2 ** Source tree organization
4 ******************************************************
6 There is at least 5 sub-projects in the tree:
8 - XBT: eXtended Bundle of Tools (low-level toolbox: logging, datatypes).
9 - SURF: a SimUlation aRtiFact. This is the simulation kernel.
10 - MSG: originally MetaSimGrid, MSG is a simple distributed application
12 - SMPI: Simulated MPI, to run MPI application using emulation technics;
14 - SIMIX: basix interface for simulated processes. This layer defines simcalls
15 (simulation calls) exposed to the simulated processes by the SIMIX "kernel".
16 This interface is used to implement the MSG, SMPI layers.
19 They are all in the same tree because they are complementary tools and
20 having all of them in the same package makes the installation easier
21 for end-users. Moreover, it enables to share the compilation chain and
22 eases the development.
24 The tree is not split on projects, but on file finality:
25 include/ -> all *public* headers
26 include/xbt/*.h -> one file per module
28 src/include -> another location for protected headers. Used by SURF, and
29 other should be converted, since this is the Right Thing.
31 examples/ -> Supposed to be copy/pastable by the user, so keep it clear and
32 avoid any kind of trick. In particular, do only include the
35 teshsuite/ -> The more test the better. Put in there any strange test
36 doing things that the users are not supposed to do,
37 just to see if our framework is robust to incorrect and
38 unusual behaviors. All tests written in this section
39 should leverage our tesh(1) utility.
42 ** Indentation standard
44 *****************************************************
46 Most files use the Kernighan & Ritchie coding style with 2 spaces of
47 indentation. The indent program can help you to stick to it:
49 indent -kr -l120 -nut -i2 -lps -npcs -br -brs -ce -cdw -bbo -npsl <myfile>
51 The script ./tools/internal/indent runs indent with the appropriate options.
53 If you use Eclipse, please import the settings in ./tools/internal/eclipse-formating.xml
56 ** Type naming standard
58 *****************************************************
60 It may sound strange, but the type naming convention was source of intense
61 discussion between da SimGrid posse members. The convention we came to may not
62 be the best solution, but it has the merit to exist and leave everyone work.
63 So please stick to it.
65 - ???_t is a valid type (built with typedef)
66 - s_toto_t is a structure (access to fields with .)
67 - s_toto is a structure needing 'struct' keyword to be used
69 - u_toto_t is an union
70 - u_toto is an union needing 'union' keyword to be used
71 - toto_t is an 'object' (struct*)
73 Please to not call toto_t something else than an 'object' (ie, something you
74 have to call _new and _free on it).
77 typedef struct s_toto {} s_toto_t, *toto_t;
78 typedef enum {} e_toto_t;
80 Moreover, only toto_t (and e_toto_t) are public. The rest (mainly s_toto_t)
83 If you see any part of the code not following this convention, this is a
84 bug. Please report it (or fix it yourself if you can).
87 ** Random bits about coding standards and portability
89 *****************************************************
92 Don't use it, or you'll have to check the result (and do some dirty stuff
93 on AIX). Use xbt_malloc (or even better, xbt_new) instead.
95 SIZE_T (FIXME: obsolete?)
96 If possible, avoid size_t and use unsigned long instead. If not,
97 #include <sys/types.h> in all files manipulating size_t
98 do cast it to unsigned long before printing (and use %lu),
102 Please avoid to use long ints. This is the source of many compatibility
103 problems between 32 bits and 64 bits archs. Either use plain ints (generally
104 32 bits wide) or long long ints (64 bits wide, at least). At last resort
105 consider using integer types defined in C99 by <stdint.h>.
107 PRINTF pointer difference (FIXME: advertise %td instead?)
108 printf ("diff = %ld\n", (long) (pointer2 - pointer1));
111 ** Commenting the source: doxygen
113 ****************************************************
115 The global structure of the documentation is in doc/modules.doc
117 The structure of each module (xbt, msg, etc) is in doc/module-<module>.doc
119 The structure of a module is in its public header. This way, you're sure to
120 see all the public interface (and only it). The different parts of the
121 interface are grouped using the @name construct, even if it's buggy. Since
122 parts often get reordered, it's better to add numbers to the parts (so that
123 users can see the intended order).
125 The documentation of each type and macro are also in the public header since
126 this is were they live.
128 The documentation of each function must be in the C file were it lives.
130 Any public element (function, type and macro) must have a @brief part.
133 ** XBT virtualization mechanism (FIXME: this section is deprecated)
135 ****************************************************
137 There is some functionalities that we want to virtualize in XBT. We
138 want xbt_time to give the simulated clock when running on top of the
139 simulator, and the host clock when running on a real system. This
140 could be placed in GRAS (and was, historically), but there is some
141 reason to lower it down to XBT.
143 Here is the used naming scheme:
145 - xbt_<module>_<func>(): functions working both in SG and RL
146 - xbt_os_<module>_<func>(): RL functions usable even in simulator
148 That way, in libsimgrid, we still can use native functions if we
149 want to. It may for example be useful to get the real time when
150 implementing the simulator. Think of the SIGINT handler, which
151 wants to see if the user pressed the key twice in a 5 seconds
152 interval. This is of little use to check the simulated time here.
154 Here is the file layout:
156 - xbt_rl_<module>.c: native implementation (xbt_<module>_<func>()).
157 Simply call the corresponding xbt_os_<module>_<func>.
158 Part only of libgras.so
160 - xbt_sg_<module>.c: SIMIX implementation xbt_<module>_<func>()).
161 Simply call the corresponding SIMIX implementation.
162 Part only of libsimgrid.so
164 - xbt_os_<module>.c: body of the functions implementing natively the
165 stuff (xbt_os_<module>_<func>()).
166 Part of both libgras.so and libsimgrid.so
168 Since there is almost nothing in xbt_rl_module.c and xbt_sg_module.c,
169 it'd be better to use symbol aliasing here (to declare in the object
170 code that the same function have two names), but I'm still
171 investigating the portability of the thing to windows.
175 * SimGrid Hacker Survival Guide (FIXME: should be betterly placed)
176 ********************************
178 * Before pushing any change, don't forget to check if the compilation
179 passes with compiler optimizations and warnings turned on:
180 cmake -Denable_compile_optimizations=ON \
181 -Denable_compile_warnings=ON
183 * Your commit message should follow the git habits, explained eg here:
184 http://tbaggery.com/2008/04/19/a-note-about-git-commit-messages.html
186 * When you add/remove files, and/or make changes in the lists of files to build,
187 please check that "make distcheck" still succeeds. This is needed to ensure
188 that the generated archive is consistent.
190 * If you want to debug memory allocation problems, here are a few hints:
191 - disable compiler optimizations, to have better backtraces;
192 - disable the mallocators, or it will be hard to match malloc's with
194 - disable model checking, unless your problem lies in the model
195 checker part of SimGrid (MC brings its own malloc implementation,
196 which valgrind doesn't understand).
197 All this is configured with:
198 cmake -Denable_model-checking=OFF \
199 -Denable_mallocators=OFF \
200 -Denable_compile_optimizations=OFF
202 * If you break the logs (for example while hacking in the dynars), you
203 want to define XBT_LOG_MAYDAY at the beginning of log.h. It will
204 deactivate the whole logging mechanism, switching to printfs
205 instead. SimGrid becomes incredibly verbose when doing so, but it
206 you let you fixing the dynars.