Before sending a fix for #18991, this adds an integration test that
locks in that the proxymap WhoIs code works with two nodes running as
different users, with the second node running a localhost service and
able to use its local tailscaled to identify a Tailscale connection
from the other tailscaled.
Updates #18991
Change-Id: I6fbb0810204d77d2ac558f0cc786b73e3248d031
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Add two small APIs to support out-of-tree projects to exchange custom
signaling messages over DERP without requiring disco protocol
extensions:
- OnDERPRecv callback on magicsock.Options / wgengine.Config: called for
every non-disco DERP packet before the peer map lookup, allowing callers
to intercept packets from unknown peers that would otherwise be dropped.
- SendDERPPacketTo method on magicsock.Conn: sends arbitrary bytes to a
node key via a DERP region, creating the connection if needed. Thin
wrapper around the existing internal sendAddr.
Also allow netstack.Start to accept a nil LocalBackend for use cases
that wire up TCP/UDP handlers directly without a full LocalBackend.
Updates tailscale/corp#24454
Change-Id: I99a523ef281625b8c0024a963f5f5bf5d8792c17
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Fix three independent flake sources, at least as debugged by Claude,
though empirically no longer flaking as it was before:
1. Poll for connection counter data instead of reading immediately.
The conncount callback fires asynchronously on received WireGuard
traffic, so after counts.Reset() there is no guarantee the counter
has been repopulated before checkStats reads it. Use tstest.WaitFor
with a 5s timeout to retry until a matching connection appears.
2. Replace the *2 symmetry assumption in global metric assertions.
metricSendUDP and friends are AggregateCounters that sum per-conn
expvars from both magicsock instances. The old assertion assumed
both instances had identical packet counts, which breaks under
asymmetric background WireGuard activity (handshake retries, etc).
The new assertGlobalMetricsMatchPerConn computes the actual sum of
both conns' expvars and compares against the AggregateCounter value.
3. Tolerate physical stats being 0 when user metrics are non-zero.
A rebind event replaces the socket mid-measurement, resetting the
physical connection counter while user metrics still reflect packets
processed before the rebind. Log instead of failing in this case.
Also move counts.Reset() after metric reads and reorder the reset
sequence (counts before metrics) to minimize the race window.
Fixestailscale/tailscale#13420
Change-Id: I7b090a4dc229a862c1a52161b3f2547ec1d1f23f
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This makes tsnet apps not depend on x/crypto/ssh and locks that in with a test.
It also paves the wave for tsnet apps to opt-in to SSH support via a
blank feature import in the future.
Updates #12614
Change-Id: Ica85628f89c8f015413b074f5001b82b27c953a9
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Two issues caused TestCollectPanic to flake:
1. ETXTBSY: The test exec'd the tailscaled binary directly without
going through StartDaemon/awaitTailscaledRunnable, so it lacked
the retry loop that other tests use to work around a mysterious
ETXTBSY on GitHub Actions.
2. Shared filch files: The test didn't pass --statedir or TS_LOGS_DIR,
so all parallel test instances wrote panic logs to the shared system
state directory (~/.local/share/tailscale). Concurrent runs would
clobber each other's filch log files, causing the second run to not
find the panic data from the first.
Fix both by adding awaitTailscaledRunnable before the first exec, and
passing --statedir and TS_LOGS_DIR to isolate each test's log files,
matching what StartDaemon does.
It now passes x/tools/cmd/stress.
Fixes#15865
Change-Id: If18b9acf8dbe9a986446a42c5d98de7ad8aae098
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Raw byte accessors for key types, mirroring existing patterns
(NodePublic.Raw32 and DiscoPublicFromRaw32 already exist).
NodePrivate.Raw32 returns the raw 32 bytes of a node private key.
DiscoPrivateFromRaw32 parses a 32-byte raw value as a DiscoPrivate.
Updates tailscale/corp#24454
Change-Id: Ibc08bed14ab359eddefbebd811c375b6365c7919
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
We did so for Linux and macOS already, so also do so for Windows. We
only didn't already because originally we never produced binaries for
it (due to our corp repo not needing them), and later because we had
no ./tool/go wrapper. But we have both of those things now.
Updates #18884
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
When a recording upload fails mid-session, killProcessOnContextDone
writes the termination message to ss.Stderr() and kills the process.
Meanwhile, run() takes the ss.ctx.Done() path and proceeds to
ss.Exit(), which tears down the SSH channel. The termination message
write races with the channel teardown, so the client sometimes never
receives it.
Fix by adding an exitHandled channel that killProcessOnContextDone
closes when done. run() now waits on this channel after ctx.Done()
fires, ensuring the termination message is fully written before
the SSH channel is torn down.
Fixes#7707
Change-Id: Ib60116c928d3af46d553a4186a72963c2c731e3e
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
I omitted a lot of the min/max modernizers because they didn't
result in more clear code.
Some of it's older "for x := range 123".
Also: errors.AsType, any, fmt.Appendf, etc.
Updates #18682
Change-Id: I83a451577f33877f962766a5b65ce86f7696471c
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This had gotten flaky with Go 1.26.
Use synctest + AllocsPerRun to make it fast and deterministic.
Updates #18682
Change-Id: If673d6ecd8c1177f59c1b9c0f3fca42309375dff
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Otherwise it gets confused on new(123) etc.
Updates #18682
Change-Id: I9e2e93ea24f2b952b2396dceaf094b4db64424b0
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Fix its/it's, who's/whose, wether/whether, missing apostrophes
in contractions, and other misspellings across the codebase.
Updates #cleanup
Change-Id: I20453b81a7aceaa14ea2a551abba08a2e7f0a1d8
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Go 1.26's url.Parser is stricter and made our tests elsewhere fail
with this scheme because when these listen addresses get shoved
into a URL, it can't parse back out.
I verified this makes tests elsewhere pass with Go 1.26.
Updates #18682
Change-Id: I04dd3cee591aa85a9417a0bbae2b6f699d8302fa
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This is a minimal hacky fix for a case where the portlist poller extension
could miss updates to NetMap's CollectServices bool.
Updates tailscale/corp#36813
Change-Id: I9b50de8ba8b09e4a44f9fbfe90c9df4d8ab4d586
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This makes Set.MarshalJSON produce deterministic output in many cases now.
We still need to do make it deterministic for non-ordered types.
Updates #18808
Change-Id: I7f341ec039c661a8e88d07d7f4dc0f15d5d4ab86
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
For paring back build tag variant CI runs' set of packages to test.
Updates tailscale/corp#28679
Change-Id: Iba46fd1f58c1eaee1f7888ef573bc8b14fa73208
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
The old check was too aggressive and required TS_GO_NEXT=1 at runtime
as well, which is too strict and onerous.
This is a sanity check only (and an outdated one, at that); it's okay for
it to be slightly loose and permit two possible values. If either is working,
we're already way past the old bug that this was introduced to catch.
Updates tailscale/corp#36382
Change-Id: Ib9a62e10382cd889ba590c3539e6b8535c6b19fe
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This adds a new ControlKnob to make MagicDNS IPv6 registration
(telling systemd/etc) opt-out rather than opt-in.
Updates #15404
Change-Id: If008e1cb046b792c6aff7bb1d7c58638f7d650b1
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
OnPolicyChange can observe a conn in activeConns before authentication
completes. The previous `c.info == nil` guard was itself a data race
against clientAuth writing c.info, and even when c.info appeared
non-nil, c.localUser could still be nil, causing a nil pointer
dereference at c.localUser.Username.
Add an authCompleted atomic.Bool to conn, stored true after all auth
fields are written in clientAuth. OnPolicyChange checks this atomic
instead of c.info, which provides the memory barrier guaranteeing all
prior writes are visible to the concurrent reader.
Updates tailscale/corp#36268 (fixes, but we might want to cherry-pick)
Co-authored-by: Gesa Stupperich <gesa@tailscale.com>
Change-Id: I4c69843541f5f9f04add9bf431e320c65a203a39
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
The gocross-wrapper.sh bash script already checks TS_GO_NEXT (as of
a374cc344e) to select go.toolchain.next.rev over go.toolchain.rev,
but when TS_USE_GOCROSS=1 the Go binary itself was hardcoded to read
go.toolchain.rev. This makes gocross also respect the TS_GO_NEXT=1
environment variable.
Updates tailscale/corp#36382
Change-Id: I04bef25a34e7ed3ccb1bfdb33a3a1f896236c6ee
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This switches our gokrazy builds to use a new variant of cmd/gok called
opinionated about using monorepos: https://github.com/bradfitz/monogok
And with that, we can get rid of all the go.mod files and builddir forests
under gokrazy/**.
Updates #13038
Updates gokrazy/gokrazy#361
Change-Id: I9f18fbe59b8792286abc1e563d686ea9472c622d
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This updates the TS_GO_NEXT=1 (testing) toolchain to Go 1.26.0
The default one is still Go 1.25.x.
Updates #18682
Change-Id: I99747798c166ce162ee9eee74baa9ff6744a62f6
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Restore synchronous method calls from LocalBackend to magicsock.Conn
for node views, filter, and delta mutations. The eventbus delivery
introduced in 8e6f63cf1 was invalid for these updates because
subsequent operations in the same call chain depend on magicsock
already having the current state. The Synchronize/settleEventBus
workaround was fragile and kept requiring more workarounds and
introducing new mystery bugs.
Since eventbus was added, we've since learned more about when to use
eventbus, and this wasn't one of the cases.
We can take another swing at using eventbus for netmap changes in a
future change.
Fixes#16369
Updates #18575 (likely fixes)
Change-Id: I79057cc9259993368bb1e350ff0e073adf6b9a8f
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Add new "webbrowser" and "colorable" feature tags so that the
github.com/toqueteos/webbrowser and mattn/go-colorable packages
can be excluded from minbox builds.
Updates #12614
Change-Id: Iabd38b242f5a56aa10ef2050113785283f4e1fe8
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
We already had a featuretag for clientupdate, but the CLI wasn't using
it, making the "minbox" build (minimal combined tailscaled + CLI
build) larger than necessary.
Updates #12614
Change-Id: Idd7546c67dece7078f25b8f2ae9886f58d599002
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This provides a mechanism to block, waiting for Tailscale's IP to be
ready for a bind/listen, to gate the starting of other services.
It also adds a new --assert=[IP] option to "tailscale ip", for services
that want extra paranoia about what IP is in use, if they're worried about
having switched to the wrong tailnet prior to reboot or something.
Updates #3340
Updates #11504
... and many more, IIRC
Change-Id: I88ab19ac5fae58fd8c516065bab685e292395565
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
So callers can run testwrapper with -vet=off if they're already
running vet explicitly in a concurrent test job.
Updates tailscale/corp#28679
Change-Id: I74ad56e560076d187f5e3a7d7381e1dac89d860c
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This is for a future test scheduler, so it can run potentially flaky
tests separately, doing all the non-flaky ones together in one batch.
Updates tailscale/corp#28679
Change-Id: Ic4a11f9bf394528ef75792fd622f17bc01a4ec8a
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
When TS_GO_NEXT=1 is set, update/use the
go.toolchain.next.{branch,rev} files instead.
This lets us do test deploys of Go release candidates on some
backends, without affecting all backends.
Updates tailscale/corp#36382
Change-Id: I00dbde87b219b720be5ea142325c4711f101a364
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
We don't hold q.mu while running normal ExecQueue.Add funcs, so we
shouldn't in RunSync either. Otherwise code it calls can't shut down
the queue, as seen in #18502.
Updates #18052
Co-authored-by: Nick Khyl <nickk@tailscale.com>
Change-Id: Ic5e53440411eca5e9fabac7f4a68a9f6ef026de1
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
And fix up the TestAutoUpdateDefaults integration tests as they
weren't testing reality: the DefaultAutoUpdate is supposed to only be
relevant on the first MapResponse in the stream, but the tests weren't
testing that. They were instead injecting a 2nd+ MapResponse.
This changes the test control server to add a hook to modify the first
map response, and then makes the test control when the node goes up
and down to make new map responses.
Also, the test now runs on macOS where the auto-update feature being
disabled would've previously t.Skipped the whole test.
Updates #11502
Change-Id: If2319bd1f71e108b57d79fe500b2acedbc76e1a6
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
(trying to get in smaller obvious chunks ahead of later PRs to make
them smaller)
Updates #17925
Change-Id: I184002001055790484e4792af8ffe2a9a2465b2e
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
6a73c0bdf5 added a feature tag but didn't re-run go generate on ./feature/buildfeatures.
Updates #9192
Change-Id: I7819450453e6b34c60cad29d2273e3e118291643
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
For manual (human) testing, this lets the user disable control plane
map polls with "tailscale set --sync=false" (which survives restarts)
and "tailscale set --sync" to restore.
A high severity health warning is shown while this is active.
Updates #12639
Updates #17945
Change-Id: I83668fa5de3b5e5e25444df0815ec2a859153a6d
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Let's fix all the typos, which lets the code be more readable, lest we
confuse our readers.
Updates #cleanup
Change-Id: I4954601b0592b1fda40269009647bb517a4457be
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Includes adding StartPaused, which will be used in a future change to
enable netmap caching testing.
Updates #12639
Change-Id: Iec39915d33b8d75e9b8315b281b1af2f5d13a44a
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
It's an unnecessary nuisance having it. We go out of our way to redact
it in so many places when we don't even need it there anyway.
Updates #12639
Change-Id: I5fc72e19e9cf36caeb42cf80ba430873f67167c3
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
On the corp tailnet (using Mullvad exit nodes + bunch of expired
devices + subnet routers), these were generating big ~35 KB blobs of
logging regularly.
This logging shouldn't even exist at this level, and should be rate
limited at a higher level, but for now as a bandaid, make it less
spammy.
Updates #cleanup
Change-Id: I0b5e9e6e859f13df5f982cd71cd5af85b73f0c0a
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
As a baby step towards eventbus-ifying controlclient, make the
Observer optional.
This also means callers that don't care (like this network lock test,
and some tests in other repos) can omit it, rather than passing in a
no-op one.
Updates #12639
Change-Id: Ibd776b45b4425c08db19405bc3172b238e87da4e
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
They distracted me in some refactoring. They're set but never used.
Updates #17858
Change-Id: I6ec7d6841ab684a55bccca7b7cbf7da9c782694f
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
I noticed a deadlock in a test in a in-development PR where during a
shutdown storm of things (from a tsnet.Server.Close), LocalBackend was
trying to call magicsock.Conn.Synchronize but the magicsock and/or
eventbus was already shut down and no longer processing events.
Updates #16369
Change-Id: I58b1f86c8959303c3fb46e2e3b7f38f6385036f1
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Unfortunately I closed the tab and lost it in my sea of CI failures
I'm currently fighting.
Updates #cleanup
Change-Id: I4e3a652d57d52b75238f25d104fc1987add64191
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
So they're not all run N times on the sharded oss builders
and are only run one time each.
Updates tailscale/corp#28679
Change-Id: Ie21e84b06731fdc8ec3212eceb136c8fc26b0115
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
When systemd notification support was omitted from the build, or on
non-Linux systems, we were unnecessarily emitting code and generating
garbage stringifying addresses upon transition to the Running state.
Updates #12614
Change-Id: If713f47351c7922bb70e9da85bf92725b25954b9
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This removes one of the O(n=peers) allocs in getStatus, as
Engine.getStatus happens more often than Reconfig.
Updates #17814
Change-Id: I8a87fbebbecca3aedadba38e46cc418fd163c2b0
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
In 3f5c560fd4 I changed to use std net/http's HTTP/2 support,
instead of pulling in x/net/http2.
But I forgot to update DialTLSContext to DialContext, which meant it
was falling back to using the std net.Dialer for its dials, instead
of the passed-in one.
The tests only passed because they were using localhost addresses, so
the std net.Dialer worked. But in prod, where a tsnet Dialer would be
needed, it didn't work, and would time out for 10 seconds before
resorting to the old protocol.
So this fixes the tests to use an isolated in-memory network to prevent
that class of problem in the future. With the test change, the old code
fails and the new code passes.
Thanks to @jasonodonnell for debugging!
Updates #17304
Updates 3f5c560fd4
Change-Id: I3602bafd07dc6548e2c62985af9ac0afb3a0e967
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This fixes a regression from dd615c8fdd that moved the
newIPTablesRunner constructor from a any-Linux-GOARCH file to one that
was only amd64 and arm64, thus breaking iptables on other platforms
(notably 32-bit "arm", as seen on older Pis running Buster with
iptables)
Tested by hand on a Raspberry Pi 2 w/ Buster + iptables for now, for
lack of automated 32-bit arm tests at the moment. But filed #17629.
Fixes#17623
Updates #17629
Change-Id: Iac1a3d78f35d8428821b46f0fed3f3717891c1bd
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
I was debugging a customer issue and saw in their 1.88.3 logs:
TPM: error opening: stat /dev/tpm0: no such file or directory
That's unnecessary output. The lack of TPM will be reported by
them having a nil Hostinfo.TPM, which is plenty elsewhere in logs.
Let's only write out an "error opening" line if it's an interesting
error. (perhaps permissions, or EIO, etc)
Updates #cleanup
Change-Id: I3f987f6bf1d3ada03473ca3eef555e9cfafc7677
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
I got sidetracked apparently and never finished writing this Clone
code in 316afe7d02 (#17448). (It really should use views instead.)
And then I missed one of the users of "routerChanged" that was broken up
into "routerChanged" vs "dnsChanged".
This broke integration tests elsewhere.
Fixes#17506
Change-Id: I533bf0fcf3da9ac6eb4a6cdef03b8df2c1fb4c8e
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
