Prevent hang in git cat-file if the repository is not a valid repository (Partial #17991) (#17992)

* Prevent hang in git cat-file if the repository is not a valid repository (Partial #17991)

Unfortunately it appears that if git cat-file is run in an invalid
repository it will hang until stdin is closed. This will result in
deadlocked /pulls pages and dangling git cat-file calls if a broken
repository is tried to be reviewed or pulls exists for a broken
repository.

Signed-off-by: Andrew Thornton <art27@cantab.net>

* placate lint

Signed-off-by: Andrew Thornton <art27@cantab.net>

* fix compilation bug

Signed-off-by: Andrew Thornton <art27@cantab.net>

* Add the missing directories to the testrepos

* fixup! Add the missing directories to the testrepos

* and ensure that all of the other places have the objects directories too

Signed-off-by: Andrew Thornton <art27@cantab.net>

Co-authored-by: techknowlogick <techknowlogick@gitea.io>

integrations/integration_test.go

@@ -251,6 +251,26 @@ func prepareTestEnv(t testing.TB, skip ...int) func() {
 	assert.NoError(t, util.RemoveAll(setting.RepoRootPath))
 
 	assert.NoError(t, util.CopyDir(path.Join(filepath.Dir(setting.AppPath), "integrations/gitea-repositories-meta"), setting.RepoRootPath))
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	if err != nil {
+		assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+	}
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		if err != nil {
+			assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+		}
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
+
 	return deferFn
 }
 
@@ -529,4 +549,23 @@ func resetFixtures(t *testing.T) {
 	assert.NoError(t, models.LoadFixtures())
 	assert.NoError(t, util.RemoveAll(setting.RepoRootPath))
 	assert.NoError(t, util.CopyDir(path.Join(filepath.Dir(setting.AppPath), "integrations/gitea-repositories-meta"), setting.RepoRootPath))
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	if err != nil {
+		assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+	}
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		if err != nil {
+			assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+		}
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
 }

integrations/migration-test/migration_test.go

@@ -61,6 +61,25 @@ func initMigrationTest(t *testing.T) func() {
 	assert.True(t, len(setting.RepoRootPath) != 0)
 	assert.NoError(t, util.RemoveAll(setting.RepoRootPath))
 	assert.NoError(t, util.CopyDir(path.Join(filepath.Dir(setting.AppPath), "integrations/gitea-repositories-meta"), setting.RepoRootPath))
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	if err != nil {
+		assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+	}
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		if err != nil {
+			assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+		}
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
 
 	git.CheckLFSVersion()
 	setting.InitDBConfig()

models/migrations/migrations_test.go

@@ -205,6 +205,25 @@ func prepareTestEnv(t *testing.T, skip int, syncModels ...interface{}) (*xorm.En
 
 	assert.NoError(t, com.CopyDir(path.Join(filepath.Dir(setting.AppPath), "integrations/gitea-repositories-meta"),
 		setting.RepoRootPath))
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	if err != nil {
+		assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+	}
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		if err != nil {
+			assert.NoError(t, err, "unable to read the new repo root: %v\n", err)
+		}
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
 
 	if err := deleteDB(); err != nil {
 		t.Errorf("unable to reset database: %v", err)

models/unit_tests.go

@@ -87,6 +87,26 @@ func MainTest(m *testing.M, pathToGiteaRoot string) {
 		fatalTestError("util.CopyDir: %v\n", err)
 	}
 
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	if err != nil {
+		fatalTestError("unable to read the new repo root: %v\n", err)
+	}
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		if err != nil {
+			fatalTestError("unable to read the new repo root: %v\n", err)
+		}
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
+
 	exitStatus := m.Run()
 	if err = util.RemoveAll(setting.RepoRootPath); err != nil {
 		fatalTestError("util.RemoveAll: %v\n", err)
@@ -128,6 +148,23 @@ func PrepareTestEnv(t testing.TB) {
 	assert.NoError(t, util.RemoveAll(setting.RepoRootPath))
 	metaPath := filepath.Join(giteaRoot, "integrations", "gitea-repositories-meta")
 	assert.NoError(t, util.CopyDir(metaPath, setting.RepoRootPath))
+
+	ownerDirs, err := os.ReadDir(setting.RepoRootPath)
+	assert.NoError(t, err)
+	for _, ownerDir := range ownerDirs {
+		if !ownerDir.Type().IsDir() {
+			continue
+		}
+		repoDirs, err := os.ReadDir(filepath.Join(setting.RepoRootPath, ownerDir.Name()))
+		assert.NoError(t, err)
+		for _, repoDir := range repoDirs {
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "pack"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "objects", "info"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "heads"), 0755)
+			_ = os.MkdirAll(filepath.Join(setting.RepoRootPath, ownerDir.Name(), repoDir.Name(), "refs", "tag"), 0755)
+		}
+	}
+
 	base.SetupGiteaRoot() // Makes sure GITEA_ROOT is set
 }
 

modules/git/batch_reader.go

@@ -27,6 +27,20 @@ type WriteCloserError interface {
 	CloseWithError(err error) error
 }
 
+// EnsureValidGitRepository runs git rev-parse in the repository path - thus ensuring that the repository is a valid repository.
+// Run before opening git cat-file.
+// This is needed otherwise the git cat-file will hang for invalid repositories.
+func EnsureValidGitRepository(ctx context.Context, repoPath string) error {
+	stderr := strings.Builder{}
+	err := NewCommandContext(ctx, "rev-parse").
+		SetDescription(fmt.Sprintf("%s rev-parse [repo_path: %s]", GitExecutable, repoPath)).
+		RunInDirFullPipeline(repoPath, nil, &stderr, nil)
+	if err != nil {
+		return ConcatenateError(err, (&stderr).String())
+	}
+	return nil
+}
+
 // CatFileBatchCheck opens git cat-file --batch-check in the provided repo and returns a stdin pipe, a stdout reader and cancel function
 func CatFileBatchCheck(repoPath string) (WriteCloserError, *bufio.Reader, func()) {
 	batchStdinReader, batchStdinWriter := io.Pipe()

modules/git/repo_base_nogogit.go

@@ -43,6 +43,11 @@ func OpenRepository(repoPath string) (*Repository, error) {
 		return nil, errors.New("no such file or directory")
 	}
 
+	// Now because of some insanity with git cat-file not immediately failing if not run in a valid git directory we need to run git rev-parse first!
+	if err := EnsureValidGitRepository(DefaultContext, repoPath); err != nil {
+		return nil, err
+	}
+
 	repo := &Repository{
 		Path:     repoPath,
 		tagCache: newObjectCache(),

modules/git/repo_commit_nogogit.go

@@ -38,7 +38,10 @@ func (repo *Repository) ResolveReference(name string) (string, error) {
 func (repo *Repository) GetRefCommitID(name string) (string, error) {
 	wr, rd, cancel := repo.CatFileBatchCheck()
 	defer cancel()
-	_, _ = wr.Write([]byte(name + "\n"))
+	_, err := wr.Write([]byte(name + "\n"))
+	if err != nil {
+		return "", err
+	}
 	shaBs, _, _, err := ReadBatchLine(rd)
 	if IsErrNotExist(err) {
 		return "", ErrNotExist{name, ""}

modules/indexer/code/bleve.go

@@ -276,6 +276,12 @@ func (b *BleveIndexer) Index(repo *models.Repository, sha string, changes *repoC
 	batch := gitea_bleve.NewFlushingBatch(b.indexer, maxBatchSize)
 	if len(changes.Updates) > 0 {
 
+		// Now because of some insanity with git cat-file not immediately failing if not run in a valid git directory we need to run git rev-parse first!
+		if err := git.EnsureValidGitRepository(git.DefaultContext, repo.RepoPath()); err != nil {
+			log.Error("Unable to open git repo: %s for %-v: %v", repo.RepoPath(), repo, err)
+			return err
+		}
+
 		batchWriter, batchReader, cancel := git.CatFileBatch(repo.RepoPath())
 		defer cancel()
 

modules/indexer/code/elastic_search.go

@@ -248,6 +248,11 @@ func (b *ElasticSearchIndexer) addDelete(filename string, repo *models.Repositor
 func (b *ElasticSearchIndexer) Index(repo *models.Repository, sha string, changes *repoChanges) error {
 	reqs := make([]elastic.BulkableRequest, 0)
 	if len(changes.Updates) > 0 {
+		// Now because of some insanity with git cat-file not immediately failing if not run in a valid git directory we need to run git rev-parse first!
+		if err := git.EnsureValidGitRepository(git.DefaultContext, repo.RepoPath()); err != nil {
+			log.Error("Unable to open git repo: %s for %-v: %v", repo.RepoPath(), repo, err)
+			return err
+		}
 
 		batchWriter, batchReader, cancel := git.CatFileBatch(repo.RepoPath())
 		defer cancel()

services/pull/pull.go

@@ -713,7 +713,8 @@ func GetIssuesLastCommitStatus(issues models.IssueList) (map[int64]*models.Commi
 		if !ok {
 			gitRepo, err = git.OpenRepository(issue.Repo.RepoPath())
 			if err != nil {
-				return nil, err
+				log.Error("Cannot open git repository %-v for issue #%d[%d]. Error: %v", issue.Repo, issue.Index, issue.ID, err)
+				continue
 			}
 			gitRepos[issue.RepoID] = gitRepo
 		}