy.
Copying alone would stand a new trap in the same place, because an Open that
follows an Open leaves the previous generation parked on a channel no later
Close can reach, so Open now closes the outgoing channel before swapping it.
closed and closedChan are also updated together under that mutex. Read and
written apart, Close could see a stale closed and skip both close(closedChan)
and StdNetBind.Close, leaving every receive function running and wedging
closeBindLocked on device.net.stopping.Wait, or two Close calls could pass the
check together and close the same channel twice.
TestICEBindOpenDoesNotBlockOnParkedReceiver fails without this change, without
needing the race detector. The other three cover the surrounding contract and
report the state races under -race.
* [client] Make the bind lifecycle transition atomic and tighten its tests
Review caught that the previous commit moved the torn transition rather than
removing it. Open published the new generation before calling StdNetBind.Open,
so an Open rejected because the bind was already open had already signalled the
outgoing generation, and a Close arriving in that window could mark the bind
closed while the same call went on to install live sockets. Every later Close
then returned early and never shut them down.
Open now calls StdNetBind.Open first, so a failure leaves the current
generation untouched, and both Open and Close hold the lock across the whole
transition. Ordering is safe: StdNetBind.Open reaches muUDPMux through
createReceiverFn, and no path takes muUDPMux before closedChanMu.
The tests were also weaker than they read. The stress test claimed to cover a
stale channel but only ever raced two Closes, and the concurrency test left
overlap to goroutine start order. Both now gate their goroutines on a common
start, the stress test races an Open against the Closes, and both assert the
surviving generation channel is actually closed. Waiting on receive functions
to be entered replaces part of the sleep in the reopen probe, and teardown
bounds its Close so a regression fails the assertion instead of hanging.
Two of the four now fail without the fix and no race detector, the stress test
by reproducing close of a closed channel at the Close early return.
* [client] Fail the reopen probe when its teardown does not complete
closeBounded swallowed its timeout and the cleanup discarded what
receiversStopped returned, so the bounds added in the previous commit only
stopped teardown hanging. A wedged Close or a parked receiver would have left
the test green with a leaked goroutine, which is the failure this test exists
to catch.
closeBounded now reports whether Close returned, and cleanup fails the test on
either bound.
* [iface] Drop redundant device dump in kernel configure()
wgctrl.ConfigureDevice already returns an error when the interface is
missing, so the preceding wg.Device() existence check is redundant. That
check dumps the entire device (all peers) on every configure() call,
making it O(peers) per call and turning bulk peer insertion into
O(peers^2): inserting N peers one by one re-parsed the whole growing peer
list N times. Removing it keeps each peer write constant-time regardless
of how many peers are already configured.
* [iface] Cache WireGuard stats to collapse per-peer device dumps
Each peer runs a WGWatcher that polls GetStats(), and every call dumps
the whole device, so with N peers the watchers perform O(N) full dumps
per poll cycle (O(N^2) work) while each keeps only its own peer's entry.
Wrap the kernel and userspace configurer GetStats() in a short-TTL cache
with singleflight: the staggered per-peer calls share a single device
dump per window and concurrent misses collapse into one dump. The kernel
and userspace WireGuard APIs have no per-peer stats query (a get always
returns the whole device), so a shared cached snapshot avoids the
repeated full dumps.
* Ignore .claude directory
* [client] categorize root/system-mutating tests behind a privileged build tag
Tests that need root or mutate host state (nftables/iptables/DNS, TUN/WireGuard
interfaces, routes, eBPF, SSH/service install) are now gated behind a
//go:build privileged tag. The default `go test ./client/...` runs as a non-root
user with no sudo and leaves host networking untouched; mixed files were split so
pure-logic tests stay in the default suite.
A self-hosting ory/dockertest/v4 harness (client/testutil/privileged) runs the
privileged suite inside a --privileged --cap-add=NET_ADMIN container via
`make test-privileged`; a DOCKER_CI=true guard skips the spawn when already inside
the container. Added `make test-unit` for the host-safe run.
* [client] add PRIV_RUN/PRIV_PKGS filters to the privileged test harness
The dockertest harness now reads two optional env vars when building the
in-container `go test` command: PRIV_RUN adds a -run test-name filter and
PRIV_PKGS overrides the package list. Both empty reproduce the full privileged
suite, so CI and `make test-privileged` behave as before. Lets a developer run a
single privileged test in the container, e.g.:
PRIV_RUN=TestNftablesManager PRIV_PKGS=./client/firewall/nftables/... make test-privileged
* [client] fix unused-helper lint after the privileged test split
Splitting privileged tests into *_privileged_test.go left their shared helpers in
the untagged files, so in the default (no-tag) build they had no callers and
golangci-lint flagged them as unused.
Moved the privileged-only helpers into the privileged files next to their callers
(generateDummyHandler; createEngine/startSignal/startManagement/getConnectedPeers/
getPeers + kaep/kasp; (*mockDaemon).setJWTToken). Annotated the shared routing-test
fixtures that must stay untagged for cross-platform compilation with //nolint:unused
(systemops_bsd expected* vars, ensureIPv6DefaultRoute on bsd/windows,
loopbackIfaceWindows), matching the existing linux variant.
* [client] fix privileged test CI failures and run the harness on macOS
The host-safe unit run dropped sudo but two privileged test groups were
never tagged, and the Docker privileged job silently never ran the suite:
- Gate the ssh/server PrivilegeDropper command-construction tests behind
the privileged tag (they require root to target a different UID); split
them into executor_unix_privileged_test.go.
- Tag sharedsock raw-socket tests privileged (need CAP_NET_RAW).
- Fix the Docker job command: nested single quotes around the build tags
closed the sh -c wrapper early, dropping the go list package set and the
privileged tag, so go test ran on the empty repo root. Use double quotes.
Make the self-hosting harness usable from a dev Mac:
- Build it on darwin as well as linux; it only drives Docker.
- Resolve the active docker context endpoint into DOCKER_HOST when the
default /var/run/docker.sock is absent (Docker Desktop, Colima, OrbStack).
- Rename the misspelled containerGoModache constant to containerGoModCache.
* Update client/internal/engine_privileged_test.go
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
* Update client/internal/routemanager/systemops/systemops_linux_test.go
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
* Update client/internal/routemanager/systemops/systemops_windows_test.go
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
* Update client/server/server_privileged_test.go
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
* [ci] Run privileged-tagged tests on darwin, windows and freebsd
The privileged build tag split moved root/system-mutating tests behind
//go:build privileged, but only the linux docker job was given the tag.
The native darwin (sudo), windows (PsExec64 -s) and freebsd VM runners
already have the required privileges, so add the privileged tag there too
to keep CI running the same set of tests as before the split.
* [ci] Exclude dockertest harness from the darwin privileged run
The privileged tag now compiles client/testutil/privileged on darwin, whose
TestRunPrivilegedSuiteInDocker spawns a container the macOS runner has no
Docker for. Exclude the harness package from the darwin list, matching the
linux job, so the privileged tests run in place without a container spawn.
---------
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
The test writes 500 packets per family and asserted exact-count
delivery within a 5s window, even though its own comment says "Some
packet loss is acceptable for UDP". On FreeBSD/QEMU runners the writer
loops cannot always finish all 500 before the 5s deadline closes the
readers (we have seen 411/500 in CI).
The real assertion of this test is the routing check — IPv4 peer only
gets v4- packets, IPv6 peer only gets v6- packets — which remains
strict. Replace the exact-count assertions with a >=80% delivery
threshold so runner speed variance no longer causes false failures.
WGIface.Close() took w.mu and held it across w.tun.Close(). The
underlying wireguard-go device waits for its send/receive goroutines to
drain before Close() returns, and some of those goroutines re-enter
WGIface during shutdown. In particular, the userspace packet filter DNS
hook in client/internal/dns.ServiceViaMemory.filterDNSTraffic calls
s.wgInterface.GetDevice() on every packet, which also needs w.mu. With
the Close-side holding the mutex, the read goroutine blocks in
GetDevice and Close waits forever for that goroutine to exit:
goroutine N (TestDNSPermanent_updateUpstream):
WGIface.Close -> holds w.mu -> tun.Close -> sync.WaitGroup.Wait
goroutine M (wireguard read routine):
FilteredDevice.Read -> filterOutbound -> udpHooksDrop ->
filterDNSTraffic.func1 -> WGIface.GetDevice -> sync.Mutex.Lock
This surfaces as a 5 minute test timeout on the macOS Client/Unit
CI job (panic: test timed out after 5m0s, running tests:
TestDNSPermanent_updateUpstream).
Release w.mu before calling w.tun.Close(). The other Close steps
(wgProxyFactory.Free, waitUntilRemoved, Destroy) do not mutate any
fields guarded by w.mu beyond what Free() already does, so the lock
is not needed once the tun has started shutting down. A new unit test
in iface_close_test.go uses a fake WGTunDevice to reproduce the
deadlock deterministically without requiring CAP_NET_ADMIN.
In netstack (proxy) mode, the process lacks permission to create
/var/run/wireguard, making the UAPI listener unnecessary and causing
a misleading error log. Introduce NewUSPConfigurerNoUAPI and use it
for the netstack device to avoid attempting to open the UAPI socket
entirely. Also consolidate UAPI error logging to a single call site.
- Add WireguardPort option to embed.Options for custom port configuration
- Fix KernelInterface detection to account for netstack mode
- Skip SSH config updates when running in netstack mode
- Skip interface removal wait when running in netstack mode
- Use BindListener for netstack to avoid port conflicts on same host
* Add IPv6 support to UDP WireGuard proxy
Add IPv6 packet header support in UDP raw socket proxy
to handle both IPv4 and IPv6 source addresses.
Refactor error handling in proxy bind implementations
to validate endpoints before acquiring locks.
* [client] Add WGConfigurer interface
To allow Rosenpass to work both with kernel
WireGuard via wgctrl (default behavior) and
userspace WireGuard via IPC on Android/iOS
using WGUSPConfigurer
* [client] Remove Rosenpass debug logs
* [client] Return simpler peer configuration in outputKey method
ConfigureDevice, the method previously used in
outputKey via wgClient to update the device's
properties, is now defined in the WGConfigurer
interface and implemented both in kernel_unix and
usp configurers.
PresharedKey datatype was also changed from
boolean to [32]byte to compare it
to the original NetBird PSK, so that Rosenpass
may replace it with its own when necessary.
* [client] Remove unused field
* [client] Replace usage of WGConfigurer
Replaced with preshared key setter interface,
which only defines a method to set / update the preshared key.
Logic has been migrated from rosenpass/netbird_handler to client/iface.
* [client] Use same default peer keepalive value when setting preshared keys
* [client] Store PresharedKeySetter iface in rosenpass manager
To avoid no-op if SetInterface is called before generateConfig
* [client] Add mutex usage in rosenpass netbird handler
* [client] change implementation setting Rosenpass preshared key
Instead of providing a method to configure a device (device/interface.go),
it forwards the new parameters to the configurer (either
kernel_unix.go / usp.go).
This removes dependency on reading FullStats, and makes use of a common
method (buildPresharedKeyConfig in configurer/common.go) to build a
minimal WG config that only sets/updates the PSK.
netbird_handler.go now keeps s list of initializedPeers to choose whether
to set the value of "UpdateOnly" when calling iface.SetPresharedKey.
* [client] Address possible race condition
Between outputKey calls and peer removal; it
checks again if the peer still exists in the
peers map before inserting it in the
initializedPeers map.
* [client] Add psk Rosenpass-initialized check
On client/internal/peer/conn.go, the presharedKey
function would always return the current key
set in wgConfig.presharedKey.
This would eventually overwrite a key set
by Rosenpass if the feature is active.
The purpose here is to set a handler that will
check if a given peer has its psk initialized
by Rosenpass to skip updating the psk
via updatePeer (since it calls presharedKey
method in conn.go).
* Add missing updateOnly flag setup for usp peers
* Change common.go buildPresharedKeyConfig signature
PeerKey datatype changed from string to
wgTypes.Key. Callers are responsible for parsing
a peer key with string datatype.
* updates to client file writing
* numerous
* minor
* - Align OnLoginSuccess behavior with Android (only call on nil error)
- Remove verbose debug logging from WaitToken in device_flow.go
- Improve TUN FD=0 fallback comments and warning messages
- Document why config save after login differs from Android
* Add nolint directive for staticcheck SA1029 in login.go
* Fix CodeRabbit review issues for iOS/tvOS SDK
- Remove goroutine from OnLoginSuccess callback, invoke synchronously
- Stop treating PermissionDenied as success, propagate as permanent error
- Replace context.TODO() with bounded timeout context (30s) in RequestAuthInfo
- Handle DirectUpdateOrCreateConfig errors in IsLoginRequired and LoginForMobile
- Add permission enforcement to DirectUpdateOrCreateConfig for existing configs
- Fix variable shadowing in device_ios.go where err was masked by := in else block
* Address additional CodeRabbit review issues for iOS/tvOS SDK
- Make tunFd == 0 a hard error with exported ErrInvalidTunnelFD (remove dead fallback code)
- Apply defaults in ConfigFromJSON to prevent partially-initialized configs
- Add nil guards for listener/urlOpener interfaces in public SDK entry points
- Reorder config save before OnLoginSuccess to prevent teardown race
- Add explanatory comment for urlOpener.Open goroutine
* Make urlOpener.Open() synchronous in device auth flow
* When a peer disconnects, remove the endpoint address to avoid sending traffic to a non-existent address, but retain the status for the activity recorder.
The Relayed connection setup is optimistic. It does not have any confirmation of an established end-to-end connection. Peers start sending WireGuard handshake packets immediately after the successful offer-answer handshake.
Meanwhile, for successful P2P connection negotiation, we change the WireGuard endpoint address, but this change does not trigger new handshake initiation. Because the peer switched from Relayed connection to P2P, the packets from the Relay server are dropped and must wait for the next WireGuard handshake via P2P.
To avoid this scenario, the relayed WireGuard proxy no longer drops the packets. Instead, it rewrites the source address to the new P2P endpoint and continues forwarding the packets.
We still have one corner case: if the Relayed server negotiation chooses a server that has not been used before. In this case, one side of the peer connection will be slower to reach the Relay server, and the Relay server will drop the handshake packet.
If everything goes well we should see exactly 5 seconds improvements between the WireGuard configuration time and the handshake time.
- Move `util/grpc` and `util/net` to `client` so `internal` packages can be accessed
- Add methods to return the next best interface after the NetBird interface.
- Use `IP_UNICAST_IF` sock opt to force the outgoing interface for the NetBird `net.Dialer` and `net.ListenerConfig` to avoid routing loops. The interface is picked by the new route lookup method.
- Some refactoring to avoid import cycles
- Old behavior is available through `NB_USE_LEGACY_ROUTING=true` env var
Deduplicate STUN package sending.
Originally, because every peer shared the same UDP address, the library could not distinguish which STUN message was associated with which candidate. As a result, the Pion library responded from all candidates for every STUN message.
Riccardo Manfrin
