Will Norris 3ec5be3f51 all: remove AUTHORS file and references to it
This file was never truly necessary and has never actually been used in
the history of Tailscale's open source releases.

A Brief History of AUTHORS files
---

The AUTHORS file was a pattern developed at Google, originally for
Chromium, then adopted by Go and a bunch of other projects. The problem
was that Chromium originally had a copyright line only recognizing
Google as the copyright holder. Because Google (and most open source
projects) do not require copyright assignemnt for contributions, each
contributor maintains their copyright. Some large corporate contributors
then tried to add their own name to the copyright line in the LICENSE
file or in file headers. This quickly becomes unwieldy, and puts a
tremendous burden on anyone building on top of Chromium, since the
license requires that they keep all copyright lines intact.

The compromise was to create an AUTHORS file that would list all of the
copyright holders. The LICENSE file and source file headers would then
include that list by reference, listing the copyright holder as "The
Chromium Authors".

This also become cumbersome to simply keep the file up to date with a
high rate of new contributors. Plus it's not always obvious who the
copyright holder is. Sometimes it is the individual making the
contribution, but many times it may be their employer. There is no way
for the proejct maintainer to know.

Eventually, Google changed their policy to no longer recommend trying to
keep the AUTHORS file up to date proactively, and instead to only add to
it when requested: https://opensource.google/docs/releasing/authors.
They are also clear that:

> Adding contributors to the AUTHORS file is entirely within the
> project's discretion and has no implications for copyright ownership.

It was primarily added to appease a small number of large contributors
that insisted that they be recognized as copyright holders (which was
entirely their right to do). But it's not truly necessary, and not even
the most accurate way of identifying contributors and/or copyright
holders.

In practice, we've never added anyone to our AUTHORS file. It only lists
Tailscale, so it's not really serving any purpose. It also causes
confusion because Tailscalars put the "Tailscale Inc & AUTHORS" header
in other open source repos which don't actually have an AUTHORS file, so
it's ambiguous what that means.

Instead, we just acknowledge that the contributors to Tailscale (whoever
they are) are copyright holders for their individual contributions. We
also have the benefit of using the DCO (developercertificate.org) which
provides some additional certification of their right to make the
contribution.

The source file changes were purely mechanical with:

    git ls-files | xargs sed -i -e 's/\(Tailscale Inc &\) AUTHORS/\1 contributors/g'

Updates #cleanup

Change-Id: Ia101a4a3005adb9118051b3416f5a64a4a45987d
Signed-off-by: Will Norris <will@tailscale.com>
2026-01-23 15:49:45 -08:00

159 lines
5.0 KiB
Go

// Copyright (c) Tailscale Inc & contributors
// SPDX-License-Identifier: BSD-3-Clause
// Package ts2021 handles the details of the Tailscale 2021 control protocol
// that are after (above) the Noise layer. In particular, the
// "tailcfg.EarlyNoise" message and the subsequent HTTP/2 connection.
package ts2021
import (
"bytes"
"context"
"encoding/binary"
"encoding/json"
"errors"
"io"
"sync"
"tailscale.com/control/controlbase"
"tailscale.com/tailcfg"
)
// Conn is a wrapper around controlbase.Conn.
//
// It allows attaching an ID to a connection to allow cleaning up references in
// the pool when the connection is closed, properly handles an optional "early
// payload" that's sent prior to beginning the HTTP/2 session, and provides a
// way to return a connection to a pool when the connection is closed.
//
// Use [NewConn] to build a new Conn if you want [Conn.GetEarlyPayload] to work.
// Otherwise making a Conn directly, only setting Conn, is fine.
type Conn struct {
*controlbase.Conn
onClose func() // or nil
readHeaderOnce sync.Once // guards init of reader field
reader io.Reader // (effectively Conn.Reader after header)
earlyPayloadReady chan struct{} // closed after earlyPayload is set (including set to nil)
earlyPayload *tailcfg.EarlyNoise
earlyPayloadErr error
}
// NewConn creates a new Conn that wraps the given controlbase.Conn.
//
// h2t is the HTTP/2 transport to use for the connection; a new
// http2.ClientConn will be created that reads from the returned Conn.
//
// connID should be a unique ID for this connection. When the Conn is closed,
// the onClose function will be called if it is non-nil.
func NewConn(conn *controlbase.Conn, onClose func()) *Conn {
return &Conn{
Conn: conn,
earlyPayloadReady: make(chan struct{}),
onClose: sync.OnceFunc(onClose),
}
}
// GetEarlyPayload waits for the early Noise payload to arrive.
// It may return (nil, nil) if the server begins HTTP/2 without one.
//
// It is safe to call this multiple times; all callers will block until the
// early Noise payload is ready (if any) and will return the same result for
// the lifetime of the Conn.
func (c *Conn) GetEarlyPayload(ctx context.Context) (*tailcfg.EarlyNoise, error) {
if c.earlyPayloadReady == nil {
return nil, errors.New("Conn was not created with NewConn; early payload not supported")
}
select {
case <-c.earlyPayloadReady:
return c.earlyPayload, c.earlyPayloadErr
default:
go c.readHeaderOnce.Do(c.readHeader)
}
select {
case <-c.earlyPayloadReady:
return c.earlyPayload, c.earlyPayloadErr
case <-ctx.Done():
return nil, ctx.Err()
}
}
// The first 9 bytes from the server to client over Noise are either an HTTP/2
// settings frame (a normal HTTP/2 setup) or, as we added later, an "early payload"
// header that's also 9 bytes long: 5 bytes (EarlyPayloadMagic) followed by 4 bytes
// of length. Then that many bytes of JSON-encoded tailcfg.EarlyNoise.
// The early payload is optional. Some servers may not send it.
const (
hdrLen = 9 // http2 frame header size; also size of our early payload size header
)
// EarlyPayloadMagic is the 5-byte magic prefix that indicates an early payload.
const EarlyPayloadMagic = "\xff\xff\xffTS"
// returnErrReader is an io.Reader that always returns an error.
type returnErrReader struct {
err error // the error to return
}
func (r returnErrReader) Read([]byte) (int, error) { return 0, r.err }
// Read is basically the same as controlbase.Conn.Read, but it first reads the
// "early payload" header from the server which may or may not be present,
// depending on the server.
func (c *Conn) Read(p []byte) (n int, err error) {
c.readHeaderOnce.Do(c.readHeader)
return c.reader.Read(p)
}
// Close closes the connection.
func (c *Conn) Close() error {
if c.onClose != nil {
defer c.onClose()
}
return c.Conn.Close()
}
// readHeader reads the optional "early payload" from the server that arrives
// after the Noise handshake but before the HTTP/2 session begins.
//
// readHeader is responsible for reading the header (if present), initializing
// c.earlyPayload, closing c.earlyPayloadReady, and initializing c.reader for
// future reads.
func (c *Conn) readHeader() {
if c.earlyPayloadReady != nil {
defer close(c.earlyPayloadReady)
}
setErr := func(err error) {
c.reader = returnErrReader{err}
c.earlyPayloadErr = err
}
var hdr [hdrLen]byte
if _, err := io.ReadFull(c.Conn, hdr[:]); err != nil {
setErr(err)
return
}
if string(hdr[:len(EarlyPayloadMagic)]) != EarlyPayloadMagic {
// No early payload. We have to return the 9 bytes read we already
// consumed.
c.reader = io.MultiReader(bytes.NewReader(hdr[:]), c.Conn)
return
}
epLen := binary.BigEndian.Uint32(hdr[len(EarlyPayloadMagic):])
if epLen > 10<<20 {
setErr(errors.New("invalid early payload length"))
return
}
payBuf := make([]byte, epLen)
if _, err := io.ReadFull(c.Conn, payBuf); err != nil {
setErr(err)
return
}
if err := json.Unmarshal(payBuf, &c.earlyPayload); err != nil {
setErr(err)
return
}
c.reader = c.Conn
}