tests: cover debug build missing mount result IDs

build/result.go

@@ -220,16 +220,24 @@ func containerConfigFromError(solveErr *errdefs.SolveError, cfg *InvokeConfig) (
 	if err != nil {
 		return nil, err
 	}
+	// When a solve fails before its inputs/mounts have been resolved (for
+	// example, when computing the cache key for an exec op fails because an
+	// input source cannot be found), MountIDs/InputIDs may be empty or shorter
+	// than the declared mount list. Guard against that so callers (such as the
+	// debug adapter) get a clean error instead of an index-out-of-range panic.
+	ids := solveErr.MountIDs
+	if cfg.Initial {
+		ids = solveErr.InputIDs
+	}
+	if len(ids) < len(exec.Mounts) {
+		return nil, errors.Errorf("cannot start debug container: failing step has no mount results available (got %d of %d)", len(ids), len(exec.Mounts))
+	}
 	var mounts []gateway.Mount
 	for i, mnt := range exec.Mounts {
-		rid := solveErr.MountIDs[i]
-		if cfg.Initial {
-			rid = solveErr.InputIDs[i]
-		}
 		mounts = append(mounts, gateway.Mount{
 			Selector:  mnt.Selector,
 			Dest:      mnt.Dest,
-			ResultID:  rid,
+			ResultID:  ids[i],
 			Readonly:  mnt.Readonly,
 			MountType: mnt.MountType,
 			CacheOpt:  mnt.CacheOpt,

build/result_test.go

@@ -0,0 +1,132 @@
+package build_test
+
+import (
+	"context"
+	"testing"
+
+	"github.com/docker/buildx/build"
+	gateway "github.com/moby/buildkit/frontend/gateway/client"
+	"github.com/moby/buildkit/solver/errdefs"
+	"github.com/moby/buildkit/solver/pb"
+	"github.com/stretchr/testify/require"
+)
+
+func Test_Creating_new_container_from_failing_result_handle(t *testing.T) {
+	t.Parallel()
+
+	mounts := []*pb.Mount{
+		{Dest: "/", Input: 0},
+		{Dest: "/src", Input: 1},
+	}
+
+	run := func(t *testing.T, cfg build.InvokeConfig, inputIDs, mountIDs []string) (*gateway.NewContainerRequest, error) {
+		t.Helper()
+		ctx := t.Context()
+		gw := &stubGwClient{}
+		solveErr := &errdefs.SolveError{
+			Solve: &errdefs.Solve{
+				Op:       &pb.Op{Op: &pb.Op_Exec{Exec: &pb.ExecOp{Mounts: mounts}}},
+				InputIDs: inputIDs,
+				MountIDs: mountIDs,
+			},
+		}
+		rh := build.NewResultHandle(ctx, gw, nil, nil, solveErr)
+		require.NotNil(t, rh, "NewResultHandle must accept a *errdefs.SolveError")
+		defer rh.Done()
+
+		var err error
+		require.NotPanics(t, func() {
+			_, err = rh.NewContainer(ctx, &cfg)
+		}, "ResultHandle.NewContainer must not panic")
+		return gw.captured, err
+	}
+
+	t.Run("returns_an_error_when", func(t *testing.T) {
+		t.Parallel()
+
+		cases := []struct {
+			name     string
+			inputIDs []string
+			mountIDs []string
+			cfg      build.InvokeConfig
+		}{
+			{
+				name:     "mount_ids_are_missing",
+				mountIDs: nil,
+				cfg:      build.InvokeConfig{},
+			},
+			{
+				name:     "input_ids_are_missing_and_initial_is_true",
+				inputIDs: nil,
+				cfg:      build.InvokeConfig{Initial: true},
+			},
+			{
+				name:     "mount_ids_are_shorter_than_declared_mounts",
+				inputIDs: []string{"input-0"},
+				mountIDs: []string{"mount-0"},
+				cfg:      build.InvokeConfig{},
+			},
+			{
+				name:     "input_ids_are_shorter_than_declared_mounts_and_initial_is_true",
+				inputIDs: []string{"input-0"},
+				mountIDs: []string{"mount-0"},
+				cfg:      build.InvokeConfig{Initial: true},
+			},
+		}
+		for _, tc := range cases {
+			t.Run(tc.name, func(t *testing.T) {
+				t.Parallel()
+				_, err := run(t, tc.cfg, tc.inputIDs, tc.mountIDs)
+				require.Error(t, err)
+			})
+		}
+	})
+
+	t.Run("forwards_per_mount_result_ids_to_the_gateway", func(t *testing.T) {
+		t.Parallel()
+
+		cases := []struct {
+			name          string
+			cfg           build.InvokeConfig
+			wantResultIDs []string
+		}{
+			{
+				name:          "from_mount_ids",
+				cfg:           build.InvokeConfig{},
+				wantResultIDs: []string{"mount-0", "mount-1"},
+			},
+			{
+				name:          "from_input_ids_when_initial_is_true",
+				cfg:           build.InvokeConfig{Initial: true},
+				wantResultIDs: []string{"input-0", "input-1"},
+			},
+		}
+		for _, tc := range cases {
+			t.Run(tc.name, func(t *testing.T) {
+				t.Parallel()
+				captured, err := run(t, tc.cfg,
+					[]string{"input-0", "input-1"},
+					[]string{"mount-0", "mount-1"},
+				)
+				require.NoError(t, err)
+				require.NotNil(t, captured, "gateway.NewContainer should have been called")
+				require.Len(t, captured.Mounts, len(tc.wantResultIDs))
+				for i, want := range tc.wantResultIDs {
+					require.Equal(t, want, captured.Mounts[i].ResultID, "mount %d ResultID", i)
+					require.Equal(t, mounts[i].Dest, captured.Mounts[i].Dest, "mount %d Dest", i)
+				}
+			})
+		}
+	})
+}
+
+type stubGwClient struct {
+	gateway.Client
+	captured *gateway.NewContainerRequest
+}
+
+func (s *stubGwClient) NewContainer(_ context.Context, req gateway.NewContainerRequest) (gateway.Container, error) {
+	s.captured = &req
+
+	return nil, nil
+}

tests/debug.go

@@ -0,0 +1,131 @@
+package tests
+
+import (
+	"bytes"
+	"io"
+	"os/exec"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/containerd/continuity/fs/fstest"
+	"github.com/creack/pty"
+	"github.com/moby/buildkit/util/testutil/integration"
+	"github.com/stretchr/testify/require"
+)
+
+var debugTests = []func(t *testing.T, sb integration.Sandbox){
+	testDebugBuildMissingBindMountSource,
+}
+
+func testDebugBuildMissingBindMountSource(t *testing.T, sb integration.Sandbox) {
+	if !isExperimental() {
+		t.Skip("debug command is experimental")
+	}
+	if !isDockerWorker(sb) {
+		t.Skip("debug monitor test only needs a docker worker")
+	}
+
+	dockerfile := []byte(`
+FROM busybox:latest
+RUN --mount=type=bind,source=missing,target=/src true
+`)
+	dir := tmpdir(t,
+		fstest.CreateFile("Dockerfile", dockerfile, 0o600),
+	)
+
+	cmd := buildxCmd(sb, withArgs("debug", "--on=error", "build", "--progress=plain", "--output=type=cacheonly", dir))
+	f, err := pty.StartWithSize(cmd, &pty.Winsize{
+		Cols: 120,
+		Rows: 24,
+	})
+	require.NoError(t, err)
+	defer f.Close()
+
+	var output debugOutput
+	copyDone := make(chan struct{})
+	go func() {
+		_, _ = io.Copy(&output, f)
+		close(copyDone)
+	}()
+
+	waitCh := make(chan error, 1)
+	go func() {
+		waitCh <- cmd.Wait()
+	}()
+
+	exited, waitErr := waitForDebugPromptOrExit(t, &output, waitCh)
+	if !exited {
+		_, err = f.Write([]byte("exit\r"))
+		require.NoError(t, err)
+		waitErr = waitForDebugCommandExit(t, cmd, waitCh, &output)
+	}
+
+	_ = f.Close()
+	select {
+	case <-copyDone:
+	case <-time.After(time.Second):
+	}
+
+	out := output.String()
+	require.Error(t, waitErr, out)
+	require.Contains(t, out, "failed to calculate checksum")
+	require.Contains(t, out, "failed to create container")
+	require.NotContains(t, out, "panic:")
+	require.NotContains(t, out, "index out of range")
+}
+
+func waitForDebugPromptOrExit(t *testing.T, output *debugOutput, waitCh <-chan error) (bool, error) {
+	t.Helper()
+
+	timeout := time.NewTimer(time.Minute)
+	defer timeout.Stop()
+	ticker := time.NewTicker(100 * time.Millisecond)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case err := <-waitCh:
+			return true, err
+		case <-ticker.C:
+			if strings.Contains(output.String(), "(buildx) ") {
+				return false, nil
+			}
+		case <-timeout.C:
+			require.FailNow(t, "timeout waiting for debug monitor", output.String())
+		}
+	}
+}
+
+func waitForDebugCommandExit(t *testing.T, cmd *exec.Cmd, waitCh <-chan error, output *debugOutput) error {
+	t.Helper()
+
+	select {
+	case err := <-waitCh:
+		return err
+	case <-time.After(30 * time.Second):
+		if cmd.Process != nil {
+			_ = cmd.Process.Kill()
+		}
+		require.FailNow(t, "timeout waiting for debug command to exit", output.String())
+		return nil
+	}
+}
+
+type debugOutput struct {
+	mu  sync.Mutex
+	buf bytes.Buffer
+}
+
+func (b *debugOutput) Write(p []byte) (int, error) {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+	return b.buf.Write(p)
+}
+
+func (b *debugOutput) String() string {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+	return b.buf.String()
+}

tests/integration_test.go

@@ -24,6 +24,7 @@ func TestIntegration(t *testing.T) {
 	var tests []func(t *testing.T, sb integration.Sandbox)
 	tests = append(tests, commonTests...)
 	tests = append(tests, buildTests...)
+	tests = append(tests, debugTests...)
 	tests = append(tests, policyBuildTests...)
 	tests = append(tests, policyEvalTests...)
 	tests = append(tests, policyBakeTests...)