cabletest/main.go

package main

import (
	"flag"
	"fmt"
	"net"
	"os"
	"os/signal"
	"sync"
	"sync/atomic"
	"syscall"
	"time"

	"golang.org/x/sys/unix"
)

type endpoint struct {
	name string
	tag  string
	idx  int
	mac  [6]byte
	mtu  int
}

func (e endpoint) macString() string {
	return fmt.Sprintf("%02x:%02x:%02x:%02x:%02x:%02x",
		e.mac[0], e.mac[1], e.mac[2], e.mac[3], e.mac[4], e.mac[5])
}

func lookupEndpoint(name string) (endpoint, error) {
	ifi, err := net.InterfaceByName(name)
	if err != nil {
		return endpoint{}, err
	}
	if len(ifi.HardwareAddr) != 6 {
		return endpoint{}, fmt.Errorf("%s: expected 6-byte MAC, got %q", name, ifi.HardwareAddr)
	}
	var mac [6]byte
	copy(mac[:], ifi.HardwareAddr)
	return endpoint{name: name, idx: ifi.Index, mac: mac, mtu: ifi.MTU}, nil
}

// Returns the new start time, so the uptime shown alongside the totals counts
// from the reset rather than from launch; the elapsed clock restarting is the
// visible mark of the re-baseline.
func resetAll(dirs []*direction, mods []*phyModule) time.Time {
	for _, d := range dirs {
		d.reset()
	}
	for _, m := range mods {
		m.reset()
	}
	return time.Now()
}

// Nothing to hold or revert here: failures during a measure were never
// counted, so the view is always the counters as they stand.
func phyView(diag *cableDiag, modules []*phyModule) (phyDisplay, bool) {
	info, measuring := diag.snapshot()
	return phyDisplayFrom(info, measuring, modules[0].view(), modules[1].view()), measuring
}

var intervalCols = []colSpec{
	{group: "NOW", title: "bits/s", width: 9, right: true},
	{group: "NOW", title: "packets/s", width: 9, right: true},
	{group: "NOW", title: "snr", width: 6, right: true},
	{group: "NOW", title: "lost", width: 7, right: true},
	{group: "NOW", title: "corrupt", width: 7, right: true},
	{group: "NOW", title: "link", width: 7, right: true},
	{group: "NOW", title: "internal", width: 8, right: true},
	{group: "NOW", title: "corrected", width: 9, right: true},
	{group: "NOW", title: "noise", width: 7, right: true},
	{group: "OVERALL", title: "elapsed", width: 9, right: true},
	{group: "OVERALL", title: "packets", width: 9, right: true},
	{group: "OVERALL", title: "bytes", width: 9, right: true},
	{group: "OVERALL", title: "cable", width: 6, right: true},
	{group: "OVERALL", title: "corrected", width: 9, right: true},
	{group: "OVERALL", title: "lost", width: 9, right: true},
	{group: "OVERALL", title: "corrupt", width: 9, right: true},
	{group: "OVERALL", title: "link", width: 9, right: true},
	{group: "OVERALL", title: "internal", width: 9, right: true},
}

// The same figures the panel draws, in the same order: the last second as
// rates and error flags with the noise cable riding at the end of them, then
// everything since the reset.
func totalRow(elapsed time.Duration, v view, target float64, phy phyDisplay, nv noiseView) []string {
	return []string{
		rateCell(v.rxGbps*1e9, target*1e9),
		scaleSI(v.rxPPS),
		snrCell(phy),
		flagCell(v.window.lost),
		flagCell(v.window.corrupt),
		flagCell(v.window.link),
		flagCell(v.window.internal),
		correctedFlag(phy.recent),
		noiseCell(nv),
		scaleTime(elapsed),
		scaleCount(v.rxFrames),
		scaleCount(v.rxBytes),
		phy.metres,
		correctedCell(phy.corrected),
		statusCell(v.since.lost),
		statusCell(v.since.corrupt),
		statusCell(v.since.link),
		statusCell(v.since.internal),
	}
}

const (
	numStreams = 7
	batchSize  = 64

	testDriver = "ixgbe"

	// A constant rather than the negotiated speed, since this has to come up
	// with no cable in the port and nothing to negotiate.
	linkSpeed = 10.0
)

// The mac appends the fcs, so 60 and 1514 here are the smallest and largest
// standard frames, 64 and 1518 on the wire; 9014 fills the 9000 MTU.
var frameSizes = []int{60, 128, 256, 512, 1024, 1280, 1514, 9014}

func main() {
	// Left empty, the test pair is found by driver name instead: as PID 1 there
	// is no udev to pin names and no command line to pass, and which port gets
	// which ethN shifts with every driver built into the kernel.
	aName := flag.String("a", "", "first interface (default: the ixgbe pair)")
	bName := flag.String("b", "", "second interface")
	flag.Parse()

	if err := run(*aName, *bName); err != nil {
		fatal(err)
	}
	// A clean return is ctrl-alt-delete, which the kernel hands PID 1 as a
	// SIGINT. Exiting on it would panic the kernel over the reboot it was asking
	// for, so init asks for the reboot by name.
	if os.Getpid() == 1 {
		if err := unix.Reboot(unix.LINUX_REBOOT_CMD_RESTART); err != nil {
			panic(err)
		}
	}
}

const (
	reportInterval = time.Second
	// Deliberately not tied to the refresh: letting a slow or blocked draw set
	// the sampling clock would stretch the window it reports.
	sampleInterval = 16 * time.Millisecond
	// How far back the shown errors reach. The rate is not taken from this ring
	// but from the smeared receive buckets, which are keyed by the read clock.
	rateWindowSpan = time.Second
)

// One sampler for both directions, so they are read back to back on one clock
// rather than drifting apart on two.
type sampler struct {
	dirs []*direction
}

func (s *sampler) run(done *atomic.Bool, startTx <-chan struct{}) {
	<-startTx

	tick := time.NewTicker(sampleInterval)
	defer tick.Stop()

	for !done.Load() {
		<-tick.C
		for _, d := range s.dirs {
			d.sample()
		}
	}
}

func run(aName, bName string) (err error) {
	defer func() {
		if p := recover(); p != nil {
			if os.Getpid() != 1 {
				panic(p)
			}
			err = fmt.Errorf("%v", p)
		}
	}()

	if err := reportChecks("BOOT", bootstrap()); err != nil {
		return err
	}

	if aName == "" || bName == "" {
		var err error
		aName, bName, err = driverPair(testDriver)
		if err != nil {
			return err
		}
	}
	a, err := lookupEndpoint(aName)
	if err != nil {
		return err
	}
	b, err := lookupEndpoint(bName)
	if err != nil {
		return err
	}
	noise, err := newNoiser()
	if err != nil {
		return err
	}
	defer noise.close()

	a.tag, b.tag = "TEST A", "TEST B"
	noise.eps[0].tag, noise.eps[1].tag = "NOISE A", "NOISE B"
	ifnames := []string{a.name, b.name}

	ethertypes := make([]uint16, numStreams)
	for i := range ethertypes {
		ethertypes[i] = uint16(etherBase + i)
	}

	modules, moduleIDs, err := openModules([2]string{a.name, b.name})
	if err != nil {
		return err
	}

	checks := configureSystem(ifnames, ethertypes)
	checks = append(checks, moduleChecks(modules, [2]string{a.name, b.name})...)
	checks = append(checks, configureNoise(noise.names())...)
	if err := reportChecks("SETTINGS", checks); err != nil {
		return err
	}

	// The MTU check may have just raised them, so both are re-read before the
	// frame sizes are judged.
	for _, e := range []*endpoint{&a, &b} {
		fresh, err := lookupEndpoint(e.name)
		if err != nil {
			return err
		}
		e.mtu = fresh.mtu
		for _, s := range frameSizes {
			if s > e.mtu+ethHdrLen {
				return fmt.Errorf("size %d exceeds %s MTU %d (max frame %d)", s, e.name, e.mtu, e.mtu+ethHdrLen)
			}
		}
	}

	diag := newCableDiag(modules, cableInfo{})

	var dirs []*direction
	for _, p := range [][2]endpoint{{a, b}, {b, a}} {
		d, err := buildDirection(p[0].name+"->"+p[1].name, p[0], p[1], &diag.measuring)
		if err != nil {
			return err
		}
		dirs = append(dirs, d)
	}
	defer func() {
		for _, d := range dirs {
			d.close()
		}
	}()

	var linkRows [][]string
	for i, e := range []endpoint{a, b, noise.eps[0], noise.eps[1]} {
		mod := ""
		if i < len(moduleIDs) {
			mod = moduleIDs[i]
		}
		linkRows = append(linkRows, []string{
			paint(e.tag, cCyan), e.name, e.macString(), fmt.Sprintf("%d", e.mtu), mod,
		})
	}
	fmt.Println(renderBox("LINKS",
		[]string{"TAG", "INTERFACE", "MAC", "MTU", "MODULE"},
		[]bool{false, false, false, true, false}, linkRows))
	fmt.Println()

	// One row carries both directions, so line rate is both links at once.
	target := linkSpeed * float64(len(dirs))

	var done atomic.Bool
	var wg sync.WaitGroup
	var rxReady sync.WaitGroup
	startTx := make(chan struct{})
	for _, d := range dirs {
		rxReady.Add(len(d.rxFDs))
	}
	for _, d := range dirs {
		d.start(&wg, &done, &rxReady, startTx)
	}
	samp := &sampler{dirs: dirs}
	wg.Add(1)
	go func() {
		defer wg.Done()
		defer holdPanic()
		samp.run(&done, startTx)
	}()
	// Not gated on startTx: the cycle and the connected verdict are wanted the
	// moment the panel is, and nothing it does touches the measurement.
	wg.Add(1)
	go func() {
		defer wg.Done()
		defer holdPanic()
		noise.run(&done)
	}()
	for _, m := range modules {
		wg.Add(1)
		go func() {
			defer wg.Done()
			defer holdPanic()
			m.run(&done)
		}()
	}
	// Every return from here on stops the workers before the deferred closes
	// pull their sockets out from under them: otherwise the sampler panics on a
	// closed fd and can mask the error that actually ended the run. An error
	// before the gate opens closes it here, or the wait would hang on
	// goroutines still parked at startTx.
	defer func() {
		done.Store(true)
		select {
		case <-startTx:
		default:
			close(startTx)
		}
		wg.Wait()
	}()
	// A worker that panics before signalling ready would hang a bare Wait.
	ready := make(chan struct{})
	go func() {
		rxReady.Wait()
		close(ready)
	}()
	select {
	case <-ready:
	case p := <-fatalCh:
		return fmt.Errorf("%v", p)
	}

	sig := make(chan os.Signal, 1)
	signal.Notify(sig, syscall.SIGINT, syscall.SIGTERM)

	space, restoreTerm := watchSpace()
	defer restoreTerm()

	disp, err := newDisplay()
	if err != nil {
		return fmt.Errorf("display: %w", err)
	}
	defer disp.close()

	touch, err := watchTouch(disp.fb.pw, disp.fb.ph)
	if err != nil {
		return fmt.Errorf("touchscreen: %w", err)
	}

	for _, d := range dirs {
		d.primeCounters()
	}

	start := time.Now()
	// Anchored to the same clock as the noise cycle, so slot boundaries and
	// phase transitions coincide.
	hist := newHistory(rateEpochStart)
	// A reset resets at the press, then re-measures; the measure's failures
	// are suppressed at their sources while it runs, so there is nothing to
	// hide or revert afterwards.
	kickMeasure := func() {
		if diag.kick(&done) {
			start = resetAll(dirs, modules)
			hist.reset()
		}
	}
	close(startTx)
	// The first measure rides the same async path as a reset, so startup never
	// waits on it.
	kickMeasure()
	tick := time.NewTicker(reportInterval)
	defer tick.Stop()

	views := make([]view, len(dirs))
	rows := make([]view, len(dirs))
	stats := &streamTable{cols: intervalCols, headerEvery: 20}
	for {
		select {
		case p := <-fatalCh:
			return fmt.Errorf("%v", p)
		case <-sig:
			return nil
		// The verdict lives in the cable cell, not its own line.
		case <-space:
			kickMeasure()
		case <-disp.fb.flips:
			now := time.Now()
			px, py, down := touch.get()
			x, y := disp.fb.fromPanel(px, py)
			if disp.holdReset(x, y, down, now) {
				kickMeasure()
			}
			disp.showVersion = down && disp.versionSpot.contains(x, y)
			for i, d := range dirs {
				views[i] = d.displayView()
			}
			v := totalView(views)
			phy, measuring := phyView(diag, modules)
			if err := disp.render(v, now.Sub(start), phy, measuring,
				noise.view(), hist.view(), target); err != nil {
				return err
			}
		case now := <-tick.C:
			elapsed := now.Sub(start)
			for i, d := range dirs {
				rows[i] = d.displayView()
			}
			v := totalView(rows)
			phy, measuring := phyView(diag, modules)
			hist.sample(now, v, phy, noise.view(), measuring, target)
			for _, m := range modules {
				for _, n := range m.takeNotes() {
					fmt.Println(stats.rule(n))
				}
			}
			for _, line := range stats.emit(totalRow(elapsed, v, target, phy,
				noise.view())) {
				fmt.Println(line)
			}
		}
	}
}