This patch extracts all the DNS related JSON handling from the
cmd/tailscale/jsonoutput package into a new tsdnsjsonv0 package.
It adds package documentation for tsdnsjsonv0 with a big WARNING that
this is an unstable format with no backwards compatibility guarantees.
When we stabilize this format, we should spin off a new tsdnsjsonv1
package that uses jsonoutput.ResponseEnvelope to declare version 1.
Updates #13326
Updates #18750
Signed-off-by: Simon Law <sfllaw@tailscale.com>
Flatten the cmd/tailscale package hierarchy by extracting the
jsonoutput package out of the cmd/tailscale/cli package.
Updates #cleanup
Change-Id: I92f80db75b0328e82f1596b6a42f6f6ef5a94bfa
Signed-off-by: Simon Law <sfllaw@tailscale.com>
Reader.Close set r.store to nil without holding r.mu, while reload read
r.store while holding r.mu. If a policy store is closed while a
concurrent reload is in flight, reload could observe a nil store and
crash tailscaled with a nil interface method call in
readPolicySettingValue.
Nil out r.store only while holding r.mu, and make reload return the
last known policy once the reader is closing instead of reading from
a store that may no longer exist.
Fixestailscale/corp#45548Fixestailscale/triage#394
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I494cfe9ea1df67b563bc061db9e6944f87b42a4e
The fake ticker has a one-element channel buffer and drops ticks when
the probe loop goroutine isn't already blocked on the channel, so
advancing the fake clock 50 times in a tight loop didn't guarantee
that the loop observed enough ticks to start three concurrent probe
runs. Under CI load, only two of the three run goroutines could be
spawned before the convergence timeout expired.
Advance the clock inside the polling loop instead, so ticks keep
firing until all three probe goroutines have started. Verified with
flakestress: the old test failed within ~41k runs, while the fixed
test passed 175,214 runs with no failures.
See http://flakes/analyze-test?name=tailscale.com%2Fprober.TestProberConcurrency
Updates #deflake
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I673a4918bbb5fea6b650e0dc1bc491c4af922b19
Files whose names contain characters with Unicode decompositions (such
as umlauts or voiced kana) could not be opened or written over
Taildrive.
Background: keyboards and IMEs emit NFC (precomposed) characters on
every platform, so filenames on Linux (ext4 etc) and Windows (NTFS)
disks are usually NFC bytes. NFD (decomposed) names mostly come from
Apple software: HFS+ forced a variant of NFD on write, and Apple's
frameworks still decompose paths via fileSystemRepresentation. APFS
preserves whatever bytes it is given but does normalization-insensitive
lookups (it stores a hash of the normalized name), so canonically
equivalent names find the same file. ext4 and NTFS lookups, by
contrast, are byte-exact.
On the wire, the macOS WebDAV client sends paths in NFD form (they
pass through the decomposing file system representation, and unlike
Apple's NFS client there is no "nfc" mount option). Windows and Linux
WebDAV clients pass names through as the application provided them,
typically NFC. WebDAV itself mandates no normalization, and PROPFIND
hrefs reflect the server's on-disk bytes.
The two forms are canonically equivalent but byte-wise different, so a
macOS client requesting the NFD form of an NFC-named file on a Linux
or Windows host got a 404 from the exact-byte lookup. Even against an
APFS host, where the filesystem absorbs the mismatch, the client-side
StatCache could still infer a 404: a cached directory listing in one
form caused depth 0 PROPFINDs in the other form to be treated as not
found without ever reaching the server. The inverse direction (NFD
bytes on a Linux disk, copied there from a Mac, requested in NFC form
by a Windows or Linux client) was broken too.
Alternative regimes considered: normalizing names at storage time (as
Nextcloud and Syncthing's autoNormalize do) would rename user files in
shared directories as a side effect of serving them; normalizing
request paths to a fixed form on the wire is unsound because the
on-disk form is unknowable a priori (ext4 can hold either form, or
both). Instead, adopt the APFS model: preserve bytes, but make lookups
normalization-insensitive.
Concretely, wrap the remote file server's webdav.Dir in a
normalizingFS that, when an exact path lookup fails, rescans the
parent directory for an entry whose name is canonically equivalent,
comparing the NFC form of both sides (which also sidesteps Apple's
nonstandard decomposition tables). Exact matches always win, and newly
created files keep the exact bytes the client sent. Also NFC-normalize
StatCache keys so canonically equivalent names share a cache entry.
The change is covered at three levels: unit tests for the StatCache,
an in-process two-node test in drive/driveimpl, and a new TestTaildrive
VM integration test in tstest/natlab/vmtest that shares a directory
between two Ubuntu VMs and exercises the NFC/NFD cases over the real
stack with curl playing the part of a macOS WebDAV client.
Fixes#15020
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I9c2f157e604efc629828581e08d5b3191dbb7d4e
TestContainerBoot/kube_shutdown_during_state_write flaked with exit
code 1 instead of 0 when SIGTERM arrived while "tailscale up" was
still running. Two problems combined:
tailscaleUp and tailscaleSet wrapped errors with %v, flattening the
error chain, so main's errors.Is(err, context.Canceled) check could
not recognize a graceful shutdown.
Even with %w, cmd.Run under a canceled context usually reports the
death of the killed subprocess ("signal: killed") rather than the
context error that caused it, since Wait prefers the process error.
Check ctx.Err() explicitly and return it (wrapped with %w) so that
a shutdown-driven cancellation is recognized wherever it lands
relative to the subprocess lifetime.
Before: the exit-code failure reproduced 4 times in 808 stress runs
under CPU starvation. After: 0 in 1195 runs.
Fixes#19380
Change-Id: Ie15ca722d2d5ac2a3f79b2d0ab01fb71d4b9220d
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Introduce a per-tailnet shared ACME account key so that all ingress
ProxyGroup replicas on a tailnet present the same account identity to
Let's Encrypt. This lets renewals claim the ARI "replaces" exemption
from the 50-certs-per-week rate limit, surviving Pod restarts,
ProxyGroup recreation, and cluster migrations.
The operator provisions a "tailscale-acme-accounts" Secret in its
namespace, guarded by a finalizer and a deletion warning event, and
watched so it is recreated promptly if removed. Proxies migrate any
pre-existing per-pod key into the shared Secret on first boot, adopt
the shared key on subsequent boots, and restore it on cert writes if
the Secret was recreated empty. Certs are stamped with the fingerprint
of the issuing account so renewals skip the "replaces" claim when the
account doesn't match.
Opt-in per-ProxyGroup via the tailscale.com/share-acme-account
annotation, or operator-wide via OPERATOR_SHARED_ACME_ACCOUNT_KEY.
Updates #18251
Updates #20288
Signed-off-by: chaosinthecrd <tom@tmlabs.co.uk>
Changing the name to "TailnetLockKey" would be clearer but introduces
more risk; this is an easy and low-stakes improvement.
Updates tailscale/corp#37904
Change-Id: I38d804202538b8670a80e744eb4dcb689f0002df
Signed-off-by: Alex Chan <alexc@tailscale.com>
I wrote this function two hours ago, tried to use it in corp, and
immediately found myself confused about the meaning of the arguments.
Time for named parameters!
Updates tailscale/corp#40404
Change-Id: Ic2866e052ccc9f6361b8d529233df54d63abbaa1
Signed-off-by: Alex Chan <alexc@tailscale.com>
We previously identified sync failures that occur when a node falls behind
the remote, and compacts away most its local state. We fixed the underlying
issue in #19444, but that PR only tested the basic scenario where the
local chain is a direct ancestor of the remote chain.
This patch adds an explicit regression test for the case where a node is on
a fork (that is, its HEAD is not part of the remote's active chain).
Although #19444 happened to cover this case, other proposed patches did not
handle the forked state. Adding this test locks in the behaviour and prevents
future sync regressions in this area.
Also, add a shared helper for writing this sort of TKA sync test.
Updates tailscale/corp#40404
Change-Id: I78fdc6beaf71392edf11806197f126db48886f93
Signed-off-by: Alex Chan <alexc@tailscale.com>
Add a large blob check to the pre-push hook, using the same git tree
diff logic as corp's check-file-size CI workflow (the
check-git-accidental-large-file GitHub Action): diff the pushed tree
against the remote's old tree (or the merge base with the remote's
default branch for new refs) and reject any new or changed blob over
1.5 MB. Unlike the CI check, which only guards PRs into main, the hook
runs before pushing to any branch, catching mistakes before they
permanently bloat the remote repo.
Set TS_SKIP_LARGE_FILE_CHECK=1 to push a large file intentionally,
mirroring the skip-large-file-check commit message tag honored by CI.
This folds the go.mod replace check and the new check into a single
CheckPrePush entry point so both share one read of the hook's stdin;
corp's git-hook.go needs the matching call site update when it next
bumps its tailscale.com dependency.
Updates tailscale/corp#9863
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I1c8cf2a277ce854d45c0ea809bed7c06b3295374
We were early-returning when the node was using an exit node, before
Connectors 2025 split DNS routes were calculated and installed.
Now we assemble the routes first, then install them in both exit node
and non-exit-node contexts. The returned resolvers set UseWithExitNode
to true even though as of today, we believe they should be installed in
all cases without regard to that boolean value. With the boolean, we
preserve the flexibility to toggle behavior without touching ipnlocal.
We also add a TODO to turn the extra split DNS route gathering into a
feature hook (tailscale/corp#37125).
This does not affect appc connectors, which receive split DNS routes,
and the UseWithExitNode value directly from control.
Updates #16384
Signed-off-by: Michael Ben-Ami <mzb@tailscale.com>
This commit modifies the generation command for the kubernetes operator
to include the CRD for peer relays in the helm chart and static
manifests
Updates: #fixup
Signed-off-by: David Bond <davidsbond93@gmail.com>
Previously there was a mismatch between how nodes store AUMs and what
the control plane would offer during sync:
- Client compaction: Nodes aggressively compact their TKA state -- they
keep the last 24 AUMs, every AUM received in the last two weeks, and
then everything from there back to the last checkpoint. Depending on
when it compacts, a node may only have ~50 AUMs.
- Exponential sampling: To save bandwidth, the control plane would send
a SyncOffer containing ancestors at exponentially increasing intervals
(4th, 16th, 64th, 256th...).
If a node has been offline for too long, the exponential sampling skips
the node's smaller window. When the SyncOffer and local state are disjoint,
the node cannot find a common ancestor to use for synchronisation.
It enters a failure loop where it keeps polling for new TKA state, but
it cannot catch up and has an increasingly-outdated view of the tailnet.
This patch replaces the exponential sampling with a SyncOffer that sends
every checkpoint ancestor of the current HEAD. Since every node is
guaranteed to keep at least one checkpoint after compaction, we're more
likely to have an intersection for the sync process.
This patch also increases `maxSyncHeadIntersectionIter`, which in
practice means the control plane will send every checkpoint in the
current chain. This means all affected nodes will be able to find an
intersection and catch up immediately, without requiring a client update.
It's still possible for a node to be unable to sync, but these edge cases
become less likely with this change. (For example, if a node is 1000+ AUMs
behind, or if it creates a local branch and then compacts away the
intersection with the main chain.)
This patch includes a regression test with synthetic data, and I
verified the fix with customer data.
Updates https://github.com/tailscale/corp/issues/40404
Change-Id: I2174011bb23a2b5972f6d1591aadcc016e3cba35
Signed-off-by: Alex Chan <alexc@tailscale.com>
We have some client builds on the unstable track where the conn25 code
doesn't run if the TAILSCALE_USE_WIP_CODE env var is not set. But the
split DNS routes for conn25 configured domains do get installed. This
means that users running those builds would get traffic for configured
domains black holed if the env var is not set.
This issue was fixed in 425a916ce.
Bump tailcfg.CapabilityVersion, and then a corresponding change to the
control server to not send conn25 config to lower versions will
avoid this issue for those users.
Updates tailscale/corp#45363
Signed-off-by: Fran Bull <fran@tailscale.com>
Add serviceclientprefs, an optional feature that stores and loads the
desktop clients' saved service launch preferences, one file per login
profile.
- Add GET|POST /localapi/v0/prefs/service-clients to load and save the
current profile's service client prefs.
- Add local client GetServiceClientPrefs and SetServiceClientPref that
call the new local api endpoint.
- Store the prefs with the ipn/store FileStore at
TailscaleVarRoot()/profile-data/<profileID>/service-client-prefs/<hex-encoded-key>,
so DeleteProfile cleans them up for free. Fall back to an in-memory
store when there's no var root.
- Register the feature and its local api route from build tagged files
so the whole thing drops out under ts_omit_serviceclientprefs.
- Add the serviceclient package holding Pref and Prefs (saved client,
username, database name, and last used time), so the local api client
and desktop apps can import the types without the feature machinery.
Change-Id: I340a99c1b332d181fb1556fbf3e8003bb3b95a08
Updates: https://github.com/tailscale/tailscale/issues/20429
Signed-off-by: Rollie Ma <rollie@tailscale.com>
The existing test only exercised the not-found-interface path. Now that
ipForwardingEnabledLinux opens its sysctl key with os.OpenInRoot
(840c6e3d3, #20572), also verify that the global keys and the
per-interface keys for every interface actually present on the machine
can be read without error, for both IPv4 and IPv6.
Updates #20572
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: Ie204a163ab9f8670abedd79a4ac81e400f71aab7
LetsEncrypt made certificates for bare IP addresses generally
available in January 2026. They require the short-lived ACME
certificate profile and are valid for about six days.
Add a new --acme-ip-certs flag. When set (with the default
--certmode=letsencrypt), connections that arrive by IP address (no
TLS SNI, or an IP address SNI matching the connection's destination
address) get a LetsEncrypt cert for that IP, obtained on demand using
the "shortlived" profile and the HTTP-01 challenge served on derper's
plaintext HTTP port. Because the certificate is requested for
whatever address the connection actually arrived on, it works for
both IPv4 and IPv6 with no per-address configuration, and a client
can never make us request a certificate for an address that isn't
ours. Connections with a DNS name in the SNI keep using the regular
autocert manager for --hostname.
autocert can't do any of this itself, as it neither orders IP address
identifiers nor serves connections without SNI, so this adds a small
dedicated cert manager using tailscale.com/tempfork/acme instead.
Clients can then connect to https://<IP> without the DERPMap CertName
pinning that self-signed certs from --certmode=manual require.
Updates tailscale/corp#45167
Updates #11776
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I8e2d5b0a7c4f9e1b3d6a8c2f5e0b9d4a7c1f3e6d
An SNI ServerName with a trailing dot (e.g. "host.ts.net.") failed
cert lookup because stored cert names have no trailing dot. Per RFC
6066 section 3 the SNI HostName carries no trailing dot, but some
clients send a fully-qualified name with one.
Trim the trailing dot at the boundary in getCertPEMWithValidity so all
lookup paths (the GetCertificate hook, Serve, and the localapi) resolve
the dotted and dotless forms to the same certificate.
Fixes#10233
Signed-off-by: Saleh <root@lr0.org>
Previously it was conn25-state. The new name prepares for the ability to
add new endpoints behind the conn25/ prefix, and prepares for parity for
an upcoming c2n endpoint with the same name.
Updates tailscale/corp#40125
Signed-off-by: Michael Ben-Ami <mzb@tailscale.com>
And rename serveStateGet to serveLocalAPIStateGet to prepare for adding
a c2n handler that is backed by the same methods as the LocalAPI
handler.
Updates tailscale/corp#40125
Signed-off-by: Michael Ben-Ami <mzb@tailscale.com>
Revert the direct fork dependency and its regenerated depaware/flake
manifests; not ready to ship yet.
This reverts commit 745bb8507.
Updates #1866
Signed-off-by: Kristoffer Dalby <kristoffer@tailscale.com>
Revert the tailscale/breakglass fork and its access-control flags;
not ready to ship yet.
This reverts commit 1d82c1b3d.
Updates #1866
Signed-off-by: Kristoffer Dalby <kristoffer@tailscale.com>
Unsigned peers aren't covered by tailnet lock, so they must never hold peer capabilities even if the packet filter grants them. This change extends the check for unsigned-peers to ensure full coverage in capabilities.
Fixestailscale/corp#45116
Change-Id: I918af24f0b9855e55921cbdad109cc68e745e125
Signed-off-by: Mike Jensen <mikej@tailscale.com>
Point tsapp at the tailscale/breakglass fork, fetch SSH keys from EC2
IMDSv2, restrict to the sec-scan user and internal CIDRs, start on
boot, and stop after 120s idle.
Updates #1866
Signed-off-by: Kristoffer Dalby <kristoffer@tailscale.com>
Depend on the tailscale/breakglass fork directly for its new
access-control flags. The fork renamed its module path so no replace
directive (disallowed here) is needed. Upstream gokrazy/breakglass
stays for the arm64 appliances.
Regenerate depaware manifests and nix flake hashes for the pkg/sftp
bump pulled in by the fork.
Updates #1866
Signed-off-by: Kristoffer Dalby <kristoffer@tailscale.com>
Change tstest's exported functions (AssertNotParallel, Replace,
Parallel, RequireRoot, SkipOnKernelVersions, MinAllocsPerRun, FixLogs,
UnfixLogs, CheckIsZero, ResourceCheck) to take testenv.TB instead of
testing.TB or *testing.T, so importing tstest from non-test code no
longer links the testing package and its flag registration side
effects into the binary. Add testenv.Verbose to replace the one use of
testing.Verbose, and a deptest check to keep testing out of tstest's
dependency graph.
Callers are unaffected: *testing.T and testing.TB both satisfy
testenv.TB.
Updates tailscale/corp#45223
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: Ib373ff66ceff638d071582baf8367245987e9155
The TB interface exists to mirror testing.TB without importing the
testing package, but it had fallen behind: Go 1.25 added Attr and
Output, and Go 1.26 added ArtifactDir. Add the missing methods and a
reflection-based test that TB has every exported method of testing.TB,
so future additions to testing.TB fail a test instead of silently
diverging. It can't be a compile-time assertion because testing.TB has
an unexported method.
Updates #16330
Updates #18682
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I9ba093afefdf3a6311ef4648bc1a13add9af453d
* tstest/natlab/vmtest: make cloud VM boot robust without KVM
Adding heavier distro images (Fedora) surfaced several ways the cloud VM
boot path breaks under TCG software emulation (no /dev/kvm), especially
with multiple concurrent VMs on few cores.
- Add a virtio-rng device to the cloud path so early boot doesn't block in
getrandom() waiting for the CRNG to seed.
- When no hardware acceleration is available, relax the stuck-console
watchdog (tuned for KVM's ~1-2s first output) and serialize VM boots so a
heavy guest doesn't starve its siblings' emulation threads.
- Bound the bring-up context to the test deadline and dump each VM's console
on failure, so a hang surfaces as a diagnosable Fatalf instead of an
opaque `go test -timeout` panic (which skips cleanups).
Fixestailscale/corp#44794
Updates tailscale/corp#44793
Signed-off-by: Brendan Creane <bcreane@gmail.com>
* tstest/natlab/vmtest: add Fedora and DNS-backend test coverage
Add the first RHEL-family distro and the machinery to assert and provision
distinct DNS backends, so adding a distro isn't "basically equivalent" to
the others.
- Add a Fedora 43 image (NetworkManager + systemd-resolved, SELinux
enforcing). restorecon-relabel the curl'd binaries so they exec under
enforcing mode.
- Add DNSBackend/AssertDNSBackend, reading the dns_manager_linux_mode_*
clientmetric to assert which backend tailscaled selected.
- Add a WithDNSMode node option. WithDNSMode(DNSDirect) masks
systemd-resolved and writes a plain resolv.conf pointing at natlab's fake
DNS, forcing the direct backend -- so one image covers multiple backends.
Fixestailscale/corp#44796
Updates tailscale/corp#44793
Signed-off-by: Brendan Creane <bcreane@gmail.com>
---------
Signed-off-by: Brendan Creane <bcreane@gmail.com>
This bumps go.mod to the current tailscale/golang-x-crypto, picking up
its rebase onto current upstream golang.org/x/crypto and its
cherry-pick of the pending upstream change
https://go-review.googlesource.com/c/crypto/+/788000, which adds ACME
certificate profile support: a new WithOrderProfile order option and
profile discovery via the directory metadata. That change has not yet
been submitted upstream and is subject to final API changes before it
lands there.
It then re-vendors that fork's acme package into tempfork/acme as
usual (per the TestSyncedToUpstream workflow), except for upstream's
pebble_test.go, which is now excluded from the sync: its
TestWithPebble downloads the Pebble module from outside our go.mod,
then builds and runs its binaries during tests.
Profile support is needed to request LetsEncrypt IP address
certificates, which require the "shortlived" profile.
Updates tailscale/corp#45167
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I3f7c2a91e5d8b4a6c0e2f9d1b7a3c8e6f4d0a2b9
Files under tempfork are vendored copies of upstream code that we
want to keep as close to upstream as possible, so don't require them
to use httpm constants. An upcoming tempfork/acme sync brings in
upstream test files using net/http's method constants.
Updates tailscale/corp#45167
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: If2a90b1d7c5e8f3a6b4d0c9e2a7f5b8d1c4e6a3f
The client/local package doc said its API is not necessarily stable, but
that caveat was easy to miss and only a few cert methods said anything
explicit either way. People have been surprised by IPN bus changes
between releases.
Add explicit "API maturity" notes, matching the existing wording on the
cert methods, marking stable: BugReport, BugReportWithOpts, CertDomains,
CheckUpdate, CurrentDERPMap, DialTCP, UserDial, DisconnectControl,
GetPrefs, EditPrefs, Status, StatusWithoutPeers, SetUseExitNode,
SwitchProfile, UserProfile, and the WhoIs* methods. Mark unstable:
ipn.Notify, WatchIPNBus, DoLocalRequest, the Debug*, Drive*, Check*,
EventBus*, and Stream* methods, SetComponentDebugLogging,
TailDaemonLogs, ShutdownTailscaled, GetDNSOSConfig, GetEffectivePolicy,
GetServeConfig, and GetAppConnectorRouteInfo.
Also note on tailcfg.DERPMap that the type is subject to minor changes
over time though its general shape is stable, document that
ipn.Prefs.CorpDNS is the internal name for "tailscale set --accept-dns",
and add a package doc paragraph to client/local saying that methods
without an explicit API maturity note should be assumed unstable.
Updates #20406
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I9333c58ae312e392c61d7de77987282e84ce2aeb
This commit contains the Kubernetes implementation of peer relays via the new `PeerRelay` CRD. It's a mega branch consisting of the commits of other PRs gone into this work:
1. https://github.com/tailscale/tailscale/pull/20211
2. https://github.com/tailscale/tailscale/pull/20329
3. https://github.com/tailscale/tailscale/pull/20423
4. https://github.com/tailscale/tailscale/pull/20503
An instance of the `PeerRelay` CRD deploys a `StatefulSet` of containerboot instances configured to advertise themselves as peer relays using the IP addresses configured via `LoadBalancer` services on each cloud provider (with some AWS specifics as it's less automatic than its competing cloud providers).
Per replica, a `LoadBalancer` type `Service` resource is provisioned and its IP address is used to configure the respective relay.
This has been tested with success in AWS, GCP & Azure and provides additional modification to `Service` resources via the CRD for any other kinds of deployment environments. It also contains some work that may appear to be duplication of what already exists within `cmd/k8s-operator` so we can start building an appropriate migration path for `Connector`, `ProxyGroup` etc into respective `k8s-operator/reconciler/*` packages.
Closes https://github.com/tailscale/corp/issues/34524
The wireguard-go receive path could be in chanTUN.Write, selecting to
send on the Inbound channel, while test cleanup called chanTUN.Close,
which closed that same channel. The select on the closed channel in
Write did not synchronize with Close, so the race detector flagged
the send racing with the close. It could also have panicked with a
send on a closed channel.
Add a mutex serializing Write and Close. Write now checks for closed
under the lock before doing a non-blocking send, so Close can't close
Inbound mid-send.
Fixes#20541
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: I8d8d10250cef0c1931753c78eebff6e8286f7201
Add an AppName field to the DERP ClientInfo so DERP servers can
attribute connections to the application making them, primarily for
best effort stats purposes. The value is plumbed per engine instance
rather than via a process global, so a process hosting multiple stacks
can attribute each one's DERP connections separately:
wgengine.Config.DERPAppName flows through magicsock.Options and
derphttp.Client into the naclbox-sealed ClientInfo JSON. Old servers
ignore the unknown field.
There are no callers in the tree yet setting the name.
Updates tailscale/corp#24454
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Change-Id: Ia7d3e9c2b6f8140e5a9d7c3b2e6f1a8d4c0b5e9f
