mquinson [Fri, 31 Dec 2004 12:08:57 +0000 (12:08 +0000)]
Add a function I need in GRAS
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@715
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 12:05:52 +0000 (12:05 +0000)]
Automatic update, damn gtkdoc-tools
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@714
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 12:04:21 +0000 (12:04 +0000)]
SimGrid is dead, long life to SimGrid
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@713
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 12:04:11 +0000 (12:04 +0000)]
Properly check for errno and unistd; don't check for old generation SimGrid as we prepare to get funky
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@712
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 11:42:40 +0000 (11:42 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@711
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 11:39:18 +0000 (11:39 +0000)]
automatic update
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@710
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 11:29:34 +0000 (11:29 +0000)]
Let's get funky and try to settle gras on top of msg new generation
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@709
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 11:16:19 +0000 (11:16 +0000)]
remove dead code
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@708
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 11:14:30 +0000 (11:14 +0000)]
Remove the headers already loaded from somewhere else. Keep the remaining for the day I wonder how to let it work on some exotic platform
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@707
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 31 Dec 2004 10:45:10 +0000 (10:45 +0000)]
remove complications to run tests from there own directory now that the path isn't hardcoded anymore in surf/msg
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@706
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 31 Dec 2004 03:03:00 +0000 (03:03 +0000)]
surf:~/Work/GRAS/heap $ valgrind --leak-check=yes --show-reachable=yes ./msg_test les
...
[421.531] P3 | I'm done. See you!
MSG: Congratulations ! Simulation terminated : all process are over
Simulation time 421.531
==8295==
==8295== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 17 from 1)
==8295== malloc/free: in use at exit: 0 bytes in 0 blocks.
==8295== malloc/free: 195685 allocs, 195685 frees,
6498475 bytes allocated.
==8295== For counts of detected errors, rerun with: -v
==8295== No malloc'd blocks -- no leaks are possible.
Rha! The very good news is that this code (msg_test) behaves exactly in the same way as with the old MSG version. And it is a way faster !!! :)
Before :
real 0m3.130s
user 0m2.290s
sys 0m0.030s
Now :
real 0m0.392s
user 0m0.290s
sys 0m0.000s
Cool, isnt'it ? :)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@705
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 31 Dec 2004 02:49:39 +0000 (02:49 +0000)]
Mem leak hunting!
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@704
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 31 Dec 2004 02:49:07 +0000 (02:49 +0000)]
Convenient to optimize the SURF.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@703
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 29 Dec 2004 14:38:32 +0000 (14:38 +0000)]
search data files in the right directory
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@702
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 29 Dec 2004 03:15:51 +0000 (03:15 +0000)]
MSG almost works. It enabled me to find some bug in the SURF. I'm going to optimize maxmin and to fix this bug.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@701
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 29 Dec 2004 03:11:50 +0000 (03:11 +0000)]
*** empty log message ***
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@700
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 29 Dec 2004 03:11:03 +0000 (03:11 +0000)]
add a startup and a cleanup function to each context
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@699
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 23 Dec 2004 00:56:58 +0000 (00:56 +0000)]
arguments were not set.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@698
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 23 Dec 2004 00:55:07 +0000 (00:55 +0000)]
task->data (formerly callback) could not be set by end-users.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@697
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 22 Dec 2004 20:43:12 +0000 (20:43 +0000)]
API modification in context breaks MSG. Removing MSG from the compilation list until it is repaired.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@696
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 22 Dec 2004 20:41:34 +0000 (20:41 +0000)]
Modifying the API so as to prevent a use of the context that would make valgrind unhappy. See, I really care about the happiness of valgrind...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@695
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 22 Dec 2004 20:19:52 +0000 (20:19 +0000)]
No more leak: valgrind approved. However, some ways of using these context make valgrind cry and I really don't know why. Some other people have faced the same problem but "solved" it by supressing valgrind's warning...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@694
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 22 Dec 2004 20:17:22 +0000 (20:17 +0000)]
How can I have ever wrote such a thing !
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@693
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 21 Dec 2004 21:46:49 +0000 (21:46 +0000)]
Trying to make old MSG exemples run on top on SURF. Definitely need to wax it more, especially the context part.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@692
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 21 Dec 2004 21:43:22 +0000 (21:43 +0000)]
debug debug debug
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@691
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 21 Dec 2004 21:26:31 +0000 (21:26 +0000)]
Forgot to disable debug printf's. Sorry for the noise.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@690
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 21 Dec 2004 21:24:50 +0000 (21:24 +0000)]
make valgrind happy happy.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@689
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 21 Dec 2004 19:50:39 +0000 (19:50 +0000)]
Dude, that was baaad... ;)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@688
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Sat, 18 Dec 2004 05:41:25 +0000 (05:41 +0000)]
platform file is not hardcoded anymore
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@687
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Sat, 18 Dec 2004 00:17:39 +0000 (00:17 +0000)]
This function has been accidentally deleted in the last version...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@686
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 23:52:52 +0000 (23:52 +0000)]
Finished to rewrite most functions. Now let's try it and spend one week to debug... :)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@685
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 23:43:04 +0000 (23:43 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@684
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 23:41:39 +0000 (23:41 +0000)]
moved src/surf/surf_parse.h to include/surf/
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@683
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 23:30:56 +0000 (23:30 +0000)]
Add a convenient function
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@682
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 22:45:59 +0000 (22:45 +0000)]
This one was easy. :)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@681
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 22:26:59 +0000 (22:26 +0000)]
A few more functions and I'll be able to make MSG program start and crash... :)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@680
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 21:29:11 +0000 (21:29 +0000)]
Cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@679
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 21:18:52 +0000 (21:18 +0000)]
Keep rewriting MSG so that it uses SURF instead of SG. The day I'm gonna run a MSG program, it is gonna take a long time before it does not crash anymore...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@678
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 05:46:56 +0000 (05:46 +0000)]
Keeping rewriting MSG
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@677
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 05:45:05 +0000 (05:45 +0000)]
Adding function to indicate whether an action is suspended or not.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@676
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 05:43:49 +0000 (05:43 +0000)]
A few more functions usefull in MSG. They should not be written like functions though as they are used in loop. Nevermind.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@675
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Fri, 17 Dec 2004 00:04:50 +0000 (00:04 +0000)]
cosmetics : s/sim_data/simdata/g
still does not break backward compatibility. It was private before.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@674
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 23:29:25 +0000 (23:29 +0000)]
cosmetics : create -> new
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@673
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 23:12:32 +0000 (23:12 +0000)]
just a move
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@672
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 22:55:55 +0000 (22:55 +0000)]
Keep rewriting MSG
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@671
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 22:55:27 +0000 (22:55 +0000)]
MSG is not a module
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@670
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 20:50:39 +0000 (20:50 +0000)]
making some field names more explicit (e.g. void * -> sim_data_task_t) even though the corresponding data are still hiden
starting to reimplement MSG
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@669
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 20:40:06 +0000 (20:40 +0000)]
prefixing name functions by xbt_
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@668
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 15:43:41 +0000 (15:43 +0000)]
move gras_datadesc_type_cmp() out of the way (from ddt_exchange.c to ddt_create.c)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@667
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 13:08:37 +0000 (13:08 +0000)]
Ignore binaries
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@666
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 13:07:43 +0000 (13:07 +0000)]
Works on G5
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@665
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 13:03:07 +0000 (13:03 +0000)]
You evil namespace polluter
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@664
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 12:59:17 +0000 (12:59 +0000)]
Fight symbol redefinition since it breaks on G5 (don't ask why it does only break on this arch)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@663
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 12:50:34 +0000 (12:50 +0000)]
no generated files in CVS
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@662
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 12:45:21 +0000 (12:45 +0000)]
Port to G5
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@661
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:55:50 +0000 (11:55 +0000)]
My god, the AIX compiler is rather picky
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@660
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:42:00 +0000 (11:42 +0000)]
automake-1.9 hits testing
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@659
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:38:20 +0000 (11:38 +0000)]
gras now works on win32
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@658
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:37:59 +0000 (11:37 +0000)]
Cleanup library dependencies declaration
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@657
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:28:51 +0000 (11:28 +0000)]
test_rl now generated
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@656
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:22:53 +0000 (11:22 +0000)]
test_rl now generated
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@655
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 11:17:23 +0000 (11:17 +0000)]
Document last changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@654
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 10:34:32 +0000 (10:34 +0000)]
cleanup the mecanism in charge of declaring the library dependencies to libtool
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@653
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 00:56:14 +0000 (00:56 +0000)]
context_t -> xbt_context_t
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@652
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Thu, 16 Dec 2004 00:54:29 +0000 (00:54 +0000)]
works better with multiple inclusion
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@651
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 00:12:38 +0000 (00:12 +0000)]
gras works on win32
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@650
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 00:05:54 +0000 (00:05 +0000)]
yeah, yeah, linux. Don't bother. O_BINARY is a nasty flag from this nasty windows. Keep cool and ignore it
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@649
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Dec 2004 00:04:58 +0000 (00:04 +0000)]
you fucking windows. Don't even think about translating my bytes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@648
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 23:42:03 +0000 (23:42 +0000)]
Some more debug
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@647
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 22:51:07 +0000 (22:51 +0000)]
Fix the ucontext detection: search for the header, not the function in libc
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@646
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 22:50:04 +0000 (22:50 +0000)]
Add portable header loading for contexts
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@645
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 22:48:55 +0000 (22:48 +0000)]
Add the ucontext reimplementation for windows from Panagiotis
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@644
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 22:47:38 +0000 (22:47 +0000)]
Make the context cruft work on windows; mark the pthread layer in context.* as deprecated. We will need threads around, but not here
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@643
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Dec 2004 20:33:08 +0000 (20:33 +0000)]
Let's keep cool and polite with the users of prehistoric platforms
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@642
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 15 Dec 2004 06:58:26 +0000 (06:58 +0000)]
Well, some things really are deprecated. Especially those that enable to
build a platform by hand. Therefore, there is no more need for links... just
process executing on hosts and working on tasks.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@641
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 15 Dec 2004 06:20:08 +0000 (06:20 +0000)]
Starting MSG. To ensure backward compability, those two headers won't really
change.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@640
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Wed, 15 Dec 2004 06:14:07 +0000 (06:14 +0000)]
when using doubles, %lg should be used for scanf and %g for printf (which is
weird because it means printf does not make any difference between float and
double). However %lg also works for printf. So let's keep it uniform and
we'll change that if we ever find a machine where it does not work.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@639
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:29:07 +0000 (21:29 +0000)]
generate an helper script to make it work on win32
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@638
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:28:16 +0000 (21:28 +0000)]
do not configure by default since I know cross-compile using VPATH
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@637
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:26:50 +0000 (21:26 +0000)]
Make it work on win32; use the portable errno on socks; fix a bug revealed by the win port (but existing for all ports)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@636
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 14 Dec 2004 21:21:26 +0000 (21:21 +0000)]
Sorry for those stupid assert.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@635
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:19:27 +0000 (21:19 +0000)]
added a debug msg
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@634
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:18:05 +0000 (21:18 +0000)]
check for the functions we need; generate and chmod +x the helper scripts we need
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@633
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:16:14 +0000 (21:16 +0000)]
Make it work on win32 ; some more debug msgs
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@632
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:14:05 +0000 (21:14 +0000)]
cross-compile support: append '.exe' to the binary names we want to start when on win32
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@631
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:11:04 +0000 (21:11 +0000)]
support for wine testsuite execution: do not run the helper scripts from within wine. Note to self: this script becomes ways too hairly and should be 'reworked' from scratch
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@630
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:06:25 +0000 (21:06 +0000)]
initialize winsock; support multiple calls to init/exit functions (should be generalized); useless but numerous cosmetics and picky changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@629
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:03:32 +0000 (21:03 +0000)]
cleanup for the upcoming release
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@628
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:01:59 +0000 (21:01 +0000)]
Make sure HAVE_WINSOCK_H is defined even if only version 2 since it's the used sentinel; define portable way to retrieve errno and errstr with sockets
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@627
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 21:00:10 +0000 (21:00 +0000)]
some more possible logs (printf is the only way to debug cross-compile so far)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@626
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 20:59:20 +0000 (20:59 +0000)]
use gras_exit, not only xbt_exit
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@625
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Dec 2004 20:58:26 +0000 (20:58 +0000)]
Compile pas avec les warnings, sinon
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@624
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 14 Dec 2004 19:04:58 +0000 (19:04 +0000)]
Adding a context library to xbt.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@623
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sun, 12 Dec 2004 18:49:24 +0000 (18:49 +0000)]
cosmetic
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@622
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Sat, 11 Dec 2004 05:06:28 +0000 (05:06 +0000)]
No more use of abort.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@621
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Sat, 11 Dec 2004 00:12:07 +0000 (00:12 +0000)]
No more xbt_maxmin_float_t not xbt_heap_float_t. I use double everywhere. In
SimGrid 2 I had to use long double for precision reason. Probably because of
the trace integration but I should not need it anymore since there is no
more trace integration. The real question is "is double sufficient for time
values". The answer is yes since for values larger than 272 years, you still
have a precision of 1e-6 s. float is clearly not sufficient however. Let's
use doubles then !
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@620
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 10 Dec 2004 15:43:40 +0000 (15:43 +0000)]
buglet and improvement
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@619
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 10 Dec 2004 15:33:56 +0000 (15:33 +0000)]
Fix the semantic (gras_os_time returns seconds, not microseconds)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@618
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 10 Dec 2004 15:30:30 +0000 (15:30 +0000)]
Fix the semantic (gras_os_time returns seconds); Fix output (I didn't understood the subtility of the cast to double by adding 0.0)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@617
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 10 Dec 2004 15:24:58 +0000 (15:24 +0000)]
Add a blurb on cross-compiling lin->win
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@616
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
