blob: ea80b8e12fa2eecfbdca8e1502fb99ba08531ec2 [file] [log] [blame]
// Copyright 2014 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "net/socket/websocket_transport_client_socket_pool.h"
#include <algorithm>
#include "base/callback_helpers.h"
#include "base/compiler_specific.h"
#include "base/location.h"
#include "base/logging.h"
#include "base/numerics/safe_conversions.h"
#include "base/single_thread_task_runner.h"
#include "base/strings/string_util.h"
#include "base/threading/thread_task_runner_handle.h"
#include "base/time/time.h"
#include "base/trace_event/trace_event.h"
#include "base/values.h"
#include "net/base/net_errors.h"
#include "net/base/trace_constants.h"
#include "net/log/net_log_event_type.h"
#include "net/log/net_log_source.h"
#include "net/log/net_log_source_type.h"
#include "net/socket/client_socket_handle.h"
#include "net/socket/client_socket_pool_base.h"
#include "net/socket/websocket_endpoint_lock_manager.h"
#include "net/socket/websocket_transport_connect_sub_job.h"
namespace net {
WebSocketTransportConnectJob::WebSocketTransportConnectJob(
const std::string& group_name,
RequestPriority priority,
ClientSocketPool::RespectLimits respect_limits,
const scoped_refptr<TransportSocketParams>& params,
base::TimeDelta timeout_duration,
CompletionOnceCallback callback,
ClientSocketFactory* client_socket_factory,
HostResolver* host_resolver,
ClientSocketHandle* handle,
Delegate* delegate,
WebSocketEndpointLockManager* websocket_endpoint_lock_manager,
NetLog* pool_net_log,
const NetLogWithSource& request_net_log)
: ConnectJob(group_name,
timeout_duration,
priority,
SocketTag(),
respect_limits,
delegate,
NetLogWithSource::Make(
pool_net_log,
NetLogSourceType::WEB_SOCKET_TRANSPORT_CONNECT_JOB)),
params_(params),
resolver_(host_resolver),
client_socket_factory_(client_socket_factory),
next_state_(STATE_NONE),
race_result_(TransportConnectJob::RACE_UNKNOWN),
handle_(handle),
websocket_endpoint_lock_manager_(websocket_endpoint_lock_manager),
callback_(std::move(callback)),
request_net_log_(request_net_log),
had_ipv4_(false),
had_ipv6_(false) {}
WebSocketTransportConnectJob::~WebSocketTransportConnectJob() = default;
LoadState WebSocketTransportConnectJob::GetLoadState() const {
LoadState load_state = LOAD_STATE_RESOLVING_HOST;
if (ipv6_job_)
load_state = ipv6_job_->GetLoadState();
// This method should return LOAD_STATE_CONNECTING in preference to
// LOAD_STATE_WAITING_FOR_AVAILABLE_SOCKET when possible because "waiting for
// available socket" implies that nothing is happening.
if (ipv4_job_ && load_state != LOAD_STATE_CONNECTING)
load_state = ipv4_job_->GetLoadState();
return load_state;
}
void WebSocketTransportConnectJob::OnIOComplete(int result) {
result = DoLoop(result);
if (result != ERR_IO_PENDING)
NotifyDelegateOfCompletion(result); // Deletes |this|
}
int WebSocketTransportConnectJob::DoLoop(int result) {
DCHECK_NE(next_state_, STATE_NONE);
int rv = result;
do {
State state = next_state_;
next_state_ = STATE_NONE;
switch (state) {
case STATE_RESOLVE_HOST:
DCHECK_EQ(OK, rv);
rv = DoResolveHost();
break;
case STATE_RESOLVE_HOST_COMPLETE:
rv = DoResolveHostComplete(rv);
break;
case STATE_TRANSPORT_CONNECT:
DCHECK_EQ(OK, rv);
rv = DoTransportConnect();
break;
case STATE_TRANSPORT_CONNECT_COMPLETE:
rv = DoTransportConnectComplete(rv);
break;
default:
NOTREACHED();
rv = ERR_FAILED;
break;
}
} while (rv != ERR_IO_PENDING && next_state_ != STATE_NONE);
return rv;
}
int WebSocketTransportConnectJob::DoResolveHost() {
next_state_ = STATE_RESOLVE_HOST_COMPLETE;
connect_timing_.dns_start = base::TimeTicks::Now();
return resolver_->Resolve(
params_->destination(), priority(), &addresses_,
base::Bind(&WebSocketTransportConnectJob::OnIOComplete,
base::Unretained(this)),
&request_, net_log());
}
int WebSocketTransportConnectJob::DoResolveHostComplete(int result) {
TRACE_EVENT0(kNetTracingCategory,
"WebSocketTransportConnectJob::DoResolveHostComplete");
connect_timing_.dns_end = base::TimeTicks::Now();
// Overwrite connection start time, since for connections that do not go
// through proxies, |connect_start| should not include dns lookup time.
connect_timing_.connect_start = connect_timing_.dns_end;
if (result != OK)
return result;
// Invoke callback, and abort if it fails.
if (!params_->host_resolution_callback().is_null()) {
result = params_->host_resolution_callback().Run(addresses_, net_log());
if (result != OK)
return result;
}
next_state_ = STATE_TRANSPORT_CONNECT;
return result;
}
int WebSocketTransportConnectJob::DoTransportConnect() {
AddressList ipv4_addresses;
AddressList ipv6_addresses;
int result = ERR_UNEXPECTED;
next_state_ = STATE_TRANSPORT_CONNECT_COMPLETE;
for (AddressList::const_iterator it = addresses_.begin();
it != addresses_.end(); ++it) {
switch (it->GetFamily()) {
case ADDRESS_FAMILY_IPV4:
ipv4_addresses.push_back(*it);
break;
case ADDRESS_FAMILY_IPV6:
ipv6_addresses.push_back(*it);
break;
default:
DVLOG(1) << "Unexpected ADDRESS_FAMILY: " << it->GetFamily();
break;
}
}
if (!ipv4_addresses.empty()) {
had_ipv4_ = true;
ipv4_job_.reset(new WebSocketTransportConnectSubJob(
ipv4_addresses, this, SUB_JOB_IPV4, websocket_endpoint_lock_manager_));
}
if (!ipv6_addresses.empty()) {
had_ipv6_ = true;
ipv6_job_.reset(new WebSocketTransportConnectSubJob(
ipv6_addresses, this, SUB_JOB_IPV6, websocket_endpoint_lock_manager_));
result = ipv6_job_->Start();
switch (result) {
case OK:
SetSocket(ipv6_job_->PassSocket());
race_result_ = had_ipv4_ ? TransportConnectJob::RACE_IPV6_WINS
: TransportConnectJob::RACE_IPV6_SOLO;
return result;
case ERR_IO_PENDING:
if (ipv4_job_) {
// This use of base::Unretained is safe because |fallback_timer_| is
// owned by this object.
fallback_timer_.Start(
FROM_HERE, base::TimeDelta::FromMilliseconds(
TransportConnectJob::kIPv6FallbackTimerInMs),
base::Bind(&WebSocketTransportConnectJob::StartIPv4JobAsync,
base::Unretained(this)));
}
return result;
default:
ipv6_job_.reset();
}
}
DCHECK(!ipv6_job_);
if (ipv4_job_) {
result = ipv4_job_->Start();
if (result == OK) {
SetSocket(ipv4_job_->PassSocket());
race_result_ = had_ipv6_ ? TransportConnectJob::RACE_IPV4_WINS
: TransportConnectJob::RACE_IPV4_SOLO;
}
}
return result;
}
int WebSocketTransportConnectJob::DoTransportConnectComplete(int result) {
if (result == OK)
TransportConnectJob::HistogramDuration(connect_timing_, race_result_);
return result;
}
void WebSocketTransportConnectJob::OnSubJobComplete(
int result,
WebSocketTransportConnectSubJob* job) {
if (result == OK) {
switch (job->type()) {
case SUB_JOB_IPV4:
race_result_ = had_ipv6_ ? TransportConnectJob::RACE_IPV4_WINS
: TransportConnectJob::RACE_IPV4_SOLO;
break;
case SUB_JOB_IPV6:
race_result_ = had_ipv4_ ? TransportConnectJob::RACE_IPV6_WINS
: TransportConnectJob::RACE_IPV6_SOLO;
break;
}
SetSocket(job->PassSocket());
// Make sure all connections are cancelled even if this object fails to be
// deleted.
ipv4_job_.reset();
ipv6_job_.reset();
} else {
switch (job->type()) {
case SUB_JOB_IPV4:
ipv4_job_.reset();
break;
case SUB_JOB_IPV6:
ipv6_job_.reset();
if (ipv4_job_ && !ipv4_job_->started()) {
fallback_timer_.Stop();
result = ipv4_job_->Start();
if (result != ERR_IO_PENDING) {
OnSubJobComplete(result, ipv4_job_.get());
return;
}
}
break;
}
if (ipv4_job_ || ipv6_job_)
return;
}
OnIOComplete(result);
}
void WebSocketTransportConnectJob::StartIPv4JobAsync() {
DCHECK(ipv4_job_);
int result = ipv4_job_->Start();
if (result != ERR_IO_PENDING)
OnSubJobComplete(result, ipv4_job_.get());
}
int WebSocketTransportConnectJob::ConnectInternal() {
next_state_ = STATE_RESOLVE_HOST;
return DoLoop(OK);
}
WebSocketTransportClientSocketPool::WebSocketTransportClientSocketPool(
int max_sockets,
int max_sockets_per_group,
HostResolver* host_resolver,
ClientSocketFactory* client_socket_factory,
WebSocketEndpointLockManager* websocket_endpoint_lock_manager,
NetLog* net_log)
: TransportClientSocketPool(max_sockets,
max_sockets_per_group,
host_resolver,
client_socket_factory,
nullptr,
net_log),
connect_job_delegate_(this),
pool_net_log_(net_log),
client_socket_factory_(client_socket_factory),
host_resolver_(host_resolver),
websocket_endpoint_lock_manager_(websocket_endpoint_lock_manager),
max_sockets_(max_sockets),
handed_out_socket_count_(0),
flushing_(false),
weak_factory_(this) {}
WebSocketTransportClientSocketPool::~WebSocketTransportClientSocketPool() {
// Clean up any pending connect jobs.
FlushWithError(ERR_ABORTED);
DCHECK(pending_connects_.empty());
DCHECK_EQ(0, handed_out_socket_count_);
DCHECK(stalled_request_queue_.empty());
DCHECK(stalled_request_map_.empty());
}
// static
void WebSocketTransportClientSocketPool::UnlockEndpoint(
ClientSocketHandle* handle,
WebSocketEndpointLockManager* websocket_endpoint_lock_manager) {
DCHECK(handle->is_initialized());
DCHECK(handle->socket());
IPEndPoint address;
if (handle->socket()->GetPeerAddress(&address) == OK)
websocket_endpoint_lock_manager->UnlockEndpoint(address);
}
int WebSocketTransportClientSocketPool::RequestSocket(
const std::string& group_name,
const void* params,
RequestPriority priority,
const SocketTag& socket_tag,
RespectLimits respect_limits,
ClientSocketHandle* handle,
CompletionOnceCallback callback,
const NetLogWithSource& request_net_log) {
DCHECK(params);
const scoped_refptr<TransportSocketParams>& casted_params =
*static_cast<const scoped_refptr<TransportSocketParams>*>(params);
NetLogTcpClientSocketPoolRequestedSocket(request_net_log, &casted_params);
CHECK(!callback.is_null());
CHECK(handle);
request_net_log.BeginEvent(NetLogEventType::SOCKET_POOL);
DCHECK(socket_tag == SocketTag());
if (ReachedMaxSocketsLimit() &&
respect_limits == ClientSocketPool::RespectLimits::ENABLED) {
request_net_log.AddEvent(NetLogEventType::SOCKET_POOL_STALLED_MAX_SOCKETS);
stalled_request_queue_.emplace_back(casted_params, priority, handle,
std::move(callback), request_net_log);
auto iterator = stalled_request_queue_.end();
--iterator;
DCHECK_EQ(handle, iterator->handle);
// Because StalledRequestQueue is a std::list, its iterators are guaranteed
// to remain valid as long as the elements are not removed. As long as
// stalled_request_queue_ and stalled_request_map_ are updated in sync, it
// is safe to dereference an iterator in stalled_request_map_ to find the
// corresponding list element.
stalled_request_map_.insert(
StalledRequestMap::value_type(handle, iterator));
return ERR_IO_PENDING;
}
std::unique_ptr<WebSocketTransportConnectJob> connect_job(
new WebSocketTransportConnectJob(
group_name, priority, respect_limits, casted_params,
ConnectionTimeout(), std::move(callback), client_socket_factory_,
host_resolver_, handle, &connect_job_delegate_,
websocket_endpoint_lock_manager_, pool_net_log_, request_net_log));
int result = connect_job->Connect();
// Regardless of the outcome of |connect_job|, it will always be bound to
// |handle|, since this pool uses early-binding. So the binding is logged
// here, without waiting for the result.
request_net_log.AddEvent(
NetLogEventType::SOCKET_POOL_BOUND_TO_CONNECT_JOB,
connect_job->net_log().source().ToEventParametersCallback());
if (result == ERR_IO_PENDING) {
// TODO(ricea): Implement backup job timer?
AddJob(handle, std::move(connect_job));
} else {
TryHandOutSocket(result, connect_job.get());
}
return result;
}
void WebSocketTransportClientSocketPool::RequestSockets(
const std::string& group_name,
const void* params,
int num_sockets,
const NetLogWithSource& net_log) {
NOTIMPLEMENTED();
}
void WebSocketTransportClientSocketPool::SetPriority(
const std::string& group_name,
ClientSocketHandle* handle,
RequestPriority priority) {
// Since sockets requested by RequestSocket are bound early and
// stalled_request_{queue,map} don't take priorities into account, there's
// nothing to do within the pool to change priority or the request.
// TODO(rdsmith, ricea): Make stalled_request_{queue,map} take priorities
// into account.
// TODO(rdsmith): Investigate plumbing the reprioritization request to the
// connect job.
}
void WebSocketTransportClientSocketPool::CancelRequest(
const std::string& group_name,
ClientSocketHandle* handle) {
DCHECK(!handle->is_initialized());
if (DeleteStalledRequest(handle))
return;
std::unique_ptr<StreamSocket> socket = handle->PassSocket();
if (socket)
ReleaseSocket(handle->group_name(), std::move(socket), handle->id());
if (!DeleteJob(handle))
pending_callbacks_.erase(handle);
ActivateStalledRequest();
}
void WebSocketTransportClientSocketPool::ReleaseSocket(
const std::string& group_name,
std::unique_ptr<StreamSocket> socket,
int id) {
websocket_endpoint_lock_manager_->UnlockSocket(socket.get());
CHECK_GT(handed_out_socket_count_, 0);
--handed_out_socket_count_;
ActivateStalledRequest();
}
void WebSocketTransportClientSocketPool::FlushWithError(int error) {
DCHECK_NE(error, OK);
// Sockets which are in LOAD_STATE_CONNECTING are in danger of unlocking
// sockets waiting for the endpoint lock. If they connected synchronously,
// then OnConnectJobComplete(). The |flushing_| flag tells this object to
// ignore spurious calls to OnConnectJobComplete(). It is safe to ignore those
// calls because this method will delete the jobs and call their callbacks
// anyway.
flushing_ = true;
for (auto it = pending_connects_.begin(); it != pending_connects_.end();) {
InvokeUserCallbackLater(it->second->handle(),
it->second->release_callback(), error);
it = pending_connects_.erase(it);
}
for (auto it = stalled_request_queue_.begin();
it != stalled_request_queue_.end(); ++it) {
InvokeUserCallbackLater(it->handle, std::move(it->callback), error);
}
stalled_request_map_.clear();
stalled_request_queue_.clear();
flushing_ = false;
}
void WebSocketTransportClientSocketPool::CloseIdleSockets() {
// We have no idle sockets.
}
void WebSocketTransportClientSocketPool::CloseIdleSocketsInGroup(
const std::string& group_name) {
// We have no idle sockets.
}
int WebSocketTransportClientSocketPool::IdleSocketCount() const {
return 0;
}
int WebSocketTransportClientSocketPool::IdleSocketCountInGroup(
const std::string& group_name) const {
return 0;
}
LoadState WebSocketTransportClientSocketPool::GetLoadState(
const std::string& group_name,
const ClientSocketHandle* handle) const {
if (stalled_request_map_.find(handle) != stalled_request_map_.end())
return LOAD_STATE_WAITING_FOR_AVAILABLE_SOCKET;
if (pending_callbacks_.count(handle))
return LOAD_STATE_CONNECTING;
return LookupConnectJob(handle)->GetLoadState();
}
std::unique_ptr<base::DictionaryValue>
WebSocketTransportClientSocketPool::GetInfoAsValue(
const std::string& name,
const std::string& type,
bool include_nested_pools) const {
std::unique_ptr<base::DictionaryValue> dict(new base::DictionaryValue());
dict->SetString("name", name);
dict->SetString("type", type);
dict->SetInteger("handed_out_socket_count", handed_out_socket_count_);
dict->SetInteger("connecting_socket_count", pending_connects_.size());
dict->SetInteger("idle_socket_count", 0);
dict->SetInteger("max_socket_count", max_sockets_);
dict->SetInteger("max_sockets_per_group", max_sockets_);
dict->SetInteger("pool_generation_number", 0);
return dict;
}
base::TimeDelta WebSocketTransportClientSocketPool::ConnectionTimeout() const {
// TODO(ricea): For now, we implement a global timeout for compatibility with
// TransportConnectJob. Since WebSocketTransportConnectJob controls the
// address selection process more tightly, it could do something smarter here.
return base::TimeDelta::FromSeconds(TransportConnectJob::kTimeoutInSeconds);
}
bool WebSocketTransportClientSocketPool::IsStalled() const {
return !stalled_request_queue_.empty();
}
bool WebSocketTransportClientSocketPool::TryHandOutSocket(
int result,
WebSocketTransportConnectJob* job) {
DCHECK_NE(result, ERR_IO_PENDING);
std::unique_ptr<StreamSocket> socket = job->PassSocket();
ClientSocketHandle* const handle = job->handle();
NetLogWithSource request_net_log = job->request_net_log();
LoadTimingInfo::ConnectTiming connect_timing = job->connect_timing();
if (result == OK) {
DCHECK(socket);
HandOutSocket(std::move(socket), connect_timing, handle, request_net_log);
request_net_log.EndEvent(NetLogEventType::SOCKET_POOL);
return true;
}
bool handed_out_socket = false;
// If we got a socket, it must contain error information so pass that
// up so that the caller can retrieve it.
job->GetAdditionalErrorState(handle);
if (socket) {
HandOutSocket(std::move(socket), connect_timing, handle, request_net_log);
handed_out_socket = true;
}
request_net_log.EndEventWithNetErrorCode(NetLogEventType::SOCKET_POOL,
result);
return handed_out_socket;
}
void WebSocketTransportClientSocketPool::OnConnectJobComplete(
int result,
WebSocketTransportConnectJob* job) {
DCHECK_NE(ERR_IO_PENDING, result);
// See comment in FlushWithError.
if (flushing_) {
std::unique_ptr<StreamSocket> socket = job->PassSocket();
websocket_endpoint_lock_manager_->UnlockSocket(socket.get());
return;
}
bool handed_out_socket = TryHandOutSocket(result, job);
CompletionOnceCallback callback = job->release_callback();
ClientSocketHandle* const handle = job->handle();
bool delete_succeeded = DeleteJob(handle);
DCHECK(delete_succeeded);
job = nullptr;
if (!handed_out_socket)
ActivateStalledRequest();
InvokeUserCallbackLater(handle, std::move(callback), result);
}
void WebSocketTransportClientSocketPool::InvokeUserCallbackLater(
ClientSocketHandle* handle,
CompletionOnceCallback callback,
int rv) {
DCHECK(!pending_callbacks_.count(handle));
pending_callbacks_.insert(handle);
base::ThreadTaskRunnerHandle::Get()->PostTask(
FROM_HERE,
base::BindOnce(&WebSocketTransportClientSocketPool::InvokeUserCallback,
weak_factory_.GetWeakPtr(), handle, std::move(callback),
rv));
}
void WebSocketTransportClientSocketPool::InvokeUserCallback(
ClientSocketHandle* handle,
CompletionOnceCallback callback,
int rv) {
if (pending_callbacks_.erase(handle))
std::move(callback).Run(rv);
}
bool WebSocketTransportClientSocketPool::ReachedMaxSocketsLimit() const {
return handed_out_socket_count_ >= max_sockets_ ||
base::checked_cast<int>(pending_connects_.size()) >=
max_sockets_ - handed_out_socket_count_;
}
void WebSocketTransportClientSocketPool::HandOutSocket(
std::unique_ptr<StreamSocket> socket,
const LoadTimingInfo::ConnectTiming& connect_timing,
ClientSocketHandle* handle,
const NetLogWithSource& net_log) {
DCHECK(socket);
DCHECK_EQ(ClientSocketHandle::UNUSED, handle->reuse_type());
DCHECK_EQ(0, handle->idle_time().InMicroseconds());
handle->SetSocket(std::move(socket));
handle->set_pool_id(0);
handle->set_connect_timing(connect_timing);
net_log.AddEvent(
NetLogEventType::SOCKET_POOL_BOUND_TO_SOCKET,
handle->socket()->NetLog().source().ToEventParametersCallback());
++handed_out_socket_count_;
}
void WebSocketTransportClientSocketPool::AddJob(
ClientSocketHandle* handle,
std::unique_ptr<WebSocketTransportConnectJob> connect_job) {
bool inserted = pending_connects_
.insert(PendingConnectsMap::value_type(
handle, std::move(connect_job)))
.second;
DCHECK(inserted);
}
bool WebSocketTransportClientSocketPool::DeleteJob(ClientSocketHandle* handle) {
auto it = pending_connects_.find(handle);
if (it == pending_connects_.end())
return false;
// Deleting a ConnectJob which holds an endpoint lock can lead to a different
// ConnectJob proceeding to connect. If the connect proceeds synchronously
// (usually because of a failure) then it can trigger that job to be
// deleted.
pending_connects_.erase(it);
return true;
}
const WebSocketTransportConnectJob*
WebSocketTransportClientSocketPool::LookupConnectJob(
const ClientSocketHandle* handle) const {
auto it = pending_connects_.find(handle);
CHECK(it != pending_connects_.end());
return it->second.get();
}
void WebSocketTransportClientSocketPool::ActivateStalledRequest() {
// Usually we will only be able to activate one stalled request at a time,
// however if all the connects fail synchronously for some reason, we may be
// able to clear the whole queue at once.
while (!stalled_request_queue_.empty() && !ReachedMaxSocketsLimit()) {
StalledRequest request = std::move(stalled_request_queue_.front());
stalled_request_queue_.pop_front();
stalled_request_map_.erase(request.handle);
// Wrap request.callback into a copyable (repeating) callback so that it can
// be passed to RequestSocket() and yet called if RequestSocket() returns
// synchronously.
auto copyable_callback =
base::AdaptCallbackForRepeating(std::move(request.callback));
int rv =
RequestSocket("ignored", &request.params, request.priority, SocketTag(),
// Stalled requests can't have |respect_limits|
// DISABLED.
RespectLimits::ENABLED, request.handle, copyable_callback,
request.net_log);
// ActivateStalledRequest() never returns synchronously, so it is never
// called re-entrantly.
if (rv != ERR_IO_PENDING)
InvokeUserCallbackLater(request.handle, copyable_callback, rv);
}
}
bool WebSocketTransportClientSocketPool::DeleteStalledRequest(
ClientSocketHandle* handle) {
auto it = stalled_request_map_.find(handle);
if (it == stalled_request_map_.end())
return false;
stalled_request_queue_.erase(it->second);
stalled_request_map_.erase(it);
return true;
}
WebSocketTransportClientSocketPool::ConnectJobDelegate::ConnectJobDelegate(
WebSocketTransportClientSocketPool* owner)
: owner_(owner) {}
WebSocketTransportClientSocketPool::ConnectJobDelegate::~ConnectJobDelegate() =
default;
void
WebSocketTransportClientSocketPool::ConnectJobDelegate::OnConnectJobComplete(
int result,
ConnectJob* job) {
owner_->OnConnectJobComplete(result,
static_cast<WebSocketTransportConnectJob*>(job));
}
WebSocketTransportClientSocketPool::StalledRequest::StalledRequest(
const scoped_refptr<TransportSocketParams>& params,
RequestPriority priority,
ClientSocketHandle* handle,
CompletionOnceCallback callback,
const NetLogWithSource& net_log)
: params(params),
priority(priority),
handle(handle),
callback(std::move(callback)),
net_log(net_log) {}
WebSocketTransportClientSocketPool::StalledRequest::StalledRequest(
StalledRequest&& other) = default;
WebSocketTransportClientSocketPool::StalledRequest::~StalledRequest() = default;
} // namespace net