Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
simgrid.git
4 years agoMerge pull request #59 from fabienchaix/master
Martin Quinson [Sat, 6 Feb 2016 00:02:36 +0000 (01:02 +0100)]
Merge pull request #59 from fabienchaix/master

Splitting the LMM constraint element_set into enabled_element_set and disabled_element_set

4 years agothat test does not need to be sooo long
Martin Quinson [Fri, 5 Feb 2016 23:50:02 +0000 (00:50 +0100)]
that test does not need to be sooo long

4 years agoconvert Jedule bindings to C++
Martin Quinson [Fri, 5 Feb 2016 23:48:25 +0000 (00:48 +0100)]
convert Jedule bindings to C++

4 years agokill platform generation from the C sources
Martin Quinson [Fri, 5 Feb 2016 23:27:54 +0000 (00:27 +0100)]
kill platform generation from the C sources

This should be reimplemented in Lua anyway. It was not tested, and I
stumbled upon it during another refactoring and got upset. RIP.

4 years agoMerge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Fri, 5 Feb 2016 20:55:44 +0000 (21:55 +0100)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid

4 years agoremove stochastic traces for now.
Martin Quinson [Fri, 5 Feb 2016 20:46:18 +0000 (21:46 +0100)]
remove stochastic traces for now.

They were not even tested anyway. They should be reimplemented once
trace_mgr is properly objectified, this way or maybe with lua.

4 years agocleaning some oldies
Frederic Suter [Fri, 5 Feb 2016 18:23:21 +0000 (19:23 +0100)]
cleaning some oldies

ok to keep old optorsim files, but not xml that were generated with an
even older version of the provided script.

If one wants to try these GridPP and LCG description circa 2004, just
has to run the script to get the XML.

4 years agoTry to make now parameter of Model::shareResources() useless
Martin Quinson [Fri, 5 Feb 2016 16:06:59 +0000 (17:06 +0100)]
Try to make now parameter of Model::shareResources() useless

4 years agoput back some fix for a warning
degomme [Fri, 5 Feb 2016 15:02:50 +0000 (16:02 +0100)]
put back some fix for a warning

4 years ago(try to) Unify detection of windows. _WIN32 may not be defined for 64 bits systems
degomme [Fri, 5 Feb 2016 14:58:59 +0000 (15:58 +0100)]
(try to) Unify detection of windows. _WIN32 may not be defined for 64 bits systems

4 years agoThis file was not supposed to be still manually privatized
degomme [Fri, 5 Feb 2016 14:54:18 +0000 (15:54 +0100)]
This file was not supposed to be still manually privatized
There its _manual version for this case (systems without automatic privatization support).

4 years agoput back macros for manual privatization
Frederic Suter [Fri, 5 Feb 2016 14:07:14 +0000 (15:07 +0100)]
put back macros for manual privatization

4 years agoattempt to remove oudated coccinelle stuff
Frederic Suter [Fri, 5 Feb 2016 13:34:34 +0000 (14:34 +0100)]
attempt to remove oudated coccinelle stuff

4 years agoprivatization with coccinelle is outdated
Frederic Suter [Fri, 5 Feb 2016 13:00:57 +0000 (14:00 +0100)]
privatization with coccinelle is outdated

4 years agotest parsing of parallel task graphs in dot
Frederic Suter [Fri, 5 Feb 2016 09:28:35 +0000 (10:28 +0100)]
test parsing of parallel task graphs in dot

code was there for a while but not added to the tesh file.

4 years agouseless cosmetics
Martin Quinson [Thu, 4 Feb 2016 21:23:07 +0000 (22:23 +0100)]
useless cosmetics

4 years agodon't compute the amount of hosts if we don't use it
Martin Quinson [Thu, 4 Feb 2016 21:21:03 +0000 (22:21 +0100)]
don't compute the amount of hosts if we don't use it

4 years agocosmetics
Martin Quinson [Thu, 4 Feb 2016 20:49:14 +0000 (21:49 +0100)]
cosmetics

4 years agoleak--, bad indentation-- too
Frederic Suter [Thu, 4 Feb 2016 15:54:06 +0000 (16:54 +0100)]
leak--, bad indentation-- too

4 years agoleak plug and cosmetics
Frederic Suter [Thu, 4 Feb 2016 14:14:05 +0000 (15:14 +0100)]
leak plug and cosmetics

4 years agocoverage madness: use asserts here too
Frederic Suter [Thu, 4 Feb 2016 10:37:30 +0000 (11:37 +0100)]
coverage madness: use asserts here too

4 years agoUnify naming of contexts support.
degomme [Thu, 4 Feb 2016 09:22:37 +0000 (10:22 +0100)]
Unify naming of contexts support.
HAVE_RAWCTX -> HAVE_RAW_CONTEXTS
HAVE_BOOST_CONTEXT -> HAVE_BOOST_CONTEXTS
CONTEXT_UCONTEXT -> HAVE_UCONTEXT_CONTEXTS
CONTEXT_THREADS -> HAVE_THREAD_CONTEXTS

4 years agoAs we don't have thread contexts everywhere (windows, for now), don't execute tests...
degomme [Thu, 4 Feb 2016 09:05:45 +0000 (10:05 +0100)]
As we don't have thread contexts everywhere (windows, for now), don't execute tests using it

4 years agonormalize the handling of speed traces in CpuTi
Martin Quinson [Wed, 3 Feb 2016 20:41:14 +0000 (21:41 +0100)]
normalize the handling of speed traces in CpuTi

4 years agoFurther untangle trace:connect. make trace_connect_list_host_avail private
Martin Quinson [Wed, 3 Feb 2016 14:28:17 +0000 (15:28 +0100)]
Further untangle trace:connect. make trace_connect_list_host_avail private

4 years agoadd the bug number in the test case
Martin Quinson [Wed, 3 Feb 2016 09:47:12 +0000 (10:47 +0100)]
add the bug number in the test case

4 years agoupdate distrib
Frederic Suter [Wed, 3 Feb 2016 10:29:04 +0000 (11:29 +0100)]
update distrib

4 years agothese scripts are outdated and thus misleading
Frederic Suter [Wed, 3 Feb 2016 10:25:00 +0000 (11:25 +0100)]
these scripts are outdated and thus misleading

they are superseeded by https://github.com/lpouillo/topo5k
At some point we should maintain a contrib platform section with
extracts of g5k obtained with this tool

4 years agowe can live without these old and useless scripts
Frederic Suter [Wed, 3 Feb 2016 10:03:00 +0000 (11:03 +0100)]
we can live without these old and useless scripts

4 years agoduplicate of tools/MSG_visualization/colorize.pl
Frederic Suter [Wed, 3 Feb 2016 09:42:51 +0000 (10:42 +0100)]
duplicate of tools/MSG_visualization/colorize.pl

no need for two instances. that in MSG_visualization is the one
copied as $INSTALL_PATH/bin/simgrid-colorizer

4 years agofix my previous commit wrt make dist, sorry
Martin Quinson [Wed, 3 Feb 2016 08:20:13 +0000 (09:20 +0100)]
fix my previous commit wrt make dist, sorry

4 years agolittle comment improvement
Martin Quinson [Wed, 3 Feb 2016 08:05:46 +0000 (09:05 +0100)]
little comment improvement

4 years agoMerge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Wed, 3 Feb 2016 07:35:47 +0000 (08:35 +0100)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid

4 years agoadd a test case for the JVM slaying our semaphores
Martin Quinson [Wed, 3 Feb 2016 07:35:31 +0000 (08:35 +0100)]
add a test case for the JVM slaying our semaphores

4 years agoCleanup in this test case
Martin Quinson [Wed, 3 Feb 2016 07:34:29 +0000 (08:34 +0100)]
Cleanup in this test case

4 years agodon't destroy synchro in a simcall, this drives the JVM nuts
Martin Quinson [Tue, 2 Feb 2016 22:08:24 +0000 (23:08 +0100)]
don't destroy synchro in a simcall, this drives the JVM nuts

The destroy is called from the GC thread when running in the JVM. And
that thread is obviously not a simix thread, so the simcall mechanism
won't work properly.

I tried to make it a simix thread when I see it (ie, when
JavaThreadFactory::self() is about to return NULL), but it leads to
further complications, such as the following error message:

  Can't extend stack to 0x22f35958 during signal delivery for thread 9:
  no stack segment

Oha. That hurts. I've no idea of why that thread has no stack, and I
think I don't want to know. Actually, it makes no sense to delete the
synchronization objects from a simcall only. Why would this need to be
linearized in a reproductible way? So, delete these objects directly,
and everything works well.

That bug was reported in:
https://gforge.inria.fr/tracker/index.php?func=detail&aid=19893&group_id=12&atid=165

4 years agocosmetics
Martin Quinson [Tue, 2 Feb 2016 21:07:59 +0000 (22:07 +0100)]
cosmetics

4 years ago[Changelog] Reordered BC breaks for clarity
Christian Heinrich [Tue, 2 Feb 2016 16:37:04 +0000 (17:37 +0100)]
[Changelog] Reordered BC breaks for clarity

4 years ago[Changelog] Updated changelog to reflect the cfg change
Christian Heinrich [Tue, 2 Feb 2016 16:29:09 +0000 (17:29 +0100)]
[Changelog] Updated changelog to reflect the cfg change

4 years ago[XBT] Cosmetics; helps with cg/vg
Christian Heinrich [Tue, 2 Feb 2016 16:20:34 +0000 (17:20 +0100)]
[XBT] Cosmetics; helps with cg/vg

Return type and function name in one line help with
finding the right match when using the 'cgvg' tool.

4 years ago[XBT] Expanded error message for cfg values
Christian Heinrich [Tue, 2 Feb 2016 16:16:58 +0000 (17:16 +0100)]
[XBT] Expanded error message for cfg values

This should help users understand why their code breaks; we don't rename
very often so I think this is sufficient.

4 years ago[SMPI] Renamed option send_is_detached_thres to send_is_detached_thresh
Christian Heinrich [Tue, 2 Feb 2016 16:14:50 +0000 (17:14 +0100)]
[SMPI] Renamed option send_is_detached_thres to send_is_detached_thresh

4 years agoI'm told I have to care about users...
Frederic Suter [Tue, 2 Feb 2016 15:21:42 +0000 (16:21 +0100)]
I'm told I have to care about users...

4 years agoend of simplification of sd_dotloader
Frederic Suter [Tue, 2 Feb 2016 13:31:45 +0000 (14:31 +0100)]
end of simplification of sd_dotloader

4 years agostart to simplify this dot parsing stuff
Frederic Suter [Tue, 2 Feb 2016 11:07:23 +0000 (12:07 +0100)]
start to simplify this dot parsing stuff

4 years ago[INSTR] Cosmetics for tracing options.
Christian Heinrich [Sat, 30 Jan 2016 18:09:52 +0000 (19:09 +0100)]
[INSTR] Cosmetics for tracing options.

Sorting + indentation

4 years ago[DOC] Added stubs for undocumented tracing options
Christian Heinrich [Sat, 30 Jan 2016 18:07:26 +0000 (19:07 +0100)]
[DOC] Added stubs for undocumented tracing options

These options still need to be documented, but I have currently
not enough knowledge about what they all do.

4 years ago[DOC] Removed option from doc as it was duplicated
Christian Heinrich [Sat, 30 Jan 2016 18:02:17 +0000 (19:02 +0100)]
[DOC] Removed option from doc as it was duplicated

4 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr/gitroot/simgrid/simgrid
degomme [Tue, 2 Feb 2016 08:43:52 +0000 (09:43 +0100)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr/gitroot/simgrid/simgrid

4 years agoFix build with enable_latency_bound_tracking
degomme [Tue, 2 Feb 2016 08:43:44 +0000 (09:43 +0100)]
Fix build with enable_latency_bound_tracking

4 years agoMerge pull request #1 from mquinson/master
fabienchaix [Tue, 2 Feb 2016 06:14:31 +0000 (08:14 +0200)]
Merge pull request #1 from mquinson/master

Preparing to merge with new commits

4 years agoadd the lualib patch in our tarball, and update the corresponding doc
Martin Quinson [Mon, 1 Feb 2016 22:14:53 +0000 (23:14 +0100)]
add the lualib patch in our tarball, and update the corresponding doc

4 years agosome more cleanups in the documentation
Martin Quinson [Mon, 1 Feb 2016 21:58:41 +0000 (22:58 +0100)]
some more cleanups in the documentation

4 years agofurther extend the doc on writing unit tests
Martin Quinson [Mon, 1 Feb 2016 21:18:40 +0000 (22:18 +0100)]
further extend the doc on writing unit tests

4 years agocosmetics: rename some documentation files, just for clarity
Martin Quinson [Mon, 1 Feb 2016 17:59:47 +0000 (18:59 +0100)]
cosmetics: rename some documentation files, just for clarity

4 years agorewrite the doc around the tests
Martin Quinson [Mon, 1 Feb 2016 16:32:45 +0000 (17:32 +0100)]
rewrite the doc around the tests

4 years agokill dead code
Martin Quinson [Mon, 1 Feb 2016 15:24:39 +0000 (16:24 +0100)]
kill dead code

4 years agoFix a link in the doc
Martin Quinson [Mon, 1 Feb 2016 13:16:50 +0000 (14:16 +0100)]
Fix a link in the doc

4 years agoGet rid of super specific gcovr version (maybe there was a reason?)
degomme [Mon, 1 Feb 2016 13:52:07 +0000 (14:52 +0100)]
Get rid of super specific gcovr version (maybe there was a reason?)

4 years agohave the two versions of EP together
Frederic Suter [Mon, 1 Feb 2016 12:18:06 +0000 (13:18 +0100)]
have the two versions of EP together

 + avoid replication
 + solve some casting issues too

4 years agofurther cleaning in NAS
Frederic Suter [Mon, 1 Feb 2016 11:55:13 +0000 (12:55 +0100)]
further cleaning in NAS
 + fortran files are not used (and not smpified)
 + Don't need a MPI_dummy implementation, we have SMPI
 + have DT and DT-folding in the same dir to avoid replication of
   DGraph.*

4 years agoplease dist check
Frederic Suter [Mon, 1 Feb 2016 11:05:48 +0000 (12:05 +0100)]
please dist check

4 years agosome cleaning in the NAS examples
Frederic Suter [Mon, 1 Feb 2016 10:45:00 +0000 (11:45 +0100)]
some cleaning in the NAS examples

4 years agocosmetics
Frederic Suter [Fri, 29 Jan 2016 15:23:38 +0000 (16:23 +0100)]
cosmetics

4 years agosome small optimizations
Frederic Suter [Fri, 29 Jan 2016 14:47:33 +0000 (15:47 +0100)]
some small optimizations

4 years agofix build with lua
degomme [Mon, 1 Feb 2016 10:08:57 +0000 (11:08 +0100)]
fix build with lua

4 years ago[appveyor] don't get libwinptread as an artefact. It's bad and fails
Martin Quinson [Mon, 1 Feb 2016 06:59:49 +0000 (07:59 +0100)]
[appveyor] don't get libwinptread as an artefact. It's bad and fails

It reports "Packaging artifacts...Pattern contains invalid characters".

And actually, that's not very cleaver to have that lib as an artefact
of each and every commit on appveyor. Even more knowing that it should
be included in the simgrid.jar artefact by now.

4 years agorevalidate a tesh file after recent changes to the LMM
Martin Quinson [Mon, 1 Feb 2016 06:56:05 +0000 (07:56 +0100)]
revalidate a tesh file after recent changes to the LMM

Commit 7f63f68a348493dfd6c702c437aed38af4024789 forces some actions to
be serialized in this test (because there is more than 250 concurrent
operations)

4 years agoMerge pull request #57 from fabienchaix/oldstyle_element_set
Martin Quinson [Mon, 1 Feb 2016 06:40:53 +0000 (07:40 +0100)]
Merge pull request #57 from fabienchaix/oldstyle_element_set

Adding concurrency to SURF LMM, with oldstyle element set

4 years agotry another trick to get winpthread properly in the jar
Martin Quinson [Mon, 1 Feb 2016 06:36:34 +0000 (07:36 +0100)]
try another trick to get winpthread properly in the jar

4 years agoThere is no way to disable libwinpthread-1.dll on mingw-64
Martin Quinson [Sun, 31 Jan 2016 22:03:11 +0000 (23:03 +0100)]
There is no way to disable libwinpthread-1.dll on mingw-64

so let's try this workaround to load it and still build in parallel.
And just to be sure, declare that lib as an artefact on appveyor so
that we can download it.

4 years agosanitize the handling of trace:connect for host_avail
Martin Quinson [Sun, 31 Jan 2016 20:51:11 +0000 (21:51 +0100)]
sanitize the handling of trace:connect for host_avail

4 years agoconvert another file to C++ so that it can use the internal API
Martin Quinson [Sun, 31 Jan 2016 15:02:59 +0000 (16:02 +0100)]
convert another file to C++ so that it can use the internal API

4 years agowe don't support MPI_Cancel, deactivate this test as it fails sometimes
degomme [Sun, 31 Jan 2016 12:36:25 +0000 (13:36 +0100)]
we don't support MPI_Cancel, deactivate this test as it fails sometimes

4 years agospecify srcdir here
degomme [Sun, 31 Jan 2016 12:35:00 +0000 (13:35 +0100)]
specify srcdir here

4 years agoweirdly cut fails here on some systems, try this workaround
degomme [Sun, 31 Jan 2016 12:33:22 +0000 (13:33 +0100)]
weirdly cut fails here on some systems, try this workaround

4 years agoSplitted the lmm_constraint swag called element_set into enabled_element_set and...
Fabien Chaix [Sun, 31 Jan 2016 10:14:30 +0000 (12:14 +0200)]
Splitted the lmm_constraint swag called element_set into enabled_element_set and disabled_element_set.
Normally better for readability and performance.

4 years agomark as const a method that is pure
Martin Quinson [Sun, 31 Jan 2016 09:45:51 +0000 (10:45 +0100)]
mark as const a method that is pure

4 years agomark as override a method that does
Martin Quinson [Sun, 31 Jan 2016 09:43:45 +0000 (10:43 +0100)]
mark as override a method that does

4 years agolooks like a bug (doesn't fix the tests, though)
degomme [Sun, 31 Jan 2016 00:51:11 +0000 (01:51 +0100)]
looks like a bug (doesn't fix the tests, though)

4 years agoSimplifies the prototype of the surf::Link::onStateChange signal
Martin Quinson [Sun, 31 Jan 2016 00:38:46 +0000 (01:38 +0100)]
Simplifies the prototype of the surf::Link::onStateChange signal

4 years agounspecialize the trace setup for CPUs
Martin Quinson [Sun, 31 Jan 2016 00:25:45 +0000 (01:25 +0100)]
unspecialize the trace setup for CPUs

Every Cpu class exposes the same interface to set a state_trace or a
speed_trace, and every code that sets it use the interface.

As a result, every code setting the traces of a CPU is very similar
and should be factorized. For now, it's spread in the whole class tree.

Actually, the right place to do it is in the xml parser itself. The
symbol trace_connect_list_host_speed and friends should have remained
private to the xml parser.

I'll do so tomorrow, and also deal with the setup of link traces.

4 years agorename a function, and kill 2 dead prototypes
Martin Quinson [Sat, 30 Jan 2016 23:56:26 +0000 (00:56 +0100)]
rename a function, and kill 2 dead prototypes

4 years agorename the datacontainers allowing the trace:connect tag
Martin Quinson [Sat, 30 Jan 2016 23:44:34 +0000 (00:44 +0100)]
rename the datacontainers allowing the trace:connect tag

4 years ago[cpu_ti] small code untangling
Martin Quinson [Sat, 30 Jan 2016 23:38:33 +0000 (00:38 +0100)]
[cpu_ti] small code untangling

4 years ago[trace_mgr] cosmetics: only declare variables on need
Martin Quinson [Sat, 30 Jan 2016 23:27:03 +0000 (00:27 +0100)]
[trace_mgr] cosmetics: only declare variables on need

4 years agocosmetics and comments in surf_solve
Martin Quinson [Sat, 30 Jan 2016 23:19:32 +0000 (00:19 +0100)]
cosmetics and comments in surf_solve

4 years ago[trace_mgr] less void*, (some) more doc
Martin Quinson [Sat, 30 Jan 2016 23:04:15 +0000 (00:04 +0100)]
[trace_mgr] less void*, (some) more doc

4 years agoThis comparison was wrong - Thanks gcc 6.0 for the warning
degomme [Sat, 30 Jan 2016 23:52:39 +0000 (00:52 +0100)]
This comparison was wrong - Thanks gcc 6.0 for the warning

4 years agoThese consts were wrong - Thanks gcc 6.0 for the warning
degomme [Sat, 30 Jan 2016 23:50:54 +0000 (00:50 +0100)]
These consts were wrong - Thanks gcc 6.0 for the warning

4 years agoreindent to avoid a (fatal) warning on RawHide
Martin Quinson [Sat, 30 Jan 2016 22:44:50 +0000 (23:44 +0100)]
reindent to avoid a (fatal) warning on RawHide

4 years agoMake it compile on 32bits, to please @degomme
Martin Quinson [Sat, 30 Jan 2016 22:08:09 +0000 (23:08 +0100)]
Make it compile on 32bits, to please @degomme

4 years agocosmetics and reindent in ptask
Martin Quinson [Sat, 30 Jan 2016 21:58:34 +0000 (22:58 +0100)]
cosmetics and reindent in ptask

4 years agoThese methods are overriden to the same content too
Martin Quinson [Sat, 30 Jan 2016 17:45:52 +0000 (18:45 +0100)]
These methods are overriden to the same content too

4 years agoFixed a dummy bug and removed an assertion that was over-testing
Fabien Chaix [Sat, 30 Jan 2016 18:32:38 +0000 (20:32 +0200)]
Fixed a dummy bug and removed an assertion that was over-testing

4 years agoChanges to the Surf LMM:
Fabien Chaix [Sat, 30 Jan 2016 18:02:01 +0000 (20:02 +0200)]
Changes to the Surf LMM:
- Adding features to control the "concurrency" of constraints
- Optimizing and cleaning some parts (modified set update)
- Adding documentation
Need more work to expose these features to the user

4 years ago[MC] Renamed option sparse-checkpoint to sparse_checkpoint
Christian Heinrich [Sat, 30 Jan 2016 17:45:20 +0000 (18:45 +0100)]
[MC] Renamed option sparse-checkpoint to sparse_checkpoint

We need to have a common naming scheme for options in place.

4 years agos/power/speed/ in that example too
Martin Quinson [Fri, 29 Jan 2016 21:02:53 +0000 (22:02 +0100)]
s/power/speed/ in that example too

4 years agodon't count tasks in a global
Frederic Suter [Fri, 29 Jan 2016 14:31:57 +0000 (15:31 +0100)]
don't count tasks in a global

raised a bug in daxloader
reindent while at it

4 years agokill dead code: override to the same content
Martin Quinson [Fri, 29 Jan 2016 11:02:08 +0000 (12:02 +0100)]
kill dead code: override to the same content