mirror of
https://github.com/klzgrad/naiveproxy.git
synced 2024-12-01 01:36:09 +03:00
939 lines
35 KiB
C
939 lines
35 KiB
C
|
// Copyright (c) 2012 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.
|
||
|
//
|
||
|
// A ClientSocketPoolBase is used to restrict the number of sockets open at
|
||
|
// a time. It also maintains a list of idle persistent sockets for reuse.
|
||
|
// Subclasses of ClientSocketPool should compose ClientSocketPoolBase to handle
|
||
|
// the core logic of (1) restricting the number of active (connected or
|
||
|
// connecting) sockets per "group" (generally speaking, the hostname), (2)
|
||
|
// maintaining a per-group list of idle, persistent sockets for reuse, and (3)
|
||
|
// limiting the total number of active sockets in the system.
|
||
|
//
|
||
|
// ClientSocketPoolBase abstracts socket connection details behind ConnectJob,
|
||
|
// ConnectJobFactory, and SocketParams. When a socket "slot" becomes available,
|
||
|
// the ClientSocketPoolBase will ask the ConnectJobFactory to create a
|
||
|
// ConnectJob with a SocketParams. Subclasses of ClientSocketPool should
|
||
|
// implement their socket specific connection by subclassing ConnectJob and
|
||
|
// implementing ConnectJob::ConnectInternal(). They can control the parameters
|
||
|
// passed to each new ConnectJob instance via their ConnectJobFactory subclass
|
||
|
// and templated SocketParams parameter.
|
||
|
//
|
||
|
#ifndef NET_SOCKET_CLIENT_SOCKET_POOL_BASE_H_
|
||
|
#define NET_SOCKET_CLIENT_SOCKET_POOL_BASE_H_
|
||
|
|
||
|
#include <stddef.h>
|
||
|
#include <stdint.h>
|
||
|
|
||
|
#include <cstddef>
|
||
|
#include <list>
|
||
|
#include <map>
|
||
|
#include <memory>
|
||
|
#include <set>
|
||
|
#include <string>
|
||
|
#include <utility>
|
||
|
#include <vector>
|
||
|
|
||
|
#include "base/macros.h"
|
||
|
#include "base/memory/ref_counted.h"
|
||
|
#include "base/memory/weak_ptr.h"
|
||
|
#include "base/time/time.h"
|
||
|
#include "base/timer/timer.h"
|
||
|
#include "net/base/address_list.h"
|
||
|
#include "net/base/completion_callback.h"
|
||
|
#include "net/base/load_states.h"
|
||
|
#include "net/base/load_timing_info.h"
|
||
|
#include "net/base/net_errors.h"
|
||
|
#include "net/base/net_export.h"
|
||
|
#include "net/base/network_change_notifier.h"
|
||
|
#include "net/base/priority_queue.h"
|
||
|
#include "net/base/request_priority.h"
|
||
|
#include "net/log/net_log_with_source.h"
|
||
|
#include "net/socket/client_socket_handle.h"
|
||
|
#include "net/socket/client_socket_pool.h"
|
||
|
#include "net/socket/stream_socket.h"
|
||
|
|
||
|
namespace base {
|
||
|
class DictionaryValue;
|
||
|
namespace trace_event {
|
||
|
class ProcessMemoryDump;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
namespace net {
|
||
|
|
||
|
class ClientSocketHandle;
|
||
|
struct NetLogSource;
|
||
|
|
||
|
// ConnectJob provides an abstract interface for "connecting" a socket.
|
||
|
// The connection may involve host resolution, tcp connection, ssl connection,
|
||
|
// etc.
|
||
|
class NET_EXPORT_PRIVATE ConnectJob {
|
||
|
public:
|
||
|
class NET_EXPORT_PRIVATE Delegate {
|
||
|
public:
|
||
|
Delegate() {}
|
||
|
virtual ~Delegate() {}
|
||
|
|
||
|
// Alerts the delegate that the connection completed. |job| must
|
||
|
// be destroyed by the delegate. A std::unique_ptr<> isn't used because
|
||
|
// the caller of this function doesn't own |job|.
|
||
|
virtual void OnConnectJobComplete(int result,
|
||
|
ConnectJob* job) = 0;
|
||
|
|
||
|
private:
|
||
|
DISALLOW_COPY_AND_ASSIGN(Delegate);
|
||
|
};
|
||
|
|
||
|
// A |timeout_duration| of 0 corresponds to no timeout.
|
||
|
ConnectJob(const std::string& group_name,
|
||
|
base::TimeDelta timeout_duration,
|
||
|
RequestPriority priority,
|
||
|
ClientSocketPool::RespectLimits respect_limits,
|
||
|
Delegate* delegate,
|
||
|
const NetLogWithSource& net_log);
|
||
|
virtual ~ConnectJob();
|
||
|
|
||
|
// Accessors
|
||
|
const std::string& group_name() const { return group_name_; }
|
||
|
const NetLogWithSource& net_log() { return net_log_; }
|
||
|
|
||
|
// Releases ownership of the underlying socket to the caller.
|
||
|
// Returns the released socket, or NULL if there was a connection
|
||
|
// error.
|
||
|
std::unique_ptr<StreamSocket> PassSocket();
|
||
|
|
||
|
void set_motivation(HttpRequestInfo::RequestMotivation motivation) {
|
||
|
motivation_ = motivation;
|
||
|
}
|
||
|
|
||
|
// Begins connecting the socket. Returns OK on success, ERR_IO_PENDING if it
|
||
|
// cannot complete synchronously without blocking, or another net error code
|
||
|
// on error. In asynchronous completion, the ConnectJob will notify
|
||
|
// |delegate_| via OnConnectJobComplete. In both asynchronous and synchronous
|
||
|
// completion, ReleaseSocket() can be called to acquire the connected socket
|
||
|
// if it succeeded.
|
||
|
int Connect();
|
||
|
|
||
|
virtual LoadState GetLoadState() const = 0;
|
||
|
|
||
|
// If Connect returns an error (or OnConnectJobComplete reports an error
|
||
|
// result) this method will be called, allowing the pool to add
|
||
|
// additional error state to the ClientSocketHandle (post late-binding).
|
||
|
virtual void GetAdditionalErrorState(ClientSocketHandle* handle) {}
|
||
|
|
||
|
const LoadTimingInfo::ConnectTiming& connect_timing() const {
|
||
|
return connect_timing_;
|
||
|
}
|
||
|
|
||
|
const NetLogWithSource& net_log() const { return net_log_; }
|
||
|
|
||
|
// TODO(mmenke): Remove when https://crbug.com/756993 is figured out.
|
||
|
|
||
|
// Checks that |group_name_| and |group_name_copy_| are the same. If not,
|
||
|
// crashes.
|
||
|
void CheckGroupName() const;
|
||
|
|
||
|
// Copies |group_name_| and |group_name_copy_| to the stack and crashes.
|
||
|
void LogGroupNameAndCrash() const;
|
||
|
|
||
|
protected:
|
||
|
RequestPriority priority() const { return priority_; }
|
||
|
ClientSocketPool::RespectLimits respect_limits() const {
|
||
|
return respect_limits_;
|
||
|
}
|
||
|
void SetSocket(std::unique_ptr<StreamSocket> socket);
|
||
|
StreamSocket* socket() { return socket_.get(); }
|
||
|
void NotifyDelegateOfCompletion(int rv);
|
||
|
void ResetTimer(base::TimeDelta remainingTime);
|
||
|
|
||
|
// Connection establishment timing information.
|
||
|
LoadTimingInfo::ConnectTiming connect_timing_;
|
||
|
|
||
|
private:
|
||
|
virtual int ConnectInternal() = 0;
|
||
|
|
||
|
void LogConnectStart();
|
||
|
void LogConnectCompletion(int net_error);
|
||
|
|
||
|
// Alerts the delegate that the ConnectJob has timed out.
|
||
|
void OnTimeout();
|
||
|
|
||
|
const std::string group_name_;
|
||
|
// TODO(mmenke): Remove when https://crbug.com/756993 is figured out.
|
||
|
const std::string group_name_copy_;
|
||
|
const base::TimeDelta timeout_duration_;
|
||
|
// TODO(akalin): Support reprioritization.
|
||
|
const RequestPriority priority_;
|
||
|
const ClientSocketPool::RespectLimits respect_limits_;
|
||
|
// Timer to abort jobs that take too long.
|
||
|
base::OneShotTimer timer_;
|
||
|
Delegate* delegate_;
|
||
|
std::unique_ptr<StreamSocket> socket_;
|
||
|
NetLogWithSource net_log_;
|
||
|
HttpRequestInfo::RequestMotivation motivation_;
|
||
|
// A ConnectJob is idle until Connect() has been called.
|
||
|
bool idle_;
|
||
|
|
||
|
DISALLOW_COPY_AND_ASSIGN(ConnectJob);
|
||
|
};
|
||
|
|
||
|
namespace internal {
|
||
|
|
||
|
// ClientSocketPoolBaseHelper is an internal class that implements almost all
|
||
|
// the functionality from ClientSocketPoolBase without using templates.
|
||
|
// ClientSocketPoolBase adds templated definitions built on top of
|
||
|
// ClientSocketPoolBaseHelper. This class is not for external use, please use
|
||
|
// ClientSocketPoolBase instead.
|
||
|
class NET_EXPORT_PRIVATE ClientSocketPoolBaseHelper
|
||
|
: public ConnectJob::Delegate,
|
||
|
public NetworkChangeNotifier::IPAddressObserver {
|
||
|
public:
|
||
|
typedef uint32_t Flags;
|
||
|
|
||
|
// Used to specify specific behavior for the ClientSocketPool.
|
||
|
enum Flag {
|
||
|
NORMAL = 0, // Normal behavior.
|
||
|
NO_IDLE_SOCKETS = 0x1, // Do not return an idle socket. Create a new one.
|
||
|
};
|
||
|
|
||
|
class NET_EXPORT_PRIVATE Request {
|
||
|
public:
|
||
|
Request(ClientSocketHandle* handle,
|
||
|
const CompletionCallback& callback,
|
||
|
RequestPriority priority,
|
||
|
ClientSocketPool::RespectLimits respect_limits,
|
||
|
Flags flags,
|
||
|
const NetLogWithSource& net_log);
|
||
|
|
||
|
virtual ~Request();
|
||
|
|
||
|
ClientSocketHandle* handle() const { return handle_; }
|
||
|
const CompletionCallback& callback() const { return callback_; }
|
||
|
RequestPriority priority() const { return priority_; }
|
||
|
void set_priority(RequestPriority priority) { priority_ = priority; }
|
||
|
ClientSocketPool::RespectLimits respect_limits() const {
|
||
|
return respect_limits_;
|
||
|
}
|
||
|
Flags flags() const { return flags_; }
|
||
|
const NetLogWithSource& net_log() const { return net_log_; }
|
||
|
|
||
|
// TODO(eroman): Temporary until crbug.com/467797 is solved.
|
||
|
void CrashIfInvalid() const;
|
||
|
|
||
|
private:
|
||
|
// TODO(eroman): Temporary until crbug.com/467797 is solved.
|
||
|
enum Liveness {
|
||
|
ALIVE = 0xCA11AB13,
|
||
|
DEAD = 0xDEADBEEF,
|
||
|
};
|
||
|
|
||
|
ClientSocketHandle* const handle_;
|
||
|
const CompletionCallback callback_;
|
||
|
RequestPriority priority_;
|
||
|
const ClientSocketPool::RespectLimits respect_limits_;
|
||
|
const Flags flags_;
|
||
|
const NetLogWithSource net_log_;
|
||
|
|
||
|
// TODO(eroman): Temporary until crbug.com/467797 is solved.
|
||
|
Liveness liveness_ = ALIVE;
|
||
|
|
||
|
DISALLOW_COPY_AND_ASSIGN(Request);
|
||
|
};
|
||
|
|
||
|
class ConnectJobFactory {
|
||
|
public:
|
||
|
ConnectJobFactory() {}
|
||
|
virtual ~ConnectJobFactory() {}
|
||
|
|
||
|
virtual std::unique_ptr<ConnectJob> NewConnectJob(
|
||
|
const std::string& group_name,
|
||
|
const Request& request,
|
||
|
ConnectJob::Delegate* delegate) const = 0;
|
||
|
|
||
|
virtual base::TimeDelta ConnectionTimeout() const = 0;
|
||
|
|
||
|
private:
|
||
|
DISALLOW_COPY_AND_ASSIGN(ConnectJobFactory);
|
||
|
};
|
||
|
|
||
|
ClientSocketPoolBaseHelper(
|
||
|
HigherLayeredPool* pool,
|
||
|
int max_sockets,
|
||
|
int max_sockets_per_group,
|
||
|
base::TimeDelta unused_idle_socket_timeout,
|
||
|
base::TimeDelta used_idle_socket_timeout,
|
||
|
ConnectJobFactory* connect_job_factory);
|
||
|
|
||
|
~ClientSocketPoolBaseHelper() override;
|
||
|
|
||
|
// Adds a lower layered pool to |this|, and adds |this| as a higher layered
|
||
|
// pool on top of |lower_pool|.
|
||
|
void AddLowerLayeredPool(LowerLayeredPool* lower_pool);
|
||
|
|
||
|
// See LowerLayeredPool::IsStalled for documentation on this function.
|
||
|
bool IsStalled() const;
|
||
|
|
||
|
// See LowerLayeredPool for documentation on these functions. It is expected
|
||
|
// in the destructor that no higher layer pools remain.
|
||
|
void AddHigherLayeredPool(HigherLayeredPool* higher_pool);
|
||
|
void RemoveHigherLayeredPool(HigherLayeredPool* higher_pool);
|
||
|
|
||
|
// See ClientSocketPool::RequestSocket for documentation on this function.
|
||
|
int RequestSocket(const std::string& group_name,
|
||
|
std::unique_ptr<Request> request);
|
||
|
|
||
|
// See ClientSocketPool::RequestSockets for documentation on this function.
|
||
|
void RequestSockets(const std::string& group_name,
|
||
|
const Request& request,
|
||
|
int num_sockets,
|
||
|
HttpRequestInfo::RequestMotivation motivation);
|
||
|
|
||
|
// See ClientSocketPool::SetPriority for documentation on this function.
|
||
|
void SetPriority(const std::string& group_name,
|
||
|
ClientSocketHandle* handle,
|
||
|
RequestPriority priority);
|
||
|
|
||
|
// See ClientSocketPool::CancelRequest for documentation on this function.
|
||
|
void CancelRequest(const std::string& group_name,
|
||
|
ClientSocketHandle* handle);
|
||
|
|
||
|
// See ClientSocketPool::ReleaseSocket for documentation on this function.
|
||
|
void ReleaseSocket(const std::string& group_name,
|
||
|
std::unique_ptr<StreamSocket> socket,
|
||
|
int id);
|
||
|
|
||
|
// See ClientSocketPool::FlushWithError for documentation on this function.
|
||
|
void FlushWithError(int error);
|
||
|
|
||
|
// See ClientSocketPool::CloseIdleSockets for documentation on this function.
|
||
|
void CloseIdleSockets();
|
||
|
|
||
|
// See ClientSocketPool::CloseIdleSocketsInGroup for documentation.
|
||
|
void CloseIdleSocketsInGroup(const std::string& group_name);
|
||
|
|
||
|
// See ClientSocketPool::IdleSocketCount() for documentation on this function.
|
||
|
int idle_socket_count() const {
|
||
|
return idle_socket_count_;
|
||
|
}
|
||
|
|
||
|
// See ClientSocketPool::IdleSocketCountInGroup() for documentation on this
|
||
|
// function.
|
||
|
int IdleSocketCountInGroup(const std::string& group_name) const;
|
||
|
|
||
|
// See ClientSocketPool::GetLoadState() for documentation on this function.
|
||
|
LoadState GetLoadState(const std::string& group_name,
|
||
|
const ClientSocketHandle* handle) const;
|
||
|
|
||
|
base::TimeDelta ConnectRetryInterval() const {
|
||
|
// TODO(mbelshe): Make this tuned dynamically based on measured RTT.
|
||
|
// For now, just use the max retry interval.
|
||
|
return base::TimeDelta::FromMilliseconds(
|
||
|
ClientSocketPool::kMaxConnectRetryIntervalMs);
|
||
|
}
|
||
|
|
||
|
int NumUnassignedConnectJobsInGroup(const std::string& group_name) const {
|
||
|
return group_map_.find(group_name)->second->unassigned_job_count();
|
||
|
}
|
||
|
|
||
|
int NumConnectJobsInGroup(const std::string& group_name) const {
|
||
|
return group_map_.find(group_name)->second->jobs().size();
|
||
|
}
|
||
|
|
||
|
int NumActiveSocketsInGroup(const std::string& group_name) const {
|
||
|
return group_map_.find(group_name)->second->active_socket_count();
|
||
|
}
|
||
|
|
||
|
bool HasGroup(const std::string& group_name) const;
|
||
|
|
||
|
// Closes all idle sockets if |force| is true. Else, only closes idle
|
||
|
// sockets that timed out or can't be reused. Made public for testing.
|
||
|
void CleanupIdleSockets(bool force);
|
||
|
|
||
|
// Closes one idle socket. Picks the first one encountered.
|
||
|
// TODO(willchan): Consider a better algorithm for doing this. Perhaps we
|
||
|
// should keep an ordered list of idle sockets, and close them in order.
|
||
|
// Requires maintaining more state. It's not clear if it's worth it since
|
||
|
// I'm not sure if we hit this situation often.
|
||
|
bool CloseOneIdleSocket();
|
||
|
|
||
|
// Checks higher layered pools to see if they can close an idle connection.
|
||
|
bool CloseOneIdleConnectionInHigherLayeredPool();
|
||
|
|
||
|
// See ClientSocketPool::GetInfoAsValue for documentation on this function.
|
||
|
std::unique_ptr<base::DictionaryValue> GetInfoAsValue(
|
||
|
const std::string& name,
|
||
|
const std::string& type) const;
|
||
|
|
||
|
// Dumps memory allocation stats. |parent_dump_absolute_name| is the name
|
||
|
// used by the parent MemoryAllocatorDump in the memory dump hierarchy.
|
||
|
void DumpMemoryStats(base::trace_event::ProcessMemoryDump* pmd,
|
||
|
const std::string& parent_dump_absolute_name) const;
|
||
|
|
||
|
base::TimeDelta ConnectionTimeout() const {
|
||
|
return connect_job_factory_->ConnectionTimeout();
|
||
|
}
|
||
|
|
||
|
static bool connect_backup_jobs_enabled();
|
||
|
static bool set_connect_backup_jobs_enabled(bool enabled);
|
||
|
|
||
|
void EnableConnectBackupJobs();
|
||
|
|
||
|
// ConnectJob::Delegate methods:
|
||
|
void OnConnectJobComplete(int result, ConnectJob* job) override;
|
||
|
|
||
|
// NetworkChangeNotifier::IPAddressObserver methods:
|
||
|
void OnIPAddressChanged() override;
|
||
|
|
||
|
private:
|
||
|
// Entry for a persistent socket which became idle at time |start_time|.
|
||
|
struct IdleSocket {
|
||
|
IdleSocket() : socket(NULL) {}
|
||
|
|
||
|
// An idle socket can't be used if it is disconnected or has been used
|
||
|
// before and has received data unexpectedly (hence no longer idle). The
|
||
|
// unread data would be mistaken for the beginning of the next response if
|
||
|
// we were to use the socket for a new request.
|
||
|
//
|
||
|
// Note that a socket that has never been used before (like a preconnected
|
||
|
// socket) may be used even with unread data. This may be, e.g., a SPDY
|
||
|
// SETTINGS frame.
|
||
|
bool IsUsable() const;
|
||
|
|
||
|
StreamSocket* socket;
|
||
|
base::TimeTicks start_time;
|
||
|
};
|
||
|
|
||
|
typedef PriorityQueue<Request*> RequestQueue;
|
||
|
typedef std::map<const ClientSocketHandle*, Request*> RequestMap;
|
||
|
|
||
|
// A Group is allocated per group_name when there are idle sockets or pending
|
||
|
// requests. Otherwise, the Group object is removed from the map.
|
||
|
// |active_socket_count| tracks the number of sockets held by clients.
|
||
|
class Group {
|
||
|
public:
|
||
|
using JobList = std::list<std::unique_ptr<ConnectJob>>;
|
||
|
|
||
|
Group();
|
||
|
~Group();
|
||
|
|
||
|
bool IsEmpty() const {
|
||
|
return active_socket_count_ == 0 && idle_sockets_.empty() &&
|
||
|
jobs_.empty() && pending_requests_.empty();
|
||
|
}
|
||
|
|
||
|
bool HasAvailableSocketSlot(int max_sockets_per_group) const {
|
||
|
return NumActiveSocketSlots() < max_sockets_per_group;
|
||
|
}
|
||
|
|
||
|
int NumActiveSocketSlots() const {
|
||
|
return active_socket_count_ + static_cast<int>(jobs_.size()) +
|
||
|
static_cast<int>(idle_sockets_.size());
|
||
|
}
|
||
|
|
||
|
// Returns true if the group could make use of an additional socket slot, if
|
||
|
// it were given one.
|
||
|
bool CanUseAdditionalSocketSlot(int max_sockets_per_group) const {
|
||
|
return HasAvailableSocketSlot(max_sockets_per_group) &&
|
||
|
pending_requests_.size() > jobs_.size();
|
||
|
}
|
||
|
|
||
|
// Returns the priority of the top of the pending request queue
|
||
|
// (which may be less than the maximum priority over the entire
|
||
|
// queue, due to how we prioritize requests with |respect_limits|
|
||
|
// DISABLED over others).
|
||
|
RequestPriority TopPendingPriority() const {
|
||
|
// NOTE: FirstMax().value()->priority() is not the same as
|
||
|
// FirstMax().priority()!
|
||
|
return pending_requests_.FirstMax().value()->priority();
|
||
|
}
|
||
|
|
||
|
// Set a timer to create a backup job if it takes too long to
|
||
|
// create one and if a timer isn't already running.
|
||
|
void StartBackupJobTimer(const std::string& group_name,
|
||
|
ClientSocketPoolBaseHelper* pool);
|
||
|
|
||
|
bool BackupJobTimerIsRunning() const;
|
||
|
|
||
|
// If there's a ConnectJob that's never been assigned to Request,
|
||
|
// decrements |unassigned_job_count_| and returns true.
|
||
|
// Otherwise, returns false.
|
||
|
bool TryToUseUnassignedConnectJob();
|
||
|
|
||
|
void AddJob(std::unique_ptr<ConnectJob> job, bool is_preconnect);
|
||
|
// Remove |job| from this group, which must already own |job|.
|
||
|
void RemoveJob(ConnectJob* job);
|
||
|
void RemoveAllJobs();
|
||
|
|
||
|
bool has_pending_requests() const {
|
||
|
return !pending_requests_.empty();
|
||
|
}
|
||
|
|
||
|
size_t pending_request_count() const {
|
||
|
return pending_requests_.size();
|
||
|
}
|
||
|
|
||
|
// Gets (but does not remove) the next pending request. Returns
|
||
|
// NULL if there are no pending requests.
|
||
|
const Request* GetNextPendingRequest() const;
|
||
|
|
||
|
// Returns true if there is a connect job for |handle|.
|
||
|
bool HasConnectJobForHandle(const ClientSocketHandle* handle) const;
|
||
|
|
||
|
// Inserts the request into the queue based on priority
|
||
|
// order. Older requests are prioritized over requests of equal
|
||
|
// priority.
|
||
|
void InsertPendingRequest(std::unique_ptr<Request> request);
|
||
|
|
||
|
// Gets and removes the next pending request. Returns NULL if
|
||
|
// there are no pending requests.
|
||
|
std::unique_ptr<Request> PopNextPendingRequest();
|
||
|
|
||
|
// Finds the pending request for |handle| and removes it. Returns
|
||
|
// the removed pending request, or NULL if there was none.
|
||
|
std::unique_ptr<Request> FindAndRemovePendingRequest(
|
||
|
ClientSocketHandle* handle);
|
||
|
|
||
|
// Change the priority of the request named by |*handle|. |*handle|
|
||
|
// must refer to a request currently present in the group. If |priority|
|
||
|
// is the same as the current priority of the request, this is a no-op.
|
||
|
void SetPriority(ClientSocketHandle* handle, RequestPriority priority);
|
||
|
|
||
|
void IncrementActiveSocketCount() { active_socket_count_++; }
|
||
|
void DecrementActiveSocketCount() { active_socket_count_--; }
|
||
|
|
||
|
int unassigned_job_count() const { return unassigned_job_count_; }
|
||
|
const JobList& jobs() const { return jobs_; }
|
||
|
const std::list<IdleSocket>& idle_sockets() const { return idle_sockets_; }
|
||
|
int active_socket_count() const { return active_socket_count_; }
|
||
|
std::list<IdleSocket>* mutable_idle_sockets() { return &idle_sockets_; }
|
||
|
|
||
|
private:
|
||
|
// Returns the iterator's pending request after removing it from
|
||
|
// the queue.
|
||
|
std::unique_ptr<Request> RemovePendingRequest(
|
||
|
const RequestQueue::Pointer& pointer);
|
||
|
|
||
|
// Called when the backup socket timer fires.
|
||
|
void OnBackupJobTimerFired(
|
||
|
std::string group_name,
|
||
|
ClientSocketPoolBaseHelper* pool);
|
||
|
|
||
|
// Checks that |unassigned_job_count_| does not execeed the number of
|
||
|
// ConnectJobs.
|
||
|
void SanityCheck();
|
||
|
|
||
|
// Total number of ConnectJobs that have never been assigned to a Request.
|
||
|
// Since jobs use late binding to requests, which ConnectJobs have or have
|
||
|
// not been assigned to a request are not tracked. This is incremented on
|
||
|
// preconnect and decremented when a preconnect is assigned, or when there
|
||
|
// are fewer than |unassigned_job_count_| ConnectJobs. Not incremented
|
||
|
// when a request is cancelled.
|
||
|
size_t unassigned_job_count_;
|
||
|
|
||
|
std::list<IdleSocket> idle_sockets_;
|
||
|
JobList jobs_;
|
||
|
RequestQueue pending_requests_;
|
||
|
int active_socket_count_; // number of active sockets used by clients
|
||
|
// A timer for when to start the backup job.
|
||
|
base::OneShotTimer backup_job_timer_;
|
||
|
};
|
||
|
|
||
|
typedef std::map<std::string, Group*> GroupMap;
|
||
|
|
||
|
typedef std::set<ConnectJob*> ConnectJobSet;
|
||
|
|
||
|
struct CallbackResultPair {
|
||
|
CallbackResultPair();
|
||
|
CallbackResultPair(const CompletionCallback& callback_in, int result_in);
|
||
|
CallbackResultPair(const CallbackResultPair& other);
|
||
|
~CallbackResultPair();
|
||
|
|
||
|
CompletionCallback callback;
|
||
|
int result;
|
||
|
};
|
||
|
|
||
|
typedef std::map<const ClientSocketHandle*, CallbackResultPair>
|
||
|
PendingCallbackMap;
|
||
|
|
||
|
// Closes all idle sockets in |group| if |force| is true. Else, only closes
|
||
|
// idle sockets in |group| that timed out with respect to |now| or can't be
|
||
|
// reused.
|
||
|
void CleanupIdleSocketsInGroup(bool force,
|
||
|
Group* group,
|
||
|
const base::TimeTicks& now);
|
||
|
|
||
|
Group* GetOrCreateGroup(const std::string& group_name);
|
||
|
void RemoveGroup(const std::string& group_name);
|
||
|
void RemoveGroup(GroupMap::iterator it);
|
||
|
|
||
|
// Called when the number of idle sockets changes.
|
||
|
void IncrementIdleCount();
|
||
|
void DecrementIdleCount();
|
||
|
|
||
|
// Scans the group map for groups which have an available socket slot and
|
||
|
// at least one pending request. Returns true if any groups are stalled, and
|
||
|
// if so (and if both |group| and |group_name| are not NULL), fills |group|
|
||
|
// and |group_name| with data of the stalled group having highest priority.
|
||
|
bool FindTopStalledGroup(Group** group, std::string* group_name) const;
|
||
|
|
||
|
// Removes |job| from |group|, which must already own |job|.
|
||
|
void RemoveConnectJob(ConnectJob* job, Group* group);
|
||
|
|
||
|
// Tries to see if we can handle any more requests for |group|.
|
||
|
void OnAvailableSocketSlot(const std::string& group_name, Group* group);
|
||
|
|
||
|
// Process a pending socket request for a group.
|
||
|
void ProcessPendingRequest(const std::string& group_name, Group* group);
|
||
|
|
||
|
// Assigns |socket| to |handle| and updates |group|'s counters appropriately.
|
||
|
void HandOutSocket(std::unique_ptr<StreamSocket> socket,
|
||
|
ClientSocketHandle::SocketReuseType reuse_type,
|
||
|
const LoadTimingInfo::ConnectTiming& connect_timing,
|
||
|
ClientSocketHandle* handle,
|
||
|
base::TimeDelta time_idle,
|
||
|
Group* group,
|
||
|
const NetLogWithSource& net_log);
|
||
|
|
||
|
// Adds |socket| to the list of idle sockets for |group|.
|
||
|
void AddIdleSocket(std::unique_ptr<StreamSocket> socket, Group* group);
|
||
|
|
||
|
// Iterates through |group_map_|, canceling all ConnectJobs and deleting
|
||
|
// groups if they are no longer needed.
|
||
|
void CancelAllConnectJobs();
|
||
|
|
||
|
// Iterates through |group_map_|, posting |error| callbacks for all
|
||
|
// requests, and then deleting groups if they are no longer needed.
|
||
|
void CancelAllRequestsWithError(int error);
|
||
|
|
||
|
// Returns true if we can't create any more sockets due to the total limit.
|
||
|
bool ReachedMaxSocketsLimit() const;
|
||
|
|
||
|
// This is the internal implementation of RequestSocket(). It differs in that
|
||
|
// it does not handle logging into NetLog of the queueing status of
|
||
|
// |request|.
|
||
|
int RequestSocketInternal(const std::string& group_name,
|
||
|
const Request& request,
|
||
|
HttpRequestInfo::RequestMotivation motivation);
|
||
|
|
||
|
// Assigns an idle socket for the group to the request.
|
||
|
// Returns |true| if an idle socket is available, false otherwise.
|
||
|
bool AssignIdleSocketToRequest(const Request& request, Group* group);
|
||
|
|
||
|
static void LogBoundConnectJobToRequest(
|
||
|
const NetLogSource& connect_job_source,
|
||
|
const Request& request);
|
||
|
|
||
|
// Same as CloseOneIdleSocket() except it won't close an idle socket in
|
||
|
// |group|. If |group| is NULL, it is ignored. Returns true if it closed a
|
||
|
// socket.
|
||
|
bool CloseOneIdleSocketExceptInGroup(const Group* group);
|
||
|
|
||
|
// Checks if there are stalled socket groups that should be notified
|
||
|
// for possible wakeup.
|
||
|
void CheckForStalledSocketGroups();
|
||
|
|
||
|
// Posts a task to call InvokeUserCallback() on the next iteration through the
|
||
|
// current message loop. Inserts |callback| into |pending_callback_map_|,
|
||
|
// keyed by |handle|.
|
||
|
void InvokeUserCallbackLater(
|
||
|
ClientSocketHandle* handle, const CompletionCallback& callback, int rv);
|
||
|
|
||
|
// Invokes the user callback for |handle|. By the time this task has run,
|
||
|
// it's possible that the request has been cancelled, so |handle| may not
|
||
|
// exist in |pending_callback_map_|. We look up the callback and result code
|
||
|
// in |pending_callback_map_|.
|
||
|
void InvokeUserCallback(ClientSocketHandle* handle);
|
||
|
|
||
|
// Tries to close idle sockets in a higher level socket pool as long as this
|
||
|
// this pool is stalled.
|
||
|
void TryToCloseSocketsInLayeredPools();
|
||
|
|
||
|
GroupMap group_map_;
|
||
|
|
||
|
// Map of the ClientSocketHandles for which we have a pending Task to invoke a
|
||
|
// callback. This is necessary since, before we invoke said callback, it's
|
||
|
// possible that the request is cancelled.
|
||
|
PendingCallbackMap pending_callback_map_;
|
||
|
|
||
|
// The total number of idle sockets in the system.
|
||
|
int idle_socket_count_;
|
||
|
|
||
|
// Number of connecting sockets across all groups.
|
||
|
int connecting_socket_count_;
|
||
|
|
||
|
// Number of connected sockets we handed out across all groups.
|
||
|
int handed_out_socket_count_;
|
||
|
|
||
|
// The maximum total number of sockets. See ReachedMaxSocketsLimit.
|
||
|
const int max_sockets_;
|
||
|
|
||
|
// The maximum number of sockets kept per group.
|
||
|
const int max_sockets_per_group_;
|
||
|
|
||
|
// The time to wait until closing idle sockets.
|
||
|
const base::TimeDelta unused_idle_socket_timeout_;
|
||
|
const base::TimeDelta used_idle_socket_timeout_;
|
||
|
|
||
|
const std::unique_ptr<ConnectJobFactory> connect_job_factory_;
|
||
|
|
||
|
// TODO(vandebo) Remove when backup jobs move to TransportClientSocketPool
|
||
|
bool connect_backup_jobs_enabled_;
|
||
|
|
||
|
// A unique id for the pool. It gets incremented every time we
|
||
|
// FlushWithError() the pool. This is so that when sockets get released back
|
||
|
// to the pool, we can make sure that they are discarded rather than reused.
|
||
|
int pool_generation_number_;
|
||
|
|
||
|
// Used to add |this| as a higher layer pool on top of lower layer pools. May
|
||
|
// be NULL if no lower layer pools will be added.
|
||
|
HigherLayeredPool* pool_;
|
||
|
|
||
|
// Pools that create connections through |this|. |this| will try to close
|
||
|
// their idle sockets when it stalls. Must be empty on destruction.
|
||
|
std::set<HigherLayeredPool*> higher_pools_;
|
||
|
|
||
|
// Pools that this goes through. Typically there's only one, but not always.
|
||
|
// |this| will check if they're stalled when it has a new idle socket. |this|
|
||
|
// will remove itself from all lower layered pools on destruction.
|
||
|
std::set<LowerLayeredPool*> lower_pools_;
|
||
|
|
||
|
base::WeakPtrFactory<ClientSocketPoolBaseHelper> weak_factory_;
|
||
|
|
||
|
DISALLOW_COPY_AND_ASSIGN(ClientSocketPoolBaseHelper);
|
||
|
};
|
||
|
|
||
|
} // namespace internal
|
||
|
|
||
|
template <typename SocketParams>
|
||
|
class ClientSocketPoolBase {
|
||
|
public:
|
||
|
class Request : public internal::ClientSocketPoolBaseHelper::Request {
|
||
|
public:
|
||
|
Request(ClientSocketHandle* handle,
|
||
|
const CompletionCallback& callback,
|
||
|
RequestPriority priority,
|
||
|
ClientSocketPool::RespectLimits respect_limits,
|
||
|
internal::ClientSocketPoolBaseHelper::Flags flags,
|
||
|
const scoped_refptr<SocketParams>& params,
|
||
|
const NetLogWithSource& net_log)
|
||
|
: internal::ClientSocketPoolBaseHelper::Request(handle,
|
||
|
callback,
|
||
|
priority,
|
||
|
respect_limits,
|
||
|
flags,
|
||
|
net_log),
|
||
|
params_(params) {}
|
||
|
|
||
|
const scoped_refptr<SocketParams>& params() const { return params_; }
|
||
|
|
||
|
private:
|
||
|
const scoped_refptr<SocketParams> params_;
|
||
|
};
|
||
|
|
||
|
class ConnectJobFactory {
|
||
|
public:
|
||
|
ConnectJobFactory() {}
|
||
|
virtual ~ConnectJobFactory() {}
|
||
|
|
||
|
virtual std::unique_ptr<ConnectJob> NewConnectJob(
|
||
|
const std::string& group_name,
|
||
|
const Request& request,
|
||
|
ConnectJob::Delegate* delegate) const = 0;
|
||
|
|
||
|
virtual base::TimeDelta ConnectionTimeout() const = 0;
|
||
|
|
||
|
private:
|
||
|
DISALLOW_COPY_AND_ASSIGN(ConnectJobFactory);
|
||
|
};
|
||
|
|
||
|
// |max_sockets| is the maximum number of sockets to be maintained by this
|
||
|
// ClientSocketPool. |max_sockets_per_group| specifies the maximum number of
|
||
|
// sockets a "group" can have. |unused_idle_socket_timeout| specifies how
|
||
|
// long to leave an unused idle socket open before closing it.
|
||
|
// |used_idle_socket_timeout| specifies how long to leave a previously used
|
||
|
// idle socket open before closing it.
|
||
|
ClientSocketPoolBase(HigherLayeredPool* self,
|
||
|
int max_sockets,
|
||
|
int max_sockets_per_group,
|
||
|
base::TimeDelta unused_idle_socket_timeout,
|
||
|
base::TimeDelta used_idle_socket_timeout,
|
||
|
ConnectJobFactory* connect_job_factory)
|
||
|
: helper_(self,
|
||
|
max_sockets,
|
||
|
max_sockets_per_group,
|
||
|
unused_idle_socket_timeout,
|
||
|
used_idle_socket_timeout,
|
||
|
new ConnectJobFactoryAdaptor(connect_job_factory)) {}
|
||
|
|
||
|
virtual ~ClientSocketPoolBase() {}
|
||
|
|
||
|
// These member functions simply forward to ClientSocketPoolBaseHelper.
|
||
|
void AddLowerLayeredPool(LowerLayeredPool* lower_pool) {
|
||
|
helper_.AddLowerLayeredPool(lower_pool);
|
||
|
}
|
||
|
|
||
|
void AddHigherLayeredPool(HigherLayeredPool* higher_pool) {
|
||
|
helper_.AddHigherLayeredPool(higher_pool);
|
||
|
}
|
||
|
|
||
|
void RemoveHigherLayeredPool(HigherLayeredPool* higher_pool) {
|
||
|
helper_.RemoveHigherLayeredPool(higher_pool);
|
||
|
}
|
||
|
|
||
|
// RequestSocket bundles up the parameters into a Request and then forwards to
|
||
|
// ClientSocketPoolBaseHelper::RequestSocket().
|
||
|
int RequestSocket(const std::string& group_name,
|
||
|
const scoped_refptr<SocketParams>& params,
|
||
|
RequestPriority priority,
|
||
|
ClientSocketPool::RespectLimits respect_limits,
|
||
|
ClientSocketHandle* handle,
|
||
|
const CompletionCallback& callback,
|
||
|
const NetLogWithSource& net_log) {
|
||
|
std::unique_ptr<Request> request(new Request(
|
||
|
handle, callback, priority, respect_limits,
|
||
|
internal::ClientSocketPoolBaseHelper::NORMAL, params, net_log));
|
||
|
return helper_.RequestSocket(group_name, std::move(request));
|
||
|
}
|
||
|
|
||
|
// RequestSockets bundles up the parameters into a Request and then forwards
|
||
|
// to ClientSocketPoolBaseHelper::RequestSockets(). Note that it assigns the
|
||
|
// priority to IDLE and specifies the NO_IDLE_SOCKETS flag.
|
||
|
void RequestSockets(const std::string& group_name,
|
||
|
const scoped_refptr<SocketParams>& params,
|
||
|
int num_sockets,
|
||
|
const NetLogWithSource& net_log,
|
||
|
HttpRequestInfo::RequestMotivation motivation) {
|
||
|
const Request request(nullptr /* no handle */, CompletionCallback(), IDLE,
|
||
|
ClientSocketPool::RespectLimits::ENABLED,
|
||
|
internal::ClientSocketPoolBaseHelper::NO_IDLE_SOCKETS,
|
||
|
params, net_log);
|
||
|
helper_.RequestSockets(group_name, request, num_sockets, motivation);
|
||
|
}
|
||
|
|
||
|
void SetPriority(const std::string& group_name,
|
||
|
ClientSocketHandle* handle,
|
||
|
RequestPriority priority) {
|
||
|
return helper_.SetPriority(group_name, handle, priority);
|
||
|
}
|
||
|
|
||
|
void CancelRequest(const std::string& group_name,
|
||
|
ClientSocketHandle* handle) {
|
||
|
return helper_.CancelRequest(group_name, handle);
|
||
|
}
|
||
|
|
||
|
void ReleaseSocket(const std::string& group_name,
|
||
|
std::unique_ptr<StreamSocket> socket,
|
||
|
int id) {
|
||
|
return helper_.ReleaseSocket(group_name, std::move(socket), id);
|
||
|
}
|
||
|
|
||
|
void FlushWithError(int error) { helper_.FlushWithError(error); }
|
||
|
|
||
|
bool IsStalled() const { return helper_.IsStalled(); }
|
||
|
|
||
|
void CloseIdleSockets() { return helper_.CloseIdleSockets(); }
|
||
|
|
||
|
void CloseIdleSocketsInGroup(const std::string& group_name) {
|
||
|
return helper_.CloseIdleSocketsInGroup(group_name);
|
||
|
}
|
||
|
|
||
|
int idle_socket_count() const { return helper_.idle_socket_count(); }
|
||
|
|
||
|
int IdleSocketCountInGroup(const std::string& group_name) const {
|
||
|
return helper_.IdleSocketCountInGroup(group_name);
|
||
|
}
|
||
|
|
||
|
LoadState GetLoadState(const std::string& group_name,
|
||
|
const ClientSocketHandle* handle) const {
|
||
|
return helper_.GetLoadState(group_name, handle);
|
||
|
}
|
||
|
|
||
|
void DumpMemoryStats(base::trace_event::ProcessMemoryDump* pmd,
|
||
|
const std::string& parent_dump_absolute_name) const {
|
||
|
return helper_.DumpMemoryStats(pmd, parent_dump_absolute_name);
|
||
|
}
|
||
|
|
||
|
virtual void OnConnectJobComplete(int result, ConnectJob* job) {
|
||
|
return helper_.OnConnectJobComplete(result, job);
|
||
|
}
|
||
|
|
||
|
int NumUnassignedConnectJobsInGroup(const std::string& group_name) const {
|
||
|
return helper_.NumUnassignedConnectJobsInGroup(group_name);
|
||
|
}
|
||
|
|
||
|
int NumConnectJobsInGroup(const std::string& group_name) const {
|
||
|
return helper_.NumConnectJobsInGroup(group_name);
|
||
|
}
|
||
|
|
||
|
int NumActiveSocketsInGroup(const std::string& group_name) const {
|
||
|
return helper_.NumActiveSocketsInGroup(group_name);
|
||
|
}
|
||
|
|
||
|
bool HasGroup(const std::string& group_name) const {
|
||
|
return helper_.HasGroup(group_name);
|
||
|
}
|
||
|
|
||
|
void CleanupIdleSockets(bool force) {
|
||
|
return helper_.CleanupIdleSockets(force);
|
||
|
}
|
||
|
|
||
|
std::unique_ptr<base::DictionaryValue> GetInfoAsValue(
|
||
|
const std::string& name,
|
||
|
const std::string& type) const {
|
||
|
return helper_.GetInfoAsValue(name, type);
|
||
|
}
|
||
|
|
||
|
base::TimeDelta ConnectionTimeout() const {
|
||
|
return helper_.ConnectionTimeout();
|
||
|
}
|
||
|
|
||
|
void EnableConnectBackupJobs() { helper_.EnableConnectBackupJobs(); }
|
||
|
|
||
|
bool CloseOneIdleSocket() { return helper_.CloseOneIdleSocket(); }
|
||
|
|
||
|
bool CloseOneIdleConnectionInHigherLayeredPool() {
|
||
|
return helper_.CloseOneIdleConnectionInHigherLayeredPool();
|
||
|
}
|
||
|
|
||
|
private:
|
||
|
// This adaptor class exists to bridge the
|
||
|
// internal::ClientSocketPoolBaseHelper::ConnectJobFactory and
|
||
|
// ClientSocketPoolBase::ConnectJobFactory types, allowing clients to use the
|
||
|
// typesafe ClientSocketPoolBase::ConnectJobFactory, rather than having to
|
||
|
// static_cast themselves.
|
||
|
class ConnectJobFactoryAdaptor
|
||
|
: public internal::ClientSocketPoolBaseHelper::ConnectJobFactory {
|
||
|
public:
|
||
|
typedef typename ClientSocketPoolBase<SocketParams>::ConnectJobFactory
|
||
|
ConnectJobFactory;
|
||
|
|
||
|
explicit ConnectJobFactoryAdaptor(ConnectJobFactory* connect_job_factory)
|
||
|
: connect_job_factory_(connect_job_factory) {}
|
||
|
~ConnectJobFactoryAdaptor() override {}
|
||
|
|
||
|
std::unique_ptr<ConnectJob> NewConnectJob(
|
||
|
const std::string& group_name,
|
||
|
const internal::ClientSocketPoolBaseHelper::Request& request,
|
||
|
ConnectJob::Delegate* delegate) const override {
|
||
|
const Request& casted_request = static_cast<const Request&>(request);
|
||
|
return connect_job_factory_->NewConnectJob(
|
||
|
group_name, casted_request, delegate);
|
||
|
}
|
||
|
|
||
|
base::TimeDelta ConnectionTimeout() const override {
|
||
|
return connect_job_factory_->ConnectionTimeout();
|
||
|
}
|
||
|
|
||
|
const std::unique_ptr<ConnectJobFactory> connect_job_factory_;
|
||
|
};
|
||
|
|
||
|
internal::ClientSocketPoolBaseHelper helper_;
|
||
|
|
||
|
DISALLOW_COPY_AND_ASSIGN(ClientSocketPoolBase);
|
||
|
};
|
||
|
|
||
|
} // namespace net
|
||
|
|
||
|
#endif // NET_SOCKET_CLIENT_SOCKET_POOL_BASE_H_
|