ci: add WSL2 CI

Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp> Signed-off-by: Justin Alvarez <alvajus@amazon.com> Co-authored-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>

Author: pendo324

.github/workflows/test.yml

@@ -73,9 +73,35 @@ jobs:
 
  windows:
  name: "Windows tests"
- runs-on: windows-2022
+ runs-on: windows-2022-8-cores
  timeout-minutes: 30
  steps:
+ - name: Enable WSL2
+ run: |
+ wsl --set-default-version 2
+ # Manually install the latest kernel from MSI
+ Invoke-WebRequest -Uri "https://wslstorestorage.blob.core.windows.net/wslblob/wsl_update_x64.msi" -OutFile "wsl_update_x64.msi"
+ $pwd = (pwd).Path
+ Start-Process msiexec.exe -Wait -ArgumentList "/I $pwd\wsl_update_x64.msi /quiet"
+ wsl --update
+ wsl --status
+ wsl --list --online
+ - name: Install WSL2 distro
+ timeout-minutes: 3
+ run: |
+ # FIXME: At least one distro has to be installed here,
+ # otherwise `wsl --list --verbose` (called from Lima) fails:
+ # https://github.com/lima-vm/lima/pull/1826#issuecomment-1729993334
+ # The distro image itself is not consumed by Lima.
+ # ------------------------------------------------------------------
+ # Ubuntu-22.04: gets stuck in some infinite loop during adduser
+ # OracleLinux_9_1: almostly silently fails, and just prints "Usage: adduser [options] LOGIN"
+ wsl --install -d openSUSE-Leap-15.5
+ wsl --list --verbose
+ - name: Set gitconfig
+ run: |
+ git config --global core.autocrlf false
+ git config --global core.eol lf
  - uses: actions/checkout@v4
  with:
  fetch-depth: 1
@@ -86,6 +112,19 @@ jobs:
  run: go test -v ./...
  - name: Make
  run: make
+ - name: Smoke test
+ # Make sure the path is set properly and then run limactl
+ run: |
+ $env:Path = 'C:\Program Files\Git\usr\bin;' + $env:Path
+ Set-ItemProperty -Path 'Registry::HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Environment' -Name PATH -Value $env:Path
+ .\_output\bin\limactl.exe start template://experimental/wsl2
+ # TODO: run the full integration tests
+ - name: Debug
+ if: always()
+ run: type C:\Users\runneradmin\.lima\wsl2\ha.stdout.log
+ - name: Debug
+ if: always()
+ run: type C:\Users\runneradmin\.lima\wsl2\ha.stderr.log
 
  integration:
  name: Integration tests

examples/experimental/wsl2.yaml

@@ -4,9 +4,9 @@ vmType: wsl2
 
 images:
 # Source: https://github.com/runfinch/finch-core/blob/main/Dockerfile
-- location: "https://deps.runfinch.com/common/x86-64/finch-rootfs-production-amd64-1690920103.tar.zst"
+- location: "https://deps.runfinch.com/common/x86-64/finch-rootfs-production-amd64-1694791577.tar.gz"
  arch: "x86_64"
- digest: "sha256:53f2e329b8da0f6a25e025d1f6cc262ae228402ba615ad095739b2f0ec6babc9"
+ digest: "sha256:2d4d2e7386450899c6d0587fd0db21afadb31d974fa744aa9365c883935c5341"
 
 mountType: wsl2
 

pkg/hostagent/hostagent.go

@@ -316,9 +316,11 @@ func (a *HostAgent) Run(ctx context.Context) error {
 return err
 }
 
+logrus.Infof("After driver.Start()")
 // WSL instance SSH address isn't known until after VM start
 if *a.y.VMType == limayaml.WSL2 {
 sshAddr, err := store.GetSSHAddress(a.instName)
+logrus.Infof("sshAddress: %s", sshAddr)
 if err != nil {
 return err
 }
@@ -378,6 +380,7 @@ func (a *HostAgent) Run(ctx context.Context) error {
 }()
 return a.driver.RunGUI()
 }
+logrus.Infof("Running startRoutinesAndWait")
 return a.startRoutinesAndWait(ctx, errCh)
 }
 

pkg/store/instance_windows.go

@@ -35,14 +35,33 @@ func inspectStatus(instDir string, inst *Instance, y *limayaml.LimaYAML) {
 }
 }
 
-// GetWslStatus runs `wsl --list --verbose` and parses its output
+// GetWslStatus runs `wsl --list --verbose` and parses its output.
+// There are several possible outputs, all listed with their whitespace preserved output below.
 //
-// Expected output (whitespace preserved):
+// (1) Expected output if at least one distro is installed:
 // PS > wsl --list --verbose
 //
 //	NAME STATE VERSION
 //
 // * Ubuntu Stopped 2
+//
+// (2) Expected output when no distros are installed, but WSL is configured properly:
+// PS > wsl --list --verbose
+// Windows Subsystem for Linux has no installed distributions.
+//
+// Use 'wsl.exe --list --online' to list available distributions
+// and 'wsl.exe --install <Distro>' to install.
+//
+// Distributions can also be installed by visiting the Microsoft Store:
+// https://aka.ms/wslstore
+// Error code: Wsl/WSL_E_DEFAULT_DISTRO_NOT_FOUND
+//
+// (3) Expected output when no distros are installed, and WSL2 has no kernel installed:
+//
+// PS > wsl --list --verbose
+// Windows Subsystem for Linux has no installed distributions.
+// Distributions can be installed by visiting the Microsoft Store:
+// https://aka.ms/wslstore
 func GetWslStatus(instName string) (string, error) {
 distroName := "lima-" + instName
 out, err := executil.RunUTF16leCommand([]string{
@@ -51,11 +70,25 @@ func GetWslStatus(instName string) (string, error) {
 "--verbose",
 })
 if err != nil {
-return "", fmt.Errorf("failed to run `wsl --list --verbose`, err: %w", err)
+return "", fmt.Errorf("failed to run `wsl --list --verbose`, err: %w (out=%q)", err, string(out))
 }
 
 if len(out) == 0 {
-return StatusBroken, fmt.Errorf("failed to read instance state for instance %s, try running `wsl --list --verbose` to debug, err: %w", instName, err)
+return StatusBroken, fmt.Errorf("failed to read instance state for instance %q, try running `wsl --list --verbose` to debug, err: %w", instName, err)
+}
+
+// Check for edge cases first
+outString := string(out)
+if strings.Contains(outString, "Windows Subsystem for Linux has no installed distributions.") {
+if strings.Contains(outString, "Wsl/WSL_E_DEFAULT_DISTRO_NOT_FOUND") {
+return StatusBroken, fmt.Errorf(
+"failed to read instance state for instance %q because no distro is installed,"+
+"try running `wsl --install -d Ubuntu` and then re-running Lima", instName)
+}
+return StatusBroken, fmt.Errorf(
+"failed to read instance state for instance %q because there is no WSL kernel installed,"+
+"this usually happens when WSL was installed for another user, but never for your user."+
+"Try running `wsl --install -d Ubuntu` and `wsl --update`, and then re-running Lima", instName)
 }
 
 var instState string
@@ -94,7 +127,7 @@ func getWslSSHAddress(instName string) (string, error) {
 cmd := exec.Command("wsl.exe", "-d", distroName, "bash", "-c", `hostname -I | cut -d ' ' -f1`)
 out, err := cmd.CombinedOutput()
 if err != nil {
-return "", fmt.Errorf("failed to get hostname for instance %s, err: %w", instName, err)
+return "", fmt.Errorf("failed to get hostname for instance %q, err: %w (out=%q)", instName, err, string(out))
 }
 
 return strings.TrimSpace(string(out)), nil

pkg/wsl2/lima-init.TEMPLATE

@@ -1,5 +1,6 @@
-set -eu; \
-export LOG_FILE=/var/log/lima-init.log; \
-exec > >(tee \$LOG_FILE) 2>&1; \
-export LIMA_CIDATA_MNT="$(/usr/bin/wslpath '{{.CIDataPath}}')"; \
-exec "\$LIMA_CIDATA_MNT/boot.sh";
+#!/bin/bash
+set -eu
+export LOG_FILE=/var/log/lima-init.log
+exec > >(tee $LOG_FILE) 2>&1
+export LIMA_CIDATA_MNT="$(/usr/bin/wslpath '{{.CIDataPath}}')"
+exec "$LIMA_CIDATA_MNT/boot.sh"

pkg/wsl2/vm_windows.go

@@ -7,8 +7,10 @@ import (
 "context"
 _ "embed"
 "fmt"
+"os"
 "os/exec"
 "path/filepath"
+"strconv"
 "strings"
 
 "github.com/lima-vm/lima/pkg/executil"
@@ -20,13 +22,14 @@ import (
 
 // startVM calls WSL to start a VM.
 func startVM(ctx context.Context, distroName string) error {
-_, err := executil.RunUTF16leCommand([]string{
+out, err := executil.RunUTF16leCommand([]string{
 "wsl.exe",
 "--distribution",
 distroName,
 }, executil.WithContext(&ctx))
 if err != nil {
-return err
+return fmt.Errorf("failed to run `wsl.exe --distribution %s`: %w (out=%q)",
+distroName, err, string(out))
 }
 return nil
 }
@@ -35,28 +38,30 @@ func startVM(ctx context.Context, distroName string) error {
 func initVM(ctx context.Context, instanceDir, distroName string) error {
 baseDisk := filepath.Join(instanceDir, filenames.BaseDisk)
 logrus.Infof("Importing distro from %q to %q", baseDisk, instanceDir)
-_, err := executil.RunUTF16leCommand([]string{
+out, err := executil.RunUTF16leCommand([]string{
 "wsl.exe",
 "--import",
 distroName,
 instanceDir,
 baseDisk,
 }, executil.WithContext(&ctx))
 if err != nil {
-return err
+return fmt.Errorf("failed to run `wsl.exe --import %s %s %s`: %w (out=%q)",
+distroName, instanceDir, baseDisk, err, string(out))
 }
 return nil
 }
 
 // stopVM calls WSL to stop a running VM.
 func stopVM(ctx context.Context, distroName string) error {
-_, err := executil.RunUTF16leCommand([]string{
+out, err := executil.RunUTF16leCommand([]string{
 "wsl.exe",
 "--terminate",
 distroName,
 }, executil.WithContext(&ctx))
 if err != nil {
-return err
+return fmt.Errorf("failed to run `wsl.exe --terminate %s`: %w (out=%q)",
+distroName, err, string(out))
 }
 return nil
 }
@@ -70,12 +75,39 @@ func provisionVM(ctx context.Context, instanceDir, instanceName, distroName stri
 m := map[string]string{
 "CIDataPath": ciDataPath,
 }
-out, err := textutil.ExecuteTemplate(limaBoot, m)
+limaBootB, err := textutil.ExecuteTemplate(limaBoot, m)
 if err != nil {
 return fmt.Errorf("failed to construct wsl boot.sh script: %w", err)
 }
-outString := strings.Replace(string(out), `\r\n`, `\n`, -1)
-
+limaBootFile, err := os.CreateTemp("", "lima-wsl2-boot-*.sh")
+if err != nil {
+return err
+}
+if _, err = limaBootFile.Write(limaBootB); err != nil {
+return err
+}
+limaBootFileWinPath := limaBootFile.Name()
+if err = limaBootFile.Close(); err != nil {
+return err
+}
+// path should be quoted and use \\ as separator
+bootFileWSLPath := strconv.Quote(limaBootFileWinPath)
+limaBootFilePathOnLinuxB, err := exec.Command(
+"wsl.exe",
+"-d",
+distroName,
+"bash",
+"-c",
+fmt.Sprintf("wslpath -u %s", bootFileWSLPath),
+bootFileWSLPath,
+).Output()
+if err != nil {
+os.RemoveAll(limaBootFileWinPath)
+// this can return an error with an exit code, which causes it not to be logged
+// because main.handleExitCoder() traps it, so wrap the error
+return fmt.Errorf("failed to run wslpath command: %w", err)
+}
+limaBootFileLinuxPath := strings.TrimSpace(string(limaBootFilePathOnLinuxB))
 go func() {
 cmd := exec.CommandContext(
 ctx,
@@ -84,12 +116,15 @@ func provisionVM(ctx context.Context, instanceDir, instanceName, distroName stri
 distroName,
 "bash",
 "-c",
-outString,
+limaBootFileLinuxPath,
 )
-if _, err := cmd.CombinedOutput(); err != nil {
+out, err := cmd.CombinedOutput()
+os.RemoveAll(limaBootFileWinPath)
+logrus.Debugf("%v: %q", cmd.Args, string(out))
+if err != nil {
 *errCh <- fmt.Errorf(
-"error running wslCommand that executes boot.sh: %w, "+
-"check /var/log/lima-init.log for more details", err)
+"error running wslCommand that executes boot.sh (%v): %w, "+
+"check /var/log/lima-init.log for more details (out=%q)", cmd.Args, err, string(out))
 }
 
 for {
@@ -130,13 +165,14 @@ func keepAlive(ctx context.Context, distroName string, errCh *chan error) {
 // unregisterVM calls WSL to unregister a VM.
 func unregisterVM(ctx context.Context, distroName string) error {
 logrus.Info("Unregistering WSL2 VM")
-_, err := executil.RunUTF16leCommand([]string{
+out, err := executil.RunUTF16leCommand([]string{
 "wsl.exe",
 "--unregister",
 distroName,
 }, executil.WithContext(&ctx))
 if err != nil {
-return err
+return fmt.Errorf("failed to run `wsl.exe --unregister %s`: %w (out=%q)",
+distroName, err, string(out))
 }
 return nil
 }