tstest/natlab/vmtest: add macOS VM snapshot caching for fast test starts

Cache a pre-booted macOS VM snapshot on disk so subsequent test runs
restore from the snapshot instead of cold-booting. The snapshot is keyed
by the Tart base image digest and a code version constant
(macOSSnapshotCodeVersion); bumping either invalidates the cache.

Snapshot preparation (one-time):
- Boot the Tart base image with a NAT NIC (--nat-nic flag)
- Wait for SSH, compile and install cmd/tta as a LaunchDaemon
- TTA polls the host via AF_VSOCK for an IP assignment; during prep
  the host replies "wait"
- Disconnect NIC, save VM state via SIGINT

Test fast path (cached, ~7s to agent connected):
- APFS clone the snapshot, write test-specific config.json
- Launch Host.app with --disconnected-nic --attach-network --assign-ip
- VZ restores from SaveFile.vzvmsave (~5s with 4GB RAM)
- TTA's vsock poll gets the IP config, sets static IP via ifconfig
  (bypasses DHCP entirely), switches driver addr to the IP directly
  (bypasses DNS), and resets the dial context so the reverse-dial
  reconnects immediately
- TTA agent connects to test driver within ~2s of IP assignment

Key optimizations:
- 4GB RAM instead of 8GB: halves SaveFile.vzvmsave (1.4GB vs 2.4GB),
  halves restore time (5.5s vs 11s)
- AF_VSOCK IP assignment: bypasses macOS DHCP (~5-7s saved)
- Direct IP dial: bypasses DNS resolution for test-driver.tailscale
- Dial context reset: cancels stale in-flight dials from snapshot
- Kill instead of SIGINT for test VM cleanup (no state save needed)
- Parallel VM launches

Also:
- Add TestDriverIPv4/TestDriverPort constants to vnet
- Add --nat-nic and --assign-ip flags to Host.app
- Fix SIGINT handler: retain DispatchSource globally, use dispatchMain()
- Add vsock listener (port 51011) to Host.app for IP config protocol
- Add disconnectNetwork() to VMController for clean snapshot state
- Fix Makefile: set -o pipefail so xcodebuild failures aren't swallowed

Updates #13038

Change-Id: Icbab73b57af7df3ae96136fb49cda2536310f31b
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This commit is contained in:
Brad Fitzpatrick
2026-04-28 14:38:15 -07:00
committed by Brad Fitzpatrick
parent 7b53550fe6
commit 02ffe5baa8
11 changed files with 1063 additions and 318 deletions
+32 -12
View File
@@ -22,8 +22,10 @@ extension HostCli {
@Option var share: String?
@Flag(help: "Run without GUI (for automated testing)") var headless: Bool = false
@Flag(help: "Create NIC with no attachment (for later hot-swap)") var disconnectedNic: Bool = false
@Flag(help: "Use NAT NIC instead of socket NIC (for snapshot prep)") var natNic: Bool = false
@Option(help: "Hot-swap NIC to this dgram socket path after boot/restore") var attachNetwork: String?
@Option(help: "Serve screenshots on this localhost port (0 = auto)") var screenshotPort: Int?
@Option(help: "Assign IP/mask/gw to guest via vsock (e.g. 192.168.1.2/255.255.255.0/192.168.1.1)") var assignIp: String?
mutating func run() {
config = Config(id)
@@ -32,19 +34,38 @@ extension HostCli {
if headless {
let attachSocket = attachNetwork
let disconnected = disconnectedNic || attachSocket != nil
let useNatNIC = natNic
let disconnected = !useNatNIC && (disconnectedNic || attachSocket != nil)
let wantScreenshots = screenshotPort != nil
let requestedPort = UInt16(screenshotPort ?? 0)
let ipConfig = assignIp
// Set up SIGINT handler before entering the event loop.
// The dispatch source must be stored in a global to prevent ARC deallocation.
signal(SIGINT, SIG_IGN)
let sigintSource = DispatchSource.makeSignalSource(signal: SIGINT, queue: .main)
retainedSigintSource = sigintSource
DispatchQueue.main.async {
let controller = VMController()
controller.createVirtualMachine(headless: true, disconnectedNIC: disconnected)
controller.createVirtualMachine(headless: true, disconnectedNIC: disconnected, natNIC: useNatNIC)
// Start vsock listener for IP assignment.
// If --assign-ip is set, the listener replies with the IP config JSON.
// If not set (snapshot prep), it replies "wait" so TTA keeps polling.
if let ipCfg = ipConfig {
let parts = ipCfg.split(separator: "/")
if parts.count == 3 {
let response = "{\"ip\":\"\(parts[0])\",\"mask\":\"\(parts[1])\",\"gw\":\"\(parts[2])\"}"
controller.startIPConfigListener(response: response)
}
} else {
controller.startIPConfigListener(response: "wait")
}
// Handle SIGINT (from test cleanup) by saving VM state before exit.
let sigintSource = DispatchSource.makeSignalSource(signal: SIGINT, queue: .main)
signal(SIGINT, SIG_IGN) // Let DispatchSource handle it
sigintSource.setEventHandler {
print("SIGINT received, saving VM state...")
print("SIGINT received, disconnecting NIC and saving VM state...")
controller.disconnectNetwork()
controller.pauseAndSaveVirtualMachine {
print("VM state saved, exiting.")
Foundation.exit(0)
@@ -79,11 +100,7 @@ extension HostCli {
let doAttach = {
if let sock = attachSocket {
// Give macOS a moment to settle after boot/restore,
// then hot-swap the NIC attachment.
DispatchQueue.main.asyncAfter(deadline: .now() + 1.0) {
controller.attachNetwork(serverSocket: sock, clientID: config.vmID)
}
controller.attachNetwork(serverSocket: sock, clientID: config.vmID)
}
}
@@ -107,7 +124,9 @@ extension HostCli {
fflush(stdout)
app.run()
} else {
RunLoop.main.run()
// Use dispatchMain() instead of RunLoop.main.run() so that
// GCD dispatch sources (like the SIGINT handler) are processed.
dispatchMain()
}
} else {
_ = NSApplicationMain(CommandLine.argc, CommandLine.unsafeArgv)
@@ -119,6 +138,7 @@ extension HostCli {
// startScreenshotServer starts a localhost HTTP server that serves VM display
// screenshots on GET /screenshot as JPEG. The port is printed to stdout as
// "SCREENSHOT_PORT=<port>" so the Go test harness can discover it.
var retainedSigintSource: DispatchSourceSignal? // prevent ARC deallocation
var screenshotServer: ScreenshotHTTPServer? // prevent GC
func startScreenshotServer(view: NSView, port: UInt16) {