Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
simgrid.git
3 years agoa few less verbose descriptions in instr
Frederic Suter [Fri, 8 Jun 2018 07:07:23 +0000 (09:07 +0200)]
a few less verbose descriptions in instr

3 years agoFix build for lua bindings.
Arnaud Giersch [Wed, 6 Jun 2018 14:37:03 +0000 (16:37 +0200)]
Fix build for lua bindings.

3 years agoyes another attempt
Augustin Degomme [Wed, 6 Jun 2018 22:34:10 +0000 (00:34 +0200)]
yes another attempt

3 years agosnake case storage
Frederic Suter [Wed, 6 Jun 2018 18:20:32 +0000 (20:20 +0200)]
snake case storage

3 years agonow useless, bummer
Frederic Suter [Wed, 6 Jun 2018 12:17:33 +0000 (14:17 +0200)]
now useless, bummer

3 years agofix build (and please valgrind)
Frederic Suter [Wed, 6 Jun 2018 12:15:47 +0000 (14:15 +0200)]
fix build (and please valgrind)

3 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Frederic Suter [Wed, 6 Jun 2018 12:15:28 +0000 (14:15 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

3 years agoremove trace-resource from trace comparison, as results are unreliable and too large...
Augustin Degomme [Wed, 6 Jun 2018 12:06:38 +0000 (14:06 +0200)]
remove trace-resource from trace comparison, as results are unreliable and too large. Need to test it elsewhere.

3 years agouseless private
Frederic Suter [Wed, 6 Jun 2018 10:58:00 +0000 (12:58 +0200)]
useless private

3 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Frederic Suter [Wed, 6 Jun 2018 10:16:54 +0000 (12:16 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

3 years agomv link map to the engine
Frederic Suter [Wed, 6 Jun 2018 10:16:37 +0000 (12:16 +0200)]
mv link map to the engine

3 years agoActually test SMPI tracing - high probability of failure
Augustin Degomme [Wed, 6 Jun 2018 09:47:05 +0000 (11:47 +0200)]
Actually test SMPI tracing - high probability of failure

3 years agoreduce time taken by a test (maybe for nothing)
Augustin Degomme [Wed, 6 Jun 2018 09:27:04 +0000 (11:27 +0200)]
reduce time taken by a test (maybe for nothing)

3 years agoadd back tracing of communications in wait/waitany/waitall, should fix #269
Augustin Degomme [Wed, 6 Jun 2018 09:26:42 +0000 (11:26 +0200)]
add back tracing of communications in wait/waitany/waitall, should fix #269

3 years agofactor attachment of state traces to resources
Frederic Suter [Wed, 6 Jun 2018 08:20:15 +0000 (10:20 +0200)]
factor attachment of state traces to resources

It also opens the way to state traces for other resources, such as
storage

3 years agosnake case for better consistency
Frederic Suter [Wed, 6 Jun 2018 08:05:02 +0000 (10:05 +0200)]
snake case for better consistency

3 years agosnake_case LinkImpl (but the part that should to the engine)
Frederic Suter [Wed, 6 Jun 2018 07:55:03 +0000 (09:55 +0200)]
snake_case LinkImpl (but the part that should to the engine)

3 years agoAvoid hard coded -lboost_unit_test_framework.
Arnaud Giersch [Tue, 5 Jun 2018 08:33:20 +0000 (10:33 +0200)]
Avoid hard coded -lboost_unit_test_framework.

3 years agoUpdate .mailmap file.
Arnaud Giersch [Fri, 1 Jun 2018 21:26:57 +0000 (23:26 +0200)]
Update .mailmap file.

3 years agotry to fix how we deal with rpath on binaries
Martin Quinson [Mon, 4 Jun 2018 20:43:50 +0000 (22:43 +0200)]
try to fix how we deal with rpath on binaries

3 years agotry to avoid a false positive on sonar saying that on_exit() is deprecated
Martin Quinson [Mon, 4 Jun 2018 07:02:33 +0000 (09:02 +0200)]
try to avoid a false positive on sonar saying that on_exit() is deprecated

3 years agohost-filtering example: improve doc, and add to documentation TOC
Martin Quinson [Sun, 3 Jun 2018 21:08:59 +0000 (23:08 +0200)]
host-filtering example: improve doc, and add to documentation TOC

3 years agosonar: Mark constructors as explicit
Martin Quinson [Sun, 3 Jun 2018 20:46:46 +0000 (22:46 +0200)]
sonar: Mark constructors as explicit

3 years agocmake: cosmetics around the detection of sem_init
Martin Quinson [Sun, 3 Jun 2018 19:53:33 +0000 (21:53 +0200)]
cmake: cosmetics around the detection of sem_init

3 years agoTry to fix the link to Boost_UNIT_TEST_FRAMEWORK
Martin Quinson [Sun, 3 Jun 2018 19:16:18 +0000 (21:16 +0200)]
Try to fix the link to Boost_UNIT_TEST_FRAMEWORK

Thanks to Olivier for the patch. It (fix: #267).

3 years agos4u::NetZone: move deprecation wrappers to the end
Martin Quinson [Thu, 31 May 2018 15:41:20 +0000 (17:41 +0200)]
s4u::NetZone: move deprecation wrappers to the end

3 years agosnake_case the s4u::NetZone signals
Martin Quinson [Thu, 31 May 2018 15:38:05 +0000 (17:38 +0200)]
snake_case the s4u::NetZone signals

Deprecation wrapper are impossible

3 years agoSnake casing simgrid::xbt::Path.
Arnaud Giersch [Fri, 1 Jun 2018 20:12:06 +0000 (22:12 +0200)]
Snake casing simgrid::xbt::Path.

3 years agoBe more liberal when matching extension for library name.
Arnaud Giersch [Fri, 1 Jun 2018 19:56:08 +0000 (21:56 +0200)]
Be more liberal when matching extension for library name.

Letters are sometimes used in version number (e.g. libcrypto.so.1.1.0h).

3 years agofix build on 32 bit systems
Augustin Degomme [Fri, 1 Jun 2018 20:29:53 +0000 (22:29 +0200)]
fix build on 32 bit systems

3 years ago[S4U] Engine: Support retrieving actors via the engine
Christian Heinrich [Fri, 1 Jun 2018 17:59:09 +0000 (19:59 +0200)]
[S4U] Engine: Support retrieving actors via the engine

3 years ago[S4U] Engine: Add Engine::get_filtered_links
Christian Heinrich [Fri, 1 Jun 2018 17:58:07 +0000 (19:58 +0200)]
[S4U] Engine: Add Engine::get_filtered_links

3 years ago[S4U] Mark s4u::Engine constructor as explicit
Christian Heinrich [Fri, 1 Jun 2018 16:21:18 +0000 (18:21 +0200)]
[S4U] Mark s4u::Engine constructor as explicit

3 years ago[S4U] Delete Engine move & copy constructors
Christian Heinrich [Fri, 1 Jun 2018 16:19:09 +0000 (18:19 +0200)]
[S4U] Delete Engine move & copy constructors

3 years ago[S4U] Implement Engine::get_filtered_hosts
Christian Heinrich [Fri, 1 Jun 2018 08:20:35 +0000 (10:20 +0200)]
[S4U] Implement Engine::get_filtered_hosts

This method takes any callable object (lambda, function, functor = function object)
that returns a boolean value (true = use this host, false = don't use it)
and that takes a Host* as parameter.

This allows to filter within the Engine rather than returning
a list with all hosts

3 years ago[S4U] Add some tests for upcoming filtering feature
Christian Heinrich [Fri, 1 Jun 2018 08:15:10 +0000 (10:15 +0200)]
[S4U] Add some tests for upcoming filtering feature

These tests are currently neither active nor compilable,
because we haven't implemented this feature yet. See next commit.

3 years agoFix type mismatch for function pointer.
Arnaud Giersch [Fri, 1 Jun 2018 13:50:31 +0000 (15:50 +0200)]
Fix type mismatch for function pointer.

3 years agoUse memcpy instead of strncpy.
Arnaud Giersch [Fri, 1 Jun 2018 13:41:18 +0000 (15:41 +0200)]
Use memcpy instead of strncpy.

Fix build error with gcc 8.1:
error: 'strncpy' specified bound depends on the length of the source argument [-Werror=stringop-overflow=]

3 years agoEnsure that the terminating nul byte is also copied.
Arnaud Giersch [Fri, 1 Jun 2018 13:27:12 +0000 (15:27 +0200)]
Ensure that the terminating nul byte is also copied.

3 years agoMark deprecated functions.
Arnaud Giersch [Fri, 1 Jun 2018 13:03:22 +0000 (15:03 +0200)]
Mark deprecated functions.

3 years agoUse new style Actor::on_exit().
Arnaud Giersch [Fri, 1 Jun 2018 12:59:06 +0000 (14:59 +0200)]
Use new style Actor::on_exit().

3 years agoUse new style SIMIX_process_on_exit().
Arnaud Giersch [Fri, 1 Jun 2018 12:42:56 +0000 (14:42 +0200)]
Use new style SIMIX_process_on_exit().

3 years agoDefine Actor::on_exit() taking a std::function.
Arnaud Giersch [Fri, 1 Jun 2018 12:37:21 +0000 (14:37 +0200)]
Define Actor::on_exit() taking a std::function.

3 years agoDefine SIMIX_process_on_exit() taking a std::function.
Arnaud Giersch [Fri, 1 Jun 2018 12:29:40 +0000 (14:29 +0200)]
Define SIMIX_process_on_exit() taking a std::function.

3 years agoUse std::function for s_smx_process_exit_fun_t.
Arnaud Giersch [Fri, 1 Jun 2018 12:23:16 +0000 (14:23 +0200)]
Use std::function for s_smx_process_exit_fun_t.

3 years ago[SMPI] Cosmetics: ! -> not
Christian Heinrich [Wed, 16 May 2018 13:23:59 +0000 (15:23 +0200)]
[SMPI] Cosmetics: ! -> not

3 years ago[INSTR] StateEvent::popEvent() calls now StateEvent::popEvent(nullptr)
Christian Heinrich [Thu, 26 Apr 2018 09:03:51 +0000 (11:03 +0200)]
[INSTR] StateEvent::popEvent() calls now StateEvent::popEvent(nullptr)

3 years agouse try_run for sem_open instead of firing the compiler ourselves
Martin Quinson [Thu, 31 May 2018 09:50:32 +0000 (11:50 +0200)]
use try_run for sem_open instead of firing the compiler ourselves

Experience shows that we don't do it right on MacOSX, at least.

3 years agostop checking for HAVE_GNU_LD now that we don't use the result
Martin Quinson [Thu, 31 May 2018 09:17:08 +0000 (11:17 +0200)]
stop checking for HAVE_GNU_LD now that we don't use the result

The detection sometimes breaks (eg on MacOSX), and the result of this
check is not used since 2014.
See commit a79b8df1d7ee2bb82aa46cdd42cd8ae344d0f32e for why the result
is not used anymore.

3 years agoStart snake_casing s4u::NetZone
Martin Quinson [Wed, 30 May 2018 16:30:16 +0000 (18:30 +0200)]
Start snake_casing s4u::NetZone

3 years agoConstructor priorities are still not supported on Apple.
Arnaud Giersch [Thu, 31 May 2018 09:41:39 +0000 (11:41 +0200)]
Constructor priorities are still not supported on Apple.

Disable constructor priorities since gcc gives an error.

Partially reverts commit f12101aa49174017c26e19bc8076f5c7124f9dd0.

3 years agosimplification
Frederic Suter [Wed, 30 May 2018 08:48:18 +0000 (10:48 +0200)]
simplification
the previous state is NEVER used

3 years agocosmetics
Frederic Suter [Wed, 30 May 2018 07:27:48 +0000 (09:27 +0200)]
cosmetics

3 years agoConvert enum class SmpiPrivStrategies tu upper case.
Arnaud Giersch [Fri, 25 May 2018 20:19:09 +0000 (22:19 +0200)]
Convert enum class SmpiPrivStrategies tu upper case.

3 years agoConvert enum shared_malloc_type to enum class.
Arnaud Giersch [Fri, 25 May 2018 19:41:16 +0000 (21:41 +0200)]
Convert enum shared_malloc_type to enum class.

3 years agoConvert enum smpi_process_state to enum class.
Arnaud Giersch [Fri, 25 May 2018 19:37:05 +0000 (21:37 +0200)]
Convert enum smpi_process_state to enum class.

3 years agofix issue where some end of recv events were not generated whe MPI_STATUS_IGNORE...
Augustin Degomme [Sun, 27 May 2018 10:00:49 +0000 (12:00 +0200)]
fix issue where some end of recv events were not generated whe MPI_STATUS_IGNORE was sent.

3 years agomove all plugins to their new home directory
Martin Quinson [Sat, 26 May 2018 15:07:54 +0000 (17:07 +0200)]
move all plugins to their new home directory

3 years agorename the plugins from the command line, and document it
Martin Quinson [Sat, 26 May 2018 14:35:35 +0000 (16:35 +0200)]
rename the plugins from the command line, and document it

3 years agoallow to activate the link energy plugin from the command line
Martin Quinson [Fri, 25 May 2018 07:42:27 +0000 (09:42 +0200)]
allow to activate the link energy plugin from the command line

3 years agoenum class values should be UPPER_CASE
Martin Quinson [Thu, 24 May 2018 23:59:06 +0000 (01:59 +0200)]
enum class values should be UPPER_CASE

3 years agomake all_existing_models visible from k:r:Model.hpp, which is public
Martin Quinson [Thu, 24 May 2018 23:51:55 +0000 (01:51 +0200)]
make all_existing_models visible from k:r:Model.hpp, which is public

3 years agotrivial snake_casing of xbt::PropertyHolder
Martin Quinson [Thu, 24 May 2018 06:58:52 +0000 (08:58 +0200)]
trivial snake_casing of xbt::PropertyHolder

No backward compat necessary as all subclasses are in the kernel (and
not exposed publicly yet)

3 years agofix NS3 builds
Martin Quinson [Wed, 23 May 2018 18:50:37 +0000 (20:50 +0200)]
fix NS3 builds

3 years agofix a gramatical error all over our variables
Martin Quinson [Wed, 23 May 2018 16:17:25 +0000 (18:17 +0200)]
fix a gramatical error all over our variables

3 years agosnake_case some easy parts of s4u::Host
Martin Quinson [Wed, 23 May 2018 16:04:28 +0000 (18:04 +0200)]
snake_case some easy parts of s4u::Host

This will make the interesting parts of the commit more readable.

3 years agotiny doc improvements
Martin Quinson [Tue, 22 May 2018 20:53:53 +0000 (22:53 +0200)]
tiny doc improvements

3 years agocleanings in k:r:Cpu
Martin Quinson [Tue, 22 May 2018 06:29:33 +0000 (08:29 +0200)]
cleanings in k:r:Cpu

3 years agoone more legacy MSG function
Frederic Suter [Tue, 22 May 2018 00:46:23 +0000 (02:46 +0200)]
one more legacy MSG function

3 years agosnake_case k:r:CpuModel and sub-classes
Martin Quinson [Mon, 21 May 2018 19:23:49 +0000 (21:23 +0200)]
snake_case k:r:CpuModel and sub-classes

3 years agomake 2 more fields of k:r:Cpu protected (should be private)
Martin Quinson [Mon, 21 May 2018 19:11:10 +0000 (21:11 +0200)]
make 2 more fields of k:r:Cpu protected (should be private)

3 years agomake k:r:Cpu::speed_per_pstate_ private as it should
Martin Quinson [Mon, 21 May 2018 19:04:55 +0000 (21:04 +0200)]
make k:r:Cpu::speed_per_pstate_ private as it should

3 years agoMerge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Sun, 20 May 2018 20:24:55 +0000 (22:24 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid

3 years agoprivatize some fields of kernel::resource::Cpu
Martin Quinson [Sun, 20 May 2018 13:36:55 +0000 (15:36 +0200)]
privatize some fields of kernel::resource::Cpu

3 years agosnake_case some parts of kernel::resource::Cpu
Martin Quinson [Sun, 20 May 2018 13:21:44 +0000 (15:21 +0200)]
snake_case some parts of kernel::resource::Cpu

3 years agoplug another memleak by not mallocating in the first place
Martin Quinson [Sat, 19 May 2018 14:03:26 +0000 (16:03 +0200)]
plug another memleak by not mallocating in the first place

3 years agoremove last bits of tracing from cpu_interface \o/
Frederic Suter [Sat, 19 May 2018 12:17:13 +0000 (14:17 +0200)]
remove last bits of tracing from cpu_interface \o/

Appears to just produce useless entries (same begin/end time with no
actual resource usage change) in the trace.

3 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Frederic Suter [Fri, 18 May 2018 22:27:10 +0000 (00:27 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

3 years agosnake_case some methods in s4u::Host
Martin Quinson [Fri, 18 May 2018 22:03:48 +0000 (00:03 +0200)]
snake_case some methods in s4u::Host

3 years agoprocess->actor
Frederic Suter [Fri, 18 May 2018 12:16:45 +0000 (14:16 +0200)]
process->actor

3 years agounsecure commit
Frederic Suter [Fri, 18 May 2018 12:11:02 +0000 (14:11 +0200)]
unsecure commit

3 years agosonar: redundent forward declaration
Martin Quinson [Fri, 18 May 2018 08:09:49 +0000 (10:09 +0200)]
sonar: redundent forward declaration

3 years agoplug a memleak -- bummer
Martin Quinson [Fri, 18 May 2018 07:59:14 +0000 (09:59 +0200)]
plug a memleak -- bummer

3 years agomore signals and more lambdas
Frederic Suter [Fri, 18 May 2018 08:10:12 +0000 (10:10 +0200)]
more signals and more lambdas

actor sleep/wake up is now tracable in S4U

3 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Frederic Suter [Fri, 18 May 2018 07:21:02 +0000 (09:21 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

3 years agocleanup the resource::Action::State::IGNORED thing
Martin Quinson [Thu, 17 May 2018 20:43:44 +0000 (22:43 +0200)]
cleanup the resource::Action::State::IGNORED thing

Factorize stuff of CpuCas01Model and CpuTiModel into Model directly.
This way, Action::get_state() always knows in which set the action is.
No shit, no magic here.

3 years agoUPPER_CASE the Action::State values (+ uniformity with s4u::Activity)
Martin Quinson [Thu, 17 May 2018 20:26:52 +0000 (22:26 +0200)]
UPPER_CASE the Action::State values (+ uniformity with s4u::Activity)

3 years agochange macro names to avoid nameclashes on FINISHED
Martin Quinson [Thu, 17 May 2018 20:15:15 +0000 (22:15 +0200)]
change macro names to avoid nameclashes on FINISHED

3 years agodoc improvement
Martin Quinson [Thu, 17 May 2018 20:01:44 +0000 (22:01 +0200)]
doc improvement

3 years agoEnum values are in UPPER_SNAKE_CASE
Martin Quinson [Thu, 17 May 2018 19:54:16 +0000 (21:54 +0200)]
Enum values are in UPPER_SNAKE_CASE

3 years agocleanups in Action::State
Martin Quinson [Thu, 17 May 2018 19:30:15 +0000 (21:30 +0200)]
cleanups in Action::State

- Rename "ready" to "inited", for symmetry to s4u::Activity
- Remove the unused state "to_free"
- Rename the funky state "not_in_the_system" into "ignored", and
  improve its documentation.

Further cleanups could include not having any action set beyond the
ones existing in kernel::resource::Model, so that
kernel::resource::Action::get_state() see them all.

CpuCas01 is the one heavily using ignored actions, as failure detectors.

3 years agoReduce the foreign usages of s4u::Host::pimpl_cpu, which should be private
Martin Quinson [Thu, 17 May 2018 12:43:57 +0000 (14:43 +0200)]
Reduce the foreign usages of s4u::Host::pimpl_cpu, which should be private

3 years agoinline SIMIX_host_off into s4u::Host->turnOff()
Martin Quinson [Thu, 17 May 2018 12:56:22 +0000 (14:56 +0200)]
inline SIMIX_host_off into s4u::Host->turnOff()

3 years agokill a couple more of the long descriptions in instr
Frederic Suter [Thu, 17 May 2018 10:39:28 +0000 (12:39 +0200)]
kill a couple more of the long descriptions in instr

3 years agokill a not so useful tracing option
Frederic Suter [Wed, 16 May 2018 15:26:44 +0000 (17:26 +0200)]
kill a not so useful tracing option

3 years agos4u::Host: move signals to the top and deprecated functions to the end
Martin Quinson [Wed, 16 May 2018 15:07:48 +0000 (17:07 +0200)]
s4u::Host: move signals to the top and deprecated functions to the end

3 years agothis should allow the tracing of resource usage with this model
Frederic Suter [Wed, 16 May 2018 14:29:20 +0000 (16:29 +0200)]
this should allow the tracing of resource usage with this model

3 years agono need to set tracing/buffer to its default value
Frederic Suter [Wed, 16 May 2018 12:03:08 +0000 (14:03 +0200)]
no need to set tracing/buffer to its default value

3 years agocosmetics
Frederic Suter [Wed, 16 May 2018 11:39:10 +0000 (13:39 +0200)]
cosmetics

3 years agooutput long descriptions on a single line
Frederic Suter [Wed, 16 May 2018 11:05:21 +0000 (13:05 +0200)]
output long descriptions on a single line