@ttsc/win32-arm64 0.33.0 → 0.34.0

package/bin/go/src/net/http/internal/http2/server.go

@@ -0,0 +1,3238 @@
+// Copyright 2014 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// TODO: turn off the serve goroutine when idle, so
+// an idle conn only has the readFrames goroutine active. (which could
+// also be optimized probably to pin less memory in crypto/tls). This
+// would involve tracking when the serve goroutine is active (atomic
+// int32 read/CAS probably?) and starting it up when frames arrive,
+// and shutting it down when all handlers exit. the occasional PING
+// packets could use time.AfterFunc to call sc.wakeStartServeLoop()
+// (which is a no-op if already running) and then queue the PING write
+// as normal. The serve loop would then exit in most cases (if no
+// Handlers running) and not be woken up again until the PING packet
+// returns.
+
+// TODO (maybe): add a mechanism for Handlers to going into
+// half-closed-local mode (rw.(io.Closer) test?) but not exit their
+// handler, and continue to be able to read from the
+// Request.Body. This would be a somewhat semantic change from HTTP/1
+// (or at least what we expose in net/http), so I'd probably want to
+// add it there too. For now, this package says that returning from
+// the Handler ServeHTTP function means you're both done reading and
+// done writing, without a way to stop just one or the other.
+
+package http2
+
+import (
+	"bufio"
+	"bytes"
+	"context"
+	"crypto/rand"
+	"crypto/tls"
+	"errors"
+	"fmt"
+	"io"
+	"log"
+	"math"
+	"net"
+	"net/http/internal"
+	"net/http/internal/httpcommon"
+	"net/textproto"
+	"net/url"
+	"os"
+	"reflect"
+	"runtime"
+	"slices"
+	"strconv"
+	"strings"
+	"sync"
+	"time"
+
+	"golang.org/x/net/http/httpguts"
+	"golang.org/x/net/http2/hpack"
+)
+
+const (
+	prefaceTimeout        = 10 * time.Second
+	firstSettingsTimeout  = 2 * time.Second // should be in-flight with preface anyway
+	handlerChunkWriteSize = 4 << 10
+	defaultMaxStreams     = 250 // TODO: make this 100 as the GFE seems to?
+
+	// maxQueuedControlFrames is the maximum number of control frames like
+	// SETTINGS, PING and RST_STREAM that will be queued for writing before
+	// the connection is closed to prevent memory exhaustion attacks.
+	maxQueuedControlFrames = 10000
+)
+
+var (
+	errClientDisconnected = errors.New("client disconnected")
+	errClosedBody         = errors.New("body closed by handler")
+	errHandlerComplete    = errors.New("http2: request body closed due to handler exiting")
+	errStreamClosed       = errors.New("http2: stream closed")
+)
+
+var responseWriterStatePool = sync.Pool{
+	New: func() any {
+		rws := &responseWriterState{}
+		rws.bw = bufio.NewWriterSize(chunkWriter{rws}, handlerChunkWriteSize)
+		return rws
+	},
+}
+
+// Test hooks.
+var (
+	testHookOnConn    func()
+	testHookOnPanicMu *sync.Mutex // nil except in tests
+	testHookOnPanic   func(sc *serverConn, panicVal any) (rePanic bool)
+)
+
+// Server is an HTTP/2 server.
+type Server struct {
+	mu          sync.Mutex
+	activeConns map[*serverConn]struct{}
+
+	// Pool of error channels. This is per-Server rather than global
+	// because channels can't be reused across synctest bubbles.
+	errChanPool sync.Pool
+}
+
+func (s *Server) registerConn(sc *serverConn) {
+	if s == nil {
+		return // if the Server was used without calling ConfigureServer
+	}
+	s.mu.Lock()
+	s.activeConns[sc] = struct{}{}
+	s.mu.Unlock()
+}
+
+func (s *Server) unregisterConn(sc *serverConn) {
+	if s == nil {
+		return // if the Server was used without calling ConfigureServer
+	}
+	s.mu.Lock()
+	delete(s.activeConns, sc)
+	s.mu.Unlock()
+}
+
+func (s *Server) startGracefulShutdown() {
+	if s == nil {
+		return // if the Server was used without calling ConfigureServer
+	}
+	s.mu.Lock()
+	for sc := range s.activeConns {
+		sc.startGracefulShutdown()
+	}
+	s.mu.Unlock()
+}
+
+// Global error channel pool used for uninitialized Servers.
+// We use a per-Server pool when possible to avoid using channels across synctest bubbles.
+var errChanPool = sync.Pool{
+	New: func() any { return make(chan error, 1) },
+}
+
+func (s *Server) getErrChan() chan error {
+	if s == nil {
+		return errChanPool.Get().(chan error) // Server used without calling ConfigureServer
+	}
+	return s.errChanPool.Get().(chan error)
+}
+
+func (s *Server) putErrChan(ch chan error) {
+	if s == nil {
+		errChanPool.Put(ch) // Server used without calling ConfigureServer
+		return
+	}
+	s.errChanPool.Put(ch)
+}
+
+func (s *Server) Configure(conf ServerConfig, tcfg *tls.Config) error {
+	s.activeConns = make(map[*serverConn]struct{})
+	s.errChanPool = sync.Pool{New: func() any { return make(chan error, 1) }}
+
+	if tcfg.CipherSuites != nil && tcfg.MinVersion < tls.VersionTLS13 {
+		// If they already provided a TLS 1.0–1.2 CipherSuite list, return an
+		// error if it is missing ECDHE_RSA_WITH_AES_128_GCM_SHA256 or
+		// ECDHE_ECDSA_WITH_AES_128_GCM_SHA256.
+		haveRequired := false
+		for _, cs := range tcfg.CipherSuites {
+			switch cs {
+			case tls.TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256,
+				// Alternative MTI cipher to not discourage ECDSA-only servers.
+				// See http://golang.org/cl/30721 for further information.
+				tls.TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256:
+				haveRequired = true
+			}
+		}
+		if !haveRequired {
+			return fmt.Errorf("http2: TLSConfig.CipherSuites is missing an HTTP/2-required AES_128_GCM_SHA256 cipher (need at least one of TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 or TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256)")
+		}
+	}
+
+	// Note: not setting MinVersion to tls.VersionTLS12,
+	// as we don't want to interfere with HTTP/1.1 traffic
+	// on the user's server. We enforce TLS 1.2 later once
+	// we accept a connection. Ideally this should be done
+	// during next-proto selection, but using TLS <1.2 with
+	// HTTP/2 is still the client's bug.
+
+	return nil
+}
+
+func (s *Server) GracefulShutdown() {
+	s.startGracefulShutdown()
+}
+
+// ServeConnOpts are options for the Server.ServeConn method.
+type ServeConnOpts struct {
+	// Context is the base context to use.
+	// If nil, context.Background is used.
+	Context context.Context
+
+	// BaseConfig optionally sets the base configuration
+	// for values. If nil, defaults are used.
+	BaseConfig ServerConfig
+
+	// Handler specifies which handler to use for processing
+	// requests. If nil, BaseConfig.Handler is used. If BaseConfig
+	// or BaseConfig.Handler is nil, http.DefaultServeMux is used.
+	Handler Handler
+
+	// Settings is the decoded contents of the HTTP2-Settings header
+	// in an h2c upgrade request.
+	Settings []byte
+
+	UpgradeRequest *ServerRequest
+
+	// SawClientPreface is set if the HTTP/2 connection preface
+	// has already been read from the connection.
+	SawClientPreface bool
+}
+
+func (o *ServeConnOpts) context() context.Context {
+	if o != nil && o.Context != nil {
+		return o.Context
+	}
+	return context.Background()
+}
+
+// ServeConn serves HTTP/2 requests on the provided connection and
+// blocks until the connection is no longer readable.
+//
+// ServeConn starts speaking HTTP/2 assuming that c has not had any
+// reads or writes. It writes its initial settings frame and expects
+// to be able to read the preface and settings frame from the
+// client. If c has a ConnectionState method like a *tls.Conn, the
+// ConnectionState is used to verify the TLS ciphersuite and to set
+// the Request.TLS field in Handlers.
+//
+// ServeConn does not support h2c by itself. Any h2c support must be
+// implemented in terms of providing a suitably-behaving net.Conn.
+//
+// The opts parameter is optional. If nil, default values are used.
+func (s *Server) ServeConn(c net.Conn, opts *ServeConnOpts) {
+	if opts == nil {
+		opts = &ServeConnOpts{}
+	}
+
+	var newf func(*serverConn)
+	if inTests {
+		// Fetch NewConnContextKey if set, leave newf as nil otherwise.
+		newf, _ = opts.Context.Value(NewConnContextKey).(func(*serverConn))
+	}
+
+	s.serveConn(c, opts, newf)
+}
+
+type contextKey string
+
+var (
+	NewConnContextKey         = new("NewConnContextKey")
+	ConnectionStateContextKey = new("ConnectionStateContextKey")
+)
+
+func (s *Server) serveConn(c net.Conn, opts *ServeConnOpts, newf func(*serverConn)) {
+	baseCtx, cancel := serverConnBaseContext(c, opts)
+	defer cancel()
+
+	conf := configFromServer(opts.BaseConfig)
+	sc := &serverConn{
+		srv:                         s,
+		hs:                          opts.BaseConfig,
+		conn:                        c,
+		baseCtx:                     baseCtx,
+		remoteAddrStr:               c.RemoteAddr().String(),
+		bw:                          newBufferedWriter(c, conf.WriteByteTimeout),
+		handler:                     opts.Handler,
+		streams:                     make(map[uint32]*stream),
+		readFrameCh:                 make(chan readFrameResult),
+		wantWriteFrameCh:            make(chan FrameWriteRequest, 8),
+		serveMsgCh:                  make(chan any, 8),
+		wroteFrameCh:                make(chan frameWriteResult, 1), // buffered; one send in writeFrameAsync
+		bodyReadCh:                  make(chan bodyReadMsg),         // buffering doesn't matter either way
+		doneServing:                 make(chan struct{}),
+		clientMaxStreams:            math.MaxUint32, // Section 6.5.2: "Initially, there is no limit to this value"
+		advMaxStreams:               uint32(conf.MaxConcurrentStreams),
+		initialStreamRecvWindowSize: int32(conf.MaxReceiveBufferPerStream),
+		maxFrameSize:                initialMaxFrameSize,
+		pingTimeout:                 conf.PingTimeout,
+		countErrorFunc:              conf.CountError,
+		serveG:                      newGoroutineLock(),
+		pushEnabled:                 true,
+		sawClientPreface:            opts.SawClientPreface,
+	}
+	if newf != nil {
+		newf(sc)
+	}
+
+	s.registerConn(sc)
+	defer s.unregisterConn(sc)
+
+	switch {
+	case sc.hs.DisableClientPriority():
+		sc.writeSched = newRoundRobinWriteScheduler()
+	default:
+		sc.writeSched = newPriorityWriteSchedulerRFC9218()
+	}
+
+	// These start at the RFC-specified defaults. If there is a higher
+	// configured value for inflow, that will be updated when we send a
+	// WINDOW_UPDATE shortly after sending SETTINGS.
+	sc.flow.init()
+	sc.inflow.init(initialWindowSize)
+	sc.hpackEncoder = hpack.NewEncoder(&sc.headerWriteBuf)
+	sc.hpackEncoder.SetMaxDynamicTableSizeLimit(uint32(conf.MaxEncoderHeaderTableSize))
+
+	fr := NewFramer(sc.bw, c)
+	if conf.CountError != nil {
+		fr.countError = conf.CountError
+	}
+	fr.ReadMetaHeaders = hpack.NewDecoder(uint32(conf.MaxDecoderHeaderTableSize), nil)
+	fr.MaxHeaderListSize = sc.maxHeaderListSize()
+	fr.MaxHeaderValueCount = sc.hs.MaxHeaderValueCount()
+	fr.SetMaxReadFrameSize(uint32(conf.MaxReadFrameSize))
+	sc.framer = fr
+
+	if tc, ok := c.(connectionStater); ok {
+		sc.tlsState = new(tls.ConnectionState)
+		*sc.tlsState = tc.ConnectionState()
+
+		// Optionally override the ConnectionState in tests.
+		if inTests {
+			f, ok := opts.Context.Value(ConnectionStateContextKey).(func() tls.ConnectionState)
+			if ok {
+				*sc.tlsState = f()
+			}
+		}
+
+		// 9.2 Use of TLS Features
+		// An implementation of HTTP/2 over TLS MUST use TLS
+		// 1.2 or higher with the restrictions on feature set
+		// and cipher suite described in this section. Due to
+		// implementation limitations, it might not be
+		// possible to fail TLS negotiation. An endpoint MUST
+		// immediately terminate an HTTP/2 connection that
+		// does not meet the TLS requirements described in
+		// this section with a connection error (Section
+		// 5.4.1) of type INADEQUATE_SECURITY.
+		if sc.tlsState.Version < tls.VersionTLS12 {
+			sc.rejectConn(ErrCodeInadequateSecurity, "TLS version too low")
+			return
+		}
+
+		if sc.tlsState.ServerName == "" {
+			// Client must use SNI, but we don't enforce that anymore,
+			// since it was causing problems when connecting to bare IP
+			// addresses during development.
+			//
+			// TODO: optionally enforce? Or enforce at the time we receive
+			// a new request, and verify the ServerName matches the :authority?
+			// But that precludes proxy situations, perhaps.
+			//
+			// So for now, do nothing here again.
+		}
+
+		if !conf.PermitProhibitedCipherSuites && isBadCipher(sc.tlsState.CipherSuite) {
+			// "Endpoints MAY choose to generate a connection error
+			// (Section 5.4.1) of type INADEQUATE_SECURITY if one of
+			// the prohibited cipher suites are negotiated."
+			//
+			// We choose that. In my opinion, the spec is weak
+			// here. It also says both parties must support at least
+			// TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 so there's no
+			// excuses here. If we really must, we could allow an
+			// "AllowInsecureWeakCiphers" option on the server later.
+			// Let's see how it plays out first.
+			sc.rejectConn(ErrCodeInadequateSecurity, fmt.Sprintf("Prohibited TLS 1.2 Cipher Suite: %x", sc.tlsState.CipherSuite))
+			return
+		}
+	}
+
+	if opts.Settings != nil {
+		fr := &SettingsFrame{
+			FrameHeader: FrameHeader{valid: true},
+			p:           opts.Settings,
+		}
+		if err := fr.ForeachSetting(sc.processSetting); err != nil {
+			sc.rejectConn(ErrCodeProtocol, "invalid settings")
+			return
+		}
+		opts.Settings = nil
+	}
+
+	if opts.UpgradeRequest != nil {
+		sc.upgradeRequest(opts.UpgradeRequest)
+		opts.UpgradeRequest = nil
+	}
+
+	sc.serve(conf)
+}
+
+func serverConnBaseContext(c net.Conn, opts *ServeConnOpts) (ctx context.Context, cancel func()) {
+	return context.WithCancel(opts.context())
+}
+
+func (sc *serverConn) rejectConn(err ErrCode, debug string) {
+	sc.vlogf("http2: server rejecting conn: %v, %s", err, debug)
+	// ignoring errors. hanging up anyway.
+	sc.framer.WriteGoAway(0, err, []byte(debug))
+	sc.bw.Flush()
+	sc.conn.Close()
+}
+
+type serverConn struct {
+	// Immutable:
+	srv              *Server
+	hs               ServerConfig
+	conn             net.Conn
+	bw               *bufferedWriter // writing to conn
+	handler          Handler
+	baseCtx          context.Context
+	framer           *Framer
+	doneServing      chan struct{}          // closed when serverConn.serve ends
+	readFrameCh      chan readFrameResult   // written by serverConn.readFrames
+	wantWriteFrameCh chan FrameWriteRequest // from handlers -> serve
+	wroteFrameCh     chan frameWriteResult  // from writeFrameAsync -> serve, tickles more frame writes
+	bodyReadCh       chan bodyReadMsg       // from handlers -> serve
+	serveMsgCh       chan any               // misc messages & code to send to / run on the serve loop
+	flow             connOutflow            // conn-wide (not stream-specific) outbound flow control
+	inflow           inflow                 // conn-wide inbound flow control
+	tlsState         *tls.ConnectionState   // shared by all handlers, like net/http
+	remoteAddrStr    string
+	writeSched       WriteScheduler
+	countErrorFunc   func(errType string)
+
+	// Everything following is owned by the serve loop; use serveG.check():
+	serveG                      goroutineLock // used to verify funcs are on serve()
+	pushEnabled                 bool
+	sawClientPreface            bool // preface has already been read, used in h2c upgrade
+	sawFirstSettings            bool // got the initial SETTINGS frame after the preface
+	needToSendSettingsAck       bool
+	unackedSettings             int    // how many SETTINGS have we sent without ACKs?
+	pendingEncoderTableSize     bool   // peer changed SETTINGS_HEADER_TABLE_SIZE; apply to hpackEncoder before the next frame write
+	encoderTableSizeMin         uint32 // smallest SETTINGS_HEADER_TABLE_SIZE since the last apply
+	encoderTableSize            uint32 // latest SETTINGS_HEADER_TABLE_SIZE
+	queuedControlFrames         int    // control frames in the writeSched queue
+	clientMaxStreams            uint32 // SETTINGS_MAX_CONCURRENT_STREAMS from client (our PUSH_PROMISE limit)
+	advMaxStreams               uint32 // our SETTINGS_MAX_CONCURRENT_STREAMS advertised the client
+	curClientStreams            uint32 // number of open streams initiated by the client
+	curPushedStreams            uint32 // number of open streams initiated by server push
+	curHandlers                 uint32 // number of running handler goroutines
+	maxClientStreamID           uint32 // max ever seen from client (odd), or 0 if there have been no client requests
+	maxPushPromiseID            uint32 // ID of the last push promise (even), or 0 if there have been no pushes
+	streams                     map[uint32]*stream
+	unstartedHandlers           []unstartedHandler
+	initialStreamRecvWindowSize int32
+	maxFrameSize                int32
+	peerMaxHeaderListSize       uint32            // zero means unknown (default)
+	canonHeader                 map[string]string // http2-lower-case -> Go-Canonical-Case
+	canonHeaderKeysSize         int               // canonHeader keys size in bytes
+	writingFrame                bool              // started writing a frame (on serve goroutine or separate)
+	writingFrameAsync           bool              // started a frame on its own goroutine but haven't heard back on wroteFrameCh
+	needsFrameFlush             bool              // last frame write wasn't a flush
+	inGoAway                    bool              // we've started to or sent GOAWAY
+	inFrameScheduleLoop         bool              // whether we're in the scheduleFrameWrite loop
+	needToSendGoAway            bool              // we need to schedule a GOAWAY frame write
+	pingSent                    bool
+	sentPingData                [8]byte
+	goAwayCode                  ErrCode
+	shutdownTimer               *time.Timer // nil until used
+	idleTimer                   *time.Timer // nil if unused
+	readIdleTimeout             time.Duration
+	pingTimeout                 time.Duration
+	readIdleTimer               *time.Timer // nil if unused
+
+	// Owned by the writeFrameAsync goroutine:
+	headerWriteBuf bytes.Buffer
+	hpackEncoder   *hpack.Encoder
+
+	// Used by startGracefulShutdown.
+	shutdownOnce sync.Once
+
+	// Used for RFC 9218 prioritization.
+	hasIntermediary bool // connection is done via an intermediary / proxy
+	priorityAware   bool // the client has sent priority signal, meaning that it is aware of it.
+}
+
+func (sc *serverConn) writeSchedIgnoresRFC7540() bool {
+	switch sc.writeSched.(type) {
+	case *priorityWriteSchedulerRFC9218:
+		return true
+	case *roundRobinWriteScheduler:
+		return true
+	default:
+		return false
+	}
+}
+
+const DefaultMaxHeaderBytes = 1 << 20 // keep this in sync with net/http
+
+func (sc *serverConn) maxHeaderListSize() uint32 {
+	n := sc.hs.MaxHeaderBytes()
+	if n <= 0 {
+		n = DefaultMaxHeaderBytes
+	}
+	return uint32(adjustHTTP1MaxHeaderSize(int64(n)))
+}
+
+func (sc *serverConn) curOpenStreams() uint32 {
+	sc.serveG.check()
+	return sc.curClientStreams + sc.curPushedStreams
+}
+
+// stream represents a stream. This is the minimal metadata needed by
+// the serve goroutine. Most of the actual stream state is owned by
+// the http.Handler's goroutine in the responseWriter. Because the
+// responseWriter's responseWriterState is recycled at the end of a
+// handler, this struct intentionally has no pointer to the
+// *responseWriter{,State} itself, as the Handler ending nils out the
+// responseWriter's state field.
+type stream struct {
+	// immutable:
+	sc        *serverConn
+	id        uint32
+	body      *pipe // non-nil if expecting DATA frames
+	reqBody   *requestBody
+	cw        closeWaiter // closed wait stream transitions to closed state
+	ctx       context.Context
+	cancelCtx func()
+
+	// owned by serverConn's serve loop:
+	bodyBytes        int64   // body bytes seen so far
+	declBodyBytes    int64   // or -1 if undeclared
+	flow             outflow // limits writing from Handler to client
+	inflow           inflow  // what the client is allowed to POST/etc to us
+	state            streamState
+	resetQueued      bool        // RST_STREAM queued for write; set by sc.resetStream
+	gotTrailerHeader bool        // HEADER frame for trailers was seen
+	wroteHeaders     bool        // whether we wrote headers (not status 100)
+	readDeadline     *time.Timer // nil if unused
+	writeDeadline    *time.Timer // nil if unused
+	closeErr         error       // set before cw is closed
+
+	trailer    Header // accumulated trailers
+	reqTrailer Header // handler's Request.Trailer
+}
+
+func (sc *serverConn) Framer() *Framer  { return sc.framer }
+func (sc *serverConn) CloseConn() error { return sc.conn.Close() }
+func (sc *serverConn) Flush() error     { return sc.bw.Flush() }
+func (sc *serverConn) HeaderEncoder() (*hpack.Encoder, *bytes.Buffer) {
+	return sc.hpackEncoder, &sc.headerWriteBuf
+}
+
+func (sc *serverConn) state(streamID uint32) (streamState, *stream) {
+	sc.serveG.check()
+	// http://tools.ietf.org/html/rfc7540#section-5.1
+	if st, ok := sc.streams[streamID]; ok {
+		return st.state, st
+	}
+	// "The first use of a new stream identifier implicitly closes all
+	// streams in the "idle" state that might have been initiated by
+	// that peer with a lower-valued stream identifier. For example, if
+	// a client sends a HEADERS frame on stream 7 without ever sending a
+	// frame on stream 5, then stream 5 transitions to the "closed"
+	// state when the first frame for stream 7 is sent or received."
+	if streamID%2 == 1 {
+		if streamID <= sc.maxClientStreamID {
+			return stateClosed, nil
+		}
+	} else {
+		if streamID <= sc.maxPushPromiseID {
+			return stateClosed, nil
+		}
+	}
+	return stateIdle, nil
+}
+
+// setConnState calls the net/http ConnState hook for this connection, if configured.
+// Note that the net/http package does StateNew and StateClosed for us.
+// There is currently no plan for StateHijacked or hijacking HTTP/2 connections.
+func (sc *serverConn) setConnState(state ConnState) {
+	sc.hs.ConnState(sc.conn, state)
+}
+
+func (sc *serverConn) vlogf(format string, args ...any) {
+	if VerboseLogs {
+		sc.logf(format, args...)
+	}
+}
+
+func (sc *serverConn) logf(format string, args ...any) {
+	if lg := sc.hs.ErrorLog(); lg != nil {
+		lg.Printf(format, args...)
+	} else {
+		log.Printf(format, args...)
+	}
+}
+
+// errno returns v's underlying uintptr, else 0.
+//
+// TODO: remove this helper function once http2 can use build
+// tags. See comment in isClosedConnError.
+func errno(v error) uintptr {
+	if rv := reflect.ValueOf(v); rv.Kind() == reflect.Uintptr {
+		return uintptr(rv.Uint())
+	}
+	return 0
+}
+
+// isClosedConnError reports whether err is an error from use of a closed
+// network connection.
+func isClosedConnError(err error) bool {
+	if err == nil {
+		return false
+	}
+
+	if errors.Is(err, net.ErrClosed) {
+		return true
+	}
+
+	// TODO(bradfitz): x/tools/cmd/bundle doesn't really support
+	// build tags, so I can't make an http2_windows.go file with
+	// Windows-specific stuff. Fix that and move this, once we
+	// have a way to bundle this into std's net/http somehow.
+	if runtime.GOOS == "windows" {
+		if oe, ok := err.(*net.OpError); ok && oe.Op == "read" {
+			if se, ok := oe.Err.(*os.SyscallError); ok && se.Syscall == "wsarecv" {
+				const WSAECONNABORTED = 10053
+				const WSAECONNRESET = 10054
+				if n := errno(se.Err); n == WSAECONNRESET || n == WSAECONNABORTED {
+					return true
+				}
+			}
+		}
+	}
+	return false
+}
+
+func (sc *serverConn) condlogf(err error, format string, args ...any) {
+	if err == nil {
+		return
+	}
+	if err == io.EOF || err == io.ErrUnexpectedEOF || isClosedConnError(err) || err == errPrefaceTimeout {
+		// Boring, expected errors.
+		sc.vlogf(format, args...)
+	} else {
+		sc.logf(format, args...)
+	}
+}
+
+// maxCachedCanonicalHeadersKeysSize is an arbitrarily-chosen limit on the size
+// of the entries in the canonHeader cache.
+// This should be larger than the size of unique, uncommon header keys likely to
+// be sent by the peer, while not so high as to permit unreasonable memory usage
+// if the peer sends an unbounded number of unique header keys.
+const maxCachedCanonicalHeadersKeysSize = 2048
+
+func (sc *serverConn) canonicalHeader(v string) string {
+	sc.serveG.check()
+	cv, ok := httpcommon.CachedCanonicalHeader(v)
+	if ok {
+		return cv
+	}
+	cv, ok = sc.canonHeader[v]
+	if ok {
+		return cv
+	}
+	if sc.canonHeader == nil {
+		sc.canonHeader = make(map[string]string)
+	}
+	cv = textproto.CanonicalMIMEHeaderKey(v)
+	size := 100 + len(v)*2 // 100 bytes of map overhead + key + value
+	if sc.canonHeaderKeysSize+size <= maxCachedCanonicalHeadersKeysSize {
+		sc.canonHeader[v] = cv
+		sc.canonHeaderKeysSize += size
+	}
+	return cv
+}
+
+type readFrameResult struct {
+	f   Frame // valid until readMore is called
+	err error
+
+	// readMore should be called once the consumer no longer needs or
+	// retains f. After readMore, f is invalid and more frames can be
+	// read.
+	readMore func()
+}
+
+// readFrames is the loop that reads incoming frames.
+// It takes care to only read one frame at a time, blocking until the
+// consumer is done with the frame.
+// It's run on its own goroutine.
+func (sc *serverConn) readFrames() {
+	gate := make(chan struct{})
+	gateDone := func() { gate <- struct{}{} }
+	for {
+		f, err := sc.framer.ReadFrame()
+		select {
+		case sc.readFrameCh <- readFrameResult{f, err, gateDone}:
+		case <-sc.doneServing:
+			return
+		}
+		select {
+		case <-gate:
+		case <-sc.doneServing:
+			return
+		}
+		if terminalReadFrameError(err) {
+			return
+		}
+	}
+}
+
+// frameWriteResult is the message passed from writeFrameAsync to the serve goroutine.
+type frameWriteResult struct {
+	_   incomparable
+	wr  FrameWriteRequest // what was written (or attempted)
+	err error             // result of the writeFrame call
+}
+
+// writeFrameAsync runs in its own goroutine and writes a single frame
+// and then reports when it's done.
+// At most one goroutine can be running writeFrameAsync at a time per
+// serverConn.
+func (sc *serverConn) writeFrameAsync(wr FrameWriteRequest, wd *writeData) {
+	var err error
+	if wd == nil {
+		err = wr.write.writeFrame(sc)
+	} else {
+		err = sc.framer.endWrite()
+	}
+	sc.wroteFrameCh <- frameWriteResult{wr: wr, err: err}
+}
+
+func (sc *serverConn) closeAllStreamsOnConnClose() {
+	sc.serveG.check()
+	for _, st := range sc.streams {
+		sc.closeStream(st, errClientDisconnected)
+	}
+}
+
+func (sc *serverConn) stopShutdownTimer() {
+	sc.serveG.check()
+	if t := sc.shutdownTimer; t != nil {
+		t.Stop()
+	}
+}
+
+func (sc *serverConn) notePanic() {
+	// Note: this is for serverConn.serve panicking, not http.Handler code.
+	if testHookOnPanicMu != nil {
+		testHookOnPanicMu.Lock()
+		defer testHookOnPanicMu.Unlock()
+	}
+	if testHookOnPanic != nil {
+		if e := recover(); e != nil {
+			if testHookOnPanic(sc, e) {
+				panic(e)
+			}
+		}
+	}
+}
+
+func (sc *serverConn) serve(conf Config) {
+	sc.serveG.check()
+	defer sc.notePanic()
+	defer sc.conn.Close()
+	defer sc.closeAllStreamsOnConnClose()
+	defer sc.stopShutdownTimer()
+	defer close(sc.doneServing) // unblocks handlers trying to send
+
+	if VerboseLogs {
+		sc.vlogf("http2: server connection from %v on %p", sc.conn.RemoteAddr(), sc.hs)
+	}
+
+	settings := writeSettings{
+		{SettingMaxFrameSize, uint32(conf.MaxReadFrameSize)},
+		{SettingMaxConcurrentStreams, sc.advMaxStreams},
+		{SettingMaxHeaderListSize, sc.maxHeaderListSize()},
+		{SettingHeaderTableSize, uint32(conf.MaxDecoderHeaderTableSize)},
+		{SettingInitialWindowSize, uint32(sc.initialStreamRecvWindowSize)},
+	}
+	if !disableExtendedConnectProtocol {
+		settings = append(settings, Setting{SettingEnableConnectProtocol, 1})
+	}
+	if sc.writeSchedIgnoresRFC7540() {
+		settings = append(settings, Setting{SettingNoRFC7540Priorities, 1})
+	}
+	sc.writeFrame(FrameWriteRequest{
+		write: settings,
+	})
+	sc.unackedSettings++
+
+	// Each connection starts with initialWindowSize inflow tokens.
+	// If a higher value is configured, we add more tokens.
+	if diff := conf.MaxReceiveBufferPerConnection - initialWindowSize; diff > 0 {
+		sc.sendWindowUpdate(nil, int(diff))
+	}
+
+	if err := sc.readPreface(); err != nil {
+		sc.condlogf(err, "http2: server: error reading preface from client %v: %v", sc.conn.RemoteAddr(), err)
+		return
+	}
+	// Now that we've got the preface, get us out of the
+	// "StateNew" state. We can't go directly to idle, though.
+	// Active means we read some data and anticipate a request. We'll
+	// do another Active when we get a HEADERS frame.
+	sc.setConnState(ConnStateActive)
+	sc.setConnState(ConnStateIdle)
+
+	if idle := sc.hs.IdleTimeout(); idle > 0 {
+		sc.idleTimer = time.AfterFunc(idle, sc.onIdleTimer)
+		defer sc.idleTimer.Stop()
+	}
+
+	if conf.SendPingTimeout > 0 {
+		sc.readIdleTimeout = conf.SendPingTimeout
+		sc.readIdleTimer = time.AfterFunc(conf.SendPingTimeout, sc.onReadIdleTimer)
+		defer sc.readIdleTimer.Stop()
+	}
+
+	go sc.readFrames() // closed by defer sc.conn.Close above
+
+	settingsTimer := time.AfterFunc(firstSettingsTimeout, sc.onSettingsTimer)
+	defer settingsTimer.Stop()
+
+	lastFrameTime := time.Now()
+	loopNum := 0
+	for {
+		loopNum++
+		select {
+		case wr := <-sc.wantWriteFrameCh:
+			if se, ok := wr.write.(StreamError); ok {
+				sc.resetStream(se)
+				break
+			}
+			sc.writeFrame(wr)
+		case res := <-sc.wroteFrameCh:
+			sc.wroteFrame(res)
+		case res := <-sc.readFrameCh:
+			lastFrameTime = time.Now()
+			// Process any written frames before reading new frames from the client since a
+			// written frame could have triggered a new stream to be started.
+			if sc.writingFrameAsync {
+				select {
+				case wroteRes := <-sc.wroteFrameCh:
+					sc.wroteFrame(wroteRes)
+				default:
+				}
+			}
+			if !sc.processFrameFromReader(res) {
+				return
+			}
+			res.readMore()
+			if settingsTimer != nil {
+				settingsTimer.Stop()
+				settingsTimer = nil
+			}
+		case m := <-sc.bodyReadCh:
+			sc.noteBodyRead(m.st, m.n)
+		case msg := <-sc.serveMsgCh:
+			switch v := msg.(type) {
+			case func(int):
+				v(loopNum) // for testing
+			case *serverMessage:
+				switch v {
+				case settingsTimerMsg:
+					sc.logf("timeout waiting for SETTINGS frames from %v", sc.conn.RemoteAddr())
+					return
+				case idleTimerMsg:
+					sc.vlogf("connection is idle")
+					sc.goAway(ErrCodeNo)
+				case readIdleTimerMsg:
+					sc.handlePingTimer(lastFrameTime)
+				case shutdownTimerMsg:
+					sc.vlogf("GOAWAY close timer fired; closing conn from %v", sc.conn.RemoteAddr())
+					return
+				case gracefulShutdownMsg:
+					sc.startGracefulShutdownInternal()
+				case handlerDoneMsg:
+					sc.handlerDone()
+				default:
+					panic("unknown timer")
+				}
+			case *startPushRequest:
+				sc.startPush(v)
+			case func(*serverConn):
+				v(sc)
+			default:
+				panic(fmt.Sprintf("unexpected type %T", v))
+			}
+		}
+
+		// If the peer is causing us to generate a lot of control frames,
+		// but not reading them from us, assume they are trying to make us
+		// run out of memory.
+		if sc.queuedControlFrames > maxQueuedControlFrames {
+			sc.vlogf("http2: too many control frames in send queue, closing connection")
+			return
+		}
+
+		// Start the shutdown timer after sending a GOAWAY. When sending GOAWAY
+		// with no error code (graceful shutdown), don't start the timer until
+		// all open streams have been completed.
+		sentGoAway := sc.inGoAway && !sc.needToSendGoAway && !sc.writingFrame
+		gracefulShutdownComplete := sc.goAwayCode == ErrCodeNo && sc.curOpenStreams() == 0
+		if sentGoAway && sc.shutdownTimer == nil && (sc.goAwayCode != ErrCodeNo || gracefulShutdownComplete) {
+			sc.shutDownIn(goAwayTimeout)
+		}
+	}
+}
+
+func (sc *serverConn) handlePingTimer(lastFrameReadTime time.Time) {
+	if sc.pingSent {
+		sc.logf("timeout waiting for PING response")
+		if f := sc.countErrorFunc; f != nil {
+			f("conn_close_lost_ping")
+		}
+		sc.conn.Close()
+		return
+	}
+
+	pingAt := lastFrameReadTime.Add(sc.readIdleTimeout)
+	now := time.Now()
+	if pingAt.After(now) {
+		// We received frames since arming the ping timer.
+		// Reset it for the next possible timeout.
+		sc.readIdleTimer.Reset(pingAt.Sub(now))
+		return
+	}
+
+	sc.pingSent = true
+	// Ignore crypto/rand.Read errors: It generally can't fail, and worse case if it does
+	// is we send a PING frame containing 0s.
+	_, _ = rand.Read(sc.sentPingData[:])
+	sc.writeFrame(FrameWriteRequest{
+		write: &writePing{data: sc.sentPingData},
+	})
+	sc.readIdleTimer.Reset(sc.pingTimeout)
+}
+
+type serverMessage int
+
+// Message values sent to serveMsgCh.
+var (
+	settingsTimerMsg    = new(serverMessage)
+	idleTimerMsg        = new(serverMessage)
+	readIdleTimerMsg    = new(serverMessage)
+	shutdownTimerMsg    = new(serverMessage)
+	gracefulShutdownMsg = new(serverMessage)
+	handlerDoneMsg      = new(serverMessage)
+)
+
+func (sc *serverConn) onSettingsTimer() { sc.sendServeMsg(settingsTimerMsg) }
+func (sc *serverConn) onIdleTimer()     { sc.sendServeMsg(idleTimerMsg) }
+func (sc *serverConn) onReadIdleTimer() { sc.sendServeMsg(readIdleTimerMsg) }
+func (sc *serverConn) onShutdownTimer() { sc.sendServeMsg(shutdownTimerMsg) }
+
+func (sc *serverConn) sendServeMsg(msg any) {
+	sc.serveG.checkNotOn() // NOT
+	select {
+	case sc.serveMsgCh <- msg:
+	case <-sc.doneServing:
+	}
+}
+
+var errPrefaceTimeout = errors.New("timeout waiting for client preface")
+
+// readPreface reads the ClientPreface greeting from the peer or
+// returns errPrefaceTimeout on timeout, or an error if the greeting
+// is invalid.
+func (sc *serverConn) readPreface() error {
+	if sc.sawClientPreface {
+		return nil
+	}
+	errc := make(chan error, 1)
+	go func() {
+		// Read the client preface
+		buf := make([]byte, len(ClientPreface))
+		if _, err := io.ReadFull(sc.conn, buf); err != nil {
+			errc <- err
+		} else if !bytes.Equal(buf, clientPreface) {
+			errc <- fmt.Errorf("bogus greeting %q", buf)
+		} else {
+			errc <- nil
+		}
+	}()
+	timer := time.NewTimer(prefaceTimeout) // TODO: configurable on *Server?
+	defer timer.Stop()
+	select {
+	case <-timer.C:
+		return errPrefaceTimeout
+	case err := <-errc:
+		if err == nil {
+			if VerboseLogs {
+				sc.vlogf("http2: server: client %v said hello", sc.conn.RemoteAddr())
+			}
+		}
+		return err
+	}
+}
+
+var writeDataPool = sync.Pool{
+	New: func() any { return new(writeData) },
+}
+
+// writeDataFromHandler writes DATA response frames from a handler on
+// the given stream.
+func (sc *serverConn) writeDataFromHandler(stream *stream, data []byte, endStream bool) error {
+	ch := sc.srv.getErrChan()
+	writeArg := writeDataPool.Get().(*writeData)
+	*writeArg = writeData{stream.id, data, endStream}
+	err := sc.writeFrameFromHandler(FrameWriteRequest{
+		write:  writeArg,
+		stream: stream,
+		done:   ch,
+	})
+	if err != nil {
+		return err
+	}
+	var frameWriteDone bool // the frame write is done (successfully or not)
+	select {
+	case err = <-ch:
+		frameWriteDone = true
+	case <-sc.doneServing:
+		return errClientDisconnected
+	case <-stream.cw:
+		// If both ch and stream.cw were ready (as might
+		// happen on the final Write after an http.Handler
+		// ends), prefer the write result. Otherwise this
+		// might just be us successfully closing the stream.
+		// The writeFrameAsync and serve goroutines guarantee
+		// that the ch send will happen before the stream.cw
+		// close.
+		select {
+		case err = <-ch:
+			frameWriteDone = true
+		default:
+			return errStreamClosed
+		}
+	}
+	sc.srv.putErrChan(ch)
+	if frameWriteDone {
+		writeDataPool.Put(writeArg)
+	}
+	return err
+}
+
+// writeFrameFromHandler sends wr to sc.wantWriteFrameCh, but aborts
+// if the connection has gone away.
+//
+// This must not be run from the serve goroutine itself, else it might
+// deadlock writing to sc.wantWriteFrameCh (which is only mildly
+// buffered and is read by serve itself). If you're on the serve
+// goroutine, call writeFrame instead.
+func (sc *serverConn) writeFrameFromHandler(wr FrameWriteRequest) error {
+	sc.serveG.checkNotOn() // NOT
+	select {
+	case sc.wantWriteFrameCh <- wr:
+		return nil
+	case <-sc.doneServing:
+		// Serve loop is gone.
+		// Client has closed their connection to the server.
+		return errClientDisconnected
+	}
+}
+
+// writeFrame schedules a frame to write and sends it if there's nothing
+// already being written.
+//
+// There is no pushback here (the serve goroutine never blocks). It's
+// the http.Handlers that block, waiting for their previous frames to
+// make it onto the wire
+//
+// If you're not on the serve goroutine, use writeFrameFromHandler instead.
+func (sc *serverConn) writeFrame(wr FrameWriteRequest) {
+	sc.serveG.check()
+
+	// If true, wr will not be written and wr.done will not be signaled.
+	var ignoreWrite bool
+
+	// We are not allowed to write frames on closed streams. RFC 7540 Section
+	// 5.1.1 says: "An endpoint MUST NOT send frames other than PRIORITY on
+	// a closed stream." Our server never sends PRIORITY, so that exception
+	// does not apply.
+	//
+	// The serverConn might close an open stream while the stream's handler
+	// is still running. For example, the server might close a stream when it
+	// receives bad data from the client. If this happens, the handler might
+	// attempt to write a frame after the stream has been closed (since the
+	// handler hasn't yet been notified of the close). In this case, we simply
+	// ignore the frame. The handler will notice that the stream is closed when
+	// it waits for the frame to be written.
+	//
+	// As an exception to this rule, we allow sending RST_STREAM after close.
+	// This allows us to immediately reject new streams without tracking any
+	// state for those streams (except for the queued RST_STREAM frame). This
+	// may result in duplicate RST_STREAMs in some cases, but the client should
+	// ignore those.
+	if wr.StreamID() != 0 {
+		_, isReset := wr.write.(StreamError)
+		if state, _ := sc.state(wr.StreamID()); state == stateClosed && !isReset {
+			ignoreWrite = true
+		}
+	}
+
+	// Don't send a 100-continue response if we've already sent headers.
+	// See golang.org/issue/14030.
+	switch wr.write.(type) {
+	case *writeResHeaders:
+		wr.stream.wroteHeaders = true
+	case write100ContinueHeadersFrame:
+		if wr.stream.wroteHeaders {
+			// We do not need to notify wr.done because this frame is
+			// never written with wr.done != nil.
+			if wr.done != nil {
+				panic("wr.done != nil for write100ContinueHeadersFrame")
+			}
+			ignoreWrite = true
+		}
+	}
+
+	if !ignoreWrite {
+		if wr.isControl() {
+			sc.queuedControlFrames++
+			// For extra safety, detect wraparounds, which should not happen,
+			// and pull the plug.
+			if sc.queuedControlFrames < 0 {
+				sc.conn.Close()
+			}
+		}
+		sc.writeSched.Push(wr)
+	}
+	sc.scheduleFrameWrite()
+}
+
+// startFrameWrite starts a goroutine to write wr (in a separate
+// goroutine since that might block on the network), and updates the
+// serve goroutine's state about the world, updated from info in wr.
+func (sc *serverConn) startFrameWrite(wr FrameWriteRequest) {
+	sc.serveG.check()
+	if sc.writingFrame {
+		panic("internal error: can only be writing one frame at a time")
+	}
+
+	st := wr.stream
+	if st != nil {
+		switch st.state {
+		case stateHalfClosedLocal:
+			switch wr.write.(type) {
+			case StreamError, handlerPanicRST, writeWindowUpdate:
+				// RFC 7540 Section 5.1 allows sending RST_STREAM, PRIORITY, and WINDOW_UPDATE
+				// in this state. (We never send PRIORITY from the server, so that is not checked.)
+			default:
+				panic(fmt.Sprintf("internal error: attempt to send frame on a half-closed-local stream: %v", wr))
+			}
+		case stateClosed:
+			panic(fmt.Sprintf("internal error: attempt to send frame on a closed stream: %v", wr))
+		}
+	}
+	if wpp, ok := wr.write.(*writePushPromise); ok {
+		var err error
+		wpp.promisedID, err = wpp.allocatePromisedID()
+		if err != nil {
+			sc.writingFrameAsync = false
+			wr.replyToWriter(err)
+			return
+		}
+	}
+
+	sc.writingFrame = true
+	sc.needsFrameFlush = true
+	if sc.pendingEncoderTableSize {
+		// hpackEncoder may be in use by writeFrameAsync, so SETTINGS
+		// changes to it are deferred until no frame is being written.
+		// Replaying the smallest size before the latest one keeps the
+		// encoder's view identical to having applied every change
+		// (RFC 7541, Section 4.2).
+		sc.pendingEncoderTableSize = false
+		sc.hpackEncoder.SetMaxDynamicTableSize(sc.encoderTableSizeMin)
+		sc.hpackEncoder.SetMaxDynamicTableSize(sc.encoderTableSize)
+	}
+	if wr.write.staysWithinBuffer(sc.bw.Available()) {
+		sc.writingFrameAsync = false
+		err := wr.write.writeFrame(sc)
+		sc.wroteFrame(frameWriteResult{wr: wr, err: err})
+	} else if wd, ok := wr.write.(*writeData); ok {
+		// Encode the frame in the serve goroutine, to ensure we don't have
+		// any lingering asynchronous references to data passed to Write.
+		// See https://go.dev/issue/58446.
+		sc.framer.startWriteDataPadded(wd.streamID, wd.endStream, wd.p, nil)
+		sc.writingFrameAsync = true
+		go sc.writeFrameAsync(wr, wd)
+	} else {
+		sc.writingFrameAsync = true
+		go sc.writeFrameAsync(wr, nil)
+	}
+}
+
+// errHandlerPanicked is the error given to any callers blocked in a read from
+// Request.Body when the main goroutine panics. Since most handlers read in the
+// main ServeHTTP goroutine, this will show up rarely.
+var errHandlerPanicked = errors.New("http2: handler panicked")
+
+// wroteFrame is called on the serve goroutine with the result of
+// whatever happened on writeFrameAsync.
+func (sc *serverConn) wroteFrame(res frameWriteResult) {
+	sc.serveG.check()
+	if !sc.writingFrame {
+		panic("internal error: expected to be already writing a frame")
+	}
+	sc.writingFrame = false
+	sc.writingFrameAsync = false
+
+	if res.err != nil {
+		sc.conn.Close()
+	}
+
+	wr := res.wr
+
+	if writeEndsStream(wr.write) {
+		st := wr.stream
+		if st == nil {
+			panic("internal error: expecting non-nil stream")
+		}
+		switch st.state {
+		case stateOpen:
+			// Here we would go to stateHalfClosedLocal in
+			// theory, but since our handler is done and
+			// the net/http package provides no mechanism
+			// for closing a ResponseWriter while still
+			// reading data (see possible TODO at top of
+			// this file), we go into closed state here
+			// anyway, after telling the peer we're
+			// hanging up on them. We'll transition to
+			// stateClosed after the RST_STREAM frame is
+			// written.
+			st.state = stateHalfClosedLocal
+			// Section 8.1: a server MAY request that the client abort
+			// transmission of a request without error by sending a
+			// RST_STREAM with an error code of NO_ERROR after sending
+			// a complete response.
+			sc.resetStream(streamError(st.id, ErrCodeNo))
+		case stateHalfClosedRemote:
+			sc.closeStream(st, errHandlerComplete)
+		}
+	} else {
+		switch v := wr.write.(type) {
+		case StreamError:
+			// st may be unknown if the RST_STREAM was generated to reject bad input.
+			if st, ok := sc.streams[v.StreamID]; ok {
+				sc.closeStream(st, v)
+			}
+		case handlerPanicRST:
+			sc.closeStream(wr.stream, errHandlerPanicked)
+		}
+	}
+
+	// Reply (if requested) to unblock the ServeHTTP goroutine.
+	wr.replyToWriter(res.err)
+
+	sc.scheduleFrameWrite()
+}
+
+// scheduleFrameWrite tickles the frame writing scheduler.
+//
+// If a frame is already being written, nothing happens. This will be called again
+// when the frame is done being written.
+//
+// If a frame isn't being written and we need to send one, the best frame
+// to send is selected by writeSched.
+//
+// If a frame isn't being written and there's nothing else to send, we
+// flush the write buffer.
+func (sc *serverConn) scheduleFrameWrite() {
+	sc.serveG.check()
+	if sc.writingFrame || sc.inFrameScheduleLoop {
+		return
+	}
+	sc.inFrameScheduleLoop = true
+	for !sc.writingFrameAsync {
+		if sc.flow.flowErr && (!sc.inGoAway || sc.goAwayCode == ErrCodeNo) {
+			sc.inGoAway = true
+			sc.needToSendGoAway = true
+			sc.goAwayCode = ErrCodeFlowControl
+		}
+		if sc.needToSendGoAway {
+			sc.needToSendGoAway = false
+			sc.startFrameWrite(FrameWriteRequest{
+				write: &writeGoAway{
+					maxStreamID: sc.maxClientStreamID,
+					code:        sc.goAwayCode,
+				},
+			})
+			continue
+		}
+		if sc.needToSendSettingsAck {
+			sc.needToSendSettingsAck = false
+			sc.startFrameWrite(FrameWriteRequest{write: writeSettingsAck{}})
+			continue
+		}
+		if !sc.inGoAway || sc.goAwayCode == ErrCodeNo {
+			if wr, ok := sc.writeSched.Pop(); ok {
+				if wr.isControl() {
+					sc.queuedControlFrames--
+				}
+				sc.startFrameWrite(wr)
+				continue
+			}
+			if sc.flow.flowErr {
+				continue
+			}
+		}
+		if sc.needsFrameFlush {
+			sc.startFrameWrite(FrameWriteRequest{write: flushFrameWriter{}})
+			sc.needsFrameFlush = false // after startFrameWrite, since it sets this true
+			continue
+		}
+		break
+	}
+	sc.inFrameScheduleLoop = false
+}
+
+// startGracefulShutdown gracefully shuts down a connection. This
+// sends GOAWAY with ErrCodeNo to tell the client we're gracefully
+// shutting down. The connection isn't closed until all current
+// streams are done.
+//
+// startGracefulShutdown returns immediately; it does not wait until
+// the connection has shut down.
+func (sc *serverConn) startGracefulShutdown() {
+	sc.serveG.checkNotOn() // NOT
+	sc.shutdownOnce.Do(func() { sc.sendServeMsg(gracefulShutdownMsg) })
+}
+
+// After sending GOAWAY with an error code (non-graceful shutdown), the
+// connection will close after goAwayTimeout.
+//
+// If we close the connection immediately after sending GOAWAY, there may
+// be unsent data in our kernel receive buffer, which will cause the kernel
+// to send a TCP RST on close() instead of a FIN. This RST will abort the
+// connection immediately, whether or not the client had received the GOAWAY.
+//
+// Ideally we should delay for at least 1 RTT + epsilon so the client has
+// a chance to read the GOAWAY and stop sending messages. Measuring RTT
+// is hard, so we approximate with 1 second. See golang.org/issue/18701.
+//
+// This is a var so it can be shorter in tests, where all requests uses the
+// loopback interface making the expected RTT very small.
+//
+// TODO: configurable?
+var goAwayTimeout = 1 * time.Second
+
+func (sc *serverConn) startGracefulShutdownInternal() {
+	sc.goAway(ErrCodeNo)
+}
+
+func (sc *serverConn) goAway(code ErrCode) {
+	sc.serveG.check()
+	if sc.inGoAway {
+		if sc.goAwayCode == ErrCodeNo {
+			sc.goAwayCode = code
+		}
+		return
+	}
+	sc.inGoAway = true
+	sc.needToSendGoAway = true
+	sc.goAwayCode = code
+	sc.scheduleFrameWrite()
+}
+
+func (sc *serverConn) shutDownIn(d time.Duration) {
+	sc.serveG.check()
+	sc.shutdownTimer = time.AfterFunc(d, sc.onShutdownTimer)
+}
+
+func (sc *serverConn) resetStream(se StreamError) {
+	sc.serveG.check()
+	sc.writeFrame(FrameWriteRequest{write: se})
+	if st, ok := sc.streams[se.StreamID]; ok {
+		st.resetQueued = true
+	}
+}
+
+// processFrameFromReader processes the serve loop's read from readFrameCh from the
+// frame-reading goroutine.
+// processFrameFromReader returns whether the connection should be kept open.
+func (sc *serverConn) processFrameFromReader(res readFrameResult) bool {
+	sc.serveG.check()
+	err := res.err
+	if err != nil {
+		if err == ErrFrameTooLarge {
+			sc.goAway(ErrCodeFrameSize)
+			return true // goAway will close the loop
+		}
+		clientGone := err == io.EOF || err == io.ErrUnexpectedEOF || isClosedConnError(err)
+		if clientGone {
+			// TODO: could we also get into this state if
+			// the peer does a half close
+			// (e.g. CloseWrite) because they're done
+			// sending frames but they're still wanting
+			// our open replies?  Investigate.
+			// TODO: add CloseWrite to crypto/tls.Conn first
+			// so we have a way to test this? I suppose
+			// just for testing we could have a non-TLS mode.
+			return false
+		}
+	} else {
+		f := res.f
+		if VerboseLogs {
+			sc.vlogf("http2: server read frame %v", summarizeFrame(f))
+		}
+		err = sc.processFrame(f)
+		if err == nil {
+			return true
+		}
+	}
+
+	switch ev := err.(type) {
+	case StreamError:
+		sc.resetStream(ev)
+		return true
+	case goAwayFlowError:
+		sc.goAway(ErrCodeFlowControl)
+		return true
+	case ConnectionError:
+		if res.f != nil {
+			if id := res.f.Header().StreamID; id > sc.maxClientStreamID {
+				sc.maxClientStreamID = id
+			}
+		}
+		sc.logf("http2: server connection error from %v: %v", sc.conn.RemoteAddr(), ev)
+		sc.goAway(ErrCode(ev))
+		return true // goAway will handle shutdown
+	default:
+		if res.err != nil {
+			sc.vlogf("http2: server closing client connection; error reading frame from client %s: %v", sc.conn.RemoteAddr(), err)
+		} else {
+			sc.logf("http2: server closing client connection: %v", err)
+		}
+		return false
+	}
+}
+
+func (sc *serverConn) processFrame(f Frame) error {
+	sc.serveG.check()
+
+	// First frame received must be SETTINGS.
+	if !sc.sawFirstSettings {
+		if _, ok := f.(*SettingsFrame); !ok {
+			return sc.countError("first_settings", ConnectionError(ErrCodeProtocol))
+		}
+		sc.sawFirstSettings = true
+	}
+
+	// Discard frames for streams initiated after the identified last
+	// stream sent in a GOAWAY, or all frames after sending an error.
+	// We still need to return connection-level flow control for DATA frames.
+	// RFC 9113 Section 6.8.
+	if sc.inGoAway && (sc.goAwayCode != ErrCodeNo || f.Header().StreamID > sc.maxClientStreamID) {
+
+		if f, ok := f.(*DataFrame); ok {
+			if !sc.inflow.take(f.Length) {
+				return sc.countError("data_flow", streamError(f.Header().StreamID, ErrCodeFlowControl))
+			}
+			sc.sendWindowUpdate(nil, int(f.Length)) // conn-level
+		}
+		return nil
+	}
+
+	switch f := f.(type) {
+	case *SettingsFrame:
+		return sc.processSettings(f)
+	case *MetaHeadersFrame:
+		return sc.processHeaders(f)
+	case *WindowUpdateFrame:
+		return sc.processWindowUpdate(f)
+	case *PingFrame:
+		return sc.processPing(f)
+	case *DataFrame:
+		return sc.processData(f)
+	case *RSTStreamFrame:
+		return sc.processResetStream(f)
+	case *PriorityFrame:
+		return sc.processPriority(f)
+	case *GoAwayFrame:
+		return sc.processGoAway(f)
+	case *PushPromiseFrame:
+		// A client cannot push. Thus, servers MUST treat the receipt of a PUSH_PROMISE
+		// frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.
+		return sc.countError("push_promise", ConnectionError(ErrCodeProtocol))
+	case *PriorityUpdateFrame:
+		return sc.processPriorityUpdate(f)
+	default:
+		sc.vlogf("http2: server ignoring frame: %v", f.Header())
+		return nil
+	}
+}
+
+func (sc *serverConn) processPing(f *PingFrame) error {
+	sc.serveG.check()
+	if f.IsAck() {
+		if sc.pingSent && sc.sentPingData == f.Data {
+			// This is a response to a PING we sent.
+			sc.pingSent = false
+			sc.readIdleTimer.Reset(sc.readIdleTimeout)
+		}
+		// 6.7 PING: " An endpoint MUST NOT respond to PING frames
+		// containing this flag."
+		return nil
+	}
+	if f.StreamID != 0 {
+		// "PING frames are not associated with any individual
+		// stream. If a PING frame is received with a stream
+		// identifier field value other than 0x0, the recipient MUST
+		// respond with a connection error (Section 5.4.1) of type
+		// PROTOCOL_ERROR."
+		return sc.countError("ping_on_stream", ConnectionError(ErrCodeProtocol))
+	}
+	sc.writeFrame(FrameWriteRequest{write: writePingAck{f}})
+	return nil
+}
+
+func (sc *serverConn) processWindowUpdate(f *WindowUpdateFrame) error {
+	sc.serveG.check()
+	switch {
+	case f.StreamID != 0: // stream-level flow control
+		state, st := sc.state(f.StreamID)
+		if state == stateIdle {
+			// Section 5.1: "Receiving any frame other than HEADERS
+			// or PRIORITY on a stream in this state MUST be
+			// treated as a connection error (Section 5.4.1) of
+			// type PROTOCOL_ERROR."
+			return sc.countError("stream_idle", ConnectionError(ErrCodeProtocol))
+		}
+		if st == nil {
+			// "WINDOW_UPDATE can be sent by a peer that has sent a
+			// frame bearing the END_STREAM flag. This means that a
+			// receiver could receive a WINDOW_UPDATE frame on a "half
+			// closed (remote)" or "closed" stream. A receiver MUST
+			// NOT treat this as an error, see Section 5.1."
+			return nil
+		}
+		if !st.flow.add(int32(f.Increment)) {
+			if st.flow.conn.flowErr {
+				// This is a lazily-detected connection-level flow control error.
+				return sc.countError("bad_flow", ConnectionError(ErrCodeFlowControl))
+			}
+			return sc.countError("bad_flow", streamError(f.StreamID, ErrCodeFlowControl))
+		}
+	default: // connection-level flow control
+		if !sc.flow.add(int32(f.Increment)) {
+			return goAwayFlowError{}
+		}
+	}
+	sc.scheduleFrameWrite()
+	return nil
+}
+
+func (sc *serverConn) processResetStream(f *RSTStreamFrame) error {
+	sc.serveG.check()
+
+	state, st := sc.state(f.StreamID)
+	if state == stateIdle {
+		// 6.4 "RST_STREAM frames MUST NOT be sent for a
+		// stream in the "idle" state. If a RST_STREAM frame
+		// identifying an idle stream is received, the
+		// recipient MUST treat this as a connection error
+		// (Section 5.4.1) of type PROTOCOL_ERROR.
+		return sc.countError("reset_idle_stream", ConnectionError(ErrCodeProtocol))
+	}
+	if st != nil {
+		st.cancelCtx()
+		sc.closeStream(st, streamError(f.StreamID, f.ErrCode))
+	}
+	return nil
+}
+
+func (sc *serverConn) closeStream(st *stream, err error) {
+	sc.serveG.check()
+	if st.state == stateIdle || st.state == stateClosed {
+		panic(fmt.Sprintf("invariant; can't close stream in state %v", st.state))
+	}
+	st.state = stateClosed
+	if st.readDeadline != nil {
+		st.readDeadline.Stop()
+	}
+	if st.writeDeadline != nil {
+		st.writeDeadline.Stop()
+	}
+	if st.isPushed() {
+		sc.curPushedStreams--
+	} else {
+		sc.curClientStreams--
+	}
+	delete(sc.streams, st.id)
+	if len(sc.streams) == 0 {
+		sc.setConnState(ConnStateIdle)
+		idleTimeout := sc.hs.IdleTimeout()
+		if idleTimeout > 0 && sc.idleTimer != nil {
+			sc.idleTimer.Reset(idleTimeout)
+		}
+		if h1ServerKeepAlivesDisabled(sc.hs) {
+			sc.startGracefulShutdownInternal()
+		}
+	}
+	if p := st.body; p != nil {
+		p.CloseWithError(err)
+	}
+	if e, ok := err.(StreamError); ok {
+		if e.Cause != nil {
+			err = e.Cause
+		} else {
+			err = errStreamClosed
+		}
+	}
+	st.closeErr = err
+	st.cancelCtx()
+	st.cw.Close() // signals Handler's CloseNotifier, unblocks writes, etc
+	sc.writeSched.CloseStream(st.id)
+}
+
+func (sc *serverConn) processSettings(f *SettingsFrame) error {
+	sc.serveG.check()
+	if f.IsAck() {
+		sc.unackedSettings--
+		if sc.unackedSettings < 0 {
+			// Why is the peer ACKing settings we never sent?
+			// The spec doesn't mention this case, but
+			// hang up on them anyway.
+			return sc.countError("ack_mystery", ConnectionError(ErrCodeProtocol))
+		}
+		return nil
+	}
+	if f.NumSettings() > 100 || f.HasDuplicates() {
+		// This isn't actually in the spec, but hang up on
+		// suspiciously large settings frames or those with
+		// duplicate entries.
+		return sc.countError("settings_big_or_dups", ConnectionError(ErrCodeProtocol))
+	}
+	if err := f.ForeachSetting(sc.processSetting); err != nil {
+		return err
+	}
+	// TODO: judging by RFC 7540, Section 6.5.3 each SETTINGS frame should be
+	// acknowledged individually, even if multiple are received before the ACK.
+	sc.needToSendSettingsAck = true
+	sc.scheduleFrameWrite()
+	return nil
+}
+
+func (sc *serverConn) processSetting(s Setting) error {
+	sc.serveG.check()
+	if err := s.Valid(); err != nil {
+		return err
+	}
+	if VerboseLogs {
+		sc.vlogf("http2: server processing setting %v", s)
+	}
+	switch s.ID {
+	case SettingHeaderTableSize:
+		// Applied by startFrameWrite; see comment there.
+		if !sc.pendingEncoderTableSize || s.Val < sc.encoderTableSizeMin {
+			sc.encoderTableSizeMin = s.Val
+		}
+		sc.encoderTableSize = s.Val
+		sc.pendingEncoderTableSize = true
+	case SettingEnablePush:
+		sc.pushEnabled = s.Val != 0
+	case SettingMaxConcurrentStreams:
+		sc.clientMaxStreams = s.Val
+	case SettingInitialWindowSize:
+		return sc.processSettingInitialWindowSize(s.Val)
+	case SettingMaxFrameSize:
+		sc.maxFrameSize = int32(s.Val) // the maximum valid s.Val is < 2^31
+	case SettingMaxHeaderListSize:
+		sc.peerMaxHeaderListSize = s.Val
+	case SettingEnableConnectProtocol:
+		// Receipt of this parameter by a server does not
+		// have any impact
+	case SettingNoRFC7540Priorities:
+		if s.Val > 1 {
+			return ConnectionError(ErrCodeProtocol)
+		}
+	default:
+		// Unknown setting: "An endpoint that receives a SETTINGS
+		// frame with any unknown or unsupported identifier MUST
+		// ignore that setting."
+		if VerboseLogs {
+			sc.vlogf("http2: server ignoring unknown setting %v", s)
+		}
+	}
+	return nil
+}
+
+func (sc *serverConn) processSettingInitialWindowSize(val uint32) error {
+	sc.serveG.check()
+	if !sc.flow.changeInitialWindowSize(int64(val)) {
+		// 6.9.2 Initial Flow Control Window Size
+		// "An endpoint MUST treat a change to
+		// SETTINGS_INITIAL_WINDOW_SIZE that causes any flow
+		// control window to exceed the maximum size as a
+		// connection error (Section 5.4.1) of type
+		// FLOW_CONTROL_ERROR."
+		return sc.countError("setting_win_size", ConnectionError(ErrCodeFlowControl))
+	}
+	return nil
+}
+
+func (sc *serverConn) processData(f *DataFrame) error {
+	sc.serveG.check()
+	id := f.Header().StreamID
+
+	data := f.Data()
+	state, st := sc.state(id)
+	if id == 0 || state == stateIdle {
+		// Section 6.1: "DATA frames MUST be associated with a
+		// stream. If a DATA frame is received whose stream
+		// identifier field is 0x0, the recipient MUST respond
+		// with a connection error (Section 5.4.1) of type
+		// PROTOCOL_ERROR."
+		//
+		// Section 5.1: "Receiving any frame other than HEADERS
+		// or PRIORITY on a stream in this state MUST be
+		// treated as a connection error (Section 5.4.1) of
+		// type PROTOCOL_ERROR."
+		return sc.countError("data_on_idle", ConnectionError(ErrCodeProtocol))
+	}
+
+	// "If a DATA frame is received whose stream is not in "open"
+	// or "half closed (local)" state, the recipient MUST respond
+	// with a stream error (Section 5.4.2) of type STREAM_CLOSED."
+	if st == nil || state != stateOpen || st.gotTrailerHeader || st.resetQueued {
+		// This includes sending a RST_STREAM if the stream is
+		// in stateHalfClosedLocal (which currently means that
+		// the http.Handler returned, so it's done reading &
+		// done writing). Try to stop the client from sending
+		// more DATA.
+
+		// But still enforce their connection-level flow control,
+		// and return any flow control bytes since we're not going
+		// to consume them.
+		if !sc.inflow.take(f.Length) {
+			return sc.countError("data_flow", streamError(id, ErrCodeFlowControl))
+		}
+		sc.sendWindowUpdate(nil, int(f.Length)) // conn-level
+
+		if st != nil && st.resetQueued {
+			// Already have a stream error in flight. Don't send another.
+			return nil
+		}
+		return sc.countError("closed", streamError(id, ErrCodeStreamClosed))
+	}
+	if st.body == nil {
+		panic("internal error: should have a body in this state")
+	}
+
+	// Sender sending more than they'd declared?
+	if st.declBodyBytes != -1 && st.bodyBytes+int64(len(data)) > st.declBodyBytes {
+		if !sc.inflow.take(f.Length) {
+			return sc.countError("data_flow", streamError(id, ErrCodeFlowControl))
+		}
+		sc.sendWindowUpdate(nil, int(f.Length)) // conn-level
+
+		st.body.CloseWithError(fmt.Errorf("sender tried to send more than declared Content-Length of %d bytes", st.declBodyBytes))
+		// RFC 7540, sec 8.1.2.6: A request or response is also malformed if the
+		// value of a content-length header field does not equal the sum of the
+		// DATA frame payload lengths that form the body.
+		return sc.countError("send_too_much", streamError(id, ErrCodeProtocol))
+	}
+	if f.Length > 0 {
+		// Check whether the client has flow control quota.
+		if !takeInflows(&sc.inflow, &st.inflow, f.Length) {
+			return sc.countError("flow_on_data_length", streamError(id, ErrCodeFlowControl))
+		}
+
+		if len(data) > 0 {
+			st.bodyBytes += int64(len(data))
+			wrote, err := st.body.Write(data)
+			if err != nil {
+				// The handler has closed the request body.
+				// Return the connection-level flow control for the discarded data,
+				// but not the stream-level flow control.
+				sc.sendWindowUpdate(nil, int(f.Length)-wrote)
+				return nil
+			}
+			if wrote != len(data) {
+				panic("internal error: bad Writer")
+			}
+		}
+
+		// Return any padded flow control now, since we won't
+		// refund it later on body reads.
+		// Call sendWindowUpdate even if there is no padding,
+		// to return buffered flow control credit if the sent
+		// window has shrunk.
+		pad := int32(f.Length) - int32(len(data))
+		sc.sendWindowUpdate32(nil, pad)
+		sc.sendWindowUpdate32(st, pad)
+	}
+	if f.StreamEnded() {
+		st.endStream()
+	}
+	return nil
+}
+
+func (sc *serverConn) processGoAway(f *GoAwayFrame) error {
+	sc.serveG.check()
+	if f.ErrCode != ErrCodeNo {
+		sc.logf("http2: received GOAWAY %+v, starting graceful shutdown", f)
+	} else {
+		sc.vlogf("http2: received GOAWAY %+v, starting graceful shutdown", f)
+	}
+	sc.startGracefulShutdownInternal()
+	// http://tools.ietf.org/html/rfc7540#section-6.8
+	// We should not create any new streams, which means we should disable push.
+	sc.pushEnabled = false
+	return nil
+}
+
+// isPushed reports whether the stream is server-initiated.
+func (st *stream) isPushed() bool {
+	return st.id%2 == 0
+}
+
+// endStream closes a Request.Body's pipe. It is called when a DATA
+// frame says a request body is over (or after trailers).
+func (st *stream) endStream() {
+	sc := st.sc
+	sc.serveG.check()
+
+	if st.declBodyBytes != -1 && st.declBodyBytes != st.bodyBytes {
+		st.body.CloseWithError(fmt.Errorf("request declared a Content-Length of %d but only wrote %d bytes",
+			st.declBodyBytes, st.bodyBytes))
+	} else {
+		st.body.closeWithErrorAndCode(io.EOF, st.copyTrailersToHandlerRequest)
+		st.body.CloseWithError(io.EOF)
+	}
+	st.state = stateHalfClosedRemote
+}
+
+// copyTrailersToHandlerRequest is run in the Handler's goroutine in
+// its Request.Body.Read just before it gets io.EOF.
+func (st *stream) copyTrailersToHandlerRequest() {
+	for k, vv := range st.trailer {
+		if _, ok := st.reqTrailer[k]; ok {
+			// Only copy it over it was pre-declared.
+			st.reqTrailer[k] = vv
+		}
+	}
+}
+
+// onReadTimeout is run on its own goroutine (from time.AfterFunc)
+// when the stream's ReadTimeout has fired.
+func (st *stream) onReadTimeout() {
+	if st.body != nil {
+		// Wrap the ErrDeadlineExceeded to avoid callers depending on us
+		// returning the bare error.
+		st.body.CloseWithError(fmt.Errorf("%w", os.ErrDeadlineExceeded))
+	}
+}
+
+// onWriteTimeout is run on its own goroutine (from time.AfterFunc)
+// when the stream's WriteTimeout has fired.
+func (st *stream) onWriteTimeout() {
+	st.sc.writeFrameFromHandler(FrameWriteRequest{write: StreamError{
+		StreamID: st.id,
+		Code:     ErrCodeInternal,
+		Cause:    os.ErrDeadlineExceeded,
+	}})
+}
+
+func (sc *serverConn) processHeaders(f *MetaHeadersFrame) error {
+	sc.serveG.check()
+	id := f.StreamID
+	// http://tools.ietf.org/html/rfc7540#section-5.1.1
+	// Streams initiated by a client MUST use odd-numbered stream
+	// identifiers. [...] An endpoint that receives an unexpected
+	// stream identifier MUST respond with a connection error
+	// (Section 5.4.1) of type PROTOCOL_ERROR.
+	if id%2 != 1 {
+		return sc.countError("headers_even", ConnectionError(ErrCodeProtocol))
+	}
+	// A HEADERS frame can be used to create a new stream or
+	// send a trailer for an open one. If we already have a stream
+	// open, let it process its own HEADERS frame (trailers at this
+	// point, if it's valid).
+	if st := sc.streams[f.StreamID]; st != nil {
+		if st.resetQueued {
+			// We're sending RST_STREAM to close the stream, so don't bother
+			// processing this frame.
+			return nil
+		}
+		// RFC 7540, sec 5.1: If an endpoint receives additional frames, other than
+		// WINDOW_UPDATE, PRIORITY, or RST_STREAM, for a stream that is in
+		// this state, it MUST respond with a stream error (Section 5.4.2) of
+		// type STREAM_CLOSED.
+		if st.state == stateHalfClosedRemote {
+			return sc.countError("headers_half_closed", streamError(id, ErrCodeStreamClosed))
+		}
+		return st.processTrailerHeaders(f)
+	}
+
+	// [...] The identifier of a newly established stream MUST be
+	// numerically greater than all streams that the initiating
+	// endpoint has opened or reserved. [...]  An endpoint that
+	// receives an unexpected stream identifier MUST respond with
+	// a connection error (Section 5.4.1) of type PROTOCOL_ERROR.
+	if id <= sc.maxClientStreamID {
+		return sc.countError("stream_went_down", ConnectionError(ErrCodeProtocol))
+	}
+	sc.maxClientStreamID = id
+
+	if sc.idleTimer != nil {
+		sc.idleTimer.Stop()
+	}
+
+	// http://tools.ietf.org/html/rfc7540#section-5.1.2
+	// [...] Endpoints MUST NOT exceed the limit set by their peer. An
+	// endpoint that receives a HEADERS frame that causes their
+	// advertised concurrent stream limit to be exceeded MUST treat
+	// this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR
+	// or REFUSED_STREAM.
+	if sc.curClientStreams+1 > sc.advMaxStreams {
+		if sc.unackedSettings == 0 {
+			// They should know better.
+			return sc.countError("over_max_streams", streamError(id, ErrCodeProtocol))
+		}
+		// Assume it's a network race, where they just haven't
+		// received our last SETTINGS update. But actually
+		// this can't happen yet, because we don't yet provide
+		// a way for users to adjust server parameters at
+		// runtime.
+		return sc.countError("over_max_streams_race", streamError(id, ErrCodeRefusedStream))
+	}
+
+	initialState := stateOpen
+	if f.StreamEnded() {
+		initialState = stateHalfClosedRemote
+	}
+
+	// We are handling two special cases here:
+	// 1. When a request is sent via an intermediary, we force priority to be
+	// u=3,i. This is essentially a round-robin behavior, and is done to ensure
+	// fairness between, for example, multiple clients using the same proxy.
+	// 2. Until a client has shown that it is aware of RFC 9218, we make its
+	// streams non-incremental by default. This is done to preserve the
+	// historical behavior of handling streams in a round-robin manner, rather
+	// than one-by-one to completion.
+	initialPriority := defaultRFC9218Priority(sc.priorityAware && !sc.hasIntermediary)
+	if _, ok := sc.writeSched.(*priorityWriteSchedulerRFC9218); ok && !sc.hasIntermediary {
+		headerPriority, priorityAware, hasIntermediary := f.rfc9218Priority(sc.priorityAware)
+		initialPriority = headerPriority
+		sc.hasIntermediary = hasIntermediary
+		if priorityAware {
+			sc.priorityAware = true
+		}
+	}
+	st := sc.newStream(id, 0, initialState, initialPriority)
+
+	if f.HasPriority() {
+		if err := sc.checkPriority(f.StreamID, f.Priority); err != nil {
+			return err
+		}
+		if !sc.writeSchedIgnoresRFC7540() {
+			sc.writeSched.AdjustStream(st.id, f.Priority)
+		}
+	}
+
+	rw, req, err := sc.newWriterAndRequest(st, f)
+	if err != nil {
+		return err
+	}
+	st.reqTrailer = req.Trailer
+	if st.reqTrailer != nil {
+		st.trailer = make(Header)
+	}
+	st.body = st.reqBody.pipe // may be nil
+	st.declBodyBytes = req.ContentLength
+
+	handler := sc.handler.ServeHTTP
+	if f.Truncated {
+		// Their header list was too long. Send a 431 error.
+		handler = handleHeaderListTooLong
+	} else if err := checkValidHTTP2RequestHeaders(req.Header); err != nil {
+		handler = serve400Handler{err}.ServeHTTP
+	}
+
+	if sc.hs.ReadTimeout() > 0 {
+		st.readDeadline = time.AfterFunc(sc.hs.ReadTimeout(), st.onReadTimeout)
+	}
+
+	return sc.scheduleHandler(st, rw, req, handler)
+}
+
+func (sc *serverConn) upgradeRequest(req *ServerRequest) {
+	sc.serveG.check()
+	id := uint32(1)
+	sc.maxClientStreamID = id
+	st := sc.newStream(id, 0, stateHalfClosedRemote, defaultRFC9218Priority(sc.priorityAware && !sc.hasIntermediary))
+	st.reqTrailer = req.Trailer
+	if st.reqTrailer != nil {
+		st.trailer = make(Header)
+	}
+	rw := sc.newResponseWriter(st)
+	rw.rws.req = *req
+	req = &rw.rws.req
+
+	// This is the first request on the connection,
+	// so start the handler directly rather than going
+	// through scheduleHandler.
+	sc.curHandlers++
+	go sc.runHandler(rw, req, sc.handler.ServeHTTP)
+}
+
+func (st *stream) processTrailerHeaders(f *MetaHeadersFrame) error {
+	sc := st.sc
+	sc.serveG.check()
+	if st.gotTrailerHeader {
+		return sc.countError("dup_trailers", ConnectionError(ErrCodeProtocol))
+	}
+	st.gotTrailerHeader = true
+	if !f.StreamEnded() {
+		return sc.countError("trailers_not_ended", streamError(st.id, ErrCodeProtocol))
+	}
+
+	if len(f.PseudoFields()) > 0 {
+		return sc.countError("trailers_pseudo", streamError(st.id, ErrCodeProtocol))
+	}
+	if f.Truncated {
+		return sc.countError("trailers_too_large", streamError(st.id, ErrCodeProtocol))
+	}
+	if st.trailer != nil {
+		for _, hf := range f.RegularFields() {
+			key := sc.canonicalHeader(hf.Name)
+			if !httpguts.ValidTrailerHeader(key) {
+				// TODO: send more details to the peer somehow. But http2 has
+				// no way to send debug data at a stream level. Discuss with
+				// HTTP folk.
+				return sc.countError("trailers_bogus", streamError(st.id, ErrCodeProtocol))
+			}
+			st.trailer[key] = append(st.trailer[key], hf.Value)
+		}
+	}
+	st.endStream()
+	return nil
+}
+
+func (sc *serverConn) checkPriority(streamID uint32, p PriorityParam) error {
+	if streamID == p.StreamDep {
+		// Section 5.3.1: "A stream cannot depend on itself. An endpoint MUST treat
+		// this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR."
+		// Section 5.3.3 says that a stream can depend on one of its dependencies,
+		// so it's only self-dependencies that are forbidden.
+		return sc.countError("priority", streamError(streamID, ErrCodeProtocol))
+	}
+	return nil
+}
+
+func (sc *serverConn) processPriority(f *PriorityFrame) error {
+	if err := sc.checkPriority(f.StreamID, f.PriorityParam); err != nil {
+		return err
+	}
+	// We need to avoid calling AdjustStream when using the RFC 9218 write
+	// scheduler. Otherwise, incremental's zero value in PriorityParam will
+	// unexpectedly make all streams non-incremental. This causes us to process
+	// streams one-by-one to completion rather than doing it in a round-robin
+	// manner (the historical behavior), which might be unexpected to users.
+	if sc.writeSchedIgnoresRFC7540() {
+		return nil
+	}
+	sc.writeSched.AdjustStream(f.StreamID, f.PriorityParam)
+	return nil
+}
+
+func (sc *serverConn) processPriorityUpdate(f *PriorityUpdateFrame) error {
+	sc.priorityAware = true
+	if _, ok := sc.writeSched.(*priorityWriteSchedulerRFC9218); !ok {
+		return nil
+	}
+	p, ok := parseRFC9218Priority(f.Priority, sc.priorityAware)
+	if !ok {
+		return sc.countError("unparsable_priority_update", streamError(f.PrioritizedStreamID, ErrCodeProtocol))
+	}
+	sc.writeSched.AdjustStream(f.PrioritizedStreamID, p)
+	return nil
+}
+
+func (sc *serverConn) newStream(id, pusherID uint32, state streamState, priority PriorityParam) *stream {
+	sc.serveG.check()
+	if id == 0 {
+		panic("internal error: cannot create stream with id 0")
+	}
+
+	ctx, cancelCtx := context.WithCancel(sc.baseCtx)
+	st := &stream{
+		sc:        sc,
+		id:        id,
+		state:     state,
+		ctx:       ctx,
+		cancelCtx: cancelCtx,
+	}
+	st.cw.Init()
+	st.flow.conn = &sc.flow // link to conn-level counter
+	st.inflow.init(sc.initialStreamRecvWindowSize)
+	if writeTimeout := sc.hs.WriteTimeout(); writeTimeout > 0 {
+		st.writeDeadline = time.AfterFunc(writeTimeout, st.onWriteTimeout)
+	}
+
+	sc.streams[id] = st
+	sc.writeSched.OpenStream(st.id, OpenStreamOptions{PusherID: pusherID, priority: priority})
+	if st.isPushed() {
+		sc.curPushedStreams++
+	} else {
+		sc.curClientStreams++
+	}
+	if sc.curOpenStreams() == 1 {
+		sc.setConnState(ConnStateActive)
+	}
+
+	return st
+}
+
+func (sc *serverConn) newWriterAndRequest(st *stream, f *MetaHeadersFrame) (*responseWriter, *ServerRequest, error) {
+	sc.serveG.check()
+
+	rp := httpcommon.ServerRequestParam{
+		Method:    f.PseudoValue("method"),
+		Scheme:    f.PseudoValue("scheme"),
+		Authority: f.PseudoValue("authority"),
+		Path:      f.PseudoValue("path"),
+		Protocol:  f.PseudoValue("protocol"),
+	}
+
+	// extended connect is disabled, so we should not see :protocol
+	if disableExtendedConnectProtocol && rp.Protocol != "" {
+		return nil, nil, sc.countError("bad_connect", streamError(f.StreamID, ErrCodeProtocol))
+	}
+
+	isConnect := rp.Method == "CONNECT"
+	if isConnect {
+		if rp.Protocol == "" && (rp.Path != "" || rp.Scheme != "" || rp.Authority == "") {
+			return nil, nil, sc.countError("bad_connect", streamError(f.StreamID, ErrCodeProtocol))
+		}
+	} else if rp.Method == "" || rp.Path == "" || (rp.Scheme != "https" && rp.Scheme != "http") {
+		// See 8.1.2.6 Malformed Requests and Responses:
+		//
+		// Malformed requests or responses that are detected
+		// MUST be treated as a stream error (Section 5.4.2)
+		// of type PROTOCOL_ERROR."
+		//
+		// 8.1.2.3 Request Pseudo-Header Fields
+		// "All HTTP/2 requests MUST include exactly one valid
+		// value for the :method, :scheme, and :path
+		// pseudo-header fields"
+		return nil, nil, sc.countError("bad_path_method", streamError(f.StreamID, ErrCodeProtocol))
+	}
+
+	header := make(Header)
+	rp.Header = header
+	for _, hf := range f.RegularFields() {
+		header.Add(sc.canonicalHeader(hf.Name), hf.Value)
+	}
+	if rp.Authority == "" {
+		rp.Authority = header.Get("Host")
+	}
+	if rp.Protocol != "" {
+		header.Set(":protocol", rp.Protocol)
+	}
+
+	rw, req, err := sc.newWriterAndRequestNoBody(st, rp)
+	if err != nil {
+		return nil, nil, err
+	}
+	bodyOpen := !f.StreamEnded()
+	if clens, ok := rp.Header["Content-Length"]; ok {
+		if cl, err := strconv.ParseUint(clens[0], 10, 63); err == nil {
+			req.ContentLength = int64(cl)
+		} else {
+			req.ContentLength = 0
+		}
+		if !bodyOpen && req.ContentLength != 0 {
+			return nil, nil, sc.countError("bodyless_content_length", streamError(f.StreamID, ErrCodeProtocol))
+		}
+		if len(clens) > 1 {
+			for _, dup := range clens[1:] {
+				if clens[0] != dup {
+					return nil, nil, sc.countError("duplicate_content_length", streamError(f.StreamID, ErrCodeProtocol))
+				}
+			}
+			rp.Header["Content-Length"] = clens[:1]
+		}
+	}
+	if bodyOpen {
+		if _, ok := rp.Header["Content-Length"]; !ok {
+			req.ContentLength = -1
+		}
+		st.reqBody.pipe = &pipe{
+			b: &dataBuffer{expected: req.ContentLength},
+		}
+	}
+	return rw, req, nil
+}
+
+func (sc *serverConn) newWriterAndRequestNoBody(st *stream, rp httpcommon.ServerRequestParam) (*responseWriter, *ServerRequest, error) {
+	sc.serveG.check()
+
+	res := httpcommon.NewServerRequest(rp)
+	if res.InvalidReason != "" {
+		return nil, nil, sc.countError(res.InvalidReason, streamError(st.id, ErrCodeProtocol))
+	}
+
+	st.reqBody = &requestBody{
+		conn:          sc,
+		stream:        st,
+		needsContinue: res.NeedsContinue,
+	}
+	rw := sc.newResponseWriter(st)
+	rw.rws.req = ServerRequest{
+		Context:    st.ctx,
+		Method:     rp.Method,
+		URL:        res.URL,
+		RemoteAddr: sc.remoteAddrStr,
+		Header:     rp.Header,
+		RequestURI: res.RequestURI,
+		Proto:      "HTTP/2.0",
+		ProtoMajor: 2,
+		ProtoMinor: 0,
+		TLS:        sc.tlsState,
+		Host:       rp.Authority,
+		Body:       st.reqBody,
+		Trailer:    res.Trailer,
+	}
+	return rw, &rw.rws.req, nil
+}
+
+func (sc *serverConn) newResponseWriter(st *stream) *responseWriter {
+	rws := responseWriterStatePool.Get().(*responseWriterState)
+	bwSave := rws.bw
+	*rws = responseWriterState{} // zero all the fields
+	rws.conn = sc
+	rws.bw = bwSave
+	rws.bw.Reset(chunkWriter{rws})
+	rws.stream = st
+	return &responseWriter{rws: rws}
+}
+
+type unstartedHandler struct {
+	streamID uint32
+	rw       *responseWriter
+	req      *ServerRequest
+	handler  func(*ResponseWriter, *ServerRequest)
+	body     *pipe
+}
+
+// scheduleHandler starts a handler goroutine,
+// or schedules one to start as soon as an existing handler finishes.
+func (sc *serverConn) scheduleHandler(st *stream, rw *responseWriter, req *ServerRequest, handler func(*ResponseWriter, *ServerRequest)) error {
+	sc.serveG.check()
+	maxHandlers := sc.advMaxStreams
+	if sc.curHandlers < maxHandlers {
+		sc.curHandlers++
+		go sc.runHandler(rw, req, handler)
+		return nil
+	}
+	if len(sc.unstartedHandlers) > int(4*sc.advMaxStreams) {
+		return sc.countError("too_many_early_resets", ConnectionError(ErrCodeEnhanceYourCalm))
+	}
+	sc.unstartedHandlers = append(sc.unstartedHandlers, unstartedHandler{
+		streamID: st.id,
+		rw:       rw,
+		req:      req,
+		handler:  handler,
+		body:     st.body,
+	})
+	return nil
+}
+
+func (sc *serverConn) handlerDone() {
+	sc.serveG.check()
+	sc.curHandlers--
+	i := 0
+	maxHandlers := sc.advMaxStreams
+	for ; i < len(sc.unstartedHandlers); i++ {
+		u := sc.unstartedHandlers[i]
+		if sc.streams[u.streamID] == nil {
+			// This stream was reset before its goroutine had a chance to start.
+			if u.body != nil {
+				u.body.BreakWithError(errClosedBody)
+				sc.sendWindowUpdate(nil, u.body.Len())
+			}
+			continue
+		}
+		if sc.curHandlers >= maxHandlers {
+			break
+		}
+		sc.curHandlers++
+		go sc.runHandler(u.rw, u.req, u.handler)
+		sc.unstartedHandlers[i] = unstartedHandler{} // don't retain references
+	}
+	sc.unstartedHandlers = sc.unstartedHandlers[i:]
+	if len(sc.unstartedHandlers) == 0 {
+		sc.unstartedHandlers = nil
+	}
+}
+
+// Run on its own goroutine.
+func (sc *serverConn) runHandler(rw *responseWriter, req *ServerRequest, handler func(*ResponseWriter, *ServerRequest)) {
+	defer sc.sendServeMsg(handlerDoneMsg)
+	didPanic := true
+	defer func() {
+		rw.rws.stream.cancelCtx()
+		if req.MultipartForm != nil {
+			req.MultipartForm.RemoveAll()
+		}
+		if b := rw.rws.stream.reqBody; b != nil {
+			// Closing the body refunds flow control credit for any unconsumed data.
+			// (reqBody is nil for Upgrade: h2c requests, but those do not use flow
+			// control for the request body.)
+			b.Close()
+		}
+		if didPanic {
+			e := recover()
+			sc.writeFrameFromHandler(FrameWriteRequest{
+				write:  handlerPanicRST{rw.rws.stream.id},
+				stream: rw.rws.stream,
+			})
+			// Same as net/http:
+			if e != nil && e != ErrAbortHandler {
+				const size = 64 << 10
+				buf := make([]byte, size)
+				buf = buf[:runtime.Stack(buf, false)]
+				sc.logf("http2: panic serving %v: %v\n%s", sc.conn.RemoteAddr(), e, buf)
+			}
+			return
+		}
+		rw.handlerDone()
+	}()
+	handler(rw, req)
+	didPanic = false
+}
+
+func handleHeaderListTooLong(w *ResponseWriter, r *ServerRequest) {
+	// 10.5.1 Limits on Header Block Size:
+	// .. "A server that receives a larger header block than it is
+	// willing to handle can send an HTTP 431 (Request Header Fields Too
+	// Large) status code"
+	const statusRequestHeaderFieldsTooLarge = 431 // only in Go 1.6+
+	w.WriteHeader(statusRequestHeaderFieldsTooLarge)
+	io.WriteString(w, "<h1>HTTP Error 431</h1><p>Request Header Field(s) Too Large</p>")
+}
+
+// called from handler goroutines.
+// h may be nil.
+func (sc *serverConn) writeHeaders(st *stream, headerData *writeResHeaders) error {
+	sc.serveG.checkNotOn() // NOT on
+	var errc chan error
+	if headerData.h != nil {
+		// If there's a header map (which we don't own), so we have to block on
+		// waiting for this frame to be written, so an http.Flush mid-handler
+		// writes out the correct value of keys, before a handler later potentially
+		// mutates it.
+		errc = sc.srv.getErrChan()
+	}
+	if err := sc.writeFrameFromHandler(FrameWriteRequest{
+		write:  headerData,
+		stream: st,
+		done:   errc,
+	}); err != nil {
+		return err
+	}
+	if errc != nil {
+		select {
+		case err := <-errc:
+			sc.srv.putErrChan(errc)
+			return err
+		case <-sc.doneServing:
+			return errClientDisconnected
+		case <-st.cw:
+			return errStreamClosed
+		}
+	}
+	return nil
+}
+
+// called from handler goroutines.
+func (sc *serverConn) write100ContinueHeaders(st *stream) {
+	sc.writeFrameFromHandler(FrameWriteRequest{
+		write:  write100ContinueHeadersFrame{st.id},
+		stream: st,
+	})
+}
+
+// A bodyReadMsg tells the server loop that the http.Handler read n
+// bytes of the DATA from the client on the given stream.
+type bodyReadMsg struct {
+	st *stream
+	n  int
+}
+
+// called from handler goroutines.
+// Notes that the handler for the given stream ID read n bytes of its body
+// and schedules flow control tokens to be sent.
+func (sc *serverConn) noteBodyReadFromHandler(st *stream, n int, err error) {
+	sc.serveG.checkNotOn() // NOT on
+	if n > 0 {
+		select {
+		case sc.bodyReadCh <- bodyReadMsg{st, n}:
+		case <-sc.doneServing:
+		}
+	}
+}
+
+func (sc *serverConn) noteBodyRead(st *stream, n int) {
+	sc.serveG.check()
+	sc.sendWindowUpdate(nil, n) // conn-level
+	if st != nil && st.state != stateHalfClosedRemote && st.state != stateClosed {
+		// Don't send this WINDOW_UPDATE if the stream is closed
+		// remotely.
+		sc.sendWindowUpdate(st, n)
+	}
+}
+
+// st may be nil for conn-level
+func (sc *serverConn) sendWindowUpdate32(st *stream, n int32) {
+	sc.sendWindowUpdate(st, int(n))
+}
+
+// st may be nil for conn-level
+func (sc *serverConn) sendWindowUpdate(st *stream, n int) {
+	sc.serveG.check()
+	var streamID uint32
+	var send int32
+	if st == nil {
+		send = sc.inflow.add(n)
+	} else {
+		streamID = st.id
+		send = st.inflow.add(n)
+	}
+	if send == 0 {
+		return
+	}
+	sc.writeFrame(FrameWriteRequest{
+		write:  writeWindowUpdate{streamID: streamID, n: uint32(send)},
+		stream: st,
+	})
+}
+
+// requestBody is the Handler's Request.Body type.
+// Read and Close may be called concurrently.
+type requestBody struct {
+	_             incomparable
+	stream        *stream
+	conn          *serverConn
+	closeOnce     sync.Once // for use by Close only
+	sawEOF        bool      // for use by Read only
+	pipe          *pipe     // non-nil if we have an HTTP entity message body
+	needsContinue bool      // need to send a 100-continue
+}
+
+func (b *requestBody) Close() error {
+	b.closeOnce.Do(func() {
+		if b.pipe != nil {
+			b.pipe.BreakWithError(errClosedBody)
+			if unread := b.pipe.Len(); unread > 0 {
+				b.conn.noteBodyReadFromHandler(nil, unread, errClosedBody)
+			}
+		}
+	})
+	return nil
+}
+
+func (b *requestBody) Read(p []byte) (n int, err error) {
+	if b.needsContinue {
+		b.needsContinue = false
+		b.conn.write100ContinueHeaders(b.stream)
+	}
+	if b.pipe == nil || b.sawEOF {
+		return 0, io.EOF
+	}
+	n, err = b.pipe.Read(p)
+	if err == io.EOF {
+		b.sawEOF = true
+	}
+	b.conn.noteBodyReadFromHandler(b.stream, n, err)
+	return
+}
+
+// responseWriter is the http.ResponseWriter implementation. It's
+// intentionally small (1 pointer wide) to minimize garbage. The
+// responseWriterState pointer inside is zeroed at the end of a
+// request (in handlerDone) and calls on the responseWriter thereafter
+// simply crash (caller's mistake), but the much larger responseWriterState
+// and buffers are reused between multiple requests.
+type responseWriter struct {
+	rws *responseWriterState
+}
+
+type responseWriterState struct {
+	// immutable within a request:
+	stream *stream
+	req    ServerRequest
+	conn   *serverConn
+
+	// TODO: adjust buffer writing sizes based on server config, frame size updates from peer, etc
+	bw *bufio.Writer // writing to a chunkWriter{this *responseWriterState}
+
+	// mutated by http.Handler goroutine:
+	handlerHeader Header   // nil until called
+	snapHeader    Header   // snapshot of handlerHeader at WriteHeader time
+	trailers      []string // set in writeChunk
+	status        int      // status code passed to WriteHeader
+	wroteHeader   bool     // WriteHeader called (explicitly or implicitly). Not necessarily sent to user yet.
+	sentHeader    bool     // have we sent the header frame?
+	handlerDone   bool     // handler has finished
+
+	sentContentLen int64 // non-zero if handler set a Content-Length header
+	wroteBytes     int64
+
+	closeNotifierMu sync.Mutex // guards closeNotifierCh
+	closeNotifierCh chan bool  // nil until first used
+}
+
+type chunkWriter struct{ rws *responseWriterState }
+
+func (cw chunkWriter) Write(p []byte) (n int, err error) {
+	n, err = cw.rws.writeChunk(p)
+	if err == errStreamClosed {
+		// If writing failed because the stream has been closed,
+		// return the reason it was closed.
+		err = cw.rws.stream.closeErr
+	}
+	return n, err
+}
+
+func (rws *responseWriterState) hasTrailers() bool { return len(rws.trailers) > 0 }
+
+func (rws *responseWriterState) hasNonemptyTrailers() bool {
+	for _, trailer := range rws.trailers {
+		if _, ok := rws.handlerHeader[trailer]; ok {
+			return true
+		}
+	}
+	return false
+}
+
+// declareTrailer is called for each Trailer header when the
+// response header is written. It notes that a header will need to be
+// written in the trailers at the end of the response.
+func (rws *responseWriterState) declareTrailer(k string) {
+	k = textproto.CanonicalMIMEHeaderKey(k)
+	if !httpguts.ValidTrailerHeader(k) {
+		// Forbidden by RFC 7230, section 4.1.2.
+		rws.conn.logf("ignoring invalid trailer %q", k)
+		return
+	}
+	if !slices.Contains(rws.trailers, k) {
+		rws.trailers = append(rws.trailers, k)
+	}
+}
+
+const TimeFormat = "Mon, 02 Jan 2006 15:04:05 GMT" // keep in sync with net/http
+
+// writeChunk writes chunks from the bufio.Writer. But because
+// bufio.Writer may bypass its chunking, sometimes p may be
+// arbitrarily large.
+//
+// writeChunk is also responsible (on the first chunk) for sending the
+// HEADER response.
+func (rws *responseWriterState) writeChunk(p []byte) (n int, err error) {
+	if !rws.wroteHeader {
+		rws.writeHeader(200)
+	}
+
+	if rws.handlerDone {
+		rws.promoteUndeclaredTrailers()
+	}
+
+	isHeadResp := rws.req.Method == "HEAD"
+	if !rws.sentHeader {
+		rws.sentHeader = true
+		var ctype, clen string
+		if clen = rws.snapHeader.Get("Content-Length"); clen != "" {
+			rws.snapHeader.Del("Content-Length")
+			if cl, err := strconv.ParseUint(clen, 10, 63); err == nil {
+				rws.sentContentLen = int64(cl)
+			} else {
+				clen = ""
+			}
+		}
+		_, hasContentLength := rws.snapHeader["Content-Length"]
+		if !hasContentLength && clen == "" && rws.handlerDone && bodyAllowedForStatus(rws.status) && (len(p) > 0 || !isHeadResp) {
+			clen = strconv.Itoa(len(p))
+		}
+		_, hasContentType := rws.snapHeader["Content-Type"]
+		// If the Content-Encoding is non-blank, we shouldn't
+		// sniff the body. See Issue golang.org/issue/31753.
+		ce := rws.snapHeader.Get("Content-Encoding")
+		hasCE := len(ce) > 0
+		if !hasCE && !hasContentType && bodyAllowedForStatus(rws.status) && len(p) > 0 {
+			ctype = internal.DetectContentType(p)
+		}
+		var date string
+		if _, ok := rws.snapHeader["Date"]; !ok {
+			// TODO(bradfitz): be faster here, like net/http? measure.
+			date = time.Now().UTC().Format(TimeFormat)
+		}
+
+		for _, v := range rws.snapHeader["Trailer"] {
+			foreachHeaderElement(v, rws.declareTrailer)
+		}
+
+		// "Connection" headers aren't allowed in HTTP/2 (RFC 7540, 8.1.2.2),
+		// but respect "Connection" == "close" to mean sending a GOAWAY and tearing
+		// down the TCP connection when idle, like we do for HTTP/1.
+		// TODO: remove more Connection-specific header fields here, in addition
+		// to "Connection".
+		if _, ok := rws.snapHeader["Connection"]; ok {
+			v := rws.snapHeader.Get("Connection")
+			delete(rws.snapHeader, "Connection")
+			if v == "close" {
+				rws.conn.startGracefulShutdown()
+			}
+		}
+
+		endStream := (rws.handlerDone && !rws.hasTrailers() && len(p) == 0) || isHeadResp
+		err = rws.conn.writeHeaders(rws.stream, &writeResHeaders{
+			streamID:      rws.stream.id,
+			httpResCode:   rws.status,
+			h:             rws.snapHeader,
+			endStream:     endStream,
+			contentType:   ctype,
+			contentLength: clen,
+			date:          date,
+		})
+		if err != nil {
+			return 0, err
+		}
+		if endStream {
+			return 0, nil
+		}
+	}
+	if isHeadResp {
+		return len(p), nil
+	}
+	if len(p) == 0 && !rws.handlerDone {
+		return 0, nil
+	}
+
+	// only send trailers if they have actually been defined by the
+	// server handler.
+	hasNonemptyTrailers := rws.hasNonemptyTrailers()
+	endStream := rws.handlerDone && !hasNonemptyTrailers
+	if len(p) > 0 || endStream {
+		// only send a 0 byte DATA frame if we're ending the stream.
+		if err := rws.conn.writeDataFromHandler(rws.stream, p, endStream); err != nil {
+			return 0, err
+		}
+	}
+
+	if rws.handlerDone && hasNonemptyTrailers {
+		err = rws.conn.writeHeaders(rws.stream, &writeResHeaders{
+			streamID:  rws.stream.id,
+			h:         rws.handlerHeader,
+			trailers:  rws.trailers,
+			endStream: true,
+		})
+		return len(p), err
+	}
+	return len(p), nil
+}
+
+// TrailerPrefix is a magic prefix for ResponseWriter.Header map keys
+// that, if present, signals that the map entry is actually for
+// the response trailers, and not the response headers. The prefix
+// is stripped after the ServeHTTP call finishes and the values are
+// sent in the trailers.
+//
+// This mechanism is intended only for trailers that are not known
+// prior to the headers being written. If the set of trailers is fixed
+// or known before the header is written, the normal Go trailers mechanism
+// is preferred:
+//
+//	https://golang.org/pkg/net/http/#ResponseWriter
+//	https://golang.org/pkg/net/http/#example_ResponseWriter_trailers
+const TrailerPrefix = "Trailer:"
+
+// promoteUndeclaredTrailers permits http.Handlers to set trailers
+// after the header has already been flushed. Because the Go
+// ResponseWriter interface has no way to set Trailers (only the
+// Header), and because we didn't want to expand the ResponseWriter
+// interface, and because nobody used trailers, and because RFC 7230
+// says you SHOULD (but not must) predeclare any trailers in the
+// header, the official ResponseWriter rules said trailers in Go must
+// be predeclared, and then we reuse the same ResponseWriter.Header()
+// map to mean both Headers and Trailers. When it's time to write the
+// Trailers, we pick out the fields of Headers that were declared as
+// trailers. That worked for a while, until we found the first major
+// user of Trailers in the wild: gRPC (using them only over http2),
+// and gRPC libraries permit setting trailers mid-stream without
+// predeclaring them. So: change of plans. We still permit the old
+// way, but we also permit this hack: if a Header() key begins with
+// "Trailer:", the suffix of that key is a Trailer. Because ':' is an
+// invalid token byte anyway, there is no ambiguity. (And it's already
+// filtered out) It's mildly hacky, but not terrible.
+//
+// This method runs after the Handler is done and promotes any Header
+// fields to be trailers.
+func (rws *responseWriterState) promoteUndeclaredTrailers() {
+	for k, vv := range rws.handlerHeader {
+		if !strings.HasPrefix(k, TrailerPrefix) {
+			continue
+		}
+		trailerKey := strings.TrimPrefix(k, TrailerPrefix)
+		rws.declareTrailer(trailerKey)
+		rws.handlerHeader[textproto.CanonicalMIMEHeaderKey(trailerKey)] = vv
+	}
+
+	if len(rws.trailers) > 1 {
+		slices.Sort(rws.trailers)
+	}
+}
+
+func (w *responseWriter) SetReadDeadline(deadline time.Time) error {
+	st := w.rws.stream
+	if !deadline.IsZero() && deadline.Before(time.Now()) {
+		// If we're setting a deadline in the past, reset the stream immediately
+		// so writes after SetWriteDeadline returns will fail.
+		st.onReadTimeout()
+		return nil
+	}
+	w.rws.conn.sendServeMsg(func(sc *serverConn) {
+		if st.readDeadline != nil {
+			if !st.readDeadline.Stop() {
+				// Deadline already exceeded, or stream has been closed.
+				return
+			}
+		}
+		if deadline.IsZero() {
+			st.readDeadline = nil
+		} else if st.readDeadline == nil {
+			st.readDeadline = time.AfterFunc(deadline.Sub(time.Now()), st.onReadTimeout)
+		} else {
+			st.readDeadline.Reset(deadline.Sub(time.Now()))
+		}
+	})
+	return nil
+}
+
+func (w *responseWriter) SetWriteDeadline(deadline time.Time) error {
+	st := w.rws.stream
+	if !deadline.IsZero() && deadline.Before(time.Now()) {
+		// If we're setting a deadline in the past, reset the stream immediately
+		// so writes after SetWriteDeadline returns will fail.
+		st.onWriteTimeout()
+		return nil
+	}
+	w.rws.conn.sendServeMsg(func(sc *serverConn) {
+		if st.writeDeadline != nil {
+			if !st.writeDeadline.Stop() {
+				// Deadline already exceeded, or stream has been closed.
+				return
+			}
+		}
+		if deadline.IsZero() {
+			st.writeDeadline = nil
+		} else if st.writeDeadline == nil {
+			st.writeDeadline = time.AfterFunc(deadline.Sub(time.Now()), st.onWriteTimeout)
+		} else {
+			st.writeDeadline.Reset(deadline.Sub(time.Now()))
+		}
+	})
+	return nil
+}
+
+func (w *responseWriter) EnableFullDuplex() error {
+	// We always support full duplex responses, so this is a no-op.
+	return nil
+}
+
+func (w *responseWriter) Flush() {
+	w.FlushError()
+}
+
+func (w *responseWriter) FlushError() error {
+	rws := w.rws
+	if rws == nil {
+		panic("Header called after Handler finished")
+	}
+	var err error
+	if rws.bw.Buffered() > 0 {
+		err = rws.bw.Flush()
+	} else {
+		// The bufio.Writer won't call chunkWriter.Write
+		// (writeChunk with zero bytes), so we have to do it
+		// ourselves to force the HTTP response header and/or
+		// final DATA frame (with END_STREAM) to be sent.
+		_, err = chunkWriter{rws}.Write(nil)
+		if err == nil {
+			select {
+			case <-rws.stream.cw:
+				err = rws.stream.closeErr
+			default:
+			}
+		}
+	}
+	return err
+}
+
+func (w *responseWriter) CloseNotify() <-chan bool {
+	rws := w.rws
+	if rws == nil {
+		panic("CloseNotify called after Handler finished")
+	}
+	rws.closeNotifierMu.Lock()
+	ch := rws.closeNotifierCh
+	if ch == nil {
+		ch = make(chan bool, 1)
+		rws.closeNotifierCh = ch
+		cw := rws.stream.cw
+		go func() {
+			cw.Wait() // wait for close
+			ch <- true
+		}()
+	}
+	rws.closeNotifierMu.Unlock()
+	return ch
+}
+
+func (w *responseWriter) Header() Header {
+	rws := w.rws
+	if rws == nil {
+		panic("Header called after Handler finished")
+	}
+	if rws.handlerHeader == nil {
+		rws.handlerHeader = make(Header)
+	}
+	return rws.handlerHeader
+}
+
+// checkWriteHeaderCode is a copy of net/http's checkWriteHeaderCode.
+func checkWriteHeaderCode(code int) {
+	// Issue 22880: require valid WriteHeader status codes.
+	// For now we only enforce that it's three digits.
+	// In the future we might block things over 599 (600 and above aren't defined
+	// at http://httpwg.org/specs/rfc7231.html#status.codes).
+	// But for now any three digits.
+	//
+	// We used to send "HTTP/1.1 000 0" on the wire in responses but there's
+	// no equivalent bogus thing we can realistically send in HTTP/2,
+	// so we'll consistently panic instead and help people find their bugs
+	// early. (We can't return an error from WriteHeader even if we wanted to.)
+	if code < 100 || code > 999 {
+		panic(fmt.Sprintf("invalid WriteHeader code %v", code))
+	}
+}
+
+func (w *responseWriter) WriteHeader(code int) {
+	rws := w.rws
+	if rws == nil {
+		panic("WriteHeader called after Handler finished")
+	}
+	rws.writeHeader(code)
+}
+
+func (rws *responseWriterState) writeHeader(code int) {
+	if rws.wroteHeader {
+		return
+	}
+
+	checkWriteHeaderCode(code)
+
+	// Handle informational headers
+	if code >= 100 && code <= 199 {
+		// Per RFC 8297 we must not clear the current header map
+		h := rws.handlerHeader
+
+		_, cl := h["Content-Length"]
+		_, te := h["Transfer-Encoding"]
+		if cl || te {
+			h = cloneHeader(h)
+			h.Del("Content-Length")
+			h.Del("Transfer-Encoding")
+		}
+
+		rws.conn.writeHeaders(rws.stream, &writeResHeaders{
+			streamID:    rws.stream.id,
+			httpResCode: code,
+			h:           h,
+			endStream:   rws.handlerDone && !rws.hasTrailers(),
+		})
+
+		return
+	}
+
+	rws.wroteHeader = true
+	rws.status = code
+	if len(rws.handlerHeader) > 0 {
+		rws.snapHeader = cloneHeader(rws.handlerHeader)
+	}
+}
+
+func cloneHeader(h Header) Header {
+	h2 := make(Header, len(h))
+	for k, vv := range h {
+		vv2 := make([]string, len(vv))
+		copy(vv2, vv)
+		h2[k] = vv2
+	}
+	return h2
+}
+
+// The Life Of A Write is like this:
+//
+// * Handler calls w.Write or w.WriteString ->
+// * -> rws.bw (*bufio.Writer) ->
+// * (Handler might call Flush)
+// * -> chunkWriter{rws}
+// * -> responseWriterState.writeChunk(p []byte)
+// * -> responseWriterState.writeChunk (most of the magic; see comment there)
+func (w *responseWriter) Write(p []byte) (n int, err error) {
+	return w.write(len(p), p, "")
+}
+
+func (w *responseWriter) WriteString(s string) (n int, err error) {
+	return w.write(len(s), nil, s)
+}
+
+// either dataB or dataS is non-zero.
+func (w *responseWriter) write(lenData int, dataB []byte, dataS string) (n int, err error) {
+	rws := w.rws
+	if rws == nil {
+		panic("Write called after Handler finished")
+	}
+	if !rws.wroteHeader {
+		w.WriteHeader(200)
+	}
+	if !bodyAllowedForStatus(rws.status) {
+		return 0, ErrBodyNotAllowed
+	}
+	rws.wroteBytes += int64(len(dataB)) + int64(len(dataS)) // only one can be set
+	if rws.sentContentLen != 0 && rws.wroteBytes > rws.sentContentLen {
+		// TODO: send a RST_STREAM
+		return 0, errors.New("http2: handler wrote more than declared Content-Length")
+	}
+
+	if dataB != nil {
+		return rws.bw.Write(dataB)
+	} else {
+		return rws.bw.WriteString(dataS)
+	}
+}
+
+func (w *responseWriter) handlerDone() {
+	rws := w.rws
+	rws.handlerDone = true
+	w.Flush()
+	w.rws = nil
+	responseWriterStatePool.Put(rws)
+}
+
+// Push errors.
+var (
+	ErrRecursivePush    = errors.New("http2: recursive push not allowed")
+	ErrPushLimitReached = errors.New("http2: push would exceed peer's SETTINGS_MAX_CONCURRENT_STREAMS")
+)
+
+func (w *responseWriter) Push(target, method string, header Header) error {
+	st := w.rws.stream
+	sc := st.sc
+	sc.serveG.checkNotOn()
+
+	// No recursive pushes: "PUSH_PROMISE frames MUST only be sent on a peer-initiated stream."
+	// http://tools.ietf.org/html/rfc7540#section-6.6
+	if st.isPushed() {
+		return ErrRecursivePush
+	}
+
+	// Default options.
+	if method == "" {
+		method = "GET"
+	}
+	if header == nil {
+		header = Header{}
+	}
+	wantScheme := "http"
+	if w.rws.req.TLS != nil {
+		wantScheme = "https"
+	}
+
+	// Validate the request.
+	u, err := url.Parse(target)
+	if err != nil {
+		return err
+	}
+	if u.Scheme == "" {
+		if !strings.HasPrefix(target, "/") {
+			return fmt.Errorf("target must be an absolute URL or an absolute path: %q", target)
+		}
+		u.Scheme = wantScheme
+		u.Host = w.rws.req.Host
+	} else {
+		if u.Scheme != wantScheme {
+			return fmt.Errorf("cannot push URL with scheme %q from request with scheme %q", u.Scheme, wantScheme)
+		}
+		if u.Host == "" {
+			return errors.New("URL must have a host")
+		}
+	}
+	for k := range header {
+		if strings.HasPrefix(k, ":") {
+			return fmt.Errorf("promised request headers cannot include pseudo header %q", k)
+		}
+		// These headers are meaningful only if the request has a body,
+		// but PUSH_PROMISE requests cannot have a body.
+		// http://tools.ietf.org/html/rfc7540#section-8.2
+		// Also disallow Host, since the promised URL must be absolute.
+		if asciiEqualFold(k, "content-length") ||
+			asciiEqualFold(k, "content-encoding") ||
+			asciiEqualFold(k, "trailer") ||
+			asciiEqualFold(k, "te") ||
+			asciiEqualFold(k, "expect") ||
+			asciiEqualFold(k, "host") {
+			return fmt.Errorf("promised request headers cannot include %q", k)
+		}
+	}
+	if err := checkValidHTTP2RequestHeaders(header); err != nil {
+		return err
+	}
+
+	// The RFC effectively limits promised requests to GET and HEAD:
+	// "Promised requests MUST be cacheable [GET, HEAD, or POST], and MUST be safe [GET or HEAD]"
+	// http://tools.ietf.org/html/rfc7540#section-8.2
+	if method != "GET" && method != "HEAD" {
+		return fmt.Errorf("method %q must be GET or HEAD", method)
+	}
+
+	msg := &startPushRequest{
+		parent: st,
+		method: method,
+		url:    u,
+		header: cloneHeader(header),
+		done:   sc.srv.getErrChan(),
+	}
+
+	select {
+	case <-sc.doneServing:
+		return errClientDisconnected
+	case <-st.cw:
+		return errStreamClosed
+	case sc.serveMsgCh <- msg:
+	}
+
+	select {
+	case <-sc.doneServing:
+		return errClientDisconnected
+	case <-st.cw:
+		return errStreamClosed
+	case err := <-msg.done:
+		sc.srv.putErrChan(msg.done)
+		return err
+	}
+}
+
+type startPushRequest struct {
+	parent *stream
+	method string
+	url    *url.URL
+	header Header
+	done   chan error
+}
+
+func (sc *serverConn) startPush(msg *startPushRequest) {
+	sc.serveG.check()
+
+	// http://tools.ietf.org/html/rfc7540#section-6.6.
+	// PUSH_PROMISE frames MUST only be sent on a peer-initiated stream that
+	// is in either the "open" or "half-closed (remote)" state.
+	if msg.parent.state != stateOpen && msg.parent.state != stateHalfClosedRemote {
+		// responseWriter.Push checks that the stream is peer-initiated.
+		msg.done <- errStreamClosed
+		return
+	}
+
+	// http://tools.ietf.org/html/rfc7540#section-6.6.
+	if !sc.pushEnabled {
+		msg.done <- ErrNotSupported
+		return
+	}
+
+	// PUSH_PROMISE frames must be sent in increasing order by stream ID, so
+	// we allocate an ID for the promised stream lazily, when the PUSH_PROMISE
+	// is written. Once the ID is allocated, we start the request handler.
+	allocatePromisedID := func() (uint32, error) {
+		sc.serveG.check()
+
+		// Check this again, just in case. Technically, we might have received
+		// an updated SETTINGS by the time we got around to writing this frame.
+		if !sc.pushEnabled {
+			return 0, ErrNotSupported
+		}
+		// http://tools.ietf.org/html/rfc7540#section-6.5.2.
+		if sc.curPushedStreams+1 > sc.clientMaxStreams {
+			return 0, ErrPushLimitReached
+		}
+
+		// http://tools.ietf.org/html/rfc7540#section-5.1.1.
+		// Streams initiated by the server MUST use even-numbered identifiers.
+		// A server that is unable to establish a new stream identifier can send a GOAWAY
+		// frame so that the client is forced to open a new connection for new streams.
+		if sc.maxPushPromiseID+2 >= 1<<31 {
+			sc.startGracefulShutdownInternal()
+			return 0, ErrPushLimitReached
+		}
+		sc.maxPushPromiseID += 2
+		promisedID := sc.maxPushPromiseID
+
+		// http://tools.ietf.org/html/rfc7540#section-8.2.
+		// Strictly speaking, the new stream should start in "reserved (local)", then
+		// transition to "half closed (remote)" after sending the initial HEADERS, but
+		// we start in "half closed (remote)" for simplicity.
+		// See further comments at the definition of stateHalfClosedRemote.
+		promised := sc.newStream(promisedID, msg.parent.id, stateHalfClosedRemote, defaultRFC9218Priority(sc.priorityAware && !sc.hasIntermediary))
+		rw, req, err := sc.newWriterAndRequestNoBody(promised, httpcommon.ServerRequestParam{
+			Method:    msg.method,
+			Scheme:    msg.url.Scheme,
+			Authority: msg.url.Host,
+			Path:      msg.url.RequestURI(),
+			Header:    cloneHeader(msg.header), // clone since handler runs concurrently with writing the PUSH_PROMISE
+		})
+		if err != nil {
+			// Should not happen, since we've already validated msg.url.
+			panic(fmt.Sprintf("newWriterAndRequestNoBody(%+v): %v", msg.url, err))
+		}
+
+		sc.curHandlers++
+		go sc.runHandler(rw, req, sc.handler.ServeHTTP)
+		return promisedID, nil
+	}
+
+	sc.writeFrame(FrameWriteRequest{
+		write: &writePushPromise{
+			streamID:           msg.parent.id,
+			method:             msg.method,
+			url:                msg.url,
+			h:                  msg.header,
+			allocatePromisedID: allocatePromisedID,
+		},
+		stream: msg.parent,
+		done:   msg.done,
+	})
+}
+
+// foreachHeaderElement splits v according to the "#rule" construction
+// in RFC 7230 section 7 and calls fn for each non-empty element.
+func foreachHeaderElement(v string, fn func(string)) {
+	v = textproto.TrimString(v)
+	if v == "" {
+		return
+	}
+	if !strings.Contains(v, ",") {
+		fn(v)
+		return
+	}
+	for f := range strings.SplitSeq(v, ",") {
+		if f = textproto.TrimString(f); f != "" {
+			fn(f)
+		}
+	}
+}
+
+// From http://httpwg.org/specs/rfc7540.html#rfc.section.8.1.2.2
+var connHeaders = []string{
+	"Connection",
+	"Keep-Alive",
+	"Proxy-Connection",
+	"Transfer-Encoding",
+	"Upgrade",
+}
+
+// checkValidHTTP2RequestHeaders checks whether h is a valid HTTP/2 request,
+// per RFC 7540 Section 8.1.2.2.
+// The returned error is reported to users.
+func checkValidHTTP2RequestHeaders(h Header) error {
+	for _, k := range connHeaders {
+		if _, ok := h[k]; ok {
+			return fmt.Errorf("request header %q is not valid in HTTP/2", k)
+		}
+	}
+	te := h["Te"]
+	if len(te) > 0 && (len(te) > 1 || (te[0] != "trailers" && te[0] != "")) {
+		return errors.New(`request header "TE" may only be "trailers" in HTTP/2`)
+	}
+	return nil
+}
+
+type serve400Handler struct {
+	err error
+}
+
+func (handler serve400Handler) ServeHTTP(w *ResponseWriter, r *ServerRequest) {
+	const statusBadRequest = 400
+
+	// TODO: Dedup with http.Error?
+	h := w.Header()
+	h.Del("Content-Length")
+	h.Set("Content-Type", "text/plain; charset=utf-8")
+	h.Set("X-Content-Type-Options", "nosniff")
+	w.WriteHeader(statusBadRequest)
+	fmt.Fprintln(w, handler.err.Error())
+}
+
+// h1ServerKeepAlivesDisabled reports whether hs has its keep-alives
+// disabled. See comments on h1ServerShutdownChan above for why
+// the code is written this way.
+func h1ServerKeepAlivesDisabled(hs ServerConfig) bool {
+	return !hs.DoKeepAlives()
+}
+
+func (sc *serverConn) countError(name string, err error) error {
+	if sc == nil || sc.srv == nil {
+		return err
+	}
+	f := sc.countErrorFunc
+	if f == nil {
+		return err
+	}
+	var typ string
+	var code ErrCode
+	switch e := err.(type) {
+	case ConnectionError:
+		typ = "conn"
+		code = ErrCode(e)
+	case StreamError:
+		typ = "stream"
+		code = ErrCode(e.Code)
+	default:
+		return err
+	}
+	codeStr := errCodeName[code]
+	if codeStr == "" {
+		codeStr = strconv.Itoa(int(code))
+	}
+	f(fmt.Sprintf("%s_%s_%s", typ, codeStr, name))
+	return err
+}