Frederic Suter [Fri, 27 Apr 2012 10:11:30 +0000 (12:11 +0200)]
global_routing doesn't exist anymore. If I'm correct, it's
routing_platf now
Frederic Suter [Fri, 27 Apr 2012 10:10:06 +0000 (12:10 +0200)]
remove unused variable that raises warning (when treated as error ...)
navarro [Fri, 27 Apr 2012 09:09:35 +0000 (11:09 +0200)]
Add a specific test for arguments start_time and kill_time of process tag.
Arnaud Legrand [Thu, 26 Apr 2012 23:54:18 +0000 (01:54 +0200)]
Tesh update.
The previous commit has added a few more simcals and the process now
gets killed before getting a chance to execute a task. Both behavior are
valid so let's take this one.
Arnaud Legrand [Thu, 26 Apr 2012 23:39:55 +0000 (01:39 +0200)]
Process that suspend themselves now create the corresponding surf action
A process that suspends himself did not create any surf action. When
processing the simcall, the maestro only set the process as suspended
and did not answer the simcall. Likewise a resume would simply wake up
the process. A problem with this is that if the corresponding host ever
went down, the suspended process would never have realized.
For the moment, just die whenever a process is suspended and it is not
waiting on an action that can be be suspended. The consequence is that
there is no reason to wake up a process that did not have any waiting
action. Again, DIE if this ever happens for the moment.
Arnaud Legrand [Thu, 26 Apr 2012 22:27:38 +0000 (00:27 +0200)]
Let's not assume that an unsigned long int always behave as a char *...
When it comes to such pointer arithmetic, char * is the way to go.
Plus plug a dumb memleak.
Sorry about the previous commit that did not work with full compilation flags.
Arnaud Legrand [Thu, 26 Apr 2012 22:24:22 +0000 (00:24 +0200)]
This function has no reason to go public.
Sorry about the previous buggy commit that did not worked with full warning flags.
Samuel Lepetit [Thu, 26 Apr 2012 15:24:24 +0000 (17:24 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Samuel Lepetit [Thu, 26 Apr 2012 15:24:15 +0000 (17:24 +0200)]
Fix liblua location on some linux distributions, like Archlinux.
Duco van Amstel [Thu, 26 Apr 2012 08:47:58 +0000 (10:47 +0200)]
Fixed some compilation issues (warnings) in last commits
Arnaud Legrand [Wed, 25 Apr 2012 23:09:02 +0000 (01:09 +0200)]
Hide pretty nasty issue.
In this example, a process is killed right at the same time as it starts
executing somthing. It creates the surf action but is killed before it
gets the opportunity to wait on it. Consequently, the kill does not clean
up the surf action (so technically, the corresponding resources are still
wasted).
This could be "solved" by doing the kill before the simcall that created
the surf action. But more generally, it raises the issue of all the
data that have been allocated during the MSG_task_execute. There is
something really annoying with the kill. The current MSG implementation
is clearly not protected against kills. In case of kill, when doing a
simcall, we should throw an exception and TRY CATCH everywhere. We're far
from this for now so I leave it for another time.
Arnaud Legrand [Wed, 25 Apr 2012 22:59:06 +0000 (00:59 +0200)]
Cosmetics.
Arnaud Legrand [Wed, 25 Apr 2012 22:56:11 +0000 (00:56 +0200)]
Minor tracing/debug output improvement.
Arnaud Legrand [Wed, 25 Apr 2012 22:16:22 +0000 (00:16 +0200)]
Do not try to resume a dying process. The corresponding surf structure
would already have been freed, and this would result in a mess. That's
also why we make sure that resume simcalls occurs before kill simcalls
in SIMIX_run.
Arnaud Legrand [Wed, 25 Apr 2012 22:14:37 +0000 (00:14 +0200)]
Ensure that the same process never appears twice in the list of process_to_run.
Otherwise, it would be awaken twice. The first time, it would execute then
wait on the completion of a new simcall. The second time it would appear
to return from the new simcall but the simcall would not have been
executed.
Since resume and kill are particular functions/simcalls that change
the behavior of other process, it is ok if the process was already in the
process_to_run list (although it must not appear twice). For other simcalls,
I think it is invalid if two simcalls lead the scheduling of the same
process.
Arnaud Legrand [Wed, 25 Apr 2012 22:09:03 +0000 (00:09 +0200)]
Make sure that kill simcalls are processed (with SIMIX_simcall_pre) at the end, right after resume simcalls, which are also done after any other type of simcall.
Arnaud Legrand [Wed, 25 Apr 2012 22:03:38 +0000 (00:03 +0200)]
Add a convenient dynar function to sort the dynar using the Dutch national flag technique.
Arnaud Legrand [Wed, 25 Apr 2012 20:04:59 +0000 (22:04 +0200)]
Fix a very old typo...
Arnaud Legrand [Mon, 2 Apr 2012 23:16:51 +0000 (01:16 +0200)]
Lame attempt to make cleanup and kill coexist
Arnaud Legrand [Mon, 2 Apr 2012 23:16:32 +0000 (01:16 +0200)]
On a timeout we do not wait anymore on a task.
-> Avoid keeping a pointer to an invalid object.
Arnaud Legrand [Mon, 2 Apr 2012 23:15:38 +0000 (01:15 +0200)]
This should indeed be useless so let's remove it.
Arnaud Legrand [Mon, 2 Apr 2012 14:19:31 +0000 (16:19 +0200)]
This commit partially reverts commit
4e0c8570051df3f7114e0097bf888a99946b6138.
Such a lazy resume seem to clash with the use of synchronization
structures. The problem is well illustrated with simgrid-boinc.
Surprisingly, simgrid/valgrind go completely crazy and segfaults in
stuff like pthread_getspecific.
schnorr [Wed, 25 Apr 2012 16:33:56 +0000 (18:33 +0200)]
[trace] let user code declare/set/push/pop states for hosts
details:
- useful for gantt-based views
navarro [Wed, 25 Apr 2012 10:38:38 +0000 (12:38 +0200)]
Add tesh for storag.
Martin Quinson [Wed, 25 Apr 2012 09:39:07 +0000 (11:39 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Wed, 25 Apr 2012 09:38:51 +0000 (11:38 +0200)]
add a link to the scientific tutorial in the 101 section of the documentatino
Arnaud Giersch [Mon, 23 Apr 2012 12:40:00 +0000 (14:40 +0200)]
Fix printf format strings.
Martin Quinson [Sun, 22 Apr 2012 19:18:42 +0000 (21:18 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Sun, 22 Apr 2012 19:17:24 +0000 (21:17 +0200)]
rewrite the documentation of MSG_task_get_from_host to reflect the fact that it is not implemented anymore
Martin Quinson [Sun, 22 Apr 2012 09:14:25 +0000 (11:14 +0200)]
add a word on the logo licensing: CC-BY-SA-3.0
Arnaud Giersch [Sat, 21 Apr 2012 16:51:28 +0000 (18:51 +0200)]
Add full references for the articles.
Marion Guthmuller [Fri, 20 Apr 2012 17:17:54 +0000 (19:17 +0200)]
model-checker : get pointer address for bytes detected as different in heap comparison algorithm
Marion Guthmuller [Fri, 20 Apr 2012 17:15:20 +0000 (19:15 +0200)]
model-checker : change XBT_DEBUG for fprintf
Marion Guthmuller [Fri, 20 Apr 2012 17:08:53 +0000 (19:08 +0200)]
model-checker : reindent, no change
Marion Guthmuller [Fri, 20 Apr 2012 17:03:20 +0000 (19:03 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot//simgrid/simgrid
Marion Guthmuller [Fri, 20 Apr 2012 17:02:58 +0000 (19:02 +0200)]
model-checker : separate bugged1_liveness (deleted) example in two cases : finite CS requests (bugged1_for_liveness) and infinite CS requests (bugged1_while_liveness)
Arnaud Giersch [Fri, 20 Apr 2012 13:11:42 +0000 (15:11 +0200)]
Avoid double free()s.
Arnaud Giersch [Fri, 20 Apr 2012 13:07:12 +0000 (15:07 +0200)]
Consolidate loops. Also remove useless initializations.
Arnaud Giersch [Fri, 20 Apr 2012 07:20:36 +0000 (09:20 +0200)]
Break recursion a bit earlier.
Compute the expected LCS length, and use it to stop as soon as the LCS
is complete.
Arnaud Giersch [Thu, 19 Apr 2012 09:08:53 +0000 (11:08 +0200)]
Add a --verbose option for testall.
Arnaud Giersch [Thu, 19 Apr 2012 08:59:52 +0000 (10:59 +0200)]
Fix parsing of options for testall.
Options such as --helpless or --dumpfoobar were wrongly treated as synonyms
for --help or --dump.
Arnaud Giersch [Wed, 18 Apr 2012 15:39:39 +0000 (17:39 +0200)]
Use pointer arithmetic for accessing to dynar elements.
It is more efficient than the xbt_dynar_get_* functions.
Assumes that dynar's contents is contiguous.
Arnaud Giersch [Wed, 18 Apr 2012 15:39:39 +0000 (17:39 +0200)]
Use a smarter algorithm for xbt_str_diff.
The previous one had a time and space complexity of O(n.m), with n and m
being the length of the inputs, and reached the memory limits pretty fast.
Implement the algorithm described in "An O(NP) Sequence Comparison Algorithm",
by Sun Wu, Udi Manber, Gene Myers, and Webb Miller (Information Processing
Letters 35(6):317-323, 1990), with the linear-space divide-and-conquer
strategy described in "An O(ND) Difference Algorithm and Its Variations",
by Eugene W. Myers (Algorithmica 1:251-266, 1986).
Christophe Thiéry [Tue, 17 Apr 2012 15:44:17 +0000 (17:44 +0200)]
Respect the coding style
Christophe Thiéry [Tue, 17 Apr 2012 15:41:51 +0000 (17:41 +0200)]
Remove tabs
navarro [Wed, 11 Apr 2012 10:13:00 +0000 (12:13 +0200)]
Fix documentation navbar and page for tour dynar.
navarro [Wed, 11 Apr 2012 09:22:58 +0000 (11:22 +0200)]
Fix ns3 and gtnets compilation
navarro [Wed, 11 Apr 2012 08:47:15 +0000 (10:47 +0200)]
Need those files to make dist
navarro [Wed, 11 Apr 2012 08:14:23 +0000 (10:14 +0200)]
Remove old files and add new files.
navarro [Tue, 10 Apr 2012 10:32:39 +0000 (12:32 +0200)]
Fix jedule compilation
navarro [Tue, 10 Apr 2012 10:24:05 +0000 (12:24 +0200)]
Fix compilation error with gcc4.4
error: redefinition of typedef 'surf_stat_t'
Martin Quinson [Mon, 9 Apr 2012 18:44:47 +0000 (08:44 -1000)]
[routing] rename routing_global_t -> routing_platf_t (and friends)
One day maybe, this could be usefull to enable the declaration (and
simulation) of several platform within one simgrid run. But for now, I
mainly want to make the sg/platf interface less context sensitive and
more usual by expressing directly the AS to which elements are added.
Martin Quinson [Mon, 9 Apr 2012 02:42:50 +0000 (16:42 -1000)]
[routing] kill supurious checks, and reformulate error messages
Martin Quinson [Mon, 9 Apr 2012 02:28:51 +0000 (16:28 -1000)]
[routing] clarify why we rely on such a dirty hack
Martin Quinson [Mon, 9 Apr 2012 02:14:32 +0000 (16:14 -1000)]
[route parsing] make surf_network_resource_by_name public under name sg_routing_edge_by_name_or_null
* This function existed already, but were almost never used, its
content being inlined in ways too much location.
* Use that function were relevant. That's much cleaner this way.
* Making it public is one step further to the C interface for routing
creation interface (drop in to XML)
Martin Quinson [Mon, 9 Apr 2012 01:58:21 +0000 (15:58 -1000)]
[route parsing] rename network_element_t into sg_routing_edge_t
Also, make this type visible from simgrid/platf.h since we need to
manipulate it to create routes efficiently
Martin Quinson [Mon, 9 Apr 2012 01:53:56 +0000 (15:53 -1000)]
ignorable reindent
Martin Quinson [Mon, 9 Apr 2012 01:43:28 +0000 (15:43 -1000)]
[route parsing] kill global_routing->size_of_link
- Actually, all links we use are pointers to structure (how
surprising). This strange thing survived until now because the
constant model declared to have a size of links being the size of a
double. But actually, since there is no link in constant routing, we
could use 0 as size of links in that case.
- Also introduce a sg_routing_link_t that is simply an alias to void*
for now. It is intended to represent what is stored into link_lib at
the surf level (that is the core-most link concept)
Martin Quinson [Mon, 9 Apr 2012 01:30:53 +0000 (15:30 -1000)]
[route parsing] greatly streamline the generic_new_extended_route function
Martin Quinson [Mon, 9 Apr 2012 01:16:46 +0000 (15:16 -1000)]
[route parsing] reduce the amount of dupplicated code
the generic_new_extended_route() can do the work of
generic_new_route(), so kill the simple one and use the complete one
instead.
Martin Quinson [Mon, 9 Apr 2012 01:12:01 +0000 (15:12 -1000)]
avoid void* pointers since we know the pointed type in all usage location of this function
schnorr [Thu, 5 Apr 2012 14:55:33 +0000 (16:55 +0200)]
[trace] bugfix
schnorr [Thu, 5 Apr 2012 12:57:43 +0000 (14:57 +0200)]
[trace] we need to close the file
schnorr [Thu, 5 Apr 2012 12:50:19 +0000 (14:50 +0200)]
file presumed to be installed by other header files
navarro [Thu, 5 Apr 2012 12:08:33 +0000 (14:08 +0200)]
Fix some documenttion bar
schnorr [Thu, 5 Apr 2012 11:22:54 +0000 (13:22 +0200)]
[trace] updating tesh due to buffering by default
schnorr [Thu, 5 Apr 2012 11:19:41 +0000 (13:19 +0200)]
[trace] it seems that buffering affects trace alias generation, fix tesh with no harm
schnorr [Thu, 5 Apr 2012 11:10:24 +0000 (13:10 +0200)]
[trace] example of creating a customized graph configuration file
schnorr [Thu, 5 Apr 2012 10:54:03 +0000 (12:54 +0200)]
[trace] functions needed to create graph configuration files for Triva by hand
schnorr [Thu, 5 Apr 2012 10:45:54 +0000 (12:45 +0200)]
[trace] fix comment
schnorr [Thu, 5 Apr 2012 10:43:02 +0000 (12:43 +0200)]
[trace] not tracing to test the absense of output
schnorr [Thu, 5 Apr 2012 10:42:19 +0000 (12:42 +0200)]
[trace] pay attention to the return value of get functions, they might be null
schnorr [Thu, 5 Apr 2012 10:38:51 +0000 (12:38 +0200)]
[trace] factor code
schnorr [Thu, 5 Apr 2012 10:38:26 +0000 (12:38 +0200)]
[trace] let code compile without tracing
schnorr [Thu, 5 Apr 2012 10:35:01 +0000 (12:35 +0200)]
[trace] test the TRACE_get_[host|link]_variables
schnorr [Thu, 5 Apr 2012 10:29:04 +0000 (12:29 +0200)]
[trace] functions to get already declared user host/link tracing variables
schnorr [Thu, 5 Apr 2012 10:20:59 +0000 (12:20 +0200)]
[trace] check if user variables are being correctly used
schnorr [Thu, 5 Apr 2012 09:11:47 +0000 (11:11 +0200)]
[trace] update example to show how to get declared marks
schnorr [Thu, 5 Apr 2012 09:10:07 +0000 (11:10 +0200)]
[trace] new function to get the declared marks
schnorr [Thu, 5 Apr 2012 08:58:23 +0000 (10:58 +0200)]
[trace] move categories dictionary declaration to interface file
schnorr [Thu, 5 Apr 2012 08:52:49 +0000 (10:52 +0200)]
[trace] update example to show how to use TRACE_get_categories
schnorr [Thu, 5 Apr 2012 08:47:52 +0000 (10:47 +0200)]
[trace] new public function to get declared categories
schnorr [Wed, 4 Apr 2012 15:20:55 +0000 (17:20 +0200)]
[trace] register power and bandwidth for hosts/links when --cfg=tracing/platform:1
Martin Quinson [Thu, 5 Apr 2012 01:05:49 +0000 (15:05 -1000)]
make clear to simgrider how to edit that tutorial, and regenerate all files
Henri Casanova [Thu, 5 Apr 2012 00:06:40 +0000 (14:06 -1000)]
doc rewording
Henri Casanova [Thu, 5 Apr 2012 00:00:39 +0000 (14:00 -1000)]
Doc rewording
navarro [Wed, 4 Apr 2012 11:57:35 +0000 (13:57 +0200)]
Add default Layout
navarro [Wed, 4 Apr 2012 11:57:23 +0000 (13:57 +0200)]
Fix warning by doxygen
navarro [Wed, 4 Apr 2012 11:27:02 +0000 (13:27 +0200)]
Use of bypassASroute.xml instead of three_clusters_bypass.xml
navarro [Wed, 4 Apr 2012 11:26:22 +0000 (13:26 +0200)]
This was necessary. Remove previously
navarro [Wed, 4 Apr 2012 11:19:57 +0000 (13:19 +0200)]
Made two platforms for bypass and add test to tesh
navarro [Wed, 4 Apr 2012 10:44:37 +0000 (12:44 +0200)]
Implement bypassRoute and bypassASroute. Add an example.
Martin Quinson [Wed, 4 Apr 2012 08:26:52 +0000 (22:26 -1000)]
ben si
Martin Quinson [Wed, 4 Apr 2012 05:38:31 +0000 (19:38 -1000)]
also include the 101 presentation into the latex documentation
Martin Quinson [Wed, 4 Apr 2012 05:21:14 +0000 (19:21 -1000)]
add a SimGrid 101 presentation as a quick starter for potential users
Martin Quinson [Wed, 4 Apr 2012 04:46:51 +0000 (18:46 -1000)]
[DOC] hide a bit the structure list, nobody cares
Martin Quinson [Wed, 4 Apr 2012 04:39:04 +0000 (18:39 -1000)]
[DOC] cosmetics: rename some tabs in menus
Martin Quinson [Wed, 4 Apr 2012 04:30:17 +0000 (18:30 -1000)]
[DOC] kill the datatype specific page
all the data are now in their home module, alongside with the function
modifying them.
Martin Quinson [Wed, 4 Apr 2012 04:21:33 +0000 (18:21 -1000)]
Doc improvement: rename GOS into task usage
who remember of Globus GOS anyway? And Operating System-like names
sound really strange now that the 21 century finally came.
Martin Quinson [Wed, 4 Apr 2012 04:06:05 +0000 (18:06 -1000)]
improve the doc of MSG datatypes (and dispatch them in the relevant modules)