fix(actions): reject jobs without runs-on (#39480)

Align job and `runs-on` validation with github.com, as implemented by
the parser in https://github.com/actions/runner. A job without `runs-on`
could be claimed by any runner, so a job meant for a container could run
on the host.

- Jobs without `runs-on` fail with `Required property is missing:
runs-on`, called workflows included
- Unknown job keys and callers (`uses:`) mixed with steps-only keys like
`runs-on` are rejected
- Empty, null and nested `runs-on` values are rejected
- A `runs-on` evaluating to such a value fails only that job
- Called workflows are validated at run creation, an invalid one fails
the run as an invalid workflow file
- Zero labels (`runs-on: []` or `{}`) never match a runner, including
jobs queued before upgrading

<img width="960" alt="image"
src="https://github.com/user-attachments/assets/e746ce5a-b711-4e8b-aab8-81336ff53d86"
/>

**Behavior Change:** workflows that omit `runs-on`, use unknown job keys
or mix `uses` with `runs-on` stop running until fixed.

---------

Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>

modules/actions/jobparser/jobparser.go

@@ -225,7 +225,6 @@ func replaceScalars(node *yaml.Node, replace func(string) string) {
 // buildMatrixCombos builds one Job per matrix combination from src, baking the combination into the
 // strategy and interpolating the name, runs-on and continue-on-error with it.
 func buildMatrixCombos(jobID string, src *Job, matrixes []map[string]any, gitCtx *model.GithubContext, results map[string]*JobResult, vars map[string]string, inputs map[string]any) ([]*Job, error) {
-	srcRunsOn := model.RunsOnFromNode(src.RawRunsOn)
 	order, names := make([]int, len(matrixes)), make([]string, len(matrixes))
 	for index, matrix := range matrixes {
 		order[index], names[index] = index, matrixName(matrix)
@@ -259,14 +258,15 @@ func buildMatrixCombos(jobID string, src *Job, matrixes []map[string]any, gitCtx
 			if err := evaluator.EvaluateYamlNode(&rawRunsOn); err != nil {
 				return nil, fmt.Errorf("interpolate runs-on for job %q: %w", jobID, err)
 			}
-			runsOn := model.RunsOnFromNode(rawRunsOn)
-			if len(runsOn) == 0 && len(srcRunsOn) > 0 { // match no runner rather than every runner
-				runsOn = []string{""}
+			if rawRunsOn.Kind != 0 && runsOnProblem(&rawRunsOn) != "" {
+				combo.RawRunsOn = rawRunsOn
+			} else {
+				runsOn := model.RunsOnFromNode(rawRunsOn)
+				for i := range runsOn {
+					runsOn[i] = escapeExpressions(runsOn[i])
+				}
+				combo.RawRunsOn = model.RunsOnNode(runsOn, "")
 			}
-			for i := range runsOn {
-				runsOn[i] = escapeExpressions(runsOn[i])
-			}
-			combo.RawRunsOn = model.RunsOnNode(runsOn, "")
 		}
 		if err := evaluator.EvaluateYamlNode(&combo.RawContinueOnError); err != nil {
 			return nil, fmt.Errorf("evaluate continue-on-error for job %q: %w", jobID, err)

modules/actions/jobparser/jobparser_test.go

@@ -290,17 +290,26 @@ func TestParseInterpolatesRunName(t *testing.T) {
 	assert.Empty(t, result[0].RunName)
 }
 
-func TestParseRunsOnFromJSONArray(t *testing.T) {
+func TestParseRunsOnFromJSONKeepsWhatGitHubRejectsForTheJobToFail(t *testing.T) {
 	content := []byte("on: push\njobs:\n  build:\n    runs-on: ${{ fromJSON(vars.RUNNER) }}\n    steps: [{run: echo}]\n")
 	_, err := Parse(content)
 	require.NoError(t, err)
-	for runner, want := range map[string][]string{`["self-hosted", "linux"]`: {"self-hosted", "linux"}, "[]": {""}} {
+	for runner, want := range map[string][]string{`["self-hosted", "linux"]`: {"self-hosted", "linux"}, "[]": {}, "{}": {}} {
 		result, err := Parse(content, WithGitContext(&model.GithubContext{}), WithVars(map[string]string{"RUNNER": runner}))
 		require.NoError(t, err)
 		require.Len(t, result, 1)
 		_, job := result[0].Job()
 		assert.Equal(t, want, job.RunsOn(), runner)
 	}
+	for runner, problem := range map[string]string{`["a"]`: "", `""`: "Unexpected value ''", `[["a"]]`: "A sequence was not expected"} {
+		result, err := Parse(content, WithGitContext(&model.GithubContext{}), WithVars(map[string]string{"RUNNER": runner}))
+		require.NoError(t, err)
+		payload, err := result[0].Marshal()
+		require.NoError(t, err)
+		_, job, err := ParseRawSingleWorkflow(payload)
+		require.NoError(t, err)
+		assert.Equal(t, problem, job.RunsOnProblem(), runner)
+	}
 }
 
 func TestJobFieldsWithoutMatrix(t *testing.T) {
@@ -458,6 +467,37 @@ func TestReadWorkflowJobConditionContexts(t *testing.T) {
 	}
 }
 
+func TestValidateWorkflowStaticJobKindAndRunsOnLikeGitHub(t *testing.T) {
+	for job, want := range map[string]string{
+		"{runs-on: x, steps: [{run: echo}]}":      "",
+		"{uses: o/r/.gitea/workflows/c.yml@main}": "",
+		"{runs-on: []}": "",
+		"{runs-on: {}}": "",
+		"{runs-on: {group: org/g, labels: [a, 1]}}":           "",
+		"{runs-on: {group: '${{ vars.G }}'}}":                 "",
+		"{steps: [{run: echo}]}":                              "Required property is missing: runs-on",
+		"{with: {}}":                                          "Required property is missing: uses",
+		"{runs-on: x, uses: o/r/.gitea/workflows/c.yml@main}": "Unexpected value 'uses'",
+		"{Runs-On: x}":                                        "Unexpected value 'Runs-On'",
+		"{runs-on: ~}":                                        "runs-on: Unexpected value ''",
+		"{runs-on: ['']}":                                     "runs-on: Unexpected value ''",
+		"{runs-on: [[a]]}":                                    "runs-on: A sequence was not expected",
+		"{runs-on: {labels: {a: b}}}":                         "runs-on: A mapping was not expected",
+		"{runs-on: {foo: x}}":                                 "runs-on: Unexpected value 'foo'",
+		"{runs-on: {group: org/}}":                            "runs-on: Invalid runs-on group name 'org/'.",
+		"{runs-on: {group: a/b/c}}":                           "runs-on: Invalid runs-on group name 'a/b/c'. Please use 'organization/' or 'enterprise/' prefix to target a single runner group.",
+		"{if: true}": "There's not enough info to determine what you meant. Add one of these properties: " +
+			"cancel-timeout-minutes, container, continue-on-error, defaults, env, environment, outputs, runs-on, secrets, services, snapshot, steps, timeout-minutes, uses, with",
+	} {
+		_, err := ValidateWorkflowStatic([]byte("on: push\njobs:\n  build: " + job + "\n"))
+		if want == "" {
+			assert.NoError(t, err, job)
+		} else {
+			assert.EqualError(t, err, "job build: "+want, job)
+		}
+	}
+}
+
 func TestRejectsUnevaluatedMatrixFilters(t *testing.T) {
 	for _, filter := range []string{"include", "exclude"} {
 		t.Run(filter, func(t *testing.T) {

modules/actions/jobparser/model.go

@@ -174,6 +174,14 @@ func (j *Job) EraseNeeds() *Job {
 	return j
 }
 
+// RunsOnProblem returns github.com's error for the job's runs-on, "" if valid.
+func (j *Job) RunsOnProblem() string {
+	if j.RawRunsOn.Kind == 0 {
+		return ""
+	}
+	return runsOnProblem(&j.RawRunsOn)
+}
+
 // RunsOn returns the labels Gitea matches runners against, unescaped like DisplayName.
 func (j *Job) RunsOn() []string {
 	runsOn := model.RunsOnFromNode(j.RawRunsOn)

modules/actions/jobparser/validate.go

@@ -7,10 +7,13 @@ import (
 	"errors"
 	"fmt"
 	"slices"
+	"strings"
 
 	"gitea.dev/actionslib/pkg/expreval"
 	"gitea.dev/actionslib/pkg/exprparser"
 	"gitea.dev/actionslib/pkg/model"
+
+	"go.yaml.in/yaml/v4"
 )
 
 // jobConditionContexts are what github.com gives `jobs.<job_id>.if`, which it decides before the matrix, plus the `gitea` alias.
@@ -21,7 +24,7 @@ func ValidateWorkflowStatic(content []byte) ([]*Event, error) {
 	if err != nil {
 		return nil, err
 	}
-	// Keep unknown and case-distinct keys accepted for existing Gitea workflows.
+	// Keep unknown and case-distinct keys outside of jobs accepted for existing Gitea workflows.
 	workflow, err := readWorkflowDoc(doc)
 	if err != nil {
 		return nil, err
@@ -33,6 +36,9 @@ func ValidateWorkflowStatic(content []byte) ([]*Event, error) {
 	if err := validateWorkflowStructure(workflow); err != nil {
 		return nil, err
 	}
+	if err := validateJobKinds(doc); err != nil {
+		return nil, err
+	}
 	var header struct {
 		RunName string `yaml:"run-name"`
 	}
@@ -76,7 +82,6 @@ func validateWorkflowStructure(workflow *model.Workflow) error {
 		if job == nil {
 			return fmt.Errorf("job %q has no configuration", id)
 		}
-		// a job without runs-on is accepted and runs on any runner, github.com rejects it
 		for _, dependency := range job.Needs() {
 			if _, ok := workflow.Jobs[dependency]; !ok {
 				return fmt.Errorf("job %q needs unknown job %q", id, dependency)
@@ -110,3 +115,116 @@ func validateWorkflowStructure(workflow *model.Workflow) error {
 	}
 	return nil
 }
+
+// job keys of github.com's workflow schema, by the kind of job allowing them
+var (
+	stepsJobKeys  = []string{"cancel-timeout-minutes", "container", "continue-on-error", "defaults", "env", "environment", "outputs", "runs-on", "services", "snapshot", "steps", "timeout-minutes"}
+	callerJobKeys = []string{"secrets", "uses", "with"}
+	sharedJobKeys = []string{"concurrency", "if", "name", "needs", "permissions", "strategy"}
+)
+
+// validateJobKinds applies github.com's job kinds, decided by the first kind-specific key.
+func validateJobKinds(doc *yaml.Node) error {
+	jobs := mappingValue(doc.Content[0], "jobs")
+	for i := 0; i+1 < len(jobs.Content); i += 2 {
+		id, job := jobs.Content[i].Value, jobs.Content[i+1]
+		var required string
+		for j := 0; j+1 < len(job.Content); j += 2 {
+			key := job.Content[j].Value
+			isStepsKey, isCallerKey := slices.Contains(stepsJobKeys, key), slices.Contains(callerJobKeys, key)
+			switch {
+			case required == "runs-on" && isCallerKey, required == "uses" && isStepsKey, !isStepsKey && !isCallerKey && !slices.Contains(sharedJobKeys, key):
+				return fmt.Errorf("job %s: Unexpected value '%s'", id, key)
+			case required == "" && isStepsKey:
+				required = "runs-on"
+			case required == "" && isCallerKey:
+				required = "uses"
+			}
+		}
+		if required == "" {
+			keys := slices.Concat(stepsJobKeys, callerJobKeys)
+			slices.Sort(keys)
+			return fmt.Errorf("job %s: There's not enough info to determine what you meant. Add one of these properties: %s", id, strings.Join(keys, ", "))
+		}
+		value := mappingValue(job, required)
+		if value == nil {
+			return fmt.Errorf("job %s: Required property is missing: %s", id, required)
+		}
+		if required != "runs-on" {
+			continue
+		}
+		if problem := runsOnProblem(value); problem != "" {
+			return fmt.Errorf("job %s: runs-on: %s", id, problem)
+		}
+	}
+	return nil
+}
+
+// runsOnProblem returns github.com's schema error for a runs-on, "" if valid.
+func runsOnProblem(node *yaml.Node) string {
+	if node.Kind != yaml.MappingNode {
+		return runsOnLabelsProblem(node)
+	}
+	for i := 0; i+1 < len(node.Content); i += 2 {
+		var problem string
+		switch key := node.Content[i].Value; key {
+		case "labels":
+			problem = runsOnLabelsProblem(node.Content[i+1])
+		case "group":
+			problem = runsOnGroupProblem(node.Content[i+1])
+		default:
+			problem = fmt.Sprintf("Unexpected value '%s'", key)
+		}
+		if problem != "" {
+			return problem
+		}
+	}
+	return ""
+}
+
+func runsOnLabelsProblem(node *yaml.Node) string {
+	if node.Kind != yaml.SequenceNode {
+		return nonEmptyStringProblem(node)
+	}
+	for _, label := range node.Content {
+		if problem := nonEmptyStringProblem(label); problem != "" {
+			return problem
+		}
+	}
+	return ""
+}
+
+func runsOnGroupProblem(node *yaml.Node) string {
+	if problem := nonEmptyStringProblem(node); problem != "" || hasExpression(node.Value) {
+		return problem
+	}
+	switch prefix, name, found := strings.Cut(node.Value, "/"); {
+	case found && name == "":
+		return fmt.Sprintf("Invalid runs-on group name '%s'.", node.Value)
+	case found && (strings.Contains(name, "/") || !slices.Contains([]string{"org", "organization", "ent", "enterprise"}, prefix)):
+		return fmt.Sprintf("Invalid runs-on group name '%s'. Please use 'organization/' or 'enterprise/' prefix to target a single runner group.", node.Value)
+	}
+	return ""
+}
+
+// nonEmptyStringProblem mirrors github.com's non-empty-string, which also accepts non-string scalars.
+func nonEmptyStringProblem(node *yaml.Node) string {
+	switch {
+	case node.Kind == yaml.SequenceNode:
+		return "A sequence was not expected"
+	case node.Kind == yaml.MappingNode:
+		return "A mapping was not expected"
+	case node.Value == "" || node.ShortTag() == "!!null":
+		return "Unexpected value ''"
+	}
+	return ""
+}
+
+func mappingValue(node *yaml.Node, key string) *yaml.Node {
+	for i := 0; i+1 < len(node.Content); i += 2 {
+		if node.Content[i].Value == key {
+			return node.Content[i+1]
+		}
+	}
+	return nil
+}

modules/actions/workflows_test.go

@@ -30,18 +30,23 @@ jobs:
 
 func TestReadWorkflowEventsStaticErrors(t *testing.T) {
 	for content, static := range map[string]bool{
-		"on: push\njobs: {}":                                                                  true,
-		"on: push\njobs: {test: {needs: absent}}":                                             true,
-		"on: push\njobs: {one: {needs: two}, two: {needs: one}}":                              true,
-		"on: push\njobs: {test: {strategy: {matrix: {os: []}}}}":                              true,
-		"on: push\nrun-name: ${{ secrets.TOKEN }}\njobs: {test: {}}":                          true,
-		"on: push\nrun-name: ${{ fromJSON(inputs.x) }}\njobs: {test: {steps: [{run: echo}]}}": false,
+		"on: push\njobs: {}": true,
+		"on: push\njobs: {test: {runs-on: x, needs: absent}}":                                             true,
+		"on: push\njobs: {one: {runs-on: x, needs: two}, two: {runs-on: x, needs: one}}":                  true,
+		"on: push\njobs: {test: {runs-on: x, strategy: {matrix: {os: []}}}}":                              true,
+		"on: push\nrun-name: ${{ secrets.TOKEN }}\njobs: {test: {runs-on: x}}":                            true,
+		"on: push\njobs: {test: {steps: [{run: echo}]}}":                                                  true,
+		"on: push\nrun-name: ${{ fromJSON(inputs.x) }}\njobs: {test: {runs-on: x, steps: [{run: echo}]}}": false,
 	} {
 		_, gotStatic, err := readWorkflowEvents([]byte(content))
 		require.Error(t, err, content)
 		assert.Equal(t, static, gotStatic, content)
 	}
-	for _, content := range []string{"on: push\njobs: {test: {steps: [{run: echo}]}}", "on: push\nrun-name: ${{ github.ref }}\njobs: {test: {}}"} {
+	for _, content := range []string{
+		"on: push\njobs: {test: {runs-on: x, steps: [{run: echo}]}}",
+		"on: push\nrun-name: ${{ github.ref }}\njobs: {test: {runs-on: x}}",
+		"on: push\njobs: {call: {uses: ./.gitea/workflows/called.yml}}",
+	} {
 		_, _, err := readWorkflowEvents([]byte(content))
 		assert.NoError(t, err, content)
 	}