1 /*! \page faq Frequently Asked Questions
3 \htmlinclude .FAQ.doc.toc
5 \section faq_simgrid I'm new to SimGrid. I have some questions. Where should I start?
7 You are at the right place... Having a look to these
8 <a href="http://www.loria.fr/~quinson/articles/simgrid-tutorial.pdf">the tutorial slides</a>
9 (or to these <a href="http://graal.ens-lyon.fr/~alegrand/articles/slides_g5k_simul.pdf">old slides</a>,
11 <a href="http://graal.ens-lyon.fr/~alegrand/articles/Simgrid-Introduction.pdf">"obsolete" slides</a>)
12 may give you some insights on what SimGrid can help you to do and what
13 are its limitations. Then you definitely should read the \ref
14 MSG_examples. The \ref GRAS_tut can also help you.
16 If you are stuck at any point and if this FAQ cannot help you, please drop us a
17 mail to the user mailing list: <simgrid-user@lists.gforge.inria.fr>.
19 \subsection faq_interfaces What is the difference between MSG, SimDag, and GRAS? Do they serve the same purpose?
21 It depend on how you define "purpose", I guess ;)
23 They all allow you to build a prototype of application which you can run
24 within the simulator afterward. They all share the same simulation kernel,
25 which is the core of the SimGrid project. They differ by the way you express
28 With SimDag, you express your code as a collection of interdependent
29 parallel tasks. So, in this model, applications can be seen as a DAG of
30 tasks. This is the interface of choice for people wanting to port old
31 code designed for SimGrid v1 or v2 to the framework current version.
33 With both GRAS and MSG, your application is seen as a set of communicating
34 processes, exchanging data by the way of messages and performing computation
37 The difference between both is that MSG is somehow easier to use, but GRAS
38 is not limited to the simulator. Once you're done writing your GRAS code,
39 you can run your code both in the simulator or on a real platform. For this,
40 there is two implementations of the GRAS interface, one for simulation, one
41 for real execution. So, you just have to relink your code to chose one of
44 \subsection faq_generic First steps with SimGrid
46 If you decide to go for the MSG interface, please read carefully the
47 \ref MSG_examples. You'll find in \ref MSG_ex_master_slave a very
48 simple consisting of a master (that owns a bunch of tasks and
49 distributes them) , some slaves (that process tasks whenever they
50 receive one) and some forwarder agents (that simply pass the tasks
51 they receive to some slaves).
53 If you decide to go for the GRAS interface, you should definitively
54 read the \ref GRAS_tut. The first section constitutes an introduction
55 to the tool and presents the model we use. The second section
56 constitutes a complete step-by-step tutorial building a distributed
57 application from the beginning and exemplifying most of the GRAS
58 features in the process. The last section groups some HOWTOS
59 highlighting a given feature of the framework in a more concise way.
61 If you decide to go for another interface, I'm afraid your only sources
62 of information will be the source code and the mailing lists...
64 \subsection faq_visualization Visualizing and analyzing the results
66 It is sometime convenient to "see" how the agents are behaving. If you
67 like colors, you can use <tt>tools/MSG_visualization/colorize.pl </tt>
68 as a filter to your MSG outputs. It works directly with INFO. Beware,
69 INFO() prints on stderr. Do not forget to redirect if you want to
70 filter (e.g. with bash):
72 ./msg_test small_platform.xml small_deployment.xml 2>&1 | ../../tools/MSG_visualization/colorize.pl
75 We also have a more graphical output. Have a look at MSG_paje_output(). It
76 generates an input to <a href="http://www-id.imag.fr/Logiciels/paje/">Paje</a>.
79 <a href="Paje_MSG_screenshot.jpg"><img src="Paje_MSG_screenshot_thn.jpg"></a>
83 Visualization with Paje can be seen as a kind of postmortem
84 analysis. However, as soon as you start playing with big simulations,
85 you'll realize that processing such output is kind of tricky. There is
86 so much generic information that it is hard to find the information
89 As a matter of fact, logging really depends on simulations (e.g. what
90 kind of events is important...). That is why we do not propose a big
91 dump of your whole simulation (it would slow everything down) but give
92 you neat tools to structure you logs. Have a look at \ref XBT_log. In
93 fact, rather than a post-mortem analysis, you may want to do it on the
94 fly. The process you are running can do whatever you want. Have you
95 thought about adding a global structure where you directly compute the
96 information that are really important rather than writing everything
97 down and then processing huge files?
99 \subsection faq_C Argh! Do I really have to code in C?
101 Up until now, there is no binding for other languages. If you use C++,
102 you should be able to use the SimGrid library as a standard C library
103 and everything should work fine (simply <i>link</i> against this
104 library; recompiling SimGrid with a C++ compiler won't work and it
105 wouldn't help if you could).
107 In fact, we are currently working on Java bindings of MSG to allow
108 all the undergrad students of the world to use this tool. This is a
109 little more tricky than I would have expected, but the work is moving
110 fast forward [2006/05/13]. More languages are evaluated, but for now,
111 we do not feel a real demand for any other language. Please speak up!
113 \section faq_installation Installing the SimGrid library
115 Many people have been asking me questions on how to use SimGrid. Quite
116 often, the questions were not really about SimGrid but on the
117 installation process. This section is intended to help people that are
118 not familiar with compiling C files under UNIX. If you follow these
119 instructions and still have some troubles, drop an e-mail to
120 <simgrid-user@lists.gforge.inria.fr>.
122 \subsection faq_compiling Compiling SimGrid from a stable archive
124 First of all, you need to download the latest version of SimGrid from
125 <a href="http://gforge.inria.fr/frs/?group_id=12">here</a>.
126 Suppose you have uncompressed SimGrid in some temporary location of
127 your home directory (say <tt>/home/joe/tmp/simgrid-3.0.1 </tt>). The
128 simplest way to use SimGrid is to install it in your home
129 directory. Change your directory to
130 <tt>/home/joe/tmp/simgrid-3.0.1</tt> and type
133 ./configure --prefix=$HOME
138 If at some point, something fails, check the section \ref faq_trouble_compil .
139 If it does not help, you can report this problem to the
140 list but, please, avoid sending a laconic mail like "There is a problem. Is it
141 okay?". Send the config.log file which is automatically generated by
142 configure. Try to capture both the standard output and the error output of the
143 <tt>make</tt> command with <tt>script</tt>. There is no way for us to help you
144 without the relevant bits of information.
146 Now, the following directory should have been created :
148 \li <tt>/home/joe/doc/simgrid/html/</tt>
149 \li <tt>/home/joe/lib/</tt>
150 \li <tt>/home/joe/include/</tt>
152 SimGrid is not a binary, it is a library. Both a static and a dynamic
153 version are available. Here is what you can find if you try a <tt>ls
156 \verbatim libsimgrid.a libsimgrid.la libsimgrid.so libsimgrid.so.0 libsimgrid.so.0.0.1
159 Thus, there is two ways to link your program with SimGrid:
160 \li Either you use the static version, e.g
161 \verbatim gcc libsimgrid.a -o MainProgram MainProgram.c
163 In this case, all the SimGrid functions are directly
164 included in <tt>MainProgram</tt> (hence a bigger binary).
165 \li Either you use the dynamic version (the preferred method)
166 \verbatim gcc -lsimgrid -o MainProgram MainProgram.c
168 In this case, the SimGrid functions are not included in
169 <tt>MainProgram</tt> and you need to set your environment
170 variable in such a way that <tt>libsimgrid.so</tt> will be
171 found at runtime. This can be done by adding the following
172 line in your .bashrc (if you use bash and if you have
173 installed the SimGrid libraries in your home directory):
174 \verbatim export LD_LIBRARY_PATH=$HOME/lib/:$LD_LIBRARY_PATH
177 \subsection faq_compiling_java Java bindings don't get compiled
179 The configure script detects automatically whether you have the
180 softwares needed to use the Java bindings or not. At the end of the
181 configure, you can see the configuration picked by the script, which
182 should look similar to
183 \verbatim Configuration of package simgrid' (version 3.3.4-svn) on
186 Compiler: gcc (version: )
188 CFlags: -O3 -finline-functions -funroll-loops -fno-strict-aliasing -Wall -Wunused -Wmissing-prototypes -Wmissing-declarations -Wpointer-arith -Wchar-subscripts -Wcomment -Wformat -Wwrite-strings -Wno-unused-function -Wno-unused-parameter -Wno-strict-aliasing -Wno-format-nonliteral -Werror -g3
192 Context backend: ucontext
199 In this example, Java backends won't be compiled.
201 On Debian-like systems (which includes ubuntu), you need the following
202 packages: sun-java6-jdk libgcj10-dev. If you cannot find the
203 libgcj10-dev, try another version, like libgcj9-dev (on Ubuntu before
204 9.10) or libgcj11-dev (not released yet, but certainly one day).
205 Please note that you need to activate the contrib and non-free
206 repositories in Debian, and the universe ones in Ubuntu. Java comes at
209 \subsection faq_compiling_snapshoot SimGrid development snapshots
211 We have very high standards on software quality, and we are reluctant releasing
212 a stable release as long as there is still some known bug in the code base. In
213 addition, we added quite an extensive test base, making sure that we correctly
214 test the most important parts of the tool.
216 As an unfortunate conclusion, there may be some time between the stable
217 releases. If you want to benefit from the most recent features we introduced,
218 but don't want to take the risk of an untested version from the SVN, then
219 development snapshots are done for you.
221 These are pre-releases of SimGrid that still fail some tests about features
222 that almost nobody use, or on platforms not being in our core target (which is
223 Linux, Mac, other Unixes and Windows, from the most important to the less
224 one). That means that using this development releases should be safe for most
227 These archives can be found on
228 <a href="http://www.loria.fr/~quinson/simgrid.html">this web page</a>. Once you
229 got the lastest archive, you can compile it just like any archive (see above).
231 \subsection faq_compiling_svn Compiling SimGrid from the SVN
233 The project development takes place in the SVN, where all changes are
234 committed when they happen. Then every once in a while, we make sure that the
235 code quality meets our standard and release an archive from the code in the
236 SVN. We afterward go back to the development in the SVN. So, if you need a
237 recently added feature and can afford some little problem with the stability
238 of the lastest features, you may want to use the SVN version instead of a
241 For that, you first need to get the "simgrid" module from
242 <a href="http://gforge.inria.fr/scm/?group_id=12">here</a>.
244 You won't find any <tt>configure</tt> and a few other things
245 (<tt>Makefile.in</tt>'s, documentation, ...) will be missing as well. The
246 reason for that is that all these files have to be regenerated using the
247 latest versions of <tt>autoconf</tt>, <tt>libtool</tt>, <tt>automake</tt>
248 (>1.9) and <tt>doxygen</tt> (>1.4). To generate the <tt>configure</tt> and
249 the <tt>Makefile.in</tt>'s, you just have to launch the <tt>bootstrap</tt>
250 command that resides in the top of the source tree. Then just follow the
251 instructions of Section \ref faq_compiling.
253 We insist on the fact that you really need the latest versions of
254 autoconf, automake and libtool. Doing this step on exotic architectures/systems
255 (i.e. anything different from a recent linux distribution) may be
256 ... uncertain. If you need to compile the SVN version on a machine where all these
257 dependencies are not met, the easiest is to do <tt>make dist</tt> in the SVN
258 directory of another machine where all dependencies are met. It will create an
259 archive you may deploy on other sites just as a regular stable release.
261 In summary, the following commands will checkout the SVN, regenerate the
262 configure script and friends, configure SimGrid and build it.
264 \verbatim svn checkout svn://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk simgrid
267 ./configure --enable-maintainer-mode --prefix=<where to install SimGrid>
270 Then, if you want to install SimGrid on the current box, just do:
271 \verbatim make install \endverbatim
273 If you want to build an snapshot of the SVN to deploy it on another box (for
274 example because the other machine don't have the autotools), do:
275 \verbatim make dist \endverbatim
277 Moreover, you should never call the autotools manually since you must run
278 them in a specific order with specific arguments. Most of the times, the
279 makefiles will automatically call the tools for you. When it's not possible
280 (such as the first time you checkout the SVN), use the ./bootstrap command
281 to call them explicitly.
284 \subsection faq_setting_MSG Setting up your own MSG code
286 Do not build your simulator by modifying the SimGrid examples. Go
287 outside the SimGrid source tree and create your own working directory
288 (say <tt>/home/joe/SimGrid/MyFirstScheduler/</tt>).
290 Suppose your simulation has the following structure (remember it is
291 just an example to illustrate a possible way to compile everything;
292 feel free to organize it as you want).
294 \li <tt>sched.h</tt>: a description of the core of the
295 scheduler (i.e. which functions are can be used by the
296 agents). For example we could find the following functions
297 (master, forwarder, slave).
299 \li <tt>sched.c</tt>: a C file including <tt>sched.h</tt> and
300 implementing the core of the scheduler. Most of these
301 functions use the MSG functions defined in section \ref
304 \li <tt>masterslave.c</tt>: a C file with the main function, i.e.
305 the MSG initialization (MSG_global_init()), the platform
306 creation (e.g. with MSG_create_environment()), the
307 deployment phase (e.g. with MSG_function_register() and
308 MSG_launch_application()) and the call to
311 To compile such a program, we suggest to use the following
312 Makefile. It is a generic Makefile that we have used many times with
313 our students when we teach the C language.
317 masterslave: masterslave.o sched.o
319 INSTALL_PATH = $$HOME
321 PEDANTIC_PARANOID_FREAK = -O0 -Wshadow -Wcast-align \
322 -Waggregate-return -Wmissing-prototypes -Wmissing-declarations \
323 -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations \
324 -Wmissing-noreturn -Wredundant-decls -Wnested-externs \
325 -Wpointer-arith -Wwrite-strings -finline-functions
326 REASONABLY_CAREFUL_DUDE = -Wall
327 NO_PRAYER_FOR_THE_WICKED = -w -O2
328 WARNINGS = $(REASONABLY_CAREFUL_DUDE)
329 CFLAGS = -g $(WARNINGS)
331 INCLUDES = -I$(INSTALL_PATH)/include
332 DEFS = -L$(INSTALL_PATH)/lib/
333 LDADD = -lm -lsimgrid
337 $(CC) $(INCLUDES) $(DEFS) $(CFLAGS) $^ $(LIBS) $(LDADD) -o $@
340 $(CC) $(INCLUDES) $(DEFS) $(CFLAGS) -c -o $@ $<
343 rm -f $(BIN_FILES) *.o *~
349 The first two lines indicates what should be build when typing make
350 (<tt>masterslave</tt>) and of which files it is to be made of
351 (<tt>masterslave.o</tt> and <tt>sched.o</tt>). This makefile assumes
352 that you have set up correctly your <tt>LD_LIBRARY_PATH</tt> variable
353 (look, there is a <tt>LDADD = -lm -lsimgrid</tt>). If you prefer using
354 the static version, remove the <tt>-lsimgrid</tt> and add a
355 <tt>$(INSTALL_PATH)/lib/libsimgrid.a</tt> on the next line, right
356 after the <tt>LIBS = </tt>.
358 More generally, if you have never written a Makefile by yourself, type
359 in a terminal : <tt>info make</tt> and read the introduction. The
360 previous example should be enough for a first try but you may want to
361 perform some more complex compilations...
363 \subsection faq_setting_GRAS Setting up your own GRAS code
365 If you use the GRAS interface instead of the MSG one, then previous section
366 is not the better source of information. Instead, you should check the GRAS
367 tutorial in general, and the \ref GRAS_tut_tour_setup in particular.
369 \section faq_cmake CMAKE
371 \subsection faq_intro Some generalitty
373 \subsubsection faq_intro1 What is Cmake?
375 CMake is a family of tools designed to build, test and package software. CMake is used to control the software compilation process using simple platform and compiler independent configuration files. CMake generates native makefiles and workspaces that can be used in the compiler environment of your choice. For more information see official web site <a href="http://www.cmake.org/">here</a>.
377 \subsubsection faq_intro2 Why cmake?
379 CMake permits to developers to compil projects on different plateform. Then many tools are embedded like ctest for making test, a link to cdash for vizualise results but also test coverage and bug reports.
381 \subsubsection faq_intro3 What cmake need?
383 CMake needs some prerequists like :
385 \li c, c++ and java compiler regards to developers
386 \li ccmake for graphical used of CMake
387 \li cmake <a href="http://www.cmake.org/cmake/resources/software.html">(download page)</a>
389 \subsubsection faq_intro4 Cmake vs Autotools...
393 \subsection faq_cmakeoption Cmake options
395 \subsubsection faq_cmakeoption1 Liste of options
398 "cmake -D[name]=[value] ... ./"
400 [name] disable_gtnets [value] ON/OFF or TRUE/FALSE or 1/0
401 disable_java ON/OFF or TRUE/FALSE or 1/0
402 disable_lua ON/OFF or TRUE/FALSE or 1/0
403 disable_ruby ON/OFF or TRUE/FALSE or 1/0
405 enable_compile_optimizations ON/OFF or TRUE/FALSE or 1/0
406 enable_compile_warnings ON/OFF or TRUE/FALSE or 1/0
407 enable_maintainer_mode ON/OFF or TRUE/FALSE or 1/0
408 enable_supernovae ON/OFF or TRUE/FALSE or 1/0
409 enable_tracing ON/OFF or TRUE/FALSE or 1/0
410 enable_coverage ON/OFF or TRUE/FALSE or 1/0
411 enable_memcheck ON/OFF or TRUE/FALSE or 1/0
412 enable_print_message ON/OFF or TRUE/FALSE or 1/0
414 gtnets_path <path_to_gtnets_directory>
415 prefix <path_to_install_directory>
416 BIBTEX2HTML <path_to_bibtex2html>
417 with_context auto/ucontext/pthread/window
420 \subsubsection faq_cmakeoption2 Options explaination
422 \li disable_gtnets : set to true implie that user doesn't want to use gtnets.
424 \li disable_java : set to true implie that user doesn't want to add java langage into simgrid compilation.
426 \li disable_lua : set to true implie that user doesn't want to add lua langage into simgrid compilation.
428 \li disable_ruby : set to true implie that user doesn't want to add ruby langage into simgrid compilation.
430 \li enable_compile_optimizations : add flags "-O3 -finline-functions -funroll-loops -fno-strict-aliasing"
432 \li enable_compile_warnings : add flags "-Wall -Wunused -Wmissing-prototypes -Wmissing-declarations -Wpointer-arith -Wchar-subscripts -Wcomment -Wformat -Wwrite-strings -Wno-unused-function -Wno-unused-parameter -Wno-strict-aliasing -Wno-format-nonliteral -Werror"
434 \li enable_maintainer_mode : set to true it remakes some files.
436 include/surf/simgrid_dtd.h
437 include/xbt/graphxml.h
446 src/xbt_strbuff_unit.c
449 src/xbt_synchro_unit.c
450 src/simgrid_units_main.c
456 src/surf/simgrid_dtd.c
457 src/surf/simgrid_dtd.l
462 src/gras/DataDesc/ddt_parse.yy.c
464 \li enable_supernovae : set to true make one file for each lib and compile with those generated files.
467 /src/supernovae_gras.c
468 /src/supernovae_smpi.c
471 \li enable_tracing : To enable the generation of simulation traces for visualization
473 \li enable_coverage : When set to true this option enable code coverage by setting -fprofile-arcs -ftest-coverage flags.
475 \li enable_memcheck : When set to true this option enable tests for memcheck.
477 \li enable_print_message : This option when enable permits to see variables from gras_config.h
479 \li gtnets_path : Path to gtnets install directory (ex /usr)
481 \li prefix : Path where are installed lib/ doc/ and include/ directories (ex /usr/local)
483 \li BIBTEX2HTML : Path where is installed bibtex2html.
485 \li with context : specify which context the user wants to use.
487 \subsubsection faq_cmakeoption3 Initialisation
489 Those options are initialized the first time you launch \"cmake ./\" whithout specified option.
497 enable_compile_optimizations off
498 enable_compile_warnings off
499 enable_maintainer_mode off
500 enable_supernovae off
504 enable_print_message off
512 \subsubsection faq_cmakeoption4 Option's cache and how to reset?
514 When options have been set they are keep into a cache file named \"CMakeCache.txt\". So if you want
515 reset values you just delete this file located to the project directory.
517 \subsection faq_cmakecompilation Cmake compilation
519 \subsubsection faq_cmakecompilation1 With command line.
522 cmake -D[name]=[value] ... ./
526 \subsubsection faq_cmakecompilation2 With ccmake tool.
531 Then follow instructions.
533 \subsubsection faq_cmakecompilation3 Resume of command line
537 cmake ./ configure the project
538 make build all targets
539 make VERBOSE=1 build all targets and print build command lines
540 make test test all targets and summarize
541 make dist make the distrib
542 make distcheck check the dist (make + make dist + make test)
543 make install-simgrid install the project (doc/ lib/ include/)
544 make uninstall uninstall the project (doc/ lib/ include/)
545 make clean clean all targets
546 make java-clean clean files created by java option
547 make doc-clean clean files created for making doc
548 make supernovae-clean clean supernovae files
549 make maintainer-clean clean maintainer files
550 make all-clean execute the 5 upper clean command
551 make html Create simgrid documentation
552 make maintainer-clean Remove all files generated by mainainer mode
555 When the project have been succesfully compiling and build you can make tests.
559 ctest launch only tests
561 ctest -D Continuous(Start|Update|Configure|Build)
562 ctest -D Continuous(Test|Coverage|MemCheck|Submit)
563 ctest -D Experimental
564 ctest -D Experimental(Start|Update|Configure|Build)
565 ctest -D Experimental(Test|Coverage|MemCheck|Submit)
567 ctest -D Nightly(Start|Update|Configure|Build)
568 ctest -D Nightly(Test|Coverage|MemCheck|Submit)
569 ctest -D NightlyMemoryCheck
572 If you want to test before make a commit you can simply make "ctest -D Experimental" and then you can visualize results submitted into Cdash. <a href="http://cdash.inria.fr/CDash/index.php?project=Simgrid">(Go to Cdash site)</a>.
574 \subsubsection faq_cmakecompilation4 Examples for different mode.
578 cmake -Denable_maintainer_mode=on ./
579 \verbatim GTnetS doesn't works : set -Ddisable_gtnets=on
580 with_context auto change to ucontext
583 Make : src/simgrid.jar with : /usr/bin/javac
584 Make examples/java with : /usr/bin/javac
586 Configuration of package `simgrid' (revision 7228M) on arch (=4):
588 SITE : Linux_Ubuntu 9.10_x86_64
590 Compiler: c++ : /usr/bin/c++
591 version: c++ (Ubuntu 4.4.1-4ubuntu9) 4.4.1
593 version: gcc (Ubuntu 4.4.1-4ubuntu9) 4.4.1
599 Context backend: ucontext
609 Simgrid dependencies: dl -llua5.1
610 Gras dependencies: pthread
614 USER_PREFIX: /usr/local
615 INSTALL_PREFIX: /usr/local
619 -- Build files have been written to: /home/navarrop/Bureau/simgrid-trunk
624 cmake -Dsupernovae=on ./
625 \verbatim GTnetS doesn't works : set -Ddisable_gtnets=on
626 with_context auto change to ucontext
629 Make : src/simgrid.jar with : /usr/bin/javac
630 Make examples/java with : /usr/bin/javac
632 Configuration of package `simgrid' (revision 7228M) on arch (=4):
633 BUILDNAME : SUPERNOVAE
634 SITE : Linux_Ubuntu 9.10_x86_64
636 Compiler: c++ : /usr/bin/c++
637 version: c++ (Ubuntu 4.4.1-4ubuntu9) 4.4.1
639 version: gcc (Ubuntu 4.4.1-4ubuntu9) 4.4.1
641 CFlags: -O3 -finline-functions -funroll-loops -fno-strict-aliasing -Wall -Wunused -Wmissing-prototypes -Wmissing-declarations -Wpointer-arith -Wchar-subscripts -Wcomment -Wformat -Wwrite-strings -Wno-unused-function -Wno-unused-parameter -Wno-strict-aliasing -Wno-format-nonliteral -Werror -g3
645 Context backend: ucontext
655 Simgrid dependencies: dl -llua5.1
656 Gras dependencies: pthread
660 USER_PREFIX: /usr/local
661 INSTALL_PREFIX: /usr/local
665 -- Build files have been written to: /home/navarrop/Bureau/simgrid-trunk
670 cmake -Dgtnets_path=/home/navarrop/Bureau/usr/ ./
671 \verbatim with_context auto change to ucontext
674 Make : src/simgrid.jar with : /usr/bin/javac
675 Make examples/java with : /usr/bin/javac
677 Configuration of package `simgrid' (revision 7228M) on arch (=4):
679 SITE : Linux_Ubuntu 9.10_x86_64
681 Compiler: c++ : /usr/bin/c++
682 version: c++ (Ubuntu 4.4.1-4ubuntu9) 4.4.1
684 version: gcc (Ubuntu 4.4.1-4ubuntu9) 4.4.1
686 CFlags: -L/home/navarrop/Bureau/usr/lib -I/home/navarrop/Bureau/usr/include/gtnets -g3
687 CPPFlags: -L/home/navarrop/Bureau/usr/lib -I/home/navarrop/Bureau/usr/include/gtnets
690 Context backend: ucontext
692 path: /home/navarrop/Bureau/usr
700 Simgrid dependencies: dl -llua5.1 -lgtnets
701 Gras dependencies: pthread
705 USER_PREFIX: /usr/local
706 INSTALL_PREFIX: /usr/local
708 INFO -->> Take care to have export LD_LIBRARY_PATH before run make command for make examples with gtnets
709 copy and paste : export LD_LIBRARY_PATH=/home/navarrop/Bureau/usr/lib/:$LD_LIBRARY_PATH
714 -- Build files have been written to: /home/navarrop/Bureau/simgrid-trunk
717 \subsection faq_cmakeinstall How to install with cmake?
719 \subsubsection faq_cmakeinstall1 From svn.
722 cmake -Denable_maintainer_mode=on -Dprefix=/home/navarrop/Bureau/install_simgrid ./
727 \subsubsection faq_cmakeinstall2 From a distrib
730 cmake -Dprefix=/home/navarrop/Bureau/install_simgrid ./
735 \subsection faq_screenshot Screenshot
738 navarrop@caraja:~$ cd Bureau/simgrid-trunk/
739 navarrop@caraja:~/Bureau/simgrid-trunk$ cmake ./
741 GTnetS doesn't works : set -Ddisable_gtnets=on <-|some warnings are printed
742 with_context auto change to ucontext <-|
743 (skaddr) <--info (needed)
744 (sksize) <--info (needed)
745 Make : src/simgrid.jar with : /usr/bin/javac <--info (if java)
746 Make examples/java with : /usr/bin/javac <--info (if java)
748 Configuration of package `simgrid' (revision 7209M) on arch (=4):
749 BUILDNAME : UCONTEXT <-- name of the compilation regarding to cdash
750 SITE : Linux_Ubuntu 9.10_x86_64 <-- distribution of the local machine regarding to cdash
752 Compiler: c++ : /usr/bin/c++
753 version: c++ (Ubuntu 4.4.1-4ubuntu9) 4.4.1
755 version: gcc (Ubuntu 4.4.1-4ubuntu9) 4.4.1
761 Context backend: ucontext
771 Simgrid dependencies: -ldl -llua5.1
772 Gras dependencies: pthread
776 USER_PREFIX: /usr/local
777 INSTALL_PREFIX: /usr/local
781 -- Build files have been written to: /home/navarrop/Bureau/simgrid-trunk
783 Here all options are checked and printed. If it doesn't match with your configuration
784 it is probably due to a wrong configuration.
786 \subsection faq_cmakehowto How to modified sources files for developers
788 \subsubsection faq_cmakehowto1 Add an executable or examples.
790 If you want make an executable you have to create a CMakeList.txt to the src directory.
791 You must specified where to create the executable, source list, dependencies and the name of the binary.
794 cmake_minimum_required(VERSION 2.6)
796 set(EXECUTABLE_OUTPUT_PATH "./")
797 set(LIBRARY_OUTPUT_PATH "${PROJECT_DIRECTORY}/lib")
799 add_executable(get_sender get_sender.c) #add_executable(<name_of_target> <src list>)
801 ### Add definitions for compile
802 target_link_libraries(get_sender simgrid m pthread -fprofile-arcs) #target_link_libraries(<name_of_targe> <dependencies>)
805 Then you have to modified <project/directory>/buildtools/Cmake/src/CMakeMakeExeLib.txt and add
808 add_subdirectory(${PROJECT_DIRECTORY}/<path_where_is_CMakeList.txt>)
811 \subsubsection faq_cmakehowto2 Delete/add sources to lib.
813 If you want modified, add or delete source files from a library you have to edit <project/directory>/buildtools/Cmake/src/CMakeDefinePackages.txt
817 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/MsgException.java
818 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/JniException.java
819 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/NativeException.java
820 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/HostNotFoundException.java
821 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/ProcessNotFoundException.java
822 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/Msg.java
823 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/Process.java
824 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/Host.java
825 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/Task.java
826 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/MsgNative.java
827 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/ApplicationHandler.java
828 ${PROJECT_DIRECTORY}/src/java/simgrid/msg/Sem.java
832 \subsubsection faq_cmakehowto3 Add test
834 If you want modified, add or delete tests you have to edit <project/directory>/buildtools/Cmake/src/CMakeTest.txt
835 with this function : ADD_TEST(<name> <bin> <ARGS>)
838 add_test(test-simdag-1 ${PROJECT_DIRECTORY}/testsuite/simdag/sd_test --cfg=path:${PROJECT_DIRECTORY}/testsuite/simdag small_platform_variable.xml)
841 \subsection faq_cmakeExplain Explaination of sources files for cmake
845 Those files are the "main parts". One located at the project directory call all the cmake sources files. The others
846 are little projects called by the first for make examples.
848 \li CMakeCompleteInFiles.txt
850 Complete all .in files and define Variables for h files
854 This file make the html documentation.
856 \li CMakeMakeExeLib.txt
858 Here are callled all "CMakeLists.txt" for make executables and libraries.
860 \li CMakePrintArgs.txt
862 This file is called at the end of the build for summarize environment variables.
864 \li CMakeDefinePackages.txt
866 Here is defined sources packages for compiling libs.
870 Defined flags which are used for compiling sources.
872 \li CMakeSupernovae.txt
874 Here are made files for the supernovae mode.
878 Here is defined packages for install simgrid and make a distribution.
880 \li CMakeMaintainerMode.txt
882 Part where are generated source files for maintainer mode.
886 Here are defined options and initialized values.
890 All tests are listed.
892 \li CTestConfig.cmake
894 Properties which link tests with dashboard.
896 \subsection faq_cmakeList List of files added for cmake
898 Here is a list of files involved into cmake build (relative to project directory path) :
903 ./buildtools/Cmake/src/CMakeCompleteInFiles.txt
904 ./buildtools/Cmake/src/CMakeDocs.txt
905 ./buildtools/Cmake/src/CMakeMakeExeLib.txt
906 ./buildtools/Cmake/src/CMakePrintArgs.txt
907 ./buildtools/Cmake/src/CMakeDefinePackages.txt
908 ./buildtools/Cmake/src/CMakeFlags.txt
909 ./buildtools/Cmake/src/CMakeSupernovae.txt
910 ./buildtools/Cmake/src/CMakeDistrib.txt
911 ./buildtools/Cmake/src/CMakeMaintainerMode.txt
912 ./buildtools/Cmake/src/CMakeOption.txt
913 ./buildtools/Cmake/src/CMakeTest.txt
914 ./buildtools/Cmake/src/CTestConfig.cmake
916 Test files for define properties :
917 ./buildtools/Cmake/prog_test/prog_GRAS_ARCH.c
918 ./buildtools/Cmake/prog_test/prog_max_size.c
919 ./buildtools/Cmake/prog_test/prog_sem_init.c
920 ./buildtools/Cmake/prog_test/prog_stackgrowth.c
921 ./buildtools/Cmake/prog_test/prog_vsnprintf.c
922 ./buildtools/Cmake/prog_test/prog_AC_CHECK_MCSC.c
923 ./buildtools/Cmake/prog_test/prog_GRAS_CHECK_STRUCT_COMPACTION.c
924 ./buildtools/Cmake/prog_test/prog_mutex_timedlock.c
925 ./buildtools/Cmake/prog_test/prog_sem_timedwait.c
926 ./buildtools/Cmake/prog_test/prog_stacksetup.c
927 ./buildtools/Cmake/prog_test/prog_getline.c
928 ./buildtools/Cmake/prog_test/prog_gtnets.cpp
929 ./buildtools/Cmake/prog_test/prog_printf_null.c
930 ./buildtools/Cmake/prog_test/prog_snprintf.c
931 ./buildtools/Cmake/prog_test/prog_va_copy.c
933 CMakeLists for each binaries or examples:
936 ./teshsuite/gras/empty_main/CMakeLists.txt
937 ./teshsuite/gras/small_sleep/CMakeLists.txt
938 ./teshsuite/gras/datadesc/CMakeLists.txt
939 ./teshsuite/gras/msg_handle/CMakeLists.txt
940 ./teshsuite/simdag/CMakeLists.txt
941 ./teshsuite/simdag/partask/CMakeLists.txt
942 ./teshsuite/simdag/platforms/CMakeLists.txt
943 ./teshsuite/simdag/network/CMakeLists.txt
944 ./teshsuite/simdag/network/mxn/CMakeLists.txt
945 ./teshsuite/simdag/network/p2p/CMakeLists.txt
946 ./teshsuite/xbt/CMakeLists.txt
947 ./teshsuite/msg/CMakeLists.txt
948 ./tools/gras/CMakeLists.txt
949 ./tools/tesh/CMakeLists.txt
950 ./testsuite/simdag/CMakeLists.txt
951 ./testsuite/xbt/CMakeLists.txt
952 ./testsuite/surf/CMakeLists.txt
953 ./examples/gras/properties/CMakeLists.txt
954 ./examples/gras/ping/CMakeLists.txt
955 ./examples/gras/pmm/CMakeLists.txt
956 ./examples/gras/mmrpc/CMakeLists.txt
957 ./examples/gras/synchro/CMakeLists.txt
958 ./examples/gras/timer/CMakeLists.txt
959 ./examples/gras/mutual_exclusion/simple_token/CMakeLists.txt
960 ./examples/gras/spawn/CMakeLists.txt
961 ./examples/gras/chrono/CMakeLists.txt
962 ./examples/gras/rpc/CMakeLists.txt
963 ./examples/gras/all2all/CMakeLists.txt
964 ./examples/simdag/properties/CMakeLists.txt
965 ./examples/simdag/CMakeLists.txt
966 ./examples/simdag/metaxml/CMakeLists.txt
967 ./examples/simdag/dax/CMakeLists.txt
968 ./examples/smpi/CMakeLists.txt
969 ./examples/amok/bandwidth/CMakeLists.txt
970 ./examples/amok/saturate/CMakeLists.txt
971 ./examples/msg/priority/CMakeLists.txt
972 ./examples/msg/properties/CMakeLists.txt
973 ./examples/msg/migration/CMakeLists.txt
974 ./examples/msg/gtnets/CMakeLists.txt
975 ./examples/msg/parallel_task/CMakeLists.txt
976 ./examples/msg/trace/CMakeLists.txt
977 ./examples/msg/suspend/CMakeLists.txt
978 ./examples/msg/masterslave/CMakeLists.txt
979 ./examples/msg/actions/CMakeLists.txt
980 ./examples/msg/sendrecv/CMakeLists.txt
983 \section faq_howto Feature related questions
985 \subsection faq_MIA "Could you please add (your favorite feature here) to SimGrid?"
987 Here is the deal. The whole SimGrid project (MSG, SURF, GRAS, ...) is
988 meant to be kept as simple and generic as possible. We cannot add
989 functions for everybody's needs when these functions can easily be
990 built from the ones already in the API. Most of the time, it is
991 possible and when it was not possible we always have upgraded the API
992 accordingly. When somebody asks us a question like "How to do that?
993 Is there a function in the API to simply do this?", we're always glad
994 to answer and help. However if we don't need this code for our own
995 need, there is no chance we're going to write it... it's your job! :)
996 The counterpart to our answers is that once you come up with a neat
997 implementation of this feature (task duplication, RPC, thread
998 synchronization, ...), you should send it to us and we will be glad to
999 add it to the distribution. Thus, other people will take advantage of
1000 it (and we don't have to answer this question again and again ;).
1002 You'll find in this section a few "Missing In Action" features. Many
1003 people have asked about it and we have given hints on how to simply do
1004 it with MSG. Feel free to contribute...
1006 \subsection faq_MIA_MSG MSG features
1008 \subsubsection faq_MIA_examples I want some more complex MSG examples!
1010 Many people have come to ask me a more complex example and each time,
1011 they have realized afterward that the basics were in the previous three
1014 Of course they have often been needing more complex functions like
1015 MSG_process_suspend(), MSG_process_resume() and
1016 MSG_process_isSuspended() (to perform synchronization), or
1017 MSG_task_Iprobe() and MSG_process_sleep() (to avoid blocking
1018 receptions), or even MSG_process_create() (to design asynchronous
1019 communications or computations). But the examples are sufficient to
1022 We know. We should add some more examples, but not really some more
1023 complex ones... We should add some examples that illustrate some other
1024 functionalists (like how to simply encode asynchronous
1025 communications, RPC, process migrations, thread synchronization, ...)
1026 and we will do it when we will have a little bit more time. We have
1027 tried to document the examples so that they are understandable. Tell
1028 us if something is not clear and once again feel free to participate!
1031 \subsubsection faq_MIA_taskdup Missing in action: MSG Task duplication/replication
1033 There is no task duplication in MSG. When you create a task, you can
1034 process it or send it somewhere else. As soon as a process has sent
1035 this task, he doesn't have this task anymore. It's gone. The receiver
1036 process has got the task. However, you could decide upon receiving to
1037 create a "copy" of a task but you have to handle by yourself the
1038 semantic associated to this "duplication".
1040 As we already told, we prefer keeping the API as simple as
1041 possible. This kind of feature is rather easy to implement by users
1042 and the semantic you associate really depends on people. Having a
1043 *generic* task duplication mechanism is not that trivial (in
1044 particular because of the data field). That is why I would recommand
1045 that you write it by yourself even if I can give you advice on how to
1048 You have the following functions to get informations about a task:
1049 MSG_task_get_name(), MSG_task_get_compute_duration(),
1050 MSG_task_get_remaining_computation(), MSG_task_get_data_size(),
1051 and MSG_task_get_data().
1053 You could use a dictionary (#xbt_dict_t) of dynars (#xbt_dynar_t). If
1054 you still don't see how to do it, please come back to us...
1056 \subsubsection faq_MIA_asynchronous I want to do asynchronous communications in MSG
1058 Up until now, there is no asynchronous communications in MSG. However,
1059 you can create as many process as you want so you should be able to do
1060 whatever you want... I've written a queue module to help implementing
1061 some asynchronous communications at low cost (creating thousands of
1062 process only to handle communications may be problematic in term of
1063 performance at some point). I'll add it in the distribution asap.
1065 \subsubsection faq_MIA_thread_synchronization I need to synchronize my MSG processes
1067 You obviously cannot use pthread_mutexes of pthread_conds. The best
1068 thing would be to propose similar structures. Unfortunately, we
1069 haven't found time to do it yet. However you can try to play with
1070 MSG_process_suspend() and MSG_process_resume(). You can even do some
1071 synchronization with fake communications (using MSG_task_get(),
1072 MSG_task_put() and MSG_task_Iprobe()).
1074 \subsubsection faq_MIA_host_load Where is the get_host_load function hidden in MSG?
1076 There is no such thing because its semantic wouldn't be really
1077 clear. Of course, it is something about the amount of host throughput,
1078 but there is as many definition of "host load" as people asking for
1079 this function. First, you have to remember that resource availability
1080 may vary over time, which make any load notion harder to define.
1082 It may be instantaneous value or an average one. Moreover it may be only the
1083 power of the computer, or may take the background load into account, or may
1084 even take the currently running tasks into account. In some SURF models,
1085 communications have an influence on computational power. Should it be taken
1088 First of all, it's near to impossible to predict the load beforehands in the
1089 simulator since it depends on too much parameters (background load
1090 variation, bandwidth sharing algorithmic complexity) some of them even being
1091 not known beforehands (other task starting at the same time). So, getting
1092 this information is really hard (just like in real life). It's not just that
1093 we want MSG to be as painful as real life. But as it is in some way
1094 realistic, we face some of the same problems as we would face in real life.
1096 How would you do it for real? The most common option is to use something
1097 like NWS that performs active probes. The best solution is probably to do
1098 the same within MSG, as in next code snippet. It is very close from what you
1099 would have to do out of the simulator, and thus gives you information that
1100 you could also get in real settings to not hinder the realism of your
1104 double get_host_load() {
1105 m_task_t task = MSG_task_create("test", 0.001, 0, NULL);
1106 double date = MSG_get_clock();
1108 MSG_task_execute(task);
1109 date = MSG_get_clock() - date;
1110 MSG_task_destroy(task);
1111 return (0.001/date);
1115 Of course, it may not match your personal definition of "host load". In this
1116 case, please detail what you mean on the mailing list, and we will extend
1117 this FAQ section to fit your taste if possible.
1119 \subsubsection faq_MIA_communication_time How can I get the *real* communication time?
1121 Communications are synchronous and thus if you simply get the time
1122 before and after a communication, you'll only get the transmission
1123 time and the time spent to really communicate (it will also take into
1124 account the time spent waiting for the other party to be
1125 ready). However, getting the *real* communication time is not really
1126 hard either. The following solution is a good starting point.
1131 m_task_t task = MSG_task_create("Task", task_comp_size, task_comm_size,
1132 calloc(1,sizeof(double)));
1133 *((double*) task->data) = MSG_get_clock();
1134 MSG_task_put(task, slaves[i % slaves_count], PORT_22);
1135 INFO0("Send completed");
1140 m_task_t task = NULL;
1143 time1 = MSG_get_clock();
1144 a = MSG_task_get(&(task), PORT_22);
1145 time2 = MSG_get_clock();
1146 if(time1<*((double *)task->data))
1147 time1 = *((double *) task->data);
1148 INFO1("Communication time : \"%f\" ", time2-time1);
1150 MSG_task_destroy(task);
1155 \subsection faq_MIA_SimDag SimDag related questions
1157 \subsubsection faq_SG_comm Implementing communication delays between tasks.
1159 A classic question of SimDag newcomers is about how to express a
1160 communication delay between tasks. The thing is that in SimDag, both
1161 computation and communication are seen as tasks. So, if you want to
1162 model a data dependency between two DAG tasks t1 and t2, you have to
1163 create 3 SD_tasks: t1, t2 and c and add dependencies in the following
1167 SD_task_dependency_add(NULL, NULL, t1, c);
1168 SD_task_dependency_add(NULL, NULL, c, t2);
1171 This way task t2 cannot start before the termination of communication c
1172 which in turn cannot start before t1 ends.
1174 When creating task c, you have to associate an amount of data (in bytes)
1175 corresponding to what has to be sent by t1 to t2.
1177 Finally to schedule the communication task c, you have to build a list
1178 comprising the workstations on which t1 and t2 are scheduled (w1 and w2
1179 for example) and build a communication matrix that should look like
1182 \subsubsection faq_SG_DAG How to implement a distributed dynamic scheduler of DAGs.
1184 Distributed is somehow "contagious". If you start making distributed
1185 decisions, there is no way to handle DAGs directly anymore (unless I
1186 am missing something). You have to encode your DAGs in term of
1187 communicating process to make the whole scheduling process
1188 distributed. Here is an example of how you could do that. Assume T1
1189 has to be done before T2.
1192 int your_agent(int argc, char *argv[] {
1194 T1 = MSG_task_create(...);
1195 T2 = MSG_task_create(...);
1199 if(cond) MSG_task_execute(T1);
1201 if((MSG_task_get_remaining_computation(T1)=0.0) && (you_re_in_a_good_mood))
1202 MSG_task_execute(T2)
1204 /* do something else */
1210 If you decide that the distributed part is not that much important and that
1211 DAG is really the level of abstraction you want to work with, then you should
1212 give a try to \ref SD_API.
1214 \subsection faq_MIA_generic Generic features
1216 \subsubsection faq_more_processes Increasing the amount of simulated processes
1218 Here are a few tricks you can apply if you want to increase the amount
1219 of processes in your simulations.
1221 - <b>A few thousands of simulated processes</b> (soft tricks)\n
1222 SimGrid can use either pthreads library or the UNIX98 contextes. On
1223 most systems, the number of pthreads is limited and then your
1224 simulation may be limited for a stupid reason. This is especially
1225 true with the current linux pthreads, and I cannot get more than
1226 2000 simulated processes with pthreads on my box. The UNIX98
1227 contexts allow me to raise the limit to 25,000 simulated processes
1229 The <tt>--with-context</tt> option of the <tt>./configure</tt>
1230 script allows you to choose between UNIX98 contextes
1231 (<tt>--with-context=ucontext</tt>) and the pthread version
1232 (<tt>--with-context=pthread</tt>). The default value is ucontext
1233 when the script detect a working UNIX98 context implementation. On
1234 Windows boxes, the provided value is discarded and an adapted
1235 version is picked up.\n\n
1236 We experienced some issues with contextes on some rare systems
1237 (solaris 8 and lower or old alpha linuxes comes to mind). The main
1238 problem is that the configure script detect the contextes as being
1239 functional when it's not true. If you happen to use such a system,
1240 switch manually to the pthread version, and provide us with a good
1241 patch for the configure script so that it is done automatically ;)
1243 - <b>Hundred thousands of simulated processes</b> (hard-core tricks)\n
1244 As explained above, SimGrid can use UNIX98 contextes to represent
1245 and handle the simulated processes. Thanks to this, the main
1246 limitation to the number of simulated processes becomes the
1247 available memory.\n\n
1248 Here are some tricks I had to use in order to run a token ring
1249 between 25,000 processes on my laptop (1Gb memory, 1.5Gb swap).\n
1250 - First of all, make sure your code runs for a few hundreds
1251 processes before trying to push the limit. Make sure it's
1252 valgrind-clean, ie that valgrind does not report neither memory
1253 error nor memory leaks. Indeed, numerous simulated processes
1254 result in *fat* simulation hindering debugging.
1255 - It was really boring to write 25,000 entries in the deployment
1256 file, so I wrote a little script
1257 <tt>examples/gras/mutual_exclusion/simple_token/make_deployment.pl</tt>, which you may
1258 want to adapt to your case. You could also think about hijacking
1259 the SURFXML parser (have look at \ref faq_flexml_bypassing).
1260 - The deployment file became quite big, so I had to do what is in
1261 the FAQ entry \ref faq_flexml_limit
1262 - Each UNIX98 context has its own stack entry. As debugging this is
1263 quite hairly, the default value is a bit overestimated so that
1264 user don't get into trouble about this. You want to tune this
1265 size to increse the number of processes. This is the
1266 <tt>STACK_SIZE</tt> define in
1267 <tt>src/xbt/xbt_context_sysv.c</tt>, which is 128kb by default.
1268 Reduce this as much as you can, but be warned that if this value
1269 is too low, you'll get a segfault. The token ring example, which
1270 is quite simple, runs with 40kb stacks.
1271 - You may tweak the logs to reduce the stack size further. When
1272 logging something, we try to build the string to display in a
1273 char array on the stack. The size of this array is constant (and
1274 equal to XBT_LOG_BUFF_SIZE, defined in include/xbt/log/h). If the
1275 string is too large to fit this buffer, we move to a dynamically
1276 sized buffer. In which case, we have to traverse one time the log
1277 event arguments to compute the size we need for the buffer,
1278 malloc it, and traverse the argument list again to do the actual
1280 The idea here is to move XBT_LOG_BUFF_SIZE to 1, forcing the logs
1281 to use a dynamic array each time. This allows us to lower further
1282 the stack size at the price of some performance loss...\n
1283 This allowed me to run the reduce the stack size to ... 4k. Ie,
1284 on my 1Gb laptop, I can run more than 250,000 processes!
1286 \subsubsection faq_MIA_batch_scheduler Is there a native support for batch schedulers in SimGrid?
1288 No, there is no native support for batch schedulers and none is
1289 planned because this is a very specific need (and doing it in a
1290 generic way is thus very hard). However some people have implemented
1291 their own batch schedulers. Vincent Garonne wrote one during his PhD
1292 and put his code in the contrib directory of our SVN so that other can
1293 keep working on it. You may find inspiring ideas in it.
1295 \subsubsection faq_MIA_checkpointing I need a checkpointing thing
1297 Actually, it depends on whether you want to checkpoint the simulation, or to
1298 simulate checkpoints.
1300 The first one could help if your simulation is a long standing process you
1301 want to keep running even on hardware issues. It could also help to
1302 <i>rewind</i> the simulation by jumping sometimes on an old checkpoint to
1303 cancel recent calculations.\n
1304 Unfortunately, such thing will probably never exist in SG. One would have to
1305 duplicate all data structures because doing a rewind at the simulator level
1306 is very very hard (not talking about the malloc free operations that might
1307 have been done in between). Instead, you may be interested in the Libckpt
1308 library (http://www.cs.utk.edu/~plank/plank/www/libckpt.html). This is the
1309 checkpointing solution used in the condor project, for example. It makes it
1310 easy to create checkpoints (at the OS level, creating something like core
1311 files), and rerunning them on need.
1313 If you want to simulate checkpoints instead, it means that you want the
1314 state of an executing task (in particular, the progress made towards
1315 completion) to be saved somewhere. So if a host (and the task executing on
1316 it) fails (cf. #MSG_HOST_FAILURE), then the task can be restarted
1317 from the last checkpoint.\n
1319 Actually, such a thing does not exists in SimGrid either, but it's just
1320 because we don't think it is fundamental and it may be done in the user code
1321 at relatively low cost. You could for example use a watcher that
1322 periodically get the remaining amount of things to do (using
1323 MSG_task_get_remaining_computation()), or fragment the task in smaller
1326 \subsection faq_platform Platform building and Dynamic resources
1328 \subsubsection faq_platform_example Where can I find SimGrid platform files?
1330 There is several little examples in the archive, in the examples/msg
1331 directory. From time to time, we are asked for other files, but we
1332 don't have much at hand right now.
1334 You should refer to the Platform Description Archive
1335 (http://pda.gforge.inria.fr) project to see the other platform file we
1336 have available, as well as the Simulacrum simulator, meant to generate
1337 SimGrid platforms using all classical generation algorithms.
1339 \subsubsection faq_platform_alnem How can I automatically map an existing platform?
1341 We are working on a project called ALNeM (Application-Level Network
1342 Mapper) which goal is to automatically discover the topology of an
1343 existing network. Its output will be a platform description file
1344 following the SimGrid syntax, so everybody will get the ability to map
1345 their own lab network (and contribute them to the catalog project).
1346 This tool is not ready yet, but it move quite fast forward. Just stay
1349 \subsubsection faq_platform_synthetic Generating synthetic but realistic platforms
1351 The third possibility to get a platform file (after manual or
1352 automatic mapping of real platforms) is to generate synthetic
1353 platforms. Getting a realistic result is not a trivial task, and
1354 moreover, nobody is really able to define what "realistic" means when
1355 speaking of topology files. You can find some more thoughts on this
1357 <a href="http://graal.ens-lyon.fr/~alegrand/articles/Simgrid-Introduction.pdf">slides</a>.
1359 If you are looking for an actual tool, there we have a little tool to
1360 annotate Tiers-generated topologies. This perl-script is in
1361 <tt>tools/platform_generation/</tt> directory of the SVN. Dinda et Al.
1362 released a very comparable tool, and called it GridG.
1364 \subsubsection faq_SURF_dynamic Expressing dynamic resource availability in platform files
1366 A nice feature of SimGrid is that it enables you to seamlessly have
1367 resources whose availability change over time. When you build a
1368 platform, you generally declare hosts like that:
1371 <host id="host A" power="100.00"/>
1374 If you want the availability of "host A" to change over time, the only
1375 thing you have to do is change this definition like that:
1378 <host id="host A" power="100.00" availability_file="trace_A.txt" state_file="trace_A_failure.txt"/>
1381 For hosts, availability files are expressed in fraction of available
1382 power. Let's have a look at what "trace_A.txt" may look like:
1391 At time 0, our host will deliver 100 flop/s. At time 11.0, it will
1392 deliver only 50 flop/s until time 20.0 where it will will start
1393 delivering 90 flop/s. Last at time 21.0 (20.0 plus the periodicity
1394 1.0), we'll be back to the beginning and it will deliver 100 flop/s.
1396 Now let's look at the state file:
1403 A negative value means "off" while a positive one means "on". At time
1404 1.0, the host is on. At time 1.0, it is turned off and at time 2.0, it
1405 is turned on again until time 12 (2.0 plus the periodicity 10.0). It
1406 will be turned on again at time 13.0 until time 23.0, and so on.
1408 Now, let's look how the same kind of thing can be done for network
1409 links. A usual declaration looks like:
1412 <link id="LinkA" bandwidth="10.0" latency="0.2"/>
1415 You have at your disposal the following options: bandwidth_file,
1416 latency_file and state_file. The only difference with hosts is that
1417 bandwidth_file and latency_file do not express fraction of available
1418 power but are expressed directly in bytes per seconds and seconds.
1420 \subsubsection faq_platform_multipath How to express multipath routing in platform files?
1422 It is unfortunately impossible to express the fact that there is more
1423 than one routing path between two given hosts. Let's consider the
1424 following platform file:
1427 <route src="A" dst="B">
1430 <route src="B" dst="C">
1433 <route src="A" dst="C">
1438 Although it is perfectly valid, it does not mean that data traveling
1439 from A to C can either go directly (using link 3) or through B (using
1440 links 1 and 2). It simply means that the routing on the graph is not
1441 trivial, and that data do not following the shortest path in number of
1442 hops on this graph. Another way to say it is that there is no implicit
1443 in these routing descriptions. The system will only use the routes you
1444 declare (such as <route src="A" dst="C"><link:ctn
1445 id="3"/></route>), without trying to build new routes by aggregating
1448 You are also free to declare platform where the routing is not
1449 symmetric. For example, add the following to the previous file:
1452 <route src="C" dst="A">
1458 This makes sure that data from C to A go through B where data from A
1459 to C go directly. Don't worry about realism of such settings since
1460 we've seen ways more weird situation in real settings (in fact, that's
1461 the realism of very regular platforms which is questionable, but
1462 that's another story).
1464 \subsubsection faq_flexml_bypassing Bypassing the XML parser with your own C functions
1466 So you want to bypass the XML files parser, uh? Maybe doing some parameter
1467 sweep experiments on your simulations or so? This is possible, and
1468 it's not even really difficult (well. Such a brutal idea could be
1469 harder to implement). Here is how it goes.
1471 For this, you have to first remember that the XML parsing in SimGrid is done
1472 using a tool called FleXML. Given a DTD, this gives a flex-based parser. If
1473 you want to bypass the parser, you need to provide some code mimicking what
1474 it does and replacing it in its interactions with the SURF code. So, let's
1475 have a look at these interactions.
1477 FleXML parser are close to classical SAX parsers. It means that a
1478 well-formed SimGrid platform XML file might result in the following
1481 - start "platform_description" with attribute version="2"
1482 - start "host" with attributes id="host1" power="1.0"
1484 - start "host" with attributes id="host2" power="2.0"
1486 - start "link" with ...
1488 - start "route" with ...
1489 - start "link:ctn" with ...
1492 - end "platform_description"
1494 The communication from the parser to the SURF code uses two means:
1495 Attributes get copied into some global variables, and a surf-provided
1496 function gets called by the parser for each event. For example, the event
1497 - start "host" with attributes id="host1" power="1.0"
1499 let the parser do something roughly equivalent to:
1501 strcpy(A_host_id,"host1");
1506 In SURF, we attach callbacks to the different events by initializing the
1507 pointer functions to some the right surf functions. Since there can be
1508 more than one callback attached to the same event (if more than one
1509 model is in use, for example), they are stored in a dynar. Example in
1510 workstation_ptask_L07.c:
1512 /* Adding callback functions */
1513 surf_parse_reset_parser();
1514 surfxml_add_callback(STag_surfxml_host_cb_list, &parse_cpu_init);
1515 surfxml_add_callback(STag_surfxml_prop_cb_list, &parse_properties);
1516 surfxml_add_callback(STag_surfxml_link_cb_list, &parse_link_init);
1517 surfxml_add_callback(STag_surfxml_route_cb_list, &parse_route_set_endpoints);
1518 surfxml_add_callback(ETag_surfxml_link_c_ctn_cb_list, &parse_route_elem);
1519 surfxml_add_callback(ETag_surfxml_route_cb_list, &parse_route_set_route);
1521 /* Parse the file */
1522 surf_parse_open(file);
1523 xbt_assert1((!surf_parse()), "Parse error in %s", file);
1527 So, to bypass the FleXML parser, you need to write your own version of the
1528 surf_parse function, which should do the following:
1529 - Fill the A_<tag>_<attribute> variables with the wanted values
1530 - Call the corresponding STag_<tag>_fun function to simulate tag start
1531 - Call the corresponding ETag_<tag>_fun function to simulate tag end
1532 - (do the same for the next set of values, and loop)
1534 Then, tell SimGrid that you want to use your own "parser" instead of the stock one:
1536 surf_parse = surf_parse_bypass_environment;
1537 MSG_create_environment(NULL);
1538 surf_parse = surf_parse_bypass_application;
1539 MSG_launch_application(NULL);
1542 A set of macros are provided at the end of
1543 include/surf/surfxml_parse.h to ease the writing of the bypass
1544 functions. An example of this trick is distributed in the file
1545 examples/msg/masterslave/masterslave_bypass.c
1547 \subsection faq_simgrid_configuration Changing SimGrid's behavior
1549 A number of options can be given at runtime to change the default
1550 SimGrid behavior. In particular, you can change the default cpu and
1553 \subsubsection faq_simgrid_configuration_gtnets Using GTNetS
1555 It is possible to use a packet-level network simulator
1556 instead of the default flow-based simulation. You may want to use such
1557 an approach if you have doubts about the validity of the default model
1558 or if you want to perform some validation experiments. At the moment,
1559 we support the GTNetS simulator (it is still rather experimental
1560 though, so leave us a message if you play with it).
1564 To enable GTNetS model inside SimGrid it is needed to patch the GTNetS simulator source code
1565 and build/install it from scratch
1568 - <b>Download and enter the recent downloaded GTNetS directory</b>
1571 svn checkout svn://scm.gforge.inria.fr/svn/simgrid/contrib/trunk/GTNetS/
1576 - <b>Use the following commands to unzip and patch GTNetS package to work within SimGrid.</b>
1579 unzip gtnets-current.zip
1580 tar zxvf gtnets-current-patch.tgz
1582 cat ../00*.patch | patch -p1
1585 - <b>OPTIONALLY</b> you can use a patch for itanium 64bit processor family.
1588 cat ../AMD64-FATAL-Removed-DUL_SIZE_DIFF-Added-fPIC-compillin.patch | patch -p1
1591 - <b>Compile GTNetS</b>
1593 Due to portability issues it is possible that GTNetS does not compile in your architecture. The patches furnished in SimGrid SVN repository are intended for use in Linux architecture only. Unfortunately, we do not have the time, the money, neither the manpower to guarantee GTNetS portability. We advice you to use one of GTNetS communication channel to get more help in compiling GTNetS.
1597 ln -sf Makefile.linux Makefile
1603 - <b>NOTE</b> A lot of warnings are expected but the application should compile
1604 just fine. If the makefile insists in compiling some QT libraries
1605 please try a make clean before asking for help.
1608 - <b>To compile optimized version</b>
1615 - <b>Installing GTNetS</b>
1617 It is important to put the full path of your libgtsim-xxxx.so file when creating the symbolic link. Replace < userhome > by some path you have write access to.
1620 ln -sf /<absolute_path>/gtnets_current/libgtsim-debug.so /<userhome>/usr/lib/libgtnets.so
1621 export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/<userhome>/usr/lib/libgtnets.so
1622 mkdir /<userhome>/usr/include/gtnets
1623 cp -fr SRC/*.h /<userhome>/usr/include/gtnets
1627 - <b>Enable GTNetS support in SimGrid</b>
1630 ./configure --with-gtnets=/<userhome>/usr
1633 - <b>Once you have followed all the instructions for compiling and
1634 installing successfully you can activate this feature at
1635 runntime with the following options:</b>
1638 cd simgrid/example/msg/
1644 - <b>Or try the GTNetS model dogbone example with</b>
1647 gtnets/gtnets gtnets/onelink-p.xml gtnets/onelink-d.xml --cfg=network_model:GTNets
1651 A long version of this <a href="http://gforge.inria.fr/docman/view.php/12/6283/GTNetS HowTo.html">HowTo</a> it is available
1654 More about GTNetS simulator at <a href="http://www.ece.gatech.edu/research/labs/MANIACS/GTNetS/index.html">GTNetS Website</a>
1658 The patches provided by us worked successfully with GTNetS found
1659 <a href="http://www.ece.gatech.edu/research/labs/MANIACS/GTNetS/software/gtnets-current.zip">here</a>,
1660 dated from 12th June 2008. Due to the discontinuing development of
1661 GTNetS it is impossible to precise a version number. We STRONGLY recommend you
1662 to download and install the GTNetS version found in SimGrid repository as explained above.
1667 \subsubsection faq_simgrid_configuration_alternate_network Using alternative flow models
1669 The default simgrid network model uses a max-min based approach as
1670 explained in the research report
1671 <a href="ftp://ftp.ens-lyon.fr/pub/LIP/Rapports/RR/RR2002/RR2002-40.ps.gz">A Network Model for Simulation of Grid Application</a>.
1672 Other models have been proposed and implemented since then (see for example
1673 <a href="http://mescal.imag.fr/membres/arnaud.legrand/articles/simutools09.pdf">Accuracy Study and Improvement of Network Simulation in the SimGrid Framework</a>)
1674 and can be activated at runtime. For example:
1676 ./mycode platform.xml deployment.xml --cfg=workstation_model:compound --cfg=network_model:LV08 -cfg=cpu_model:Cas01
1679 Possible models for the network are currently "Constant", "CM02",
1680 "LegrandVelho", "GTNets", Reno", "Reno2", "Vegas". Others will
1681 probably be added in the future and many of the previous ones are
1682 experimental and are likely to disappear without notice...
1684 \subsection faq_tracing Tracing Simulations for Visualization
1686 The trace visualization is widely used to observe and understand the behavior
1687 of parallel applications and distributed algorithms. Usually, this is done in a
1688 two-step fashion: the user instruments the application and the traces are
1689 analyzed after the end of the execution. The visualization itself can highlights
1690 unexpected behaviors, bottlenecks and sometimes can be used to correct
1691 distributed algorithms. The SimGrid team is currently instrumenting the library
1692 in order to let users trace their simulations and analyze them. This part of the
1693 user manual explains how the tracing-related features can be enabled and used
1694 during the development of simulators using the SimGrid library.
1696 \subsubsection faq_tracing_howitworks How it works
1698 For now, the SimGrid library is instrumented so users can trace the <b>platform
1699 utilization</b> using the MSG interface. This means that the tracing will
1700 register how much power is used for each host and how much bandwidth is used for
1701 each link of the platform. The idea with this type of tracing is to observe the
1702 overall view of resources utilization in the first place, especially the
1703 identification of bottlenecks, load-balancing among hosts, and so on.
1705 The idea of the instrumentation is to classify the MSG tasks by category,
1707 the platform utilization (hosts and links) for each of the categories. For that,
1708 the tracing interface enables the declaration of categories and a function to
1709 mark a task with a previously declared category. <em>The tasks that are not
1710 classified according to a category are not traced</em>.
1712 \subsubsection faq_tracing_enabling Enabling using CMake
1714 With the sources of SimGrid, it is possible to enable the tracing
1715 using the parameter <b>-Dtracing=on</b> when the cmake is executed.
1716 The section \ref faq_tracing_functions describes all the functions available
1717 when this Cmake options is activated. These functions will have no effect
1718 if SimGrid is configured without this option (they are wiped-out by the
1722 $ cmake -Dtracing=on .
1726 \subsubsection faq_tracing_functions Tracing Functions
1728 \subsubsubsection Mandatory Functions
1730 \li <b>\c TRACE_start (const char *filename)</b>: This is the first function to
1731 be called. It receives a single argument as parameter that contains the name of
1732 the file that will hold the trace in the end of the simulation. It returns 0 if
1733 everything was properly initialized, 1 otherwise. All trace functions called
1734 before TRACE_start do nothing.
1736 \li <b>\c TRACE_category (const char *category)</b>: This function should be used
1737 to define a user category. The category can be used to differentiate the tasks
1738 that are created during the simulation (for example, tasks from server1,
1739 server2, or request tasks, computation tasks, communication tasks).
1740 All resource utilization (host power and link bandwidth) will be
1741 classified according to the task category. Tasks that do not belong to a
1742 category are not traced.
1744 \li <b>\c TRACE_msg_set_task_category (m_task_t task, const char *category)</b>:
1745 This function should be called after the creation of a task, to define the
1746 category of that task. The first parameter \c task must contain a task that was
1747 created with the function \c MSG_task_create. The second parameter
1748 \c category must contain a category that was previously defined by the function
1751 \li <b>\c TRACE_end ()</b>: This is the last function to be called. It closes
1752 the trace file and stops the tracing of the simulation. All tracing will be
1753 completely disabled after the calling this function. Although we recommend
1754 the use of this function somewhere in the end of program, it can be used
1755 anywhere in the code. This function returns 0 if everything is ok, 1 otherwise.
1757 \subsubsubsection Optional Functions
1759 \li <b>\c TRACE_host_variable_declare (const char *variable)</b>:
1760 Declare a user variable that will be associated to hosts. A variable can
1761 be used to trace user variables such as the number of tasks in a server,
1762 the number of clients in an application, and so on.
1764 \li <b>\c TRACE_host_variable_[set|add|sub] (const char *variable, double
1766 Set the value of a given user variable. It is important to remind that
1767 the value of this variable is always associated to the host. The host
1768 that will be used when these functions are called is the one returned by
1769 the function \c MSG_host_self().
1771 \subsubsection faq_tracing_example Example of Instrumentation
1773 A simplified example using the tracing mandatory functions.
1776 int main (int argc, char **argv)
1778 TRACE_start ("traced_simulation.trace");
1779 TRACE_category ("request");
1780 TRACE_category ("computation");
1781 TRACE_category ("finalize");
1783 MSG_global_init (&argc, &argv);
1785 //(... after deployment ...)
1787 m_task_t req1 = MSG_task_create("1st_request_task", 10, 10, NULL);
1788 m_task_t req2 = MSG_task_create("2nd_request_task", 10, 10, NULL);
1789 m_task_t req3 = MSG_task_create("3rd_request_task", 10, 10, NULL);
1790 m_task_t req4 = MSG_task_create("4th_request_task", 10, 10, NULL);
1791 TRACE_msg_set_task_category (req1, "request");
1792 TRACE_msg_set_task_category (req2, "request");
1793 TRACE_msg_set_task_category (req3, "request");
1794 TRACE_msg_set_task_category (req4, "request");
1796 m_task_t comp = MSG_task_create ("comp_task", 100, 100, NULL);
1797 TRACE_msg_set_task_category (comp, "computation");
1799 m_task_t finalize = MSG_task_create ("finalize", 0, 0, NULL);
1800 TRACE_msg_set_task_category (finalize, "finalize");
1811 \subsubsection faq_tracing_analyzing Analyzing the SimGrid Traces
1813 The SimGrid library, during an instrumented simulation, creates a trace file in
1814 the Paje file format that contains the platform utilization for the simulation
1815 that was executed. The visualization analysis of this file is performed with the
1816 visualization tool <a href="http://triva.gforge.inria.fr">Triva</a>, with
1817 special configurations tunned to SimGrid needs. This part of the documentation
1818 explains how to configure and use Triva to analyse a SimGrid trace file.
1820 - <b>Installing Triva</b>: the tool is available in the INRIAGforge,
1821 at <a href="http://triva.gforge.inria.fr">http://triva.gforge.inria.fr</a>.
1822 Use the following command to get the sources, and then check the file
1823 <i>INSTALL.simplified</i>. This file contains instructions to install
1824 the tool's dependencies in a Ubuntu/Debian Linux.
1826 $ svn checkout svn://scm.gforge.inria.fr/svn/triva
1828 $ cat INSTALL.simplified
1831 - <b>Executing Triva</b>: a binary called <i>Triva</i> is available after the
1832 installation (you can execute it passing <em>--help</em> to check its
1833 options). If the triva binary is not available after following the
1834 installation instructions, you may want to execute the following command to
1835 initialize the GNUstep environment variables (note that the location of the
1836 <i>GNUstep.sh</i> file may vary depending on your GNUstep installation - the
1837 command is known to work in Ubuntu and Debian Linux):
1839 $ source /usr/share/GNUstep/Makefiles/GNUstep.sh
1841 You should be able to see this output after the installation of triva:
1843 $ ./Triva.app/Triva --help
1844 Usage: Triva [OPTION...] TRACEFILE
1845 Trace Analysis through Visualization
1847 You need to use one of the following options:
1848 -g, --graph Graph Analysis
1849 -t, --treemap Treemap Analysis
1851 Other auxiliary options to check the trace file:
1852 -c, --check Check the integrity of trace file
1853 -h, --hierarchy Export the trace type hierarchy
1854 -l, --list List entity types
1856 -?, --help Give this help list
1857 --usage Give a short usage message
1859 Triva expects that the user choose one of the available options
1860 (currently <em>--graph</em> or <em>--treemap</em> for a visualization analysis)
1861 and the trace file from the simulation.
1863 - <b>Understanding Triva - time-slice</b>: the analysis of a trace file using
1864 the tool always takes into account the concept of the <em>time-slice</em>.
1865 This concept means that what is being visualized in the screen is always
1866 calculated considering a specific time frame, with its beggining and end
1867 timestamp. The time-slice is configured by the user and can be changed
1868 dynamically through the window called <em>Time Interval</em> that is opened
1869 whenever a trace file is being analyzed. The next figure depicts the time-slice
1870 configuration window.
1871 In the top of the window, in the space named <i>Trace Time</i>,
1872 the two fields show the beggining of the trace (which usually starts in 0) and
1873 the end (that depends on the time simulated by SimGrid). The middle of the
1874 window, in the square named <i>Time Slice Configuration</i>, contains the
1875 aspects related to the time-slice, including its <i>start</i> and its
1876 <i>size</i>. The gray rectangle in the bottom of this part indicates the
1877 <i>current time-slice</i> that is considered for the drawings. If the checkbox
1878 <i>Update Drawings on Sliders Change</i> is not selected, the button
1879 <i>Apply</i> must be clicked in order to inform triva that the
1880 new time-slice must be considered. The bottom part of the window, in the space
1881 indicated by the square <i>Time Slice Animation</i> can be used to advance
1882 the time-frame automatically. The user configures the amount of time that the
1883 time-frame will forward and how frequent this update will happen. Once this is
1884 configured, the user clicks the <i>Play</i> button in order to see the dynamic
1885 changes on the drawings.
1888 <a href="triva-time_interval.png" border=0><img src="triva-time_interval.png" width="50%" border=0></a>
1891 <b>Remarks:</b> when the trace has too many hosts or links, the computation to
1892 take into account a new time-slice can be expensive. When this happens, the
1893 <i>Frequency</i> parameter, but also updates caused by change on configurations
1894 when the checkbox <i>Update Drawings on Sliders
1895 Change</i> is selected will not be followed.
1897 - <b>Understanding Triva - graph</b>: this part of the documention explains how
1898 to analyze the traces using the graph view of Triva, when the user executes
1899 the tool passing <em>--graph</em> as parameter. Triva opens three windows when
1900 this parameter is used: the <i>Time Interval</i> window (previously described),
1901 the <i>Graph Representation</i> window, and the <em>Graph Configuration</em>
1902 window. The Graph Representation is the window where drawings take place.
1903 Initially, it is completely white waiting for a proper graph configuration input
1904 by the user. We start the description of this type of analysis by describing the
1905 <i>Graph Configuration</i> window (depicted below). By using a particular
1906 configuration, triva
1907 can be used to customize the graph drawing according to
1908 the SimGrid trace that was created with user-specific categories. Before delving
1909 into the details of this customization, let us first explain the major parts of
1910 the graph configuration window. The buttons located in the top-right corner can
1911 be used to delete, copy and create a new configuration. The checkbox in the
1912 top-middle part of the window indicates if the configuration typed in the
1913 textfield is syntactically correct (we are using the non-XML
1914 <a href="http://en.wikipedia.org/wiki/Property_list">Property List Format</a> to
1915 describe the configuration). The pop-up button located on the top-left corner
1916 indicates the selected configuration (the user can have multiple graph
1917 configurations). The bottom-left text field contains the name of the current
1918 configuration (updates on this field must be followed by typing enter on the
1919 keyboard to take into account the name change). The bottom-right <em>Apply</em>
1920 button activates the current configuration, resulting on an update on the graph
1924 <a href="triva-graph_configuration.png" border=0><img src="triva-graph_configuration.png" width="50%" border=0></a>
1927 <b>Basic SimGrid Configuration</b>: The figure shows in the big textfield the
1928 basic configuration that should be used during the analysis of a SimGrid trace
1929 file. The basic logic of the configuration is as follows:
1935 The nodes of the graph will be created based on the <i>node</i> parameter, which
1936 in this case is the different <em>"HOST"</em>s of the platform
1937 used to simulate. The <i>edge</i> parameter indicates that the edges of the
1938 graph will be created based on the <em>"LINK"</em>s of the platform. After the
1939 definition of these two parameters, the configuration must detail how
1940 <em>HOST</em>s and <em>LINK</em>s should be drawn. For that, the configuration
1941 must have an entry for each of the types used. For <em>HOST</em>, as basic
1942 configuration, we have:
1949 The parameter <em>size</em> indicates which variable from the trace file will be
1950 used to define the size of the node HOST in the visualization. If the simulation
1951 was executed with availability traces, the size of the nodes will be changed
1952 according to these traces. The parameter <em>scale</em> indicates if the value
1953 of the variable is <em>global</em> or <em>local</em>. If it is global, the value
1954 will be relative to the power of all other hosts, if it is local, the value will
1955 be relative locally.
1956 For <em>LINK</em> we have:
1966 For the types specified in the <em>edge</em> parameter (such as <em>LINK</em>),
1967 the configuration must contain two additional parameters: <em>src</em> and
1968 <em>dst</em> that are used to properly identify which nodes this edge is
1969 connecting. The values <em>SrcHost</em> and <em>DstHost</em> are always present
1970 in the SimGrid trace file and should not be changed in the configuration. The
1971 parameter <em>size</em> for the LINK, in this case, is configured as the
1972 variable <em>bandwidth</em>, with a <em>global</em> scale. The scale meaning
1973 here is exactly the same used for nodes. The last parameter is the GraphViz
1974 algorithm used to calculate the position of the nodes in the graph
1977 graphviz-algorithm = neato;
1980 <b>Customizing the Graph Representation</b>: triva is capable to handle
1981 a customized graph representation based on the variables present in the trace
1982 file. In the case of SimGrid, every time a category is created for tasks, two
1983 variables in the trace file are defined: one to indicate node utilization (how
1984 much power was used by that task category), and another to indicate link
1985 utilization (how much bandwidth was used by that category). For instance, if the
1986 user declares a category named <i>request</i>, there will be variables named
1987 <b>p</b><i>request</i> and a <b>b</b><i>request</i> (<b>p</b> for power and
1988 <b>b</b> for bandwidth). It is important to notice that the variable
1989 <i>prequest</i> in this case is only available for HOST, and
1990 <i>brequest</i> is only available for LINK. <b>Example</b>: suppose there are
1991 two categories for tasks: request and compute. To create a customized graph
1992 representation with a proportional separation of host and link utilization, use
1993 as configuration for HOST and LINK this:
2002 values = (prequest, pcomputation);
2015 values = (brequest, bcomputation);
2019 Where <i>sep_host</i> contains a composition of type <i>separation</i> where
2020 its max size is the <i>power</i> of the host and the variables <i>prequest</i>
2021 and <i>pcomputation</i> are drawn proportionally to the size of the HOST. And
2022 <i>sep_link</i> is also a separation where max is defined as the
2023 <i>bandwidth</i> of the link, and the variables <i>brequest</i> and
2024 <i>bcomputation</i> are drawn proportionally within a LINK.
2025 <i>This configuration enables the analysis of resource utilization by MSG tasks,
2026 and the identification of load-balancing issues, network bottlenecks, for
2028 <b>Other compositions</b>: besides <i>separation</i>, it is possible to use
2029 other types of compositions, such as gradients, and colors, like this:
2034 values = (numberOfTasks);
2038 values = (is_server);
2041 Where <i>gra_host</i> creates a gradient within a node of the graph, using a
2042 global scale and using as value a variable called <i>numberOfTasks</i>, that
2043 could be declared by the user using the optional tracing functions of SimGrid.
2044 If scale is global, the max and min value for the gradient will be equal to the
2045 max and min numberOfTasks among all hosts, and if scale is local, the max and
2046 min value based on the value of numberOfTasks locally in each host.
2047 And <i>color_host</i> composition draws a square based on a positive value of
2048 the variable <i>is_server</i>, that could also be defined by the user using the
2049 SimGrid tracing functions. \n
2050 <b>The Graph Visualization</b>: The next figure shows a graph visualization of a
2051 given time-slice of the masterslave_forwarder example (present in the SimGrid
2052 sources). The red color indicates tasks from the <i>compute</i> category. This
2053 visualization was generated with the following configuration:
2066 values = (pcompute, pfinalize);
2078 values = (bcompute, bfinalize);
2081 graphviz-algorithm = neato;
2086 <a href="triva-graph_visualization.png" border=0><img src="triva-graph_visualization.png" width="50%" border=0></a>
2090 - <b>Understading Triva - colors</b>: An important issue when using Triva is how
2091 to define colors. To do that, we have to know which variables are defined in
2092 the trace file generated by the SimGrid library. The parameter <em>--list</em>
2093 lists the variables for a given trace file:
2095 $ Triva -l masterslave_forwarder.trace
2113 We can see that HOST has seven variables (from power to pfinalize) and LINK has
2114 four (from bandwidth to bfinalize). To define a red color for the
2115 <i>pcompute</i> and <i>bcompute</i> (which are defined based on user category
2116 <i>compute</i>), execute:
2118 $ defaults write Triva 'pcompute Color' '1 0 0'
2119 $ defaults write Triva 'bcompute Color' '1 0 0'
2121 Where the three numbers in each line are the RGB color with values from 0 to 1.
2123 \section faq_troubleshooting Troubleshooting
2125 \subsection faq_trouble_lib_compil SimGrid compilation and installation problems
2127 \subsubsection faq_trouble_lib_config ./configure fails!
2129 We know only one reason for the configure to fail:
2131 - <b>You are using a broken build environment</b>\n
2132 If symptom is that configure complains about gcc not being able to build
2133 executables, you are probably missing the libc6-dev package. Damn Ubuntu.
2135 If you experience other kind of issue, please get in touch with us. We are
2136 always interested in improving our portability to new systems.
2138 \subsubsection faq_trouble_distcheck Dude! "make check" fails on my machine!
2140 Don't assume we never run this target, because we do. Check
2141 http://bob.loria.fr:8010 if you don't believe us.
2143 There is several reasons which may cause the make check to fail on your
2146 - <b>You are using a broken libc (probably concerning the contextes)</b>.\n
2147 The symptom is that the "make check" fails within the examples/msg directory.\n
2148 By default, SimGrid uses something called ucontexts. This is part of the
2149 libc, but it's quite undertested. For example, some (old) versions of the
2150 glibc on alpha do not implement these functions, but provide the stubs
2151 (which return ENOSYS: not implemented). It may fool our detection mechanism
2152 and leads to segfaults. There is not much we can do to fix the bug.
2153 A workaround is to compile with --with-context=pthread to avoid
2154 ucontext completely. You'll be a bit more limited in the number
2155 of simulated processes you can start concurrently, but 5000
2156 processes is still enough for most purposes, isn't it?\n
2157 This limitation is the reason why we insist on using this piece of ...
2158 software even if it's so troublesome.\n
2159 <b>=> use --with-pthread on AMD64 architecture that do not have an
2160 ultra-recent libc.</b>
2162 - <b>There is a bug in SimGrid we aren't aware of</b>.\n
2163 If none of the above apply, please drop us a mail on the mailing list so
2164 that we can check it out. Make sure to read \ref faq_bugrepport
2167 \subsection faq_trouble_compil User code compilation problems
2169 \subsubsection faq_trouble_err_logcat "gcc: _simgrid_this_log_category_does_not_exist__??? undeclared (first use in this function)"
2171 This is because you are using the log mecanism, but you didn't created
2172 any default category in this file. You should refer to \ref XBT_log
2173 for all the details, but you simply forgot to call one of
2174 XBT_LOG_NEW_DEFAULT_CATEGORY() or XBT_LOG_NEW_DEFAULT_SUBCATEGORY().
2176 \subsubsection faq_trouble_pthreadstatic "gcc: undefined reference to pthread_key_create"
2178 This indicates that one of the library SimGrid depends on (libpthread
2179 here) was missing on the linking command line. Dependencies of
2180 libsimgrid are expressed directly in the dynamic library, so it's
2181 quite impossible that you see this message when doing dynamic linking.
2183 If you compile your code statically (and if you use a pthread version
2184 of SimGrid -- see \ref faq_more_processes), you must absolutely
2185 specify <tt>-lpthread</tt> on the linker command line. As usual, this should
2186 come after <tt>-lsimgrid</tt> on this command line.
2188 \subsection faq_trouble_errors Runtime error messages
2190 \subsubsection faq_flexml_limit "surf_parse_lex: Assertion `next limit' failed."
2192 This is because your platform file is too big for the parser.
2194 Actually, the message comes directly from FleXML, the technology on top of
2195 which the parser is built. FleXML has the bad idea of fetching the whole
2196 document in memory before parsing it. And moreover, the memory buffer size
2197 must be determined at compilation time.
2199 We use a value which seems big enough for our need without bloating the
2200 simulators footprints. But of course your mileage may vary. In this case,
2201 just edit src/surf/surfxml.l modify the definition of
2202 FLEXML_BUFFERSTACKSIZE. E.g.
2205 #define FLEXML_BUFFERSTACKSIZE 1000000000
2208 Then recompile and everything should be fine, provided that your version of
2209 Flex is recent enough (>= 2.5.31). If not the compilation process should
2212 A while ago, we worked on FleXML to reduce a bit its memory consumption, but
2213 these issues remain. There is two things we should do:
2215 - use a dynamic buffer instead of a static one so that the only limit
2216 becomes your memory, not a stupid constant fixed at compilation time
2217 (maybe not so difficult).
2218 - change the parser so that it does not need to get the whole file in
2219 memory before parsing
2220 (seems quite difficult, but I'm a complete newbe wrt flex stuff).
2222 These are changes to FleXML itself, not SimGrid. But since we kinda hijacked
2223 the development of FleXML, I can grant you that any patches would be really
2224 welcome and quickly integrated.
2226 <b>Update:</b> A new version of FleXML (1.7) was released. Most of the work
2227 was done by William Dowling, who use it in his own work. The good point is
2228 that it now use a dynamic buffer, and that the memory usage was greatly
2229 improved. The downside is that William also changed some things internally,
2230 and it breaks the hack we devised to bypass the parser, as explained in
2231 \ref faq_flexml_bypassing. Indeed, this is not a classical usage of the
2232 parser, and Will didn't imagine that we may have used (and even documented)
2233 such a crude usage of FleXML. So, we now have to repair the bypassing
2234 functionality to use the lastest FleXML version and fix the memory usage in
2237 \subsubsection faq_trouble_gras_transport GRAS spits networking error messages
2239 Gras, on real platforms, naturally use regular sockets to communicate. They
2240 are deeply hidden in the gras abstraction, but when things go wrong, you may
2241 get some weird error messages. Here are some example, with the probable
2244 - <b>Transport endpoint is not connected</b>: several processes try to open
2245 a server socket on the same port number of the same machine. This is
2246 naturally bad and each process should pick its own port number for this.\n
2247 Maybe, you just have some processes remaining from a previous experiment
2249 Killing them may help, but again if you kill -KILL them, you'll have to
2250 wait for a while: they didn't close there sockets properly and the system
2251 needs a while to notice that this port is free again.
2253 - <b>Socket closed by remote side</b>: if the remote process is not
2254 supposed to close the socket at this point, it may be dead.
2256 - <b>Connection reset by peer</b>: I found this on Internet about this
2257 error. I think it's what's happening here, too:\n
2258 <i>This basically means that a network error occurred while the client was
2259 receiving data from the server. But what is really happening is that the
2260 server actually accepts the connection, processes the request, and sends
2261 a reply to the client. However, when the server closes the socket, the
2262 client believes that the connection has been terminated abnormally
2263 because the socket implementation sends a TCP reset segment telling the
2264 client to throw away the data and report an error.\n
2265 Sometimes, this problem is caused by not properly closing the
2266 input/output streams and the socket connection. Make sure you close the
2267 input/output streams and socket connection properly. If everything is
2268 closed properly, however, and the problem persists, you can work around
2269 it by adding a one-second sleep before closing the streams and the
2270 socket. This technique, however, is not reliable and may not work on all
2272 Since GRAS sockets are closed properly (repeat after me: there is no bug
2273 in GRAS), it is either that you are closing your sockets on server side
2274 before the client get a chance to read them (use gras_os_sleep() to delay
2275 the server), or the server died awfully before the client got the data.
2277 \subsubsection faq_trouble_errors_big_fat_warning I'm told that my XML files are too old.
2279 The format of the XML platform description files is sometimes
2280 improved. For example, we decided to change the units used in SimGrid
2281 from MBytes, MFlops and seconds to Bytes, Flops and seconds to ease
2282 people exchanging small messages. We also reworked the route
2283 descriptions to allow more compact descriptions.
2285 That is why the XML files are versionned using the 'version' attribute
2286 of the root tag. Currently, it should read:
2288 <platform version="2">
2291 If your files are too old, you can use the simgrid_update_xml.pl
2292 script which can be found in the tools directory of the archive.
2294 \subsection faq_trouble_valgrind Valgrind-related and other debugger issues
2296 If you don't, you really should use valgrind to debug your code, it's
2299 \subsubsection faq_trouble_vg_longjmp longjmp madness in valgrind
2301 This is when valgrind starts complaining about longjmp things, just like:
2303 \verbatim ==21434== Conditional jump or move depends on uninitialised value(s)
2304 ==21434== at 0x420DBE5: longjmp (longjmp.c:33)
2306 ==21434== Use of uninitialised value of size 4
2307 ==21434== at 0x420DC3A: __longjmp (__longjmp.S:48)
2310 This is the sign that you didn't used the exception mecanism well. Most
2311 probably, you have a <tt>return;</tt> somewhere within a <tt>TRY{}</tt>
2312 block. This is <b>evil</b>, and you must not do this. Did you read the section
2315 \subsubsection faq_trouble_vg_libc Valgrind spits tons of errors about backtraces!
2317 It may happen that valgrind, the memory debugger beloved by any decent C
2318 programmer, spits tons of warnings like the following :
2319 \verbatim ==8414== Conditional jump or move depends on uninitialised value(s)
2320 ==8414== at 0x400882D: (within /lib/ld-2.3.6.so)
2321 ==8414== by 0x414EDE9: (within /lib/tls/i686/cmov/libc-2.3.6.so)
2322 ==8414== by 0x400B105: (within /lib/ld-2.3.6.so)
2323 ==8414== by 0x414F937: _dl_open (in /lib/tls/i686/cmov/libc-2.3.6.so)
2324 ==8414== by 0x4150F4C: (within /lib/tls/i686/cmov/libc-2.3.6.so)
2325 ==8414== by 0x400B105: (within /lib/ld-2.3.6.so)
2326 ==8414== by 0x415102D: __libc_dlopen_mode (in /lib/tls/i686/cmov/libc-2.3.6.so)
2327 ==8414== by 0x412D6B9: backtrace (in /lib/tls/i686/cmov/libc-2.3.6.so)
2328 ==8414== by 0x8076446: xbt_dictelm_get_ext (dict_elm.c:714)
2329 ==8414== by 0x80764C1: xbt_dictelm_get (dict_elm.c:732)
2330 ==8414== by 0x8079010: xbt_cfg_register (config.c:208)
2331 ==8414== by 0x806821B: MSG_config (msg_config.c:42)
2334 This problem is somewhere in the libc when using the backtraces and there is
2335 very few things we can do ourselves to fix it. Instead, here is how to tell
2336 valgrind to ignore the error. Add the following to your ~/.valgrind.supp (or
2337 create this file on need). Make sure to change the obj line according to
2338 your personnal mileage (change 2.3.6 to the actual version you are using,
2339 which you can retrieve with a simple "ls /lib/ld*.so").
2342 name: Backtrace madness
2344 obj:/lib/ld-2.3.6.so
2349 fun:__libc_dlopen_mode
2352 Then, you have to specify valgrind to use this suppression file by passing
2353 the <tt>--suppressions=$HOME/.valgrind.supp</tt> option on the command line.
2354 You can also add the following to your ~/.bashrc so that it gets passed
2355 automatically. Actually, it passes a bit more options to valgrind, and this
2356 happen to be my personnal settings. Check the valgrind documentation for
2359 \verbatim export VALGRIND_OPTS="--leak-check=yes --leak-resolution=high --num-callers=40 --tool=memcheck --suppressions=$HOME/.valgrind.supp" \endverbatim
2361 \subsubsection faq_trouble_backtraces Truncated backtraces
2363 When debugging SimGrid, it's easier to pass the
2364 --disable-compiler-optimization flag to the configure if valgrind or
2365 gdb get fooled by the optimization done by the compiler. But you
2366 should remove these flag when everything works before going in
2367 production (before launching your 1252135 experiments), or everything
2368 will run only one half of the true SimGrid potential.
2370 \subsection faq_deadlock There is a deadlock in my code!!!
2372 Unfortunately, we cannot debug every code written in SimGrid. We
2373 furthermore believe that the framework provides ways enough
2374 information to debug such informations yourself. If the textual output
2375 is not enough, Make sure to check the \ref faq_visualization FAQ entry to see
2376 how to get a graphical one.
2378 Now, if you come up with a really simple example that deadlocks and
2379 you're absolutely convinced that it should not, you can ask on the
2380 list. Just be aware that you'll be severely punished if the mistake is
2381 on your side... We have plenty of FAQ entries to redact and new
2382 features to implement for the impenitents! ;)
2384 \subsection faq_surf_network_latency I get weird timings when I play with the latencies.
2386 OK, first of all, remember that units should be Bytes, Flops and
2387 Seconds. If you don't use such units, some SimGrid constants (e.g. the
2388 SG_TCP_CTE_GAMMA constant used in most network models) won't have the
2389 right unit and you'll end up with weird results.
2391 Here is what happens with a single transfer of size L on a link
2392 (bw,lat) when nothing else happens.
2395 0-----lat--------------------------------------------------t
2396 |-----|**** real_bw =min(bw,SG_TCP_CTE_GAMMA/(2*lat)) *****|
2399 In more complex situations, this min is the solution of a complex
2400 max-min linear system. Have a look
2401 <a href="http://lists.gforge.inria.fr/pipermail/simgrid-devel/2006-April/thread.html">here</a>
2402 and read the two threads "Bug in SURF?" and "Surf bug not
2403 fixed?". You'll have a few other examples of such computations. You
2404 can also read "A Network Model for Simulation of Grid Application" by
2405 Henri Casanova and Loris Marchal to have all the details. The fact
2406 that the real_bw is smaller than bw is easy to understand. The fact
2407 that real_bw is smaller than SG_TCP_CTE_GAMMA/(2*lat) is due to the
2408 window-based congestion mechanism of TCP. With TCP, you can't exploit
2409 your huge network capacity if you don't have a good round-trip-time
2410 because of the acks...
2412 Anyway, what you get is t=lat + L/min(bw,SG_TCP_CTE_GAMMA/(2*lat)).
2414 * if I you set (bw,lat)=(100 000 000, 0.00001), you get t = 1.00001 (you fully
2416 * if I you set (bw,lat)=(100 000 000, 0.0001), you get t = 1.0001 (you're on the
2418 * if I you set (bw,lat)=(100 000 000, 0.001), you get t = 10.001 (ouch!)
2420 This bound on the effective bandwidth of a flow is not the only thing
2421 that may make your result be unexpected. For example, two flows
2422 competing on a saturated link receive an amount of bandwidth inversely
2423 proportional to their round trip time.
2425 \subsection faq_bugrepport So I've found a bug in SimGrid. How to report it?
2427 We do our best to make sure to hammer away any bugs of SimGrid, but this is
2428 still an academic project so please be patient if/when you find bugs in it.
2429 If you do, the best solution is to drop an email either on the simgrid-user
2430 or the simgrid-devel mailing list and explain us about the issue. You can
2431 also decide to open a formal bug report using the
2432 <a href="https://gforge.inria.fr/tracker/?atid=165&group_id=12&func=browse">relevant
2433 interface</a>. You need to login on the server to get the ability to submit
2436 We will do our best to solve any problem repported, but you need to help us
2437 finding the issue. Just telling "it segfault" isn't enough. Telling "It
2438 segfaults when running the attached simulator" doesn't really help either.
2439 You may find the following article interesting to see how to repport
2440 informative bug repports:
2441 http://www.chiark.greenend.org.uk/~sgtatham/bugs.html (it is not SimGrid
2442 specific at all, but it's full of good advices).
2444 \author Arnaud Legrand (arnaud.legrand::imag.fr)
2445 \author Martin Quinson (martin.quinson::loria.fr)
2450 ******************************************************************
2451 * OLD CRUFT NOT USED ANYMORE *
2452 ******************************************************************
2455 \subsection faq_crosscompile Cross-compiling a Windows DLL of SimGrid from linux
2457 At the moment, we do not distribute Windows pre-compiled version of SimGrid
2458 because the support for this platform is still experimental. We know that
2459 some parts of the GRAS environment do not work, and we think that the others
2460 environments (MSG and SD) have good chances to work, but we didn't test
2461 ourselves. This section explains how we generate the SimGrid DLL so that you
2462 can build it for yourself. First of all, you need to have a version more
2463 recent than 3.1 (ie, a SVN version as time of writting).
2465 In order to cross-compile the package to windows from linux, you need to
2466 install mingw32 (minimalist gnu win32). On Debian, you can do so by
2467 installing the packages mingw32 (compiler), mingw32-binutils (linker and
2468 so), mingw32-runtime.
2470 You can use the VPATH support of configure to compile at the same time for
2471 linux and windows without dupplicating the source nor cleaning the tree
2472 between each. Just run bootstrap (if you use the SVN) to run the autotools.
2473 Then, create a linux and a win directories. Then, type:
2474 \verbatim cd linux; ../configure --srcdir=.. <usual configure flags>; make; cd ..
2475 cd win; ../configure --srcdir=.. --host=i586-mingw32msvc <flags>; make; cd ..
2477 The trick to VPATH builds is to call configure from another directory,
2478 passing it an extra --srcdir argument to tell it where all the sources are.
2479 It will understand you want to use VPATH. Then, the trick to cross-compile
2480 is simply to add a --host argument specifying the target you want to build
2481 for. The i586-mingw32msvc string is what you have to pass to use the mingw32
2482 environment as distributed in Debian.
2484 After that, you can run all make targets from both directories, and test
2485 easily that what you change for one arch does not break the other one.
2487 It is possible that this VPATH build thing breaks from time to time in the
2488 SVN since it's quite fragile, but it's granted to work in any released
2489 version. If you experience problems, drop us a mail.
2491 Another possible source of issue is that at the moment, building the
2492 examples request to use the gras_stub_generator tool, which is a compiled
2493 program, not a script. In cross-compilation, you need to cross-execute with
2494 wine for example, which is not really pleasant. We are working on this, but
2495 in the meanwhile, simply don't build the examples in cross-compilation
2496 (<tt>cd src</tt> before running make).
2498 Program (cross-)compiled with mingw32 do request an extra DLL at run-time to be
2499 usable. For example, if you want to test your build with wine, you should do
2500 the following to put this library where wine looks for DLLs.
2502 cp /usr/share/doc/mingw32-runtime/mingwm10.dll.gz ~/.wine/c/windows/system/
2503 gunzip ~/.wine/c/windows/system/mingwm10.dll.gz
2506 The DLL is built in src/.libs, and installed in the <i>prefix</i>/bin directory
2507 when you run make install.
2509 If you want to use it in a native project on windows, you need to use
2510 simgrid.dll and mingwm10.dll. For each DLL, you need to build .def file
2511 under linux (listing the defined symbols), and convert it into a .lib file
2512 under windows (specifying this in a way that windows compilers like). To
2513 generate the def files, run (under linux):
2514 \verbatim echo "LIBRARY libsimgrid-0.dll" > simgrid.def
2515 echo EXPORTS >> simgrid.def
2516 nm libsimgrid-0.dll | grep ' T _' | sed 's/.* T _//' >> simgrid.def
2517 nm libsimgrid-0.dll | grep ' D _' | sed 's/.* D _//' | sed 's/$/ DATA/' >> simgrid.def
2519 echo "LIBRARY mingwm10.dll" > mingwm10.def
2520 echo EXPORTS >> mingwm10.def
2521 nm mingwm10.dll | grep ' T _' | sed 's/.* T _//' >> mingwm10.def
2522 nm mingwm10.dll | grep ' D _' | sed 's/.* D _//' | sed 's/$/ DATA/' >> mingwm10.def
2525 To create the import .lib files, use the <tt>lib</tt> windows tool (from
2526 MSVC) the following way to produce simgrid.lib and mingwm10.lib
2527 \verbatim lib /def:simgrid.def
2528 lib /def:mingwm10.def
2531 If you happen to use Borland C Builder, the right command line is the
2532 following (note that you don't need any file.def to get this working).
2533 \verbatim implib simgrid.lib libsimgrid-0.dll
2534 implib mingwm10.lib mingwm10.dll
2537 Then, set the following parameters in Visual C++ 2005:
2538 Linker -> Input -> Additional dependencies = simgrid.lib mingwm10.lib
2540 Just in case you wonder how to generate a DLL from libtool in another
2541 project, we added -no-undefined to any lib*_la_LDFLAGS variables so that
2542 libtool accepts to generate a dynamic library under windows. Then, to make
2543 it true, we pass any dependencies (such as -lws2 under windows or -lpthread
2544 on need) on the linking line. Passing such deps is a good idea anyway so
2545 that they get noted in the library itself, avoiding the users to know about
2546 our dependencies and put them manually on their compilation line. Then we
2547 added the AC_LIBTOOL_WIN32_DLL macro just before AC_PROG_LIBTOOL in the
2548 configure.ac. It means that we exported any symbols which need to be.
2549 Nowadays, functions get automatically exported, so we don't need to load our
2550 header files with tons of __declspec(dllexport) cruft. We only need to do so
2551 for data, but there is no public data in SimGrid so we are good.