1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
|
/**
* Copyright (C) 2021-2024 Saturneric <eric@bktus.com>
*
* This file is part of GpgFrontend.
*
* GpgFrontend is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* GpgFrontend is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with GpgFrontend. If not, see <https://www.gnu.org/licenses/>.
*
* The initial version of the source code is inherited from
* the gpg4usb project, which is under GPL-3.0-or-later.
*
* All the source code of GpgFrontend was modified and released by
* Saturneric <eric@bktus.com> starting on May 12, 2021.
*
* SPDX-License-Identifier: GPL-3.0-or-later
*
*/
#pragma once
#include <QElapsedTimer>
#include <atomic>
#include <limits>
#include <memory>
// The test target defines this on the command line; the module build does
// not, so it is set here and guarded rather than assumed either way.
#ifndef VMIME_STATIC
#define VMIME_STATIC
#endif
#include <vmime/vmime.hpp>
/**
* @brief A stop request that can safely cross threads.
*
* Deliberately a bare atomic rather than a QObject. It is set from the GUI
* thread while a worker thread is blocked inside a socket call, and that is
* only safe because there is no Qt machinery involved -- no event loop to
* deliver to, no affinity to violate, nothing to destroy underneath the
* reader.
*/
class EMailCancelToken {
public:
/// Cancels the operation numbered @p seq: the one running now, or one that
/// has been queued to the worker and has not started yet.
///
/// Scoped by sequence number rather than a bare flag because both ends of
/// that range matter. A stop pressed while a request is still sitting in the
/// worker's event queue has to survive until the slot runs -- the previous
/// design cleared the flag at slot entry, so such a stop did nothing at all
/// and the operation ran to its full timeout. Equally, a stop belonging to a
/// finished operation must not cancel the next one, which a latch that was
/// never cleared would do.
void Cancel(quint64 seq) {
auto previous = cancelled_through_.load(std::memory_order_relaxed);
while (seq > previous &&
!cancelled_through_.compare_exchange_weak(
previous, seq, std::memory_order_relaxed)) {
}
}
/// Cancels the current operation and everything queued behind it, whatever
/// their numbers. For teardown, where no further work is wanted at all.
void CancelAll() {
cancelled_through_.store(std::numeric_limits<quint64>::max(),
std::memory_order_relaxed);
}
/// Announces the operation the worker is about to run.
///
/// Deliberately does NOT clear a cancel already issued for @p seq: that
/// stop was aimed at this very operation.
void Begin(quint64 seq) {
current_.store(seq, std::memory_order_relaxed);
}
[[nodiscard]] auto IsCancelled() const -> bool {
const auto current = current_.load(std::memory_order_relaxed);
return current != 0 &&
current <= cancelled_through_.load(std::memory_order_relaxed);
}
private:
/// The operation the worker is running. 0 means none has begun.
std::atomic<quint64> current_{0};
/// Every operation numbered at or below this is cancelled.
std::atomic<quint64> cancelled_through_{0};
};
using EMailCancelTokenPtr = std::shared_ptr<EMailCancelToken>;
/**
* @brief How a cancellation actually reaches a blocked vmime call.
*
* vmime has no cancel of its own, but it polls its timeout handler from inside
* every blocking read, write and connect loop -- and inside the TLS handshake.
* Answering "yes, time is up" is therefore the one way to make a blocked
* socket call return, which is why cancellation is expressed as a timeout.
*
* The cost of that trick is that a cancelled operation and a genuinely
* unresponsive server raise the same exception, so this records which it was.
* Reporting a user's own stop as "the server stopped responding" would be a
* lie, and the classifier consults @ref WasCancelled to avoid telling it.
*/
class EMailTimeoutHandler : public vmime::net::timeoutHandler {
public:
explicit EMailTimeoutHandler(EMailCancelTokenPtr token, int timeout_seconds);
auto isTimeOut() -> bool override;
void resetTimeOut() override;
auto handleTimeOut() -> bool override;
/// Whether the last firing was the user stopping rather than a timeout.
[[nodiscard]] auto WasCancelled() const -> bool { return cancelled_; }
/// Forgets a recorded stop once the operation it belonged to has ended, so
/// a later timeout is not reported as the user's own cancellation.
void ClearCancelled() { cancelled_ = false; }
private:
EMailCancelTokenPtr token_;
int timeout_seconds_;
/// Monotonic: a clock change mid-transfer must not look like a timeout.
QElapsedTimer elapsed_;
bool cancelled_{false};
};
/**
* @brief Supplies the handler above to every connection vmime opens.
*/
class EMailTimeoutHandlerFactory : public vmime::net::timeoutHandlerFactory {
public:
explicit EMailTimeoutHandlerFactory(EMailCancelTokenPtr token,
int timeout_seconds = 30);
auto create() -> vmime::shared_ptr<vmime::net::timeoutHandler> override;
/// The handler most recently created, so a caller unwinding an exception can
/// ask whether the stop was deliberate.
[[nodiscard]] auto LastWasCancelled() const -> bool;
/// Clears the recorded stop on the most recent handler. See
/// EMailTimeoutHandler::ClearCancelled.
void ClearLastCancelled();
private:
EMailCancelTokenPtr token_;
int timeout_seconds_;
vmime::shared_ptr<EMailTimeoutHandler> last_;
};
|
