1 #####################################################################
2 ########################### CORE ###################################
3 #####################################################################
5 /** \addtogroup GRAS_API
7 \section GRAS_funct Offered functionnalities
8 - <b>\ref GRAS_comm</b>: Exchanging messages between peers
9 - \ref GRAS_dd : any data which may transit on the network must be
10 described beforehand so that GRAS can handle the platform
11 heterogeneity and convert them if needed.
12 - \ref GRAS_sock : this is how to open a communication channel to
13 other processes, and retrive information about them.
14 - \ref GRAS_msg : communications are message oriented. You have to
15 describe all possible messages and their payload beforehand, and
16 can then attach callbacks to the arrival of a given kind of message.
17 - \ref GRAS_timer : this is how to program repetitive and delayed
18 tasks, not unlike cron(8) and at(1). This cannot be used to timeout
19 a function (like setitimer(2) or signal(2) games could do).
20 - <b>\ref GRAS_run</b>: Running both on top of the simulator and on
21 top of real platforms, and portability support.
22 - \ref GRAS_virtu : You naturally don't want to call the
23 gettimeofday(2) function in simulation mode since it would give
24 you the time on the host running the simulation, not the time in
25 the simulated world (you are belonging to).\n
26 This a system call virtualization layer, which also acts as a
28 - \ref GRAS_globals : The use of globals is forbidden since the
29 "processes" are threads in simulation mode. \n
30 This is how to let GRAS handle your globals properly.
31 - \ref GRAS_emul : Support to emulate code excution (ie, reporting
32 execution time into the simulator and having code sections specific
33 to simulation or to real mode).
34 - <b>\ref GRAS_code</b>: Here are some tools which may help
35 you setting up a GRAS project.\n
36 Setting up and building a GRAS application is complicated by the
37 library schizoid. The code to setup the environment differs
38 depending on whether you run on the simulator on a real platform.
39 And then, you'll have to deal with the usual distributed
40 application development difficulties.
41 - \ref GRAS_main_generation : Since processes are threads in
42 simulation mode and regular processes in the real world, GRAS does
43 generate your main functions for you.
47 \section GRAS_example Examples
49 There is for now rather few examples of GRAS, but it's better than
57 /** @defgroup GRAS_comm Communication facilities */
58 /** @defgroup GRAS_run Virtualization */
59 /** @defgroup GRAS_code Project and code management */
60 /** @defgroup GRAS_ex Examples */
62 #####################################################################
63 /** @addtogroup GRAS_comm
65 Here are the communication facilities. GRAS allows you to exchange
66 <i>messages</i> on <i>sockets</i> (which can be seen as pipes between
67 processes). On reception, messages start <i>callbacks</i> (that's the
68 default communication mode, not the only one). All messages of a given
69 type convey the same kind of data, and you have to describe it
72 Timers are also seen as a mean of communication (with yourself). It
73 allows you to run a repetitive task ("do this every N second until I tell
74 you to stop"), or to deffer a treatment ("do this in 3 sec").
77 /** @defgroup GRAS_dd Data description */
78 /** @defgroup GRAS_sock Sockets */
79 /** @defgroup GRAS_msg Messages */
80 /** @defgroup GRAS_timer Timers */
83 #####################################################################
84 /** @addtogroup GRAS_run
88 /** @defgroup GRAS_globals Globals */
89 /** @defgroup GRAS_emul Emulation support */
90 /** @defgroup GRAS_virtu Syscalls */
94 #####################################################################
95 /** @addtogroup GRAS_code
98 DOXYGEN_NAVBAR_LABEL="Project management"
99 DOXYGEN_NAVBAR_CHILD "main() and GRAS"=GRAS_main_generation.html
100 DOXYGEN_NAVBAR_CHILD "Compiling your GRAS project"=GRAS_compile.html
104 #####################################################################
105 /** @addtogroup GRAS_ex
108 DOXYGEN_NAVBAR_CHILD "Ping-Pong"=GRAS_ex_ping.html
109 DOXYGEN_NAVBAR_CHILD "RPC"=GRAS_ex_mmrpc.html
110 DOXYGEN_NAVBAR_CHILD "Timers"=GRAS_ex_timer.html
114 #####################################################################
115 ######################### EXTRA PAGES ##############################
116 #####################################################################
118 ---------------------------------------------------------------------
119 --------------------- main() generation -----------------------------
120 ---------------------------------------------------------------------
122 /** \page GRAS_main_generation main function
124 \section GRAS_maingen_toc Table of content
126 - \ref GRAS_maingen_intro
127 - \ref GRAS_maingen_script
128 - \ref GRAS_maingen_make
132 \section GRAS_maingen_intro What's the matter with main() functions in GRAS?
134 In simulation mode, all processes are run as thread of the same process
135 while they are real processes in the real life. Unfortunately, the main
136 function of a real process must be called <tt>main</tt> while this
137 function must not use this name for threads.
139 To deal with this, you should call the main function of your processes
140 with another name (usually, the process function such as client, server,
141 or such). Then GRAS can generate the wrapper functions adapted to the
142 real and simulated modes.
144 \section GRAS_maingen_script Generating the main()s automatically
146 This is done by the gras_stub_generator program, which gets installed on
147 <tt>make install</tt> (the source resides in the tools/gras/ directory).
148 Here is the calling syntax:
149 \verbatim gras_stub_generator <project_name> <deployment_file.xml>\endverbatim
151 It parses the deployment file, searching for all the kind of processes
152 you have in your project. It then generates the following C files:
153 - a <tt>_<project_name>_<process_kind>.c</tt> file for each process kind you
155 They are used to launch your project in real life. They
156 contain a main() in charge of initializing the GRAS infrastructure and
157 launching your code afterward.
158 - a <tt>_<project_name>_simulator.c</tt> file.\n
159 This file is suited to the simulation mode. It contains a main()
160 function initializing the simulator and launching your project within.
162 For this to work, the name of process described in your deployment file
163 should match the name of a function in your code, which prototype is for
164 example: \verbatim int client(int argc,char *argv[]);\endverbatim
166 Unfortunately, all this is still partially documented. I guess I ought
167 to improve this situation somehow. In the meanwhile, check the generated
168 code and maybe also the GRAS \ref GRAS_example, sorry.
170 \section GRAS_maingen_make Integration within an hand-made Makefile
172 The easiest to set it up is to add the following chunk at the end of
173 your Makefile (or Makefile.am), putting the right values into NAME and
175 \verbatim NAME=your_project_name
176 PROCESSES=list of processes type in your project
178 $(foreach proc, $(PROCESSES), _$(NAME)_$(proc).c) _$(NAME)_simulator.c: $(NAME).c $(NAME)_deployment.xml
179 path/to/gras_stub_generator $(NAME) $(NAME)_deployment.xml >/dev/null
182 Of course, your personal millage may vary. For the \ref GRAS_ex_ping, may read:
183 \verbatim _ping_client.c _ping_server.c _ping_simulator.c: ping.c ping_deployment.xml
184 $(top_srcdir)/tools/gras/gras_stub_generator ping ping_deployment.xml >/dev/null
188 Actually, gras_stub_generator also generates some makefiles both for
189 local compilation and remote code distribution and installation. See the
190 section \ref GRAS_compile for more details.
194 ---------------------------------------------------------------------
195 ------------------------- Compiling ---------------------------------
196 ---------------------------------------------------------------------
198 /** \page GRAS_compile Compiling your project
200 As explained in section \ref GRAS_main_generation, the
201 gras_stub_generator tool can be used to generate the system
202 initialization code in your projet. While we were at this, this tool
203 also generates the makefiles you will need to compile your project
206 Code source deployment and remote compilation also constitutes a
207 challenging area in distributed applications development. The GRASPE
208 (GRAS Platform Expender) tool was designed to make this less painful.
210 \section GRAS_compile_toc Table of content
212 - \ref GRAS_compile_local
213 - \ref GRAS_compile_local_install
214 - \ref GRAS_compile_local_helpfiles
215 - \ref GRAS_compile_local_makefile
216 - \ref GRAS_compile_remote
220 \section GRAS_compile_local Local compilation of GRAS projects
222 \subsection GRAS_compile_local_install Installing SimGrid and GRAS
224 To compile locally a GRAS project, you first need to install SimGrid on
225 your machine. Use the --prefix flag to the configure script to specify
226 where you want to install the toolkit (refere to section \ref
227 faq_compiling for more information)
229 \subsection GRAS_compile_local_helpfiles Simulation description files
231 Then, you will probably need to write a platform description file and
232 application deployment description file to feed the simulator with. This
233 part is unfortunatelly not documented enough. Files examples can be
234 found along with the MSG \ref MSG_ex_master_slave example.
236 \note yes, both platform and application description files are portable
237 between MSG and GRAS. Actually, there depend on the SURF, not on the
238 programming environment you use.
240 For the first try, you could probably reuse the provided platform file
241 as is while you will need to adapt the application file to fit your
244 To generate new platform files, we usually use the Tiers Topology
245 Generator (ask google about it) and annotate the generated graph with
246 home-made scripts to let them fit the SURF. Those scripts live in the
247 tools/platform_generation/ directory of the distribution.
249 \subsection GRAS_compile_local_makefile Generating a Makefile usable for your project
251 From the information contained in the application description file, the
252 gras_stub_generator tool can create a Makefile which can be used to
253 seamlessly compile your project. Just go to the directory containing all
254 your project files, and type:
256 \verbatim path/to/gras_stub_generator [project_name] [application_deployment.file] >/dev/null
259 The first argument is the name of your project, such as
260 "MyLovelyApplication" while the second one is the application deployment
263 Several files get generated by this command. One C file per kind of
264 process in your project (such as "master" and "slave") plus one C file
265 for simulating your project. All those files are (or should ;) described
266 in section \ref GRAS_main_generation.
268 The most intersting file in this context is
269 [project_name].Makefile.local (you can safely ignore the others for
270 now). To use it, simply type (from your project main directory):
272 \verbatim GRAS_ROOT=/path/to/simgrid/installation make -f [project_name].Makefile.local
275 And that's it, all the binaries are built and linked against the correct
278 \section GRAS_compile_remote Distribution and remote compilation of GRAS projects
280 Actually, there is two somehow parallel ways to do so since both Arnaud
281 and Martin gave it a try. Merging both approaches is underway. As usual,
282 if you want to help, you're welcome ;)
286 #####################################################################
287 ######################### EXAMPLES #################################
288 #####################################################################
290 ---------------------------------------------------------------------
291 ------------------------- Ping Pong ---------------------------------
292 ---------------------------------------------------------------------
294 /** \page GRAS_ex_ping The classical Ping-Pong in GRAS
296 This example implements the very classical ping-pong in GRAS. It
297 involves a client (initiating the ping-pong) and a server (answering to
300 It works the following way:
301 - Both the client and the server register all needed messages
302 - The server registers a callback to the ping message, which sends pong
304 - The client sends the ping message to the server, and waits for the
305 pong message as an answer.
307 This example resides in the <b>examples/gras/ping/ping.c</b> file. Yes, both
308 the code of the client and of the server is placed in the same file. See
309 the \ref GRAS_main_generation section if wondering.
311 \section GRAS_ex_ping_toc Table of contents of the ping example
312 - \ref GRAS_ex_ping_common
313 - \ref GRAS_ex_ping_initial
314 - \ref GRAS_ex_ping_register
315 - \ref GRAS_ex_ping_server
316 - \ref GRAS_ex_ping_serdata
317 - \ref GRAS_ex_ping_sercb
318 - \ref GRAS_ex_ping_sermain
319 - \ref GRAS_ex_ping_client
320 - \ref GRAS_ex_ping_climain
324 \dontinclude gras/ping/ping.c
326 \section GRAS_ex_ping_common 1) Common code to the client and the server
328 \subsection GRAS_ex_ping_initial 1.a) Initial settings
330 Let's first load the gras header and declare a logging category (see
331 \ref XBT_log for more info on logging).
336 \subsection GRAS_ex_ping_register 1.b) Register the messages
338 This function, called by both the client and the server is in charge of
339 declaring the existing messages to GRAS. Since the payload does not
340 involve any newly created types but only int, this is quite easy.
341 (to exchange more complicated types, see \ref GRAS_dd or
342 \ref GRAS_ex_mmrpc for an example).
344 \skip register_messages
347 [Back to \ref GRAS_ex_ping_toc]
349 \section GRAS_ex_ping_server 2) Server's code
351 \subsection GRAS_ex_ping_serdata 2.a) The server's globals
353 In order to ensure the communication between the "main" and the callback
354 of the server, we need to declare some globals. We have to put them in a
355 struct definition so that they can be handled properly in GRAS (see the
356 \ref GRAS_globals for more info).
361 \subsection GRAS_ex_ping_sercb 2.b) The callback to the ping message
363 Here is the callback run when the server receives any ping message (this
364 will be registered later by the server).
366 \skip server_cb_ping_handler
367 \until end_of_server_cb_ping_handler
369 \subsection GRAS_ex_ping_sermain 2.c) The "main" of the server
371 This is the "main" of the server. As explained in the \ref
372 GRAS_main_generation, you must not write any main()
373 function yourself. Instead, you just have to write a regular function
374 like this one which will act as a main.
379 [Back to \ref GRAS_ex_ping_toc]
381 \section GRAS_ex_ping_client 3) Client's code
383 \subsection GRAS_ex_ping_climain 3.a) Client's "main" function
385 This function is quite straightforward, and the inlined comments should
386 be enough to understand it.
391 [Back to \ref GRAS_ex_ping_toc]
394 ---------------------------------------------------------------------
395 -------------------------- MM RPC -----------------------------------
396 ---------------------------------------------------------------------
398 /** \page GRAS_ex_mmrpc A simple RPC for matrix multiplication
400 This example implements a remote matrix multiplication. It involves a client
401 (creating the matrices and sending the multiplications requests) and a server
402 (computing the multiplication on client's behalf).
404 This example also constitutes a more advanced example of data description
405 mechanisms, since the message payload type is a bit more complicated than in
406 other examples such as the ping one (\ref GRAS_ex_ping).
408 It works the following way (not very different from the ping example):
409 - Both the client and the server register all needed messages and datatypes
410 - The server registers a callback to the "request" message, which computes
411 what needs to be and returns the result to the expeditor.
412 - The client creates two matrices, ask for their multiplication and check
415 This example resides in the <b>examples/gras/mmrpc/mmrpc.c</b> file. (See
416 the \ref GRAS_main_generation section if wondering why both the server
417 and the client live in the same source file)
419 \section GRAS_ex_mmrpc_toc Table of contents of the mmrpc example
420 - \ref GRAS_ex_mmrpc_common
421 - \ref GRAS_ex_mmrpc_initial
422 - \ref GRAS_ex_mmrpc_dataregister
423 - \ref GRAS_ex_mmrpc_msgregister
424 - \ref GRAS_ex_mmrpc_server
425 - \ref GRAS_ex_mmrpc_sercb
426 - \ref GRAS_ex_mmrpc_sermain
427 - \ref GRAS_ex_mmrpc_client
428 - \ref GRAS_ex_mmrpc_climain
432 \dontinclude gras/mmrpc/mmrpc.c
434 \section GRAS_ex_mmrpc_common 1) Common code to the client and the server
436 \subsection GRAS_ex_mmrpc_initial 1.a) Initial settings
438 Let's first load the gras header, specify the matrix size and declare a
439 logging category (see \ref XBT_log for more info on logging).
444 \subsection GRAS_ex_mmrpc_dataregister 1.b) Register the data types
446 The messages involved in this example do use structures as payload,
447 so we have to declare it to GRAS. Hopefully, this can be done easily by enclosing
448 the structure declaration within a \ref GRAS_DEFINE_TYPE macro call. It will then copy this
449 declaration into an hidden string variable, which can be automatically parsed at
450 run time. Of course, the declaration is also copied unmodified by this macro, so that it
451 gets parsed by the compiler also.
453 There is some semantic that GRAS cannot guess alone and you need to <i>annotate</i>
454 your declaration to add some. For example, the ctn pointer can be a reference to an
455 object or a whole array (in which case you also has to specify its size). This is done
456 with the GRAS_ANNOTE call. It is removed from the text passed to the compiler, but it helps
457 GRAS getting some information about the semantic of your data. Here, it says that \a ctn is an
458 array, which size is the result of the operation \a rows * \a cols (with \a rows and \a cols
459 being the other fields of the structure).
461 Please note that this annotation mechanism is not as robust and cool as this example seems to
462 imply. If you want to use it yourself, you'd better use the exact right syntax, which is
463 detailed in the \ref GRAS_dd section.
465 \skip GRAS_DEFINE_TYPE
468 \subsection GRAS_ex_mmrpc_msgregister 1.c) Register the messages
470 This function, called by both the client and the server is in charge of
471 declaring the existing messages to GRAS. Note the use of the \ref gras_datadesc_by_symbol
472 function to parse and retrieve the structure declaration which were passed to \ref GRAS_DEFINE_TYPE
475 The datatype description builded that way can then be used to build an array datatype or
478 \skip register_messages
481 [Back to \ref GRAS_ex_mmrpc_toc]
483 \section GRAS_ex_mmrpc_server 2) Server's code
485 \subsection GRAS_ex_mmrpc_sercb 2.a) The callback to the mmrpc message
487 Here is the callback run when the server receives any mmrpc message (this
488 will be registered later by the server). Note the way we get the message
489 payload. In the ping example, there was one additional level of pointer
490 indirection (see \ref GRAS_ex_ping_sercb). This is because the payload is
491 an array here (ie a pointer) whereas it is a scalar in the ping example.
493 \skip server_cb_request_handler
494 \until end_of_server_cb_request_handler
496 \subsection GRAS_ex_mmrpc_sermain 2.b) The "main" of the server
498 This is the "main" of the server. As explained in the \ref
499 GRAS_main_generation, you must not write any main()
500 function yourself. Instead, you just have to write a regular function
501 like this one which will act as a main.
506 [Back to \ref GRAS_ex_mmrpc_toc]
508 \section GRAS_ex_mmrpc_client 3) Client's code
510 \subsection GRAS_ex_mmrpc_climain 3.a) Client's "main" function
512 This function is quite straightforward, and the inlined comments should
513 be enough to understand it.
518 [Back to \ref GRAS_ex_mmrpc_toc]
521 ---------------------------------------------------------------------
522 ---------------------------- Timers ---------------------------------
523 ---------------------------------------------------------------------
525 /** \page GRAS_ex_timer Some timer games
527 This example fools around with the GRAS timers (\ref GRAS_timer). It is
528 mainly a regression test, since it uses almost all timer features.
530 The main program registers a repetititive task and a delayed one, and
531 then loops until the <tt>still_to_do</tt> variables of its globals reach
532 0. The delayed task set it to 5, and the repetititive one decrease it
533 each time. Here is an example of output:
534 \verbatim Initialize GRAS
536 [1108335471] Programming the repetitive_action with a frequency of 1.000000 sec
537 [1108335471] Programming the delayed_action for after 2.000000 sec
538 [1108335471] Have a rest
539 [1108335472] Canceling the delayed_action.
540 [1108335472] Re-programming the delayed_action for after 2.000000 sec
541 [1108335472] Repetitive_action has nothing to do yet
542 [1108335473] Repetitive_action has nothing to do yet
543 [1108335473] delayed_action setting globals->still_to_do to 5
544 [1108335474] repetitive_action decrementing globals->still_to_do. New value: 4
545 [1108335475] repetitive_action decrementing globals->still_to_do. New value: 3
546 [1108335476] repetitive_action decrementing globals->still_to_do. New value: 2
547 [1108335477] repetitive_action decrementing globals->still_to_do. New value: 1
548 [1108335478] repetitive_action decrementing globals->still_to_do. New value: 0
549 Exiting GRAS\endverbatim
552 - \ref GRAS_ex_timer_decl
553 - \ref GRAS_ex_timer_delay
554 - \ref GRAS_ex_timer_repeat
555 - \ref GRAS_ex_timer_main
559 \section GRAS_ex_timer_decl 1. Declarations and headers
563 \section GRAS_ex_timer_delay 2. Source code of the delayed action
564 \skip repetitive_action
565 \until end_of_repetitive_action
567 \section GRAS_ex_timer_repeat 3. Source code of the repetitive action
569 \until end_of_delayed_action
571 \section GRAS_ex_timer_main 4. Source code of main function