donassbr [Thu, 27 Aug 2009 11:25:12 +0000 (11:25 +0000)]
Remove unnecessary changes at maxmin.c.
Error when updating max_duration of action.
Make indent.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6643
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
donassbr [Wed, 26 Aug 2009 15:34:29 +0000 (15:34 +0000)]
Put SG_BEGIN_DECL and SG_END_DECL again.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6642
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 26 Aug 2009 15:19:55 +0000 (15:19 +0000)]
Stop saying stupidities: no perf loss once well compiled
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6641
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
donassbr [Wed, 26 Aug 2009 15:05:59 +0000 (15:05 +0000)]
Merge changes of maxmin_selective_update branch into the trunk.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6640
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Tue, 25 Aug 2009 14:45:15 +0000 (14:45 +0000)]
Test for debug state before calling debug printing function in
simix_action.c. This saves a lots of unecessary calls.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6636
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 19:23:05 +0000 (19:23 +0000)]
This file does not exist anymore after xbt_dynar_foreach inlining
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6622
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 19:22:10 +0000 (19:22 +0000)]
Forgot the date in the ChangeLog
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6621
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 15:39:56 +0000 (15:39 +0000)]
release 3.3.3
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6619
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 14:18:17 +0000 (14:18 +0000)]
Also inline xbt_swag_belongs
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6618
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 13:53:32 +0000 (13:53 +0000)]
Inline the functions used in xbt_dynar_foreach
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6617
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 20 Aug 2009 13:53:29 +0000 (13:53 +0000)]
Connect the surf_routing log channel so that it becomes usable (shows a bug in log maybe, I thought it was only mandatory on windows, not linux)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6616
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 16:37:59 +0000 (16:37 +0000)]
missing file in the archive
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6612
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 16:16:49 +0000 (16:16 +0000)]
init development of 3.3.3
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6611
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 15:16:36 +0000 (15:16 +0000)]
Add a end of line to this file
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6608
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 15:14:55 +0000 (15:14 +0000)]
Release 3.3.2
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6607
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:58:31 +0000 (13:58 +0000)]
Force -O0 when the --disable-compiler-optimization flag is passed to configure; not sure what the default value is
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6606
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:51:22 +0000 (13:51 +0000)]
GARGH. Fix a *nasty* long standing bug in dictionaries (many thanks to Cristian)
The deal is that strcmp was used regardless of whether the entry was
created through the _ext interface or not. So, when you used pointers
as key (as we do in gras datadesc for cycle detection for example),
the key got truncated on the first \0 (and the rest of it was
uninitialized).
This is what caused my bug with gras datadesc tests. And it happened
only some times because the pointer does not always contain a 0. And
it happened more often on 64bits because the pointers are longer and
thus have a bigger probability to contain \0.
Actually, this bug was already fixed by Cristian on 2008-11-25, but on
the mc branch of our private git in Nancy, and never propagated to the
trunk. We are still git childrens...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6605
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:59 +0000 (13:50 +0000)]
no need for that verbose output anymore
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6604
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:57 +0000 (13:50 +0000)]
More verbose on cycle detection
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6603
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:35 +0000 (13:50 +0000)]
Reordering
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6602
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:33 +0000 (13:50 +0000)]
Great simplification in a code appearing to be broken on amd64 (cycle detection during marshalisation)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6601
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:10 +0000 (13:50 +0000)]
use the right syntax to set the path (plus shut the config module off)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6600
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 19 Aug 2009 13:50:08 +0000 (13:50 +0000)]
more explicit message on problem
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6599
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 18 Aug 2009 14:25:56 +0000 (14:25 +0000)]
Get the simple log format omit any file location when --log=no_loc is passed
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6597
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 18 Aug 2009 14:25:53 +0000 (14:25 +0000)]
More fixes to VPATH testing
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6596
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 18 Aug 2009 14:24:57 +0000 (14:24 +0000)]
Various fixes to the java building infrastructure (now working in VPATH builds)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6595
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Tue, 18 Aug 2009 14:01:09 +0000 (14:01 +0000)]
The exception associated to each process was relocated from the
context to the simix process that contains it. This is both faster and
simpler because the initialization code was factorized and exception
fetching uses one dereference less.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6594
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Tue, 18 Aug 2009 12:53:39 +0000 (12:53 +0000)]
Merge java simix process creation with the standard process creation
function.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6592
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Tue, 18 Aug 2009 09:24:56 +0000 (09:24 +0000)]
Fix bug related to the exception mechanism once the context module exits
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6589
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 16:35:07 +0000 (16:35 +0000)]
Add a missing file to the archive
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6588
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 15:49:23 +0000 (15:49 +0000)]
follow the advice of autoreconf
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6587
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 15:49:21 +0000 (15:49 +0000)]
Use the right guard for snprintf (thanks to ubuntu stack for detecting the potential buffer overflow)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6586
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 15:35:13 +0000 (15:35 +0000)]
Update the ChangeLog
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6585
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 15:00:27 +0000 (15:00 +0000)]
Kill a dupplicated target in makefile
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6584
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 15:00:25 +0000 (15:00 +0000)]
update tesh files (I assume that ordering differances in messages are due to bcast reimplementation using tree)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6583
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 14:41:18 +0000 (14:41 +0000)]
Finally the Java interface is working again! (It was broken because of
me, sorry :S)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6579
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 14:34:39 +0000 (14:34 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6578
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 14:20:46 +0000 (14:20 +0000)]
display more info during the test
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6577
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 14:18:07 +0000 (14:18 +0000)]
Trying to fix another java bug
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6576
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 13:57:57 +0000 (13:57 +0000)]
Fix another bug in the java context implementation
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6575
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 13:49:25 +0000 (13:49 +0000)]
Fixed bug in smx_context_mod_exit.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6574
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 13:47:39 +0000 (13:47 +0000)]
Output update in tesh file
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6573
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 13:37:40 +0000 (13:37 +0000)]
Kill a off by one bug appearing in some rare conditions
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6572
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 12:43:16 +0000 (12:43 +0000)]
Kill context test since contextes cannot be used without simix anymore
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6571
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
cristianrosa [Mon, 17 Aug 2009 12:40:52 +0000 (12:40 +0000)]
Fix try to fix the JAVA crap!
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6570
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 12:22:54 +0000 (12:22 +0000)]
be more friendly in tests with VPATH builds
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6569
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 11:33:27 +0000 (11:33 +0000)]
update output in tesh files (line number changes again)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6568
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 11:24:40 +0000 (11:24 +0000)]
Stupid me. I still need to put the retrieved value into the variable
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6567
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 10:22:32 +0000 (10:22 +0000)]
Context factory's API simplification: the function for creating the
context for maestro was merged with the function for the creation
of standard contexts.
The functions for managing the context module where simplified.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6566
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 10:15:29 +0000 (10:15 +0000)]
SMPI: Fix the prototype of SMPI_MPI_Type_get_extent (use MPI_Aint instead of int since types differ on 64bits)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6565
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:24:47 +0000 (09:24 +0000)]
Fixed the java's context interface to build with the context module layout [Cristian]
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6564
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:24:23 +0000 (09:24 +0000)]
Fixed some issues in logs logic due to recent changes [Cristian]
We assumed that maestro have no simix process (not true anymore)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6563
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:24:20 +0000 (09:24 +0000)]
Some polishing to the new context factory API to avoid accesing simix
data structures from the context implementations.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6562
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:23:53 +0000 (09:23 +0000)]
New simplified API for the context factory [Cristian]
It provides functionality for context creation/destruction, start,
stop, resume and suspend. The rest of the logic was factorized and
moved to the SIMIX layer, resulting in a simpler and cleaner
implementation.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6561
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:23:50 +0000 (09:23 +0000)]
Intermidiate step towards the new context mechanism [Cristian]
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6560
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:23:26 +0000 (09:23 +0000)]
Fixed the pthread context implementation, now it compiles and runs
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6559
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 17 Aug 2009 09:23:24 +0000 (09:23 +0000)]
Inline abstract interface to factories' functions (smx_context_* func) [Cristian]
Also relocate them in simix/private.h.
They are now used as macros for simplifying the code.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6558
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 14 Aug 2009 21:05:23 +0000 (21:05 +0000)]
C++ needs love
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6557
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 11 Aug 2009 12:34:12 +0000 (12:34 +0000)]
Add entries about GTNetS and alternate flow models.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6547
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
alegrand [Tue, 11 Aug 2009 12:34:09 +0000 (12:34 +0000)]
Add carriage returns to help gcc compiling... (otherwise, it fails!).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6546
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sun, 9 Aug 2009 13:10:59 +0000 (13:10 +0000)]
with GTNetS, the routing is not within to SimGrid but externalized
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6543
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Sat, 1 Aug 2009 08:19:02 +0000 (08:19 +0000)]
Alltoallv
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6539
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Fri, 31 Jul 2009 17:57:42 +0000 (17:57 +0000)]
started alltoallv(). Still some bugs.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6538
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 30 Jul 2009 20:33:52 +0000 (20:33 +0000)]
Various cleanups in the routing code, all by Silas
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6534
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 30 Jul 2009 20:02:01 +0000 (20:02 +0000)]
update work status
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6533
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 30 Jul 2009 19:51:17 +0000 (19:51 +0000)]
alltoall implemented (almost opmpi algorithms)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6532
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 30 Jul 2009 13:55:22 +0000 (13:55 +0000)]
thought this one would fix my long-standing wait_all() bug.
Not yet, but obviously better.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6530
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
velho [Tue, 28 Jul 2009 13:34:18 +0000 (13:34 +0000)]
Branch to work on maxmin selective update, consists of flaging portion of the system that change instead of computing the whole system each time a small change is detected.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6527
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
velho [Tue, 28 Jul 2009 13:06:56 +0000 (13:06 +0000)]
Cope with new tmgr (trace manager) interface.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6526
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Tue, 21 Jul 2009 13:41:23 +0000 (13:41 +0000)]
* minimum of datatype handling for alltoall tuned
=> start add files for separate topics, e.g smpi_mpi_dt.{c,h}
* started debugging alltoall basic_linear
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6525
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Mon, 20 Jul 2009 19:23:06 +0000 (19:23 +0000)]
... untested ...
* second part in tuned coll alltoall: basic_linear
* enriched a bit more datatypes (ub,lb).
MPI_Type_get_extent() . -Does not allow datatype creation.)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6524
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 19:46:02 +0000 (19:46 +0000)]
shut of, model
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6523
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 19:12:58 +0000 (19:12 +0000)]
Allow to instanciate the constant network model. It uses a newly created routing schema 'none' doing nothing, and every communication takes exactly 1s
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6522
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:53 +0000 (18:51 +0000)]
Three new routing schema by Silas De Munck
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6521
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:46 +0000 (18:51 +0000)]
Automatically switch to compound workstation model if the user changes the network or the CPU. If not, his choice is ignored
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6520
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:40 +0000 (18:51 +0000)]
config returns NULL if a string is looked for, but not available
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6519
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:33 +0000 (18:51 +0000)]
bummer
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6518
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:26 +0000 (18:51 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6517
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 16 Jul 2009 18:51:17 +0000 (18:51 +0000)]
Call surf_timer_model_init() from surf_config_models_setup() so that upper layer don't have to do it manually
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6516
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 16 Jul 2009 15:25:18 +0000 (15:25 +0000)]
* MPI_Sendrecv user+internal levels
* added another test for alltoall
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6515
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 16 Jul 2009 10:36:34 +0000 (10:36 +0000)]
update test
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6514
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Thu, 16 Jul 2009 08:49:25 +0000 (08:49 +0000)]
separate user-level and internal-level for SMPI_MPI_{Reduce,Bcast}
Allreduce ok.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6513
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:47:44 +0000 (16:47 +0000)]
Compile the sendrecv example
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6512
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:47:38 +0000 (16:47 +0000)]
Yet more debug
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6511
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:38:34 +0000 (16:38 +0000)]
todo++
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6510
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:38:32 +0000 (16:38 +0000)]
do not call smpi_bench_* from SMPI_MPI_Allreduce since it calls SMPI_MPI_* function, breaking the parenthesing of smpi_bench_* functions
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6509
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:38:09 +0000 (16:38 +0000)]
More debug
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6508
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 15 Jul 2009 16:38:07 +0000 (16:38 +0000)]
Only sleep to handle real computation if the time to sleep is large enough; do not protect the access to the timer by mutex: there is no concurency in simgrid
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6507
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Wed, 15 Jul 2009 15:35:40 +0000 (15:35 +0000)]
examples to demonstrate bugs in sendrecv
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6506
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Jul 2009 08:46:36 +0000 (08:46 +0000)]
one more comment
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6505
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Jul 2009 08:46:30 +0000 (08:46 +0000)]
gosh, this file is outdated
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6504
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 14 Jul 2009 08:44:33 +0000 (08:44 +0000)]
Do not ignore fread results to keep ubuntu compiler happy
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6503
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Mon, 13 Jul 2009 22:02:42 +0000 (22:02 +0000)]
MPI_Sendrecv
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6502
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
genaud [Mon, 13 Jul 2009 21:20:08 +0000 (21:20 +0000)]
+ MPI_Sendrecv()
+ part of optimized MPI_Alltoall() (pairwise algo)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6501
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:08:36 +0000 (18:08 +0000)]
revalidate output after recent source changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6500
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:08:31 +0000 (18:08 +0000)]
do not give maestro a name, or it will change every output, invalidating any tests
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6499
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:04:22 +0000 (18:04 +0000)]
Finish the merging of Cristian branch by fixing some of the remaining glitches preventing the stuff to compile
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6498
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:04:15 +0000 (18:04 +0000)]
Lots of renamings from xbt_ to smx_
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6497
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:04:09 +0000 (18:04 +0000)]
First running version after the relocation of the context module [Cristian]
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6496
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 13 Jul 2009 18:04:03 +0000 (18:04 +0000)]
First step into moving the context subsystem into simix. Still segfaults because of exceptions late initialization [Cristian]
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@6495
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
