fix: various security fixes (#38406)

Addresses a batch of privately reported security issues, grouped by
area:

- **SSRF** - migration PR-patch/asset fetches, OAuth2 avatar & OpenID
discovery, pull-mirror URL re-validation, and the outbound proxy path.
- **Access-token scope** - prevent scope escalation on token creation;
keep public-only tokens confined (feeds, packages, Actions listings,
star/watch lists, limited/private owners).
- **Access control / disclosure** - go-get default-branch leak, webhook
authorization-header leak, watch clearing on private transitions,
label/attachment scoping.
- **Denial of service** - input bounds for npm dist-tags, Debian control
files, Arch file lists, and SSH keys.

### 📌 Attention for site admins

Not breaking - existing configs keep working - but two changes are worth
a look:

- **New SSRF protection** Outbound requests (migrations, OAuth2 avatars,
OpenID discovery, pull mirrors, proxy path) are now validated against
the allow/block host lists. If your instance legitimately reaches
internal hosts, you may need to add them to
`[security].ALLOWED_HOST_LIST` (and the relevant `ALLOW_LOCALNETWORKS`
settings).
- **Deprecation** `[webhook].ALLOWED_HOST_LIST` is deprecated and will
be removed in a future release. Use `[security].ALLOWED_HOST_LIST`
instead; the old key still works for now.

---------

Co-authored-by: TheFox0x7 <thefox0x7@gmail.com>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Zettat123 <zettat123@gmail.com>

modules/packages/arch/metadata.go

@@ -13,6 +13,7 @@ import (
 	"strconv"
 	"strings"
 
+	"gitea.dev/modules/packages"
 	"gitea.dev/modules/util"
 	"gitea.dev/modules/validation"
 
@@ -46,6 +47,11 @@ var (
 	namePattern = regexp.MustCompile(`\A[a-zA-Z0-9@._+-]+\z`)
 	// (epoch:pkgver-pkgrel)
 	versionPattern = regexp.MustCompile(`\A(?:\d:)?[\w.+~]+(?:-[-\w.+~]+)?\z`)
+
+	// caps on the accumulated package file list (vars so tests can lower them); far above
+	// any legitimate package, but low enough to stop metadata amplification
+	maxFileEntries   = 100000
+	maxFileNameBytes = 16 * 1024 * 1024
 )
 
 type Package struct {
@@ -124,7 +130,7 @@ func ParsePackage(r io.Reader) (*Package, error) {
 	}
 
 	var p *Package
-	files := make([]string, 0, 10)
+	files := packages.NewBoundedFileList(maxFileEntries, maxFileNameBytes)
 
 	tr := tar.NewReader(inner)
 	for {
@@ -147,7 +153,9 @@ func ParsePackage(r io.Reader) (*Package, error) {
 				return nil, err
 			}
 		} else if !strings.HasPrefix(filename, ".") {
-			files = append(files, hd.Name)
+			if err := files.Add(hd.Name); err != nil {
+				return nil, err
+			}
 		}
 	}
 
@@ -155,7 +163,7 @@ func ParsePackage(r io.Reader) (*Package, error) {
 		return nil, ErrMissingPKGINFOFile
 	}
 
-	p.FileMetadata.Files = files
+	p.FileMetadata.Files = files.Files()
 	p.FileCompressionExtension = compressionType
 
 	return p, nil

modules/packages/arch/metadata_test.go

@@ -10,6 +10,9 @@ import (
 	"io"
 	"testing"
 
+	"gitea.dev/modules/test"
+	"gitea.dev/modules/util"
+
 	"github.com/klauspost/compress/zstd"
 	"github.com/stretchr/testify/assert"
 	"github.com/ulikunitz/xz"
@@ -167,3 +170,25 @@ func TestParsePackageInfo(t *testing.T) {
 		assert.ElementsMatch(t, []string{"usr/bin/paket1"}, p.FileMetadata.Backup)
 	})
 }
+
+// TestParsePackageTooManyFiles ensures the accumulated file list is bounded to prevent
+// metadata amplification from a package with a huge number of (tiny) file entries.
+func TestParsePackageTooManyFiles(t *testing.T) {
+	defer test.MockVariableValue(&maxFileEntries, 3)()
+	buf := test.WriteTarCompression(func(w io.Writer) io.WriteCloser { return gzip.NewWriter(w) }, map[string]string{
+		"file1":    "content1",
+		".PKGINFO": string(createPKGINFOContent(packageName, packageVersion)),
+	})
+	_, err := ParsePackage(buf)
+	assert.NoError(t, err)
+
+	buf = test.WriteTarCompression(func(w io.Writer) io.WriteCloser { return gzip.NewWriter(w) }, map[string]string{
+		"file1":    "content1",
+		"file2":    "content2",
+		"file3":    "content3",
+		"file4":    "content4",
+		".PKGINFO": string(createPKGINFOContent(packageName, packageVersion)),
+	})
+	_, err = ParsePackage(buf)
+	assert.ErrorIs(t, err, util.ErrInvalidArgument)
+}

modules/packages/debian/metadata.go

@@ -135,7 +135,10 @@ func ParsePackage(r io.Reader) (*Package, error) {
 				return nil, GlobalVars().ErrUnsupportedCompression
 			}
 
-			tr := tar.NewReader(inner)
+			// bound the decompressed control archive: it holds only the small control file
+			// and maintainer scripts, so a much larger stream is a decompression bomb
+			const maxControlTarSize = 32 * 1024 * 1024
+			tr := tar.NewReader(io.LimitReader(inner, maxControlTarSize))
 			for {
 				hd, err := tr.Next()
 				if err == io.EOF {
@@ -168,6 +171,7 @@ func ParseControlFile(r io.Reader) (*Package, error) {
 	key := ""
 	var depends strings.Builder
 	var control strings.Builder
+	var description strings.Builder
 
 	// https://www.debian.org/doc/debian-policy/ch-controlfields.html#syntax-of-control-files
 	s := bufio.NewScanner(r)
@@ -189,10 +193,13 @@ func ParseControlFile(r io.Reader) (*Package, error) {
 		control.WriteString(line)
 		control.WriteByte('\n')
 
+		// a leading space or tab marks a folded continuation line that belongs to the previous field
+		// (identified by key), not a new "Key: value" pair; only the multi-line fields append here.
+		// Continuation lines may themselves contain a colon, so they must not be re-split on ":".
 		if line[0] == ' ' || line[0] == '\t' {
 			switch key {
 			case "Description":
-				p.Metadata.Description += line
+				description.WriteString(line)
 			case "Depends":
 				depends.WriteString(trimmed)
 			}
@@ -219,7 +226,8 @@ func ParseControlFile(r io.Reader) (*Package, error) {
 					p.Metadata.Maintainer = a.Name
 				}
 			case "Description":
-				p.Metadata.Description = value
+				description.Reset()
+				description.WriteString(value)
 			case "Depends":
 				depends.WriteString(value)
 			case "Homepage":
@@ -243,6 +251,8 @@ func ParseControlFile(r io.Reader) (*Package, error) {
 		return nil, GlobalVars().ErrInvalidArchitecture
 	}
 
+	p.Metadata.Description = description.String()
+
 	dependencies := strings.Split(depends.String(), ",")
 	for i := range dependencies {
 		dependencies[i] = strings.TrimSpace(dependencies[i])

modules/packages/debian/metadata_test.go

@@ -232,3 +232,15 @@ func TestValidateDistributionOrComponent(t *testing.T) {
 		assert.True(t, IsValidDistributionOrComponent(name), "good=%q", name)
 	}
 }
+
+// TestParseControlFileMultilineDescription verifies a multi-line Description is assembled in order
+// (the parser accumulates it in a strings.Builder); it guards the assembled value, not its timing.
+func TestParseControlFileMultilineDescription(t *testing.T) {
+	var buf bytes.Buffer
+	buf.WriteString("Package: testpkg\nVersion: 1.0\nArchitecture: amd64\nDescription: short summary\n more details\n even more\n")
+
+	p, err := ParseControlFile(&buf)
+	assert.NoError(t, err)
+	assert.NotNil(t, p)
+	assert.Equal(t, "short summary more details even more", p.Metadata.Description)
+}

modules/packages/filelist.go

@@ -0,0 +1,37 @@
+// Copyright 2026 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package packages
+
+import "gitea.dev/modules/util"
+
+// BoundedFileList accumulates file names from a package archive while enforcing caps on the number of
+// entries and their total name length, returning an error once either cap would be exceeded.
+type BoundedFileList struct {
+	files     []string
+	nameBytes int
+	maxFiles  int
+	maxBytes  int
+}
+
+// NewBoundedFileList creates a BoundedFileList with the given caps; a non-positive cap falls back to the
+// corresponding default.
+func NewBoundedFileList(maxFiles, maxNameBytes int) *BoundedFileList {
+	return &BoundedFileList{maxFiles: maxFiles, maxBytes: maxNameBytes}
+}
+
+// Add appends name, returning util.ErrInvalidArgument once the entry count or accumulated byte length
+// would exceed the configured cap.
+func (b *BoundedFileList) Add(name string) error {
+	if len(b.files) >= b.maxFiles || b.nameBytes+len(name) > b.maxBytes {
+		return util.NewInvalidArgumentErrorf("package contains too many file entries")
+	}
+	b.nameBytes += len(name)
+	b.files = append(b.files, name)
+	return nil
+}
+
+// Files returns the accumulated file names.
+func (b *BoundedFileList) Files() []string {
+	return b.files
+}