Merge 64d415bcc5 into ba5c4ac5e3

.changes/v1.15/BUG FIXES-20260122-061500.yaml

@@ -0,0 +1,3 @@
+kind: BUG FIXES
+body: 'core: Fixed panic when duplicate check status reports occur during concurrent resource evaluation, particularly during resource recreation with preconditions'
+time: 2026-01-22T06:15:00.000000Z

internal/checks/state_report.go

@@ -64,8 +64,10 @@ func (c *State) ReportCheckableObjects(configAddr addrs.ConfigCheckable, objectA
 // or if the check index is out of bounds for the number of checks expected
 // of the given type, this method will panic to indicate a bug in the caller.
 //
-// This method will also panic if the specified check already had a known
-// status; each check should have its result reported only once.
+// If the specified check already had a known status, this method will log
+// a warning or error (depending on whether the status conflicts) and preserve
+// the original status. This is a defensive measure against race conditions or
+// duplicate reporting during concurrent graph walks.
 func (c *State) ReportCheckResult(objectAddr addrs.Checkable, checkType addrs.CheckRuleType, index int, status Status) {
 	c.mu.Lock()
 	defer c.mu.Unlock()
@@ -111,8 +113,18 @@ func (c *State) reportCheckResult(objectAddr addrs.Checkable, checkType addrs.Ch
 	if index >= len(checks[checkType]) {
 		panic(fmt.Sprintf("%s index %d out of range for %s", checkType, index, objectAddr))
 	}
-	if checks[checkType][index] != StatusUnknown {
-		panic(fmt.Sprintf("duplicate status report for %s %s %d", objectAddr, checkType, index))
+
+	existingStatus := checks[checkType][index]
+	if existingStatus != StatusUnknown {
+		// Duplicate report detected; preserve first-write-wins semantics.
+		if existingStatus == status {
+			// Same status reported twice, likely a benign race condition.
+			log.Printf("[WARN] checks: duplicate status report for %s %s %d", objectAddr, checkType, index)
+			return
+		}
+		// Conflicting status reported; keep the original and log for debugging.
+		log.Printf("[ERROR] checks: conflicting status report for %s %s %d: was %s, got %s", objectAddr, checkType, index, existingStatus, status)
+		return
 	}
 
 	checks[checkType][index] = status

internal/checks/state_test.go

@@ -226,3 +226,60 @@ func TestChecksHappyPath(t *testing.T) {
 		}
 	}
 }
+
+func TestChecksDuplicateReport(t *testing.T) {
+	const fixtureDir = "testdata/happypath"
+	loader, close := configload.NewLoaderForTests(t)
+	defer close()
+	inst := initwd.NewModuleInstaller(loader.ModulesDir(), loader, nil)
+	_, instDiags := inst.InstallModules(context.Background(), fixtureDir, "tests", true, false, initwd.ModuleInstallHooksImpl{})
+	if instDiags.HasErrors() {
+		t.Fatal(instDiags.Err())
+	}
+	if err := loader.RefreshModules(); err != nil {
+		t.Fatalf("failed to refresh modules after installation: %s", err)
+	}
+
+	/////////////////////////////////////////////////////////////////////////
+
+	cfg, hclDiags := loader.LoadConfig(fixtureDir)
+	if hclDiags.HasErrors() {
+		t.Fatalf("invalid configuration: %s", hclDiags.Error())
+	}
+
+	resourceA := addrs.Resource{
+		Mode: addrs.ManagedResourceMode,
+		Type: "null_resource",
+		Name: "a",
+	}.InModule(addrs.RootModule)
+
+	checks := NewState(cfg)
+
+	resourceInstA := resourceA.Resource.Absolute(addrs.RootModuleInstance).Instance(addrs.NoKey)
+
+	checks.ReportCheckableObjects(resourceA, addrs.MakeSet[addrs.Checkable](resourceInstA))
+
+	// Report initial check results for resource A (2 preconditions, 1 postcondition)
+	checks.ReportCheckResult(resourceInstA, addrs.ResourcePrecondition, 0, StatusPass)
+	checks.ReportCheckResult(resourceInstA, addrs.ResourcePrecondition, 1, StatusPass)
+	checks.ReportCheckResult(resourceInstA, addrs.ResourcePostcondition, 0, StatusPass)
+
+	if got, want := checks.ObjectCheckStatus(resourceInstA), StatusPass; got != want {
+		t.Fatalf("incorrect check status after first report: got %s, want %s", got, want)
+	}
+
+	// Reporting the same status again should not panic or change the result
+	checks.ReportCheckResult(resourceInstA, addrs.ResourcePrecondition, 0, StatusPass)
+
+	if got, want := checks.ObjectCheckStatus(resourceInstA), StatusPass; got != want {
+		t.Errorf("incorrect check status after duplicate report: got %s, want %s", got, want)
+	}
+
+	// Reporting a conflicting status should not panic, and should preserve
+	// the original status (first write wins)
+	checks.ReportCheckResult(resourceInstA, addrs.ResourcePrecondition, 0, StatusFail)
+
+	if got, want := checks.ObjectCheckStatus(resourceInstA), StatusPass; got != want {
+		t.Errorf("check status should be preserved after conflicting report: got %s, want %s", got, want)
+	}
+}