Files
coder/tailnet/tunnel.go
T
Garrett Delfosse 9b351c3602 fix: validate agent-supplied AllowedIPs in coordinator (#26144)
`AgentCoordinateeAuth.Authorize` validated every prefix in
`upd.Node.Addresses` (each must be a `/128` derived from the
authenticating agent's own UUID) but applied no equivalent check to
`upd.Node.AllowedIps`. Because `AllowedIPs` are installed verbatim into
the WireGuard peer config (`tailnet/configmaps.go`) and WireGuard
routing is driven by `AllowedIPs`, a malicious agent could advertise a
victim agent's `/128` and become an eligible route for that IP. With
`ServerTailnet` tunneling to many agents and routing by destination IP,
this could let an attacker intercept sessions intended for the victim
workspace.

This applies the same UUID-derivation validation to `AllowedIps` that
already guards `Addresses`, extracted into a shared
`authorizeNodePrefixes` helper. The check is the single chokepoint used
by both the in-memory coordinator (`tailnet/coordinator.go`) and the
Postgres coordinator (`enterprise/tailnet/connio.go`), so one fix covers
both. Legitimate agents are unaffected: an agent's `AllowedIPs` is a
clone of its `Addresses` (`tailnet/node.go`), which are already
UUID-derived `/128`s.

Fixes PLAT-264 (SEC-89): https://linear.app/codercom/issue/PLAT-264

<details>
<summary>Implementation notes and decision log</summary>

### Root cause

Asymmetric validation in `tailnet/tunnel.go`: `Addresses` were bound to
the agent's UUID, but `AllowedIps` were trusted as-is and propagated
into the WireGuard peer config, which drives routing.

### Why the fix is safe for legitimate agents

- `tailnet/node.go` builds the node with `AllowedIPs:
slices.Clone(u.addresses)`, identical to `Addresses`.
- `agent/agent.go` sets those addresses to
`TailscaleServicePrefix.PrefixFromUUID(agentID)` and
`CoderServicePrefix.PrefixFromUUID(agentID)` (both `/128`,
UUID-derived).
- The existing `Addresses` check already accepts exactly those prefixes
plus the legacy workspace agent IP, so identical validation of
`AllowedIPs` passes for real traffic and only rejects forged prefixes.

### Coverage: one method, both coordinators

`AgentCoordinateeAuth.Authorize` is the shared auth path. A failed
`Authorize` is wrapped as `AuthorizationError{Wrapped: err}` and closes
the agent's response stream.

### Tests

- `tailnet/tunnel_internal_test.go`: fast unit tests on `Authorize`
(valid AllowedIPs accepted; foreign `/128` rejected with
`InvalidNodeAddressError`; wrong-bits rejected with
`InvalidAddressBitsError`).
- `tailnet/coordinator_test.go`: in-memory coordinator closes the agent
stream on a forged `AllowedIp`.
- `enterprise/tailnet/pgcoord_test.go`: same regression for the Postgres
coordinator.

Verified the regression tests fail when the new `AllowedIps` check is
disabled, then pass with it enabled. Local validation: targeted tests
(in-memory, internal, and Postgres-backed enterprise), plus `make
pre-commit` (gen/fmt/lint/build) passing.

</details>

> Generated by Coder Agents on behalf of @f0ssel.
2026-06-11 15:26:58 -04:00

228 lines
5.9 KiB
Go

package tailnet
import (
"context"
"fmt"
"net/netip"
"github.com/google/uuid"
"golang.org/x/xerrors"
"github.com/coder/coder/v2/tailnet/proto"
)
var legacyWorkspaceAgentIP = netip.MustParseAddr("fd7a:115c:a1e0:49d6:b259:b7ac:b1b2:48f4")
type InvalidAddressBitsError struct {
Bits int
}
func (e InvalidAddressBitsError) Error() string {
return fmt.Sprintf("invalid address bits, expected 128, got %d", e.Bits)
}
type InvalidNodeAddressError struct {
Addr string
}
func (e InvalidNodeAddressError) Error() string {
return fmt.Sprintf("invalid node address, got %s", e.Addr)
}
type CoordinateeAuth interface {
Authorize(ctx context.Context, req *proto.CoordinateRequest) error
}
// SingleTailnetCoordinateeAuth allows all tunnels, since Coderd and wsproxy are allowed to initiate a tunnel to any agent
type SingleTailnetCoordinateeAuth struct{}
func (SingleTailnetCoordinateeAuth) Authorize(context.Context, *proto.CoordinateRequest) error {
return nil
}
// ClientCoordinateeAuth allows connecting to a single, given agent
type ClientCoordinateeAuth struct {
AgentID uuid.UUID
}
func (c ClientCoordinateeAuth) Authorize(_ context.Context, req *proto.CoordinateRequest) error {
if tun := req.GetAddTunnel(); tun != nil {
uid, err := uuid.FromBytes(tun.Id)
if err != nil {
return xerrors.Errorf("parse add tunnel id: %w", err)
}
if c.AgentID != uid {
return xerrors.Errorf("invalid agent id, expected %s, got %s", c.AgentID.String(), uid.String())
}
}
return handleClientNodeRequests(req)
}
// AgentCoordinateeAuth disallows all tunnels, since agents are not allowed to initiate their own tunnels
type AgentCoordinateeAuth struct {
ID uuid.UUID
}
func (a AgentCoordinateeAuth) Authorize(_ context.Context, req *proto.CoordinateRequest) error {
if tun := req.GetAddTunnel(); tun != nil {
return xerrors.New("agents cannot open tunnels")
}
if upd := req.GetUpdateSelf(); upd != nil {
// Both Addresses and AllowedIPs are installed into the WireGuard peer
// config and drive routing, so an agent may only advertise prefixes
// derived from its own UUID. Without this an agent could claim a victim
// agent's IP and have traffic routed to it.
if err := a.authorizeNodePrefixes(upd.Node.Addresses); err != nil {
return xerrors.Errorf("Addresses: %w", err)
}
if err := a.authorizeNodePrefixes(upd.Node.AllowedIps); err != nil {
return xerrors.Errorf("AllowedIps: %w", err)
}
}
return nil
}
// authorizeNodePrefixes verifies that every prefix is a /128 address derived
// from the agent's own UUID (or the legacy workspace agent IP).
func (a AgentCoordinateeAuth) authorizeNodePrefixes(prefixes []string) error {
for _, prefixStr := range prefixes {
pre, err := netip.ParsePrefix(prefixStr)
if err != nil {
return xerrors.Errorf("parse node address: %w", err)
}
if pre.Bits() != 128 {
return InvalidAddressBitsError{pre.Bits()}
}
if TailscaleServicePrefix.AddrFromUUID(a.ID).Compare(pre.Addr()) != 0 &&
CoderServicePrefix.AddrFromUUID(a.ID).Compare(pre.Addr()) != 0 &&
legacyWorkspaceAgentIP.Compare(pre.Addr()) != 0 {
return InvalidNodeAddressError{pre.Addr().String()}
}
}
return nil
}
type ClientUserCoordinateeAuth struct {
Auth TunnelAuthorizer
}
func (a ClientUserCoordinateeAuth) Authorize(ctx context.Context, req *proto.CoordinateRequest) error {
if tun := req.GetAddTunnel(); tun != nil {
uid, err := uuid.FromBytes(tun.Id)
if err != nil {
return xerrors.Errorf("parse add tunnel id: %w", err)
}
err = a.Auth.AuthorizeTunnel(ctx, uid)
if err != nil {
return xerrors.Errorf("workspace agent not found or you do not have permission")
}
}
return handleClientNodeRequests(req)
}
// handleClientNodeRequests validates GetUpdateSelf requests and declines ReadyForHandshake requests
func handleClientNodeRequests(req *proto.CoordinateRequest) error {
if upd := req.GetUpdateSelf(); upd != nil {
for _, addrStr := range upd.Node.Addresses {
pre, err := netip.ParsePrefix(addrStr)
if err != nil {
return xerrors.Errorf("parse node address: %w", err)
}
if pre.Bits() != 128 {
return InvalidAddressBitsError{pre.Bits()}
}
}
}
if rfh := req.GetReadyForHandshake(); rfh != nil {
return xerrors.Errorf("clients may not send ready_for_handshake")
}
return nil
}
// tunnelStore contains tunnel information and allows querying it. It is not threadsafe and all
// methods must be serialized by holding, e.g. the core mutex.
type tunnelStore struct {
bySrc map[uuid.UUID]map[uuid.UUID]struct{}
byDst map[uuid.UUID]map[uuid.UUID]struct{}
}
func newTunnelStore() *tunnelStore {
return &tunnelStore{
bySrc: make(map[uuid.UUID]map[uuid.UUID]struct{}),
byDst: make(map[uuid.UUID]map[uuid.UUID]struct{}),
}
}
func (s *tunnelStore) add(src, dst uuid.UUID) {
srcM, ok := s.bySrc[src]
if !ok {
srcM = make(map[uuid.UUID]struct{})
s.bySrc[src] = srcM
}
srcM[dst] = struct{}{}
dstM, ok := s.byDst[dst]
if !ok {
dstM = make(map[uuid.UUID]struct{})
s.byDst[dst] = dstM
}
dstM[src] = struct{}{}
}
func (s *tunnelStore) remove(src, dst uuid.UUID) {
delete(s.bySrc[src], dst)
if len(s.bySrc[src]) == 0 {
delete(s.bySrc, src)
}
delete(s.byDst[dst], src)
if len(s.byDst[dst]) == 0 {
delete(s.byDst, dst)
}
}
func (s *tunnelStore) removeAll(src uuid.UUID) {
for dst := range s.bySrc[src] {
s.remove(src, dst)
}
}
func (s *tunnelStore) findTunnelPeers(id uuid.UUID) []uuid.UUID {
set := make(map[uuid.UUID]struct{})
for dst := range s.bySrc[id] {
set[dst] = struct{}{}
}
for src := range s.byDst[id] {
set[src] = struct{}{}
}
out := make([]uuid.UUID, 0, len(set))
for id := range set {
out = append(out, id)
}
return out
}
func (s *tunnelStore) tunnelExists(src, dst uuid.UUID) bool {
_, srcOK := s.bySrc[src][dst]
_, dstOK := s.byDst[src][dst]
return srcOK || dstOK
}
func (s *tunnelStore) htmlDebug() []HTMLTunnel {
out := make([]HTMLTunnel, 0)
for src, dsts := range s.bySrc {
for dst := range dsts {
out = append(out, HTMLTunnel{Src: src, Dst: dst})
}
}
return out
}