Discard unread data of `git cat-file` (#29297) (#29310)

Backport #29297 by @KN4CK3R

Fixes #29101
Related #29298

Discard all read data to prevent misinterpreting existing data. Some
discard calls were missing in error cases.

Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
Co-authored-by: yp05327 <576951401@qq.com>
This commit is contained in:
Giteabot 2024-02-22 12:23:38 +08:00 committed by GitHub
parent 0870e0bc9b
commit ed5e0c8c27
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
10 changed files with 66 additions and 80 deletions

View File

@ -203,16 +203,7 @@ headerLoop:
}
// Discard the rest of the tag
discard := size - n + 1
for discard > math.MaxInt32 {
_, err := rd.Discard(math.MaxInt32)
if err != nil {
return id, err
}
discard -= math.MaxInt32
}
_, err := rd.Discard(int(discard))
return id, err
return id, DiscardFull(rd, size-n+1)
}
// ReadTreeID reads a tree ID from a cat-file --batch stream, throwing away the rest of the stream.
@ -238,16 +229,7 @@ headerLoop:
}
// Discard the rest of the commit
discard := size - n + 1
for discard > math.MaxInt32 {
_, err := rd.Discard(math.MaxInt32)
if err != nil {
return id, err
}
discard -= math.MaxInt32
}
_, err := rd.Discard(int(discard))
return id, err
return id, DiscardFull(rd, size-n+1)
}
// git tree files are a list:
@ -345,3 +327,21 @@ func init() {
_, filename, _, _ := runtime.Caller(0)
callerPrefix = strings.TrimSuffix(filename, "modules/git/batch_reader.go")
}
func DiscardFull(rd *bufio.Reader, discard int64) error {
if discard > math.MaxInt32 {
n, err := rd.Discard(math.MaxInt32)
discard -= int64(n)
if err != nil {
return err
}
}
for discard > 0 {
n, err := rd.Discard(int(discard))
discard -= int64(n)
if err != nil {
return err
}
}
return nil
}

View File

@ -9,7 +9,6 @@ import (
"bufio"
"bytes"
"io"
"math"
"code.gitea.io/gitea/modules/log"
)
@ -104,25 +103,6 @@ func (b *blobReader) Read(p []byte) (n int, err error) {
// Close implements io.Closer
func (b *blobReader) Close() error {
defer b.cancel()
if b.n > 0 {
for b.n > math.MaxInt32 {
n, err := b.rd.Discard(math.MaxInt32)
b.n -= int64(n)
if err != nil {
return err
}
b.n -= math.MaxInt32
}
n, err := b.rd.Discard(int(b.n))
b.n -= int64(n)
if err != nil {
return err
}
}
if b.n == 0 {
_, err := b.rd.Discard(1)
b.n--
return err
}
return nil
return DiscardFull(b.rd, b.n+1)
}

View File

@ -151,6 +151,9 @@ func GetLastCommitForPaths(ctx context.Context, commit *Commit, treePath string,
return nil, err
}
if typ != "commit" {
if err := DiscardFull(batchReader, size+1); err != nil {
return nil, err
}
return nil, fmt.Errorf("unexpected type: %s for commit id: %s", typ, commitID)
}
c, err = CommitFromReader(commit.repo, MustIDFromString(commitID), io.LimitReader(batchReader, size))

View File

@ -170,6 +170,10 @@ func FindLFSFile(repo *git.Repository, hash git.SHA1) ([]*LFSResult, error) {
} else {
break commitReadingLoop
}
default:
if err := git.DiscardFull(batchReader, size+1); err != nil {
return nil, err
}
}
}
}

View File

@ -121,8 +121,7 @@ func (repo *Repository) getCommitFromBatchReader(rd *bufio.Reader, id SHA1) (*Co
return commit, nil
default:
log.Debug("Unknown typ: %s", typ)
_, err = rd.Discard(int(size) + 1)
if err != nil {
if err := DiscardFull(rd, size+1); err != nil {
return nil, err
}
return nil, ErrNotExist{

View File

@ -6,10 +6,8 @@
package git
import (
"bufio"
"bytes"
"io"
"math"
"strings"
"code.gitea.io/gitea/modules/analyze"
@ -168,8 +166,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
return nil, err
}
content = contentBuf.Bytes()
err = discardFull(batchReader, discard)
if err != nil {
if err := DiscardFull(batchReader, discard); err != nil {
return nil, err
}
}
@ -212,21 +209,3 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
return mergeLanguageStats(sizes), nil
}
func discardFull(rd *bufio.Reader, discard int64) error {
if discard > math.MaxInt32 {
n, err := rd.Discard(math.MaxInt32)
discard -= int64(n)
if err != nil {
return err
}
}
for discard > 0 {
n, err := rd.Discard(int(discard))
discard -= int64(n)
if err != nil {
return err
}
}
return nil
}

View File

@ -103,6 +103,9 @@ func (repo *Repository) getTag(tagID SHA1, name string) (*Tag, error) {
return nil, err
}
if typ != "tag" {
if err := DiscardFull(rd, size+1); err != nil {
return nil, err
}
return nil, ErrNotExist{ID: tagID.String()}
}

View File

@ -58,6 +58,9 @@ func (repo *Repository) getTree(id SHA1) (*Tree, error) {
tree.entriesParsed = true
return tree, nil
default:
if err := DiscardFull(rd, size+1); err != nil {
return nil, err
}
return nil, ErrNotExist{
ID: id.String(),
}

View File

@ -7,7 +7,6 @@ package git
import (
"io"
"math"
"strings"
)
@ -63,19 +62,8 @@ func (t *Tree) ListEntries() (Entries, error) {
}
// Not a tree just use ls-tree instead
for sz > math.MaxInt32 {
discarded, err := rd.Discard(math.MaxInt32)
sz -= int64(discarded)
if err != nil {
return nil, err
}
}
for sz > 0 {
discarded, err := rd.Discard(int(sz))
sz -= int64(discarded)
if err != nil {
return nil, err
}
if err := DiscardFull(rd, sz+1); err != nil {
return nil, err
}
}

27
modules/git/tree_test.go Normal file
View File

@ -0,0 +1,27 @@
// Copyright 2024 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package git
import (
"path/filepath"
"testing"
"github.com/stretchr/testify/assert"
)
func TestSubTree_Issue29101(t *testing.T) {
repo, err := openRepositoryWithDefaultContext(filepath.Join(testReposDir, "repo1_bare"))
assert.NoError(t, err)
defer repo.Close()
commit, err := repo.GetCommit("ce064814f4a0d337b333e646ece456cd39fab612")
assert.NoError(t, err)
// old code could produce a different error if called multiple times
for i := 0; i < 10; i++ {
_, err = commit.SubTree("file1.txt")
assert.Error(t, err)
assert.True(t, IsErrNotExist(err))
}
}