Multithreaded Application Tutorial/zh CN: what changed

Revision 1 to 2. Removed lines have a minus sign; added lines have a plus sign.

Old	New		Line
1	1		## 概述
2	2
3	3		这个页面将试图解释如何在Free Pascal 和 Lazarus上编写和调试一个多线程应用程序。一个多线程应用程序创建两个或多个线程同时执行任务。如果你正需要使用多线程，请阅读"你需要多线程吗?"，以确定你是否真的需要；这可以避免些麻烦。
4	4
5	5		程序中的一个线程被称为主线程，也就是在应用程序启动时由操作系统创建的那个线程。主线程是唯一一个可以更新用户界面组件的线程，否则，应用程序可能会挂起。
6	6
7	7		其主要思想是：应用程序使用第二个线程在后台处其他事情，同时用户可以继续使用主线程工作。
8	8
9	9		线程的另一个用途是为了更好的响应应用程序。如，你创建了一个应用程序，在应用程序启动时，用户按下按钮启动一个大的任务……处理中……之后，程序停止响应，这让用户觉得应用程序假死了，在用户体验上这很不好。如果大的任务在另一个线程中运行，应用程序保持响应，它就像在闲置，在本例中它是个不错的选择。不过，在启用线程之前，你需要禁用表单按钮，以避用户启动多个线程。
10	10
11	11		还有，服务器应用程序在使用多线程后，它能够同时响应多个客户端的请求。
12	12
13	13		## 你需要多线程吗?
14	14
15	15		如果你刚想使用多线程，而只是想让你的应用程序长时间运行任务而不失去响应，那么，多线程可能超过你的需要。多线程应用程序难以调试并且很复杂，多数情况下你不需要多线程。单线程就足够了。如果你能将耗时的任务分隔成几块，并使用Application.ProcessMessages。这种方法让LCL处理所有等待的消息并返回。其核心是，在长时间运行任务时定期调用Application.ProcessMessages以处理用户点击了什么东西，或者重绘进度条等等。
16	16
17	17		例如：读取一个大文件并处理它。参看 examples/multithreading/singlethreadingexample1.lpi。
18	18
19	19		使用多线程
20	20
21	21		- blocking handles, like network communications(~~屏蔽处理,比如网络通信~~)
22	22		- 同时使用多个处理器 (SMP)
23	23		- 算法和库调用，必须通过一个 API 调用，如不能被分解成更小的部分。
24	24
25	25		If you want to use multi-threading to increase speed by using multiple processors simultaneously, check if your current program now uses all 100% resources of 1 core CPU (for example, your program can actively use input-output operations, e.g. writing to file; this takes a lot of time, but doesn't load CPU; in this case your program will not be faster with multiple threads). Also check if optimisation level is set to maximum (3). When switching optimisation level from 1 to 3, a program may become about 5 times faster.
26	26
27	27		~~你想使用多线程使用多个处理器来提高速度，检查当前程序使用1个核心的CPU上的100%资源（如，你的程序可以积极使用输入输出操作，如写入文件等等，这需要大量时间，但CPU没有负载；在这种情况下，你的程序将不会比多种线程快。）检查优化级别设置为最大(3)。切换优化级别1到3，程序运行可能会快5倍。~~
28	28
29	29		## 多线程单元
30	30
31	31		在Windows上你不需要任何特殊单元。
32	32
33	33		然而在Linux、Mac OS X和FreeBSD上,你需要cthreads单元，并且，必须项目中第一个引入的单元（项目源文件，通常是 .lpr文件）！
34	34
35	35		所以，你的Lazarus应用程序代码看起来应该是：
36	36
37	37		``` pascal
38	38		program MyMultiThreadedProgram;
39	39		{$mode objfpc}{$H+}
40	40		uses
41	41		{$ifdef unix}
42	42		cthreads,
43	43		cmem, // the c memory manager is on some systems much faster for multi-threading(C内存管理器在某些系统中使多线程运行更快)
44	44		{$endif}
45	45		Interfaces, // this includes the LCL widgetset 引入LCL 部件工具箱
46	46		Forms
47	47		{ you can add units here },
48	48		```
49	49
50	50		如果你忘记了这一点，使用TThread启动应用程序时，你会得到这样的错误：
51	51
52	52		` This binary has no thread support compiled in. (`~~`二进制文件编译时没有线程支持`~~`)`\
53	53		` Recompile the application with a thread-driver in the program uses clause before other units using thread.(`~~`在引用线程单元前引用了其他单元`~~`)`
54	54
55	55		> Note: 提示"mcount"未找到的错误。在你的单元里包含了多线程代码，你需要添加 cthreads单元或使用智能链接。
56	56
57	57		> Note: "Project raised exception class 'RunError(232)'" in procedure SYSTEM_NOTHREADERROR (项目引发的异常类)错误，你的代码里需要线程，你需要添加 cthreads单元。
58	58
59	59		## TThread类
60	60
61	61		下面的例子可以在 examples/multithreading/ 目录中找到。
62	62
63	63		若要创建一个多线程的应用程序，最简单的方法是使用 TThread 类。This class permits the creation of an additional thread (alongside the main thread) in a simple way.(~~这个线程允许在主线程上创建一个附加线程)。~~)通常你只需要重写 2 个方法：创建构造函数（constructor）和执行方法（Execute）。
64	64
65	65		在构造函数中，设置线程运行前初始化需要的变量，在TThread原始构造函数中有一参数Suspended，设置Suspended = True防止在创建后自动启动线程。设置为False，则在在调用Start方法后才执行。
66	66
67	67		> Note: 方法摘要在：[自FPC 2.4.4 弃用](http://wiki.freepascal.org/User_Changes_2.4.4#TThread.Suspend_and_TThread.Resume_have_been_deprecated)。它被Start代替。
68	68
69	69		FPC 2.0.1及更高版本，TThread.Create有一个隐含的堆栈大小参数。根据需要，你可以改变默认堆栈的大小。深层递归过程是个很好的例子。如果你不指定堆栈大小，将使用系统默认的大小。
70	70
71	71		重写Execute方法，编写将在线程中执行的代码。
72	72
73	73		TThread 类有一个重要属性：
74	74
75	75		`Terminated : boolean;`
76	76
77	77		If the thread has a loop (and this is typical), the loop should be exited when Terminated is true (it is false by default). Within each pass, the value of Terminated must be checked, and if it is true then the loop should be exited as quickly as is appropriate, after any necessary cleanup. Bear in mind that the Terminate method does not do anything by default: the .Execute method must explicitly implement support for it to quit its job.
78	78
79	79		(~~如果线程中包含循环（这是典型的），当 Terminated 为True时退出循环（默认为 False）。每次循环都将检查 Terminated值，如果为真将退出循环，之后做些必要的清理。牢记 Terminate方法不做任何事情，默认情况下：.Execute 方法实现退出循环后的操作。~~)
80	80
81	81		如上所述，线程不应该与可见组件交互，对可见组件操作应该在主线程的上下文中进行。
82	82
83	83		To do this, a TThread method called Synchronize exists. Synchronize requires a method within the thread (that takes no parameters) as an argument. When you call that method through Synchronize(@MyMethod), the thread execution will be paused, the code of MyMethod will be called from the main thread, and then the thread execution will be resumed.
84	84
85	85		(~~要这样做，可以在TThread 方法调用Synchronize。以同步线程中的方法（不带参数）作为参数。当你调用Synchronize(@MyMethod)，线程将暂停执行，并从主线程中调用MyMethod，之后线程恢复。~~)
86	86
87	87		The exact working of Synchronize depends on the platform, but basically it does this: (~~同步的具体工作取决于平台，它基本上是：~~)
88	88
89	89		- 发送消息到主消息队列并进入睡眠
90	90		- eventually the main thread processes the message and calls MyMethod. This way MyMethod is called without context, that means not during a mouse down event or during paint event, but after.
91	91		- 主线程执行 MyMethod后，唤醒睡眠的线程和下条消息
92	92		- 之后，线程继续执行
93	93
94	94		另外，还有一重要属性： TThread: FreeOnTerminate。为True时，线程停止(.Execute 方法)后自动释放。否则，需要手动释放。
95	95
96	96		示例：
97	97
98	98		``` pascal
99	99		Type
100	100		TMyThread = class(TThread)
101	101		private
102	102		fStatusText : string;
103	103		procedure ShowStatus;
104	104		protected
105	105		procedure Execute; override;
106	106		public
107	107		Constructor Create(CreateSuspended : boolean);
108	108		end;
109	109
110	110		constructor TMyThread.Create(CreateSuspended : boolean);
111	111		begin
112	112		FreeOnTerminate := True;
113	113		inherited Create(CreateSuspended);
114	114		end;
115	115
116	116		procedure TMyThread.ShowStatus;
117	117		// this method is executed by the mainthread and can therefore access all GUI elements.
118	118		begin
119	119		Form1.Caption := fStatusText;
120	120		end;
121	121
122	122		procedure TMyThread.Execute;
123	123		var
124	124		newStatus : string;
125	125		begin
126	126		fStatusText := 'TMyThread Starting...';
127	127		Synchronize(@Showstatus);
128	128		fStatusText := 'TMyThread Running...';
129	129		while (not Terminated) and ([any condition required]) do
130	130		begin
131	131		...
132	132		[here goes the code of the main thread loop]
133	133		...
134	134		if NewStatus <> fStatusText then
135	135		begin
136	136		fStatusText := newStatus;
137	137		Synchronize(@Showstatus);
138	138		end;
139	139		end;
140	140		end;
141	141		```
142	142
143	143		在应用程序中，
144	144
145	145		``` pascal
146	146		var
147	147		MyThread : TMyThread;
148	148		begin
149	149		MyThread := TMyThread.Create(True); // This way it doesn't start automatically
150	150		...
151	151		[Here the code initialises anything required before the threads starts executing]
152	152		...
153	153		MyThread.Start;
154	154		end;
155	155		```
156	156
157	157		If you want to make your application more flexible you can create an event for the thread; this way your synchronized method won't be tightly coupled with a specific form or class: you can attach listeners to the thread's event. Here is an example:
158	158
159	159		(~~如果想使你的应用更加灵活，可以创建一个事件的线程；这样一来,synchronized方法将不会被特定窗体或类紧耦合：你可以监听线程的事件。~~)
160	160
161	161		下面是一个例子：
162	162
163	163		``` pascal
164	164		Type
165	165		TShowStatusEvent = procedure(Status: String) of Object;
166	166
167	167		TMyThread = class(TThread)
168	168		private
169	169		fStatusText : string;
170	170		FOnShowStatus: TShowStatusEvent;
171	171		procedure ShowStatus;
172	172		protected
173	173		procedure Execute; override;
174	174		public
175	175		Constructor Create(CreateSuspended : boolean);
176	176		property OnShowStatus: TShowStatusEvent read FOnShowStatus write FOnShowStatus;
177	177		end;
178	178
179	179		constructor TMyThread.Create(CreateSuspended : boolean);
180	180		begin
181	181		FreeOnTerminate := True;
182	182		inherited Create(CreateSuspended);
183	183		end;
184	184
185	185		procedure TMyThread.ShowStatus;
186	186		// this method is executed by the mainthread and can therefore access all GUI elements.
187	187		begin
188	188		if Assigned(FOnShowStatus) then
189	189		begin
190	190		FOnShowStatus(fStatusText);
191	191		end;
192	192		end;
193	193
194	194		procedure TMyThread.Execute;
195	195		var
196	196		newStatus : string;
197	197		begin
198	198		fStatusText := 'TMyThread Starting...';
199	199		Synchronize(@Showstatus);
200	200		fStatusText := 'TMyThread Running...';
201	201		while (not Terminated) and ([any condition required]) do
202	202		begin
203	203		...
204	204		[here goes the code of the main thread loop]
205	205		...
206	206		if NewStatus <> fStatusText then
207	207		begin
208	208		fStatusText := newStatus;
209	209		Synchronize(@Showstatus);
210	210		end;
211	211		end;
212	212		end;
213	213		```
214	214
215	215		在应用程序中，
216	216
217	217		``` pascal
218	218		Type
219	219		TForm1 = class(TForm)
220	220		Button1: TButton;
221	221		Label1: TLabel;
222	222		procedure FormCreate(Sender: TObject);
223	223		procedure FormDestroy(Sender: TObject);
224	224		private
225	225		{ private declarations }
226	226		MyThread: TMyThread;
227	227		procedure ShowStatus(Status: string);
228	228		public
229	229		{ public declarations }
230	230		end;
231	231
232	232		procedure TForm1.FormCreate(Sender: TObject);
233	233		begin
234	234		inherited;
235	235		MyThread := TMyThread.Create(true);
236	236		MyThread.OnShowStatus := @ShowStatus;
237	237		end;
238	238
239	239		procedure TForm1.FormDestroy(Sender: TObject);
240	240		begin
241	241		MyThread.Terminate;
242	242
243	243		// FreeOnTerminate is true so we should not write:
244	244		// MyThread.Free;
245	245		inherited;
246	246		end;
247	247
248	248		procedure TForm1.Button1Click(Sender: TObject);
249	249		begin
250	250		MyThread.Start;
251	251		end;
252	252
253	253		procedure TForm1.ShowStatus(Status: string);
254	254		begin
255	255		Label1.Caption := Status;
256	256		end;
257	257		```
258	258
259	259		## 特殊处理
260	260
261	261		### Windows堆栈检查
262	262
263	263		There is a potential headache in Windows with Threads if you use the -Ct (stack check) switch. For reasons not so clear the stack check will "trigger" on any TThread.Create if you use the default stack size. The only work-around for the moment is to simply not use -Ct switch. Note that it does NOT cause an exception in the main thread, but in the newly created one. This "looks" like if the thread was never started.
264	264
265	265		A good code to check for this and other exceptions which can occur in thread creation is:
266	266
267		-	<strike> 如果在Windows使用-CT（堆栈检查）会引发一些问题。如果你使用默认的堆栈大小在，堆栈检查将在TThread.Create时"触发"，原因不是很清楚。唯一的解决方法是，不使用-CT。请注意，它不会导致主线程的异常，但新创建的却不一定。如果线程从未启动过，它"看起来"会。
	267	+	如果在Windows使用-CT（堆栈检查）会引发一些问题。如果你使用默认的堆栈大小在，堆栈检查将在TThread.Create时"触发"，原因不是很清楚。唯一的解决方法是，不使用-CT。请注意，它不会导致主线程的异常，但新创建的却不一定。如果线程从未启动过，它"看起来"会。
268	268
269		-	~~下面的代码来检查它，用于在线程创建时抛出异常：</strike>~~
	269	+	下面的代码来检查它，用于在线程创建时抛出异常：
270	270
271	271		``` pascal
272	272		MyThread := TThread.Create(False);
273	273		if Assigned(MyThread.FatalException) then
274	274		raise MyThread.FatalException;
275	275		```
276	276
277	277		This code will assure that any exception which occurred during thread creation will be raised in your main thread.
278	278
279	279		~~上面的代码，将在某个线程创建时抛出任何异常在主线程中。~~
280	280
281	281		### 多线程包
282	282
283	283		Packages which uses multi-threading should add the -dUseCThreads flag to the custom usage options. Open the package editor of the package, then Options \> Usage \> Custom and add -dUseCThreads. This will define this flag to all projects and packages using this package, including the IDE. The IDE and all new applications created by the IDE have already the following code in their .lpr file:
284	284
285	285		``` pascal
286	286		uses
287	287		{$IFDEF UNIX}{$IFDEF UseCThreads}
288	288		cthreads,
289	289		cmem, // the c memory manager is on some systems much faster for multi-threading
290	290		{$ENDIF}{$ENDIF}
291	291		```
292	292
293	293		### Heaptrc
294	294
295	295		You can not use the -gh switch with the cmem unit. The -gh switch uses the heaptrc unit, which extends the heap manager. Therefore the heaptrc unit must be used after the cmem unit.
296	296
297	297		~~你不能使用-gh选项与 cmem 单元。-gh使用了heaptrc单元，它扩展了堆管理器。因此heaptrc 必须在cmem单元后引用。~~
298	298
299	299		``` pascal
300	300		uses
301	301		{$IFDEF UNIX}{$IFDEF UseCThreads}
302	302		cthreads,
303	303		cmem, // the c memory manager is on some systems much faster for multi-threading
304	304		{$ENDIF}{$ENDIF}
305	305		heaptrc,
306	306		```
307	307
308	308		## SMP支持
309	309
310	310		好消息是，如果你的应用程序在多线程上工作正常，看来SMP已经启用。
311	311
312	312		## Lazarus调试多线程应用程序
313	313
314	314		Lzarus上使用[GDB](http://zh.wikipedia.org/zh-cn/GDB)进行调试，它功能齐全、稳定。不过，在一些Linux发行版上存在一些问题。
315	315
316	316		### 调试输出
317	317
318	318		In a single threaded application, you can simply write to console/terminal/whatever and the order of the lines is the same as they were written. In multi-threaded application things are more complicated. If two threads are writing, say a line is written by thread A before a line by thread B, then the lines are not necessarily written in that order. It can even happen, that a thread writes its output, while the other thread is writing a line.While under linux (maybe) you'll get proper DebugLn() output, under win32 you can get exceptions (probably DiskFull) because of DebugLn() usage outside of main thread.So, to avoid headaches use DebugLnThreadLog() mentioned below.
319	319
320		-	~~（<strike>在一个单线程的应用程序中，你可以很容易的写入（输出信息）到控制台/终端/其他，因为写入顺序是一样的。~~
	320	+	（在一个单线程的应用程序中，你可以很容易的写入（输出信息）到控制台/终端/其他，因为写入顺序是一样的。
321	321
322	322		在多线程应用程序中这样做会变得很复杂。如果两个线程在写，线程A在线程B之前输出一行，似乎输出结果并不会按写入顺序显示。甚至会发生，一个线程写入一个线程输出，其他的线程在写。在Linux下使用DebugLn()你会得到正确的输出结果。在Win32下可能得到异常（或许是DiskFull），它是因为，在主线程外使用DebugLn()。
323	323
324		-	~~所以，为了避免这类问题，使用DebugLnThreadLog()，下面将会提到：</strike>）~~
	324	+	所以，为了避免这类问题，使用DebugLnThreadLog()，下面将会提到：）
325	325
326	326		LCLProc 单元包含几个过程，让每个线程写入它自己的日志文件：
327	327
328	328		``` pascal
329	329		procedure DbgOutThreadLog(const Msg: string); overload;
330	330		procedure DebuglnThreadLog(const Msg: string); overload;
331	331		procedure DebuglnThreadLog(Args: array of const); overload;
332	332		procedure DebuglnThreadLog; overload;
333	333		```
334	334
335	335		示例：不是writeln('Some text ',123); 而是，使用
336	336
337	337		` DebuglnThreadLog(['Some text ',123]);`
338	338
339	339		这将在Log\<PID\>.txt文件中添加一行'Some text 123'。\<PID\>是当前线程的进程ID。
340	340
341	341		在每次运行前删除日志文件是个不错的选择：
342	342
343	343		` rm -f Log* && ./project1`
344	344
345	345		### Linux
346	346
347	347		如果你尝试在Linux调试多线程应用程序，你将会遇到一个很大的问题：
348	348
349	349		X服务器上的桌面管理器可能会挂掉。如应用程序已捕获鼠标/键盘gdb暂停，X服务器等待应用程序响应。
350	350
351	351		发生这种情况时，你可以使用另一个用户登录计算机杀掉gdb或按Ctrl+Alt+F3退出会话并杀掉gdb。
352	352
353	353		或者，你重新启动窗口管理器：输入 /etc/init.d/gdm restart。
354	354
355	355		这将重新启动桌面管理器，让你回到桌面。
356	356
357	357		Since it depends where gdb stops your program in some cases some tricks may help: 为Ubuntu x64设置项目选项用于调试所需的额外信息文件...
358	358
359	359		` 工程菜单 -> 工程选项 -> 编译选项中的调试项，勾选``使用外部gdb调试语法文件(-Xg)``。`
360	360
361	361		另一种选择是，打开另一个X桌面，在里面运行 IDE/GDB和其他应用程序，因此，只会在测试桌面中冻结。创建一个新的X实例：
362	362
363	363		` X :1 &`
364	364
365	365		It will open, and when you switch to another desktop (the one you are working with pressing CTRL+ALT+F7), you will be able to go back to the new graphical desktop with CTRL+ALT+F8 (if this combination does not work, try with CTRL+ALT+F2... this one worked on [Slackware](http://www.slackware.com)).
366	366
367	367		（~~这将打开，按Ctrl+Alt+F7切换到另一个桌面，Ctrl+Alt+F8切换到原桌面（桌面快捷键没效果，尝试按Ctrl+Alt+F2，它工作在[Slackware](http://www.slackware.com)）。~~）
368	368
369	369		Then you could, if you want, create a desktop session on the X started with:
370	370
371	371		（~~之后，如果你愿意，你可以在X开始时创建会话：~~）
372	372
373	373		` gnome-session --display=:1 &`
374	374
375	375		Then, in Lazarus, on the run parameters dialog for the project, check "Use display" and enter :1.
376	376
377	377		（~~之后在Lazarus运行参数对话框中检查Use display并设置为:1。~~）
378	378
379	379		现在应用程序运行在第二个X服务器上，你可以在第一个X服务器上调试它。。
380	380
381	381		在Windows及Linux上的Free Pascal 2.0 和Lazarus 0.9.10 测试。
382	382
383	383		------------------------------------------------------------------------
384	384
385	385		Instead of creating a new X session, one can use [Xnest](http://en.wikipedia.org/wiki/Xnest). Xnest is a X session on a window. Using it X server didn't lock while debugging threads, and it's much easier to debug without keeping changing terminals.
386	386
387	387		（~~而不是创建一个新的X会话，可以使用[Xnest](http://en.wikipedia.org/wiki/Xnest)。Xnest是X会话窗口，它使用X服务器不锁定调试线程，及更容易调试并改变终端。~~）
388	388
389	389		运行 Xnest 命令
390	390
391	391		` Xnest :1 -ac`
392	392
393	393		创建一个X会话:1，并禁用访问控制。
394	394
395	395		## Lazarus 部件工具箱接口
396	396
397	397		The win32, the gtk and the carbon interfaces support multi-threading. This means, TThread, critical sections and Synchronize work. But they are not thread safe. This means only one thread at a time can access the LCL. And since the main thread should never wait for another thread, it means only the main thread is allowed to access the LCL, which means anything that has to do with TControl, Application and LCL widget handles. There are some thread safe functions in the LCL. For example most of the functions in the FileUtil unit are thread safe.
398	398
399	399		(~~在Win32 中，gtk 和carbon接口支持多线程。这意味着，TThread、关键区域和同步工作，但它们不是线程安全的。也就是说只有一个线程在同一时间可以访问LCL。由于主线程不应该等待另一个线程，这使得只有主线程访问LCL，这意味着任何TControl，应用程序和LCL控件句柄。LCL中有一些线程安全函数。例如大部分的 FileUtil 单元中的函数是[线程安全](http://zh.wikipedia.org/wiki/%E7%BA%BF%E7%A8%8B%E5%AE%89%E5%85%A8)的。~~)
400	400
401	401		### 使用SendMessage/PostMessage进行线程间通信
402	402
403	403		在应用程序中只有一个线程可以调用LCL API，通常是主线程。其他线程可以通过一些间接的方法调用LCL，使用SendMessage、PostMessage是不错的选择。LCLIntf.SendMessage 和 LCLIntf.PostMessage将发送一条消息到应用程序的消息队列中。
404	404
405	405		参看这些例程文档
406	406
407	407		- [[doc:lcl/lclintf/sendmessage.html\|SendMessage]]
408	408		- [[doc:lcl/lclintf/postmessage.html\|PostMessage]]
409	409
410	410		The difference between SendMessage and PostMessage is the way that they return control to the calling thread. With SendMessage control is not returned until the window that the message was sent to has completed processing the sent message, however with PostMessage control is returned immediately.
411	411
412	412		(~~SendMessage 和 PostMessage区别在于处理调用线程的方式。SendMessage 在接收到返回值后才继续执行，而PostMessage则不会等待。~~)
413	413
414	414		Here is an example of how a secondary thread could send text to be displayed in an LCL control to the main thread:
415	415
416	416		``` pascal
417	417		const
418	418		WM_GOT_ERROR = LM_USER + 2004;
419	419		WM_VERBOSE = LM_USER + 2005;
420	420
421	421		procedure VerboseLog(Msg: string);
422	422		var
423	423		PError: PChar;
424	424		begin
425	425		if MessageHandler = 0 then Exit;
426	426		PError := StrAlloc(Length(Msg)+1);
427	427		StrCopy(PError, PChar(Msg));
428	428		PostMessage(formConsole.Handle, WM_VERBOSE, Integer(PError), 0);
429	429		end;
430	430		```
431	431
432	432		及如何从窗口中处理这个消息：
433	433
434	434		``` pascal
435	435		const
436	436		WM_GOT_ERROR = LM_USER + 2004;
437	437		WM_VERBOSE = LM_USER + 2005;
438	438
439	439		type
440	440		{ TformConsole }
441	441
442	442		TformConsole = class(TForm)
443	443		DebugList: TListView;
444	444		// ...
445	445		private
446	446		procedure HandleDebug(var Msg: TLMessage); message WM_VERBOSE;
447	447		end;
448	448
449	449		var
450	450		formConsole: TformConsole;
451	451
452	452		implementation
453	453
454	454		....
455	455
456	456		{ TformConsole }
457	457
458	458		procedure TformConsole.HandleDebug(var Msg: TLMessage);
459	459		var
460	460		Item: TListItem;
461	461		MsgStr: PChar;
462	462		MsgPasStr: string;
463	463		begin
464	464		MsgStr := PChar(Msg.wparam);
465	465		MsgPasStr := StrPas(MsgStr);
466	466		Item := DebugList.Items.Add;
467	467		Item.Caption := TimeToStr(SysUtils.Now);
468	468		Item.SubItems.Add(MsgPasStr);
469	469		Item.MakeVisible(False);
470	470		//f/TrayControl.SetError(MsgPasStr);
471	471		end;
472	472
473	473		end.
474	474		```
475	475
476	476		## 临界区
477	477
478	478		A critical section is an object used to make sure, that some part of the code is executed only by one thread at a time. A critical section needs to be created/initialized before it can be used and be freed when it is not needed anymore.
479	479
480	480		(~~[临界区](http://zh.wikipedia.org/wiki/%E4%B8%B4%E7%95%8C%E5%8C%BA)用来确保代码的某些部分同一时间只能由一个线程执行。临界区需要被创建/初始化后才能使用，不需要时释放它。~~)
481	481
482	482		临界区通常使用这种方式：
483	483
484	484		声明（全局范围内所有线程可以访问）：
485	485
486	486		` MyCriticalSection: TRTLCriticalSection;`
487	487
488	488		初始化临界区：
489	489
490	490		` InitializeCriticalSection(MyCriticalSection);`
491	491
492	492		运行一些线程，只做
493	493
494	494		``` pascal
495	495		EnterCriticalSection(MyCriticalSection);
496	496		try
497	497		// 访问变量、写入文件、发送网络数据包等等
498	498		finally
499	499		LeaveCriticalSection(MyCriticalSection);
500	500		end;
501	501		```
502	502
503	503		所有线程结束后，释放它：
504	504
505	505		` DeleteCriticalSection(MyCriticalSection);`
506	506
507	507		As an alternative, you can use a TCriticalSection object. The creation does the initialization, the Enter method does the EnterCriticalSection, the Leave method does the LeaveCriticalSection and the destruction of the object does the deletion.
508	508
509	509		(~~作为代替者，你可以使用TCriticalSection对象。创建即初始化临界区，进入临界区使用 EnterCriticalSection，离开时使用 LeaveCriticalSection，对象销毁时删除临界区。~~)
510	510
511	511		例如：5个线程递增计数器。查看 lazarus/examples/multithreading/criticalsectionexample1.lpi
512	512
513	513		注意： There are two sets of the above 4 functions. The RTL and the LCL ones. The LCL ones are defined in the unit LCLIntf and LCLType. Both work pretty much the same. You can use both at the same time in your application, but you should not use a RTL function with an LCL Critical Section and vice versus.
514	514
515	515		### 共享变量
516	516
517	517		如果一些线程共享变量，它是只读的，那么就没有什么好担心的，直接读取就可以了。
518	518
519	519		但是如果一个或多个线程想改变变量，那么你必须确保，同一时间只有一个线程访问该变量。
520	520
521	521		示例：5线程递增计数器。查看 lazarus/examples/multithreading/criticalsectionexample1.lpi
522	522
523	523		## 等待另一个线程
524	524
525	525		如果一个线程A需要线程B的结果，那么它必须等待，直到B完成。
526	526
527	527		重要：主线程不应该等待另一个线程，不是使用Synchronize（见上文）。
528	528
529	529		查看示例： lazarus/examples/multithreading/waitforexample1.lpi
530	530
531	531		``` pascal
532	532		{ TThreadA }
533	533
534	534		procedure TThreadA.Execute;
535	535		begin
536	536		Form1.ThreadB:=TThreadB.Create(false);
537	537		// create event
538	538		WaitForB:=RTLEventCreate;
539	539		while not Application.Terminated do begin
540	540		// wait infinitely (until B wakes A)
541	541		RtlEventWaitFor(WaitForB);
542	542		writeln('A: ThreadB.Counter='+IntToStr(Form1.ThreadB.Counter));
543	543		end;
544	544		end;
545	545
546	546		{ TThreadB }
547	547
548	548		procedure TThreadB.Execute;
549	549		var
550	550		i: Integer;
551	551		begin
552	552		Counter:=0;
553	553		while not Application.Terminated do begin
554	554		// B: Working ...
555	555		Sleep(1500);
556	556		inc(Counter);
557	557		// wake A
558	558		RtlEventSetEvent(Form1.ThreadA.WaitForB);
559	559		end;
560	560		end;
561	561		```
562	562
563	563		> Note: RtlEventSetEvent can be called before RtlEventWaitFor. Then RtlEventWaitFor will return immediately. Use RTLeventResetEvent to clear a flag.
564	564
565	565		> Note: ~~RtlEventSetEvent可以在RtlEventWaitFor之前被调用。之后RtlEventWaitFor将立即返回。使用RTLeventResetEvent清除标记。~~
566	566
567	567		## Fork
568	568
569	569		When forking in a multi-threaded application, be aware that any threads created and running BEFORE the fork (or fpFork) call, will NOT be running in the child process. As stated on the fork() man page, any threads that were running before the fork call, their state will be undefined.
570	570
571	571		So be aware of any threads initializing before the call (including on the initialization section). They will NOT work.
572	572
573	573		## 并行程序/循环
574	574
575	575		多线程的一个特殊情况是并行运行一个程序。参见[[Parallel procedures\|并行程序]]。
576	576
577	577		## 分布式计算
578	578
579	579		The next higher steps after multi threading is running the threads on multiple machines.
580	580
581	581		- 可以使用TCP组件像synapse、lnet 或 indy通信。这给你最大的灵活性和主要用于松散连接的客户端/服务器应用程序。
582	582		- 你可以使用消息传递库，像[[MPICH]]，用于HPC（高性能计算）集群上。
583	583
584	584		## 外部线程
585	585
586	586		To make Free Pascal's threading system work properly, each newly created FPC thread needs to be initialized (more exactly, the exception, I/O system and threadvar system per thread needs to be initialized so threadvars and heap are working). That is fully automatically done for you if you use BeginThread (or indirectly by using the TThread class). However, if you use threads that were created without BeginThread (i.e. external threads), additional work (currently) might be required. External threads also include those that were created in external C libraries (.DLL/.so).
587	587
588	588		Things to consider when using external threads (might not be needed in all or future compiler versions):
589	589
590	590		- Do not use external threads at all - use FPC threads. If can you can get control over how the thread is created, create the thread by yourself by using BeginThread.
591	591
592	592		If the calling convention doesn't fit (e.g. if your original thread function needs cdecl calling convention but BeginThread needs pascal convention, create a record, store the original required thread function in it, and call that function in your pascal thread function:
593	593
594	594		``` pascal
595	595		type
596	596		TCdeclThreadFunc = function (user_data:Pointer):Pointer;cdecl;
597	597
598	598		PCdeclThreadFuncData = ^TCdeclThreadFuncData;
599	599		TCdeclThreadFuncData = record
600	600		Func: TCdeclThreadFunc; //cdecl function
601	601		Data: Pointer; //original data
602	602		end;
603	603
604	604		// The Pascal thread calls the cdecl function
605	605		function C2P_Translator(FuncData: pointer) : ptrint;
606	606		var
607	607		ThreadData: TCdeclThreadFuncData;
608	608		begin
609	609		ThreadData := PCdeclThreadFuncData(FuncData)^;
610	610		Result := ptrint(ThreadData.Func(ThreadData.Data));
611	611		end;
612	612
613	613		procedure CreatePascalThread;
614	614		var
615	615		ThreadData: PCdeclThreadFuncData;
616	616		begin
617	617		New(ThreadData);
618	618		// this is the desired cdecl thread function
619	619		ThreadData^.Func := func;
620	620		ThreadData^.Data := user_data;
621	621		// this creates the Pascal thread
622	622		BeginThread(@C2P_Translator, ThreadData );
623	623		end;
624	624		```
625	625
626	626		- Initialize the FPC's threading system by creating a dummy thread. If you don't create any Pascal thread in your app, the thread system won't be initialized (and thus threadvars won't work and thus heap will not work correctly).
627	627
628	628		``` pascal
629	629		type
630	630		tc = class(tthread)
631	631		procedure execute;override;
632	632		end;
633	633
634	634		procedure tc.execute;
635	635		begin
636	636		end;
637	637
638	638		{ main program }
639	639		begin
640	640		{ initialise threading system }
641	641		with tc.create(false) do
642	642		begin
643	643		waitfor;
644	644		free;
645	645		end;
646	646		{ ... your code follows }
647	647		end.
648	648		```
649	649
650	650		(After the threading system is initialized, the runtime may set the system variable "IsMultiThread" to true which is used by FPC routines to perform locks here and there. You should not set this variable manually.)
651	651
652	652		- If for some reason this doesn't work for you, try this code in your external thread function:
653	653
654	654		``` pascal
655	655		function ExternalThread(param: Pointer): LongInt; stdcall;
656	656		var
657	657		tm: TThreadManager;
658	658		begin
659	659		GetThreadManager(tm);
660	660		tm.AllocateThreadVars;
661	661		InitThread(1000000); // adjust inital stack size here
662	662
663	663		{ do something threaded here ... }
664	664
665	665		Result:=0;
666	666		end;
667	667		```
668	668
669	669		### 识别外部线程
670	670
671	671		有时你甚至不知道是否要处理外部线程（如对C 库进行回调）。这个可以帮助你来分析：
672	672
673	673		1\. 询问操作系统当前应用程序线程的ID
674	674
675	675		``` pascal
676	676		GetCurrentThreadID() // Windows;
677	677		GetThreadID() // Darwin/OSX;
678	678		TThreadID(pthread_self) // Linux;
679	679		```
680	680
681	681		2\. 再次询问线程内部的当前线程的ID和步骤1结果对比
682	682
683	683		### 放弃时间片
684	684
685	685		ThreadSwitch()
686	686
687	687		> Note: 不要使用Windows的 Sleep(0)，因为它不会在所有平台上工作。
688	688
689	689		## 参见
690	690
691	691		- [[Streaming components\|流媒体组件]]
692	692		- [[Manager Worker Threads System\|管理工作线程系统]]
693	693		- [[Example of multi-threaded application: array of threads\|多线程应用示例：数组线程]]
694	694		- [[Parallel procedures\|并行程序]]
695	695		- [[Main Loop Hooks\|主循环钩子]]
696	696		- [[Asynchronous Calls\|异步调用]]
697	697
698	698		---
699	699		From the Free Pascal wiki, page "Multithreaded Application Tutorial/zh CN", as it stood on 2020-02-20 after a change by Trev, out of the wiki's dump of 2024-05-26. Copyright stays with the author who wrote it, according to the old wiki's [[Licensing]] page. If you have concerns about this material, [message the admin](/wiki-contact).

Back to history