Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
simgrid.git
5 years agoMerge branch 'mc'
Gabriel Corona [Fri, 25 Jul 2014 11:28:50 +0000 (13:28 +0200)]
Merge branch 'mc'

Conflicts:
src/mc/mc_compare.cpp

5 years ago[mc] Change the signature of mc_restore_page_snapshot_region
Gabriel Corona [Fri, 25 Jul 2014 10:24:59 +0000 (12:24 +0200)]
[mc] Change the signature of mc_restore_page_snapshot_region

Better separation of the different snapshot layers.

5 years ago[mc] Split mc_region_new_dense out of MC_region_new
Gabriel Corona [Fri, 25 Jul 2014 10:24:22 +0000 (12:24 +0200)]
[mc] Split mc_region_new_dense out of MC_region_new

MC_region_new uses the global SimGrid setting.
mc_region_new_dense uses dense snapshot.

5 years ago[mc] Expand unit test of mc_snapshot.c
Gabriel Corona [Fri, 25 Jul 2014 10:23:00 +0000 (12:23 +0200)]
[mc] Expand unit test of mc_snapshot.c

5 years ago[mc] Remove page_store.tesh (has been changed into a unit test)
Gabriel Corona [Fri, 25 Jul 2014 10:42:42 +0000 (12:42 +0200)]
[mc] Remove page_store.tesh (has been changed into a unit test)

5 years ago(stupid) warning --
Augustin Degomme [Thu, 24 Jul 2014 16:36:52 +0000 (18:36 +0200)]
(stupid) warning --

5 years agoFix msg-host-on-off tesh
Paul Bédaride [Thu, 24 Jul 2014 15:47:19 +0000 (17:47 +0200)]
Fix msg-host-on-off tesh

5 years agoAdd and use knomial reduce algorithm from mvapich
Augustin Degomme [Thu, 24 Jul 2014 15:26:37 +0000 (17:26 +0200)]
Add and use knomial reduce algorithm from mvapich

5 years agoAdd and use mvapich's scatter_dest alltoall algorithm
Augustin Degomme [Thu, 24 Jul 2014 14:41:28 +0000 (16:41 +0200)]
Add and use mvapich's scatter_dest alltoall algorithm

5 years agoMerge remote-tracking branch 'origin/mc-fastsnapshot' into mc
Gabriel Corona [Thu, 24 Jul 2014 14:11:19 +0000 (16:11 +0200)]
Merge remote-tracking branch 'origin/mc-fastsnapshot' into mc

Conflicts:
buildtools/Cmake/Flags.cmake

5 years ago[mc] Remove irrelevant comment
Gabriel Corona [Thu, 24 Jul 2014 14:10:02 +0000 (16:10 +0200)]
[mc] Remove irrelevant comment

5 years agofix daemon typos
Adrien Lebre [Thu, 24 Jul 2014 14:01:33 +0000 (16:01 +0200)]
fix daemon typos

5 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Thu, 24 Jul 2014 13:49:43 +0000 (15:49 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

5 years agofix minor issues in the management of Java Exceptions + comment process cancelled...
Adrien Lebre [Thu, 24 Jul 2014 13:49:37 +0000 (15:49 +0200)]
fix minor issues in the management of Java Exceptions + comment process cancelled exception call

5 years agofix dist
Augustin Degomme [Thu, 24 Jul 2014 13:39:18 +0000 (15:39 +0200)]
fix dist

5 years agoindent
Augustin Degomme [Thu, 24 Jul 2014 13:22:33 +0000 (15:22 +0200)]
indent

5 years agoprotect these calls against MPI_DATATYPE_NULL errors
Augustin Degomme [Thu, 24 Jul 2014 13:20:49 +0000 (15:20 +0200)]
protect these calls against MPI_DATATYPE_NULL errors

5 years agocleanup a bit the code, ensure tests do pass
Augustin Degomme [Thu, 24 Jul 2014 13:20:25 +0000 (15:20 +0200)]
cleanup a bit the code, ensure tests do pass

5 years agoadd mvapich allreduce rs algorithm
Augustin Degomme [Thu, 24 Jul 2014 09:06:29 +0000 (11:06 +0200)]
add mvapich allreduce rs algorithm

5 years agomanually privatize allred test, to have it work on non-mmap systems also
Augustin Degomme [Thu, 24 Jul 2014 08:41:00 +0000 (10:41 +0200)]
manually privatize allred test, to have it work on non-mmap systems also

5 years agoAdd last collectives from mvapich selector : bcast reduce reduce_scatter scatter
Augustin Degomme [Thu, 24 Jul 2014 00:04:40 +0000 (02:04 +0200)]
Add last collectives from mvapich selector : bcast reduce reduce_scatter scatter

bcast stilll defaults to mpich one, as they need smp support

5 years agoNew collectives for mvapich2 selector : allgatherv, allreduce, alltoallv, barrier
Augustin Degomme [Wed, 23 Jul 2014 15:35:52 +0000 (17:35 +0200)]
New collectives for mvapich2 selector : allgatherv, allreduce, alltoallv, barrier

5 years agoBegin to add a MVAPICH2 collectives selector. Alltoall, Allgather and gather done.
Augustin Degomme [Wed, 23 Jul 2014 12:26:17 +0000 (14:26 +0200)]
Begin to add a MVAPICH2 collectives selector. Alltoall, Allgather and gather done.

Problems :
- code is (for now) quite a mess, with several versions of tuning available at the same time (coll folder from mvapich has currently 145k cloc).
- code is copy-pasted directly ... So, no comments
- only Stampede calibration is imported for now. Some others are available, we should provide a mechanism to switch to another calibration.
- MVAPICH collectives are SMP aware. SMPI is not really ... A mechanism to automatically generate an "internal" communicator for processes sharing a physical node will be needed. Gather actually defaults to mpich one as a result of this.

5 years agoFix mallocator tesh
Paul Bédaride [Thu, 24 Jul 2014 13:16:12 +0000 (15:16 +0200)]
Fix mallocator tesh

5 years ago[mc] Use mc_region_contain where it could be used
Gabriel Corona [Thu, 24 Jul 2014 12:43:13 +0000 (14:43 +0200)]
[mc] Use mc_region_contain where it could be used

5 years ago[mc] Udpate doxygen comments
Gabriel Corona [Thu, 24 Jul 2014 12:09:55 +0000 (14:09 +0200)]
[mc] Udpate doxygen comments

5 years agoAdd test for host on off and some fixes
Paul Bédaride [Thu, 24 Jul 2014 09:53:48 +0000 (11:53 +0200)]
Add test for host on off and some fixes

5 years agoSet 4 tests that fail on Windows and won't ever pass as "expected to fail"
Augustin Degomme [Wed, 23 Jul 2014 09:07:42 +0000 (11:07 +0200)]
Set 4 tests that fail on Windows and won't ever pass as "expected to fail"

This will allow to better discriminate between bugs and wontfix non-issues.

One of them fails because random returns differently on win and generates different colors in the resulting trace. Duh.
The three others are expected to crash, and do crash correctly on Windows, but the return code is different from the one expected (SIGABRT). Meh

5 years ago[mc] Add tests for sparse snapshot
Gabriel Corona [Tue, 22 Jul 2014 13:42:44 +0000 (15:42 +0200)]
[mc] Add tests for sparse snapshot

5 years ago[mc] Disable soft-dirty page tracking by default
Gabriel Corona [Tue, 22 Jul 2014 13:15:17 +0000 (15:15 +0200)]
[mc] Disable soft-dirty page tracking by default

In all tests I ran, it has a negative impact on performance.

5 years agosanitize get/set_name functions for fortran use
Augustin Degomme [Mon, 21 Jul 2014 13:22:50 +0000 (15:22 +0200)]
sanitize get/set_name functions for fortran use

5 years ago[mc] Disable optimisation for xbt when using MC
Gabriel Corona [Mon, 21 Jul 2014 12:36:16 +0000 (14:36 +0200)]
[mc] Disable optimisation for xbt when using MC

In some cases, it breaks the state comparison for some reason. This
was observed with mm.c and dynar.c.

5 years ago[mc] Do not handle mc_model_checker->parent_snapshot unless it is necessary
Gabriel Corona [Mon, 21 Jul 2014 11:38:08 +0000 (13:38 +0200)]
[mc] Do not handle mc_model_checker->parent_snapshot unless it is necessary

5 years agoAdd MPI_Win_get_name and MPI_Win_set_name support
Augustin Degomme [Mon, 21 Jul 2014 09:33:08 +0000 (11:33 +0200)]
Add MPI_Win_get_name and MPI_Win_set_name support

5 years agoreactivate allred test
Augustin Degomme [Mon, 21 Jul 2014 09:32:40 +0000 (11:32 +0200)]
reactivate allred test

But have its compilation flags set to O0 to avoid issues with ci slaves (too long time to compile with optims all the macros used, and too much memory used)

5 years ago[mmalloc] Force metadata update in mmalloc/mrealloc
Gabriel Corona [Thu, 17 Jul 2014 14:30:45 +0000 (16:30 +0200)]
[mmalloc] Force metadata update in mmalloc/mrealloc

5 years ago[mmalloc] Add mmcheck() which checks mmalloc heap consistency
Gabriel Corona [Thu, 17 Jul 2014 08:41:20 +0000 (10:41 +0200)]
[mmalloc] Add mmcheck() which checks mmalloc heap consistency

5 years ago[mmalloc] Add new block type for heapinfo blocks
Gabriel Corona [Thu, 10 Jul 2014 13:48:16 +0000 (15:48 +0200)]
[mmalloc] Add new block type for heapinfo blocks

5 years agorevert changes on allgatherv4, which needed manual privatization to run on freebsd
Augustin Degomme [Thu, 17 Jul 2014 16:38:50 +0000 (18:38 +0200)]
revert changes on allgatherv4, which needed manual privatization to run on freebsd

5 years agoremove warning with mc
Augustin Degomme [Thu, 17 Jul 2014 16:15:02 +0000 (18:15 +0200)]
remove warning with mc

5 years agoremove warning
Augustin Degomme [Thu, 17 Jul 2014 16:00:47 +0000 (18:00 +0200)]
remove warning

5 years agoFinish pulling changes from mpich trunk testsuite
Augustin Degomme [Thu, 17 Jul 2014 15:52:24 +0000 (17:52 +0200)]
Finish pulling changes from mpich trunk testsuite

5 years agoUpdate comm, datatype from mpich trunk
Augustin Degomme [Thu, 17 Jul 2014 14:54:35 +0000 (16:54 +0200)]
Update comm, datatype from mpich trunk

5 years agoenforce a scatter error in some cases
Augustin Degomme [Thu, 17 Jul 2014 13:38:45 +0000 (15:38 +0200)]
enforce a scatter error in some cases

5 years agoupdate collectives teshsuite from mpich git (only minor changes)
Augustin Degomme [Thu, 17 Jul 2014 13:38:28 +0000 (15:38 +0200)]
update collectives teshsuite from mpich git (only minor changes)

5 years agotesh update for fabien's work on surf
Augustin Degomme [Thu, 17 Jul 2014 09:06:53 +0000 (11:06 +0200)]
tesh update for fabien's work on surf

5 years agofabien's work on surf
Augustin Degomme [Thu, 17 Jul 2014 09:06:12 +0000 (11:06 +0200)]
fabien's work on surf

5 years agoPush Takahiro Patch and fix cloud.tesh - Adrien
Adrien Lebre [Wed, 16 Jul 2014 17:39:34 +0000 (19:39 +0200)]
Push Takahiro Patch and fix cloud.tesh - Adrien

5 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Wed, 16 Jul 2014 16:09:41 +0000 (18:09 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

5 years agoFix a typo in the java-cloud example and add one TODO related to the migration invoca...
Adrien Lebre [Wed, 16 Jul 2014 16:09:37 +0000 (18:09 +0200)]
Fix a typo in the java-cloud example and add one TODO related to the migration invocation in VM.java - adrien

5 years agofix dist
Augustin Degomme [Wed, 16 Jul 2014 16:00:55 +0000 (18:00 +0200)]
fix dist

5 years agorename function
Augustin Degomme [Wed, 16 Jul 2014 15:44:00 +0000 (17:44 +0200)]
rename function

5 years agorequalify tesh
Augustin Degomme [Wed, 16 Jul 2014 15:43:52 +0000 (17:43 +0200)]
requalify tesh

5 years agouse global variables to store values that may be used millions of times..
Augustin Degomme [Wed, 16 Jul 2014 15:16:50 +0000 (17:16 +0200)]
use global variables to store values that may be used millions of times..

5 years agoadd time injection in MPI_Wtime and MPI_Test, to match what was done in iprobe
Augustin Degomme [Wed, 16 Jul 2014 14:52:03 +0000 (16:52 +0200)]
add time injection in MPI_Wtime and MPI_Test, to match what was done in iprobe

5 years agoadd comment for magic value
Augustin Degomme [Wed, 16 Jul 2014 12:49:18 +0000 (14:49 +0200)]
add comment for magic value

5 years agoadd F90 rma tests
Augustin Degomme [Wed, 16 Jul 2014 12:46:37 +0000 (14:46 +0200)]
add F90 rma tests

5 years agomigration: minor cleanup and update TODO
Takahiro Hirofuchi [Wed, 16 Jul 2014 07:14:23 +0000 (16:14 +0900)]
migration: minor cleanup and update TODO

Use MSG_process_create() instead of that of _with_arguments().

5 years agomigration: fix status check of migration
Takahiro Hirofuchi [Wed, 16 Jul 2014 07:04:53 +0000 (16:04 +0900)]
migration: fix status check of migration

When a migration of a VM is already ongoing, do not allow
MSG_vm_migrate() for the VM.

5 years agoremove the allred test
degomme [Tue, 15 Jul 2014 18:07:44 +0000 (20:07 +0200)]
remove the allred test
ci slaves have trouble building it (not enough memory probably)

5 years agoallred test needed unsigned char support, which was forgotten.
degomme [Tue, 15 Jul 2014 16:29:37 +0000 (18:29 +0200)]
allred test needed unsigned char support, which was forgotten.
TODO : This part of the code is ugly and should be replaced by macros asap, as it may lead to nasty bugs

5 years agoAdd MPI_Type_set_name and MPI_Type_get_name and activate tests
degomme [Tue, 15 Jul 2014 16:28:21 +0000 (18:28 +0200)]
Add MPI_Type_set_name and MPI_Type_get_name and activate tests

5 years agoactivate scatterv test, which needed Carts and Dim
degomme [Tue, 15 Jul 2014 14:14:52 +0000 (16:14 +0200)]
activate scatterv test, which needed Carts and Dim

5 years agoadd definitions
degomme [Mon, 14 Jul 2014 22:32:57 +0000 (00:32 +0200)]
add definitions

5 years agofix dist
degomme [Mon, 14 Jul 2014 21:41:35 +0000 (23:41 +0200)]
fix dist

5 years agoset default size of Aint to integer*8 ... not ideal, though.
degomme [Mon, 14 Jul 2014 21:36:43 +0000 (23:36 +0200)]
set default size of Aint to integer*8 ... not ideal, though.
Real mpich testsuite uses autoconf to configure the size

5 years agoadapt mpif.h for rma
degomme [Sat, 12 Jul 2014 00:00:35 +0000 (02:00 +0200)]
adapt mpif.h for rma

5 years agoAdd f77 RMA tests
degomme [Fri, 11 Jul 2014 23:52:13 +0000 (01:52 +0200)]
Add f77 RMA tests

5 years agompich testsuite: add f77 topo test
degomme [Fri, 11 Jul 2014 20:19:17 +0000 (22:19 +0200)]
mpich testsuite: add f77 topo test

5 years agompich testsuite: activate now working datatype test
degomme [Fri, 11 Jul 2014 20:18:21 +0000 (22:18 +0200)]
mpich testsuite: activate now working datatype test

5 years ago[mmalloc] Add documentation
Gabriel Corona [Fri, 11 Jul 2014 10:06:47 +0000 (12:06 +0200)]
[mmalloc] Add documentation

5 years ago[mmalloc] Avoid useless memset0
Gabriel Corona [Fri, 11 Jul 2014 10:04:34 +0000 (12:04 +0200)]
[mmalloc] Avoid useless memset0

5 years ago[mc] Remove 'previous' variable ind mmalloc_compare_heap()
Gabriel Corona [Fri, 11 Jul 2014 08:15:59 +0000 (10:15 +0200)]
[mc] Remove 'previous' variable ind mmalloc_compare_heap()

It was not used.

5 years ago[mc] Fix test on type in mc_diff
Gabriel Corona [Fri, 11 Jul 2014 08:02:18 +0000 (10:02 +0200)]
[mc] Fix test on type in mc_diff

5 years ago[mc] Fix name of mc_snapshot_memcmp()
Gabriel Corona [Thu, 10 Jul 2014 11:32:29 +0000 (13:32 +0200)]
[mc] Fix name of mc_snapshot_memcmp()

5 years ago[mc] Add unit tests for reading/comparing the whole region in mc_snapshot
Gabriel Corona [Thu, 10 Jul 2014 10:44:22 +0000 (12:44 +0200)]
[mc] Add unit tests for reading/comparing the whole region in mc_snapshot

5 years ago[mc] Test flat snapshots as well
Gabriel Corona [Thu, 10 Jul 2014 10:30:10 +0000 (12:30 +0200)]
[mc] Test flat snapshots as well

5 years ago[mc] Add unit test for mc_snapshot
Gabriel Corona [Thu, 10 Jul 2014 09:36:32 +0000 (11:36 +0200)]
[mc] Add unit test for mc_snapshot

5 years agoFix distcheck
Augustin Degomme [Tue, 8 Jul 2014 22:51:01 +0000 (00:51 +0200)]
Fix distcheck

5 years agofile was not in the dist
Augustin Degomme [Tue, 8 Jul 2014 22:43:50 +0000 (00:43 +0200)]
file was not in the dist

5 years agoactivate working test
Augustin Degomme [Tue, 8 Jul 2014 22:21:24 +0000 (00:21 +0200)]
activate working test

5 years agoAdd perf mpich3 tests
Augustin Degomme [Tue, 8 Jul 2014 22:03:47 +0000 (00:03 +0200)]
Add perf mpich3 tests

5 years agoreduce size of recently added tests, to avoid them taking so long to complete
Augustin Degomme [Tue, 8 Jul 2014 16:09:28 +0000 (18:09 +0200)]
reduce size of recently added tests, to avoid them taking so long to complete

5 years agoPrevious commit exposes a bug in OpenMPI, switch algorithm for bcast to avoid it
Augustin Degomme [Tue, 8 Jul 2014 15:26:15 +0000 (17:26 +0200)]
Previous commit exposes a bug in OpenMPI, switch algorithm for bcast to avoid it
The bug is that OpenMPI performs selection of algorithm based on the count on messages.
Or this count can be different from one process to another for bcast.
For example this test has a sender sending one element of 4096 bytes, and receivers receiving 1024 elements of 4 bytes
Yes, this is legal in MPI (but stupid ... but legal).
OpenMPI collective algorithm selector selects a different algorithm for these processes, thus deadlocking

5 years agoactivate two tests in collective communications
Augustin Degomme [Tue, 8 Jul 2014 15:21:30 +0000 (17:21 +0200)]
activate two tests in collective communications

5 years ago[mc] Move page store test from the integration test into the unit test infrastructure
Gabriel Corona [Tue, 8 Jul 2014 14:41:35 +0000 (16:41 +0200)]
[mc] Move page store test from the integration test into the unit test infrastructure

5 years agoAdd support for C++ in tools/sg_unit_extractor.pl
Gabriel Corona [Tue, 8 Jul 2014 14:27:51 +0000 (16:27 +0200)]
Add support for C++ in tools/sg_unit_extractor.pl

5 years agoleak --
Augustin Degomme [Tue, 8 Jul 2014 13:32:42 +0000 (15:32 +0200)]
leak --

5 years ago[mc] Use mc_snapshot_region_memcmp() in compare_areas_with_type()
Gabriel Corona [Tue, 8 Jul 2014 13:12:00 +0000 (15:12 +0200)]
[mc] Use mc_snapshot_region_memcmp() in compare_areas_with_type()

5 years ago[mc] Avoid memory allocation for flat snapshots in mc_snapshot_region_memcmp()
Gabriel Corona [Tue, 8 Jul 2014 13:04:17 +0000 (15:04 +0200)]
[mc] Avoid memory allocation for flat snapshots in mc_snapshot_region_memcmp()

5 years ago[mc] Fix name of mc_snapshot_region_memcmp
Gabriel Corona [Tue, 8 Jul 2014 12:48:42 +0000 (14:48 +0200)]
[mc] Fix name of mc_snapshot_region_memcmp

5 years agoMerge branch 'mc-fix' into mc-fastsnapshot
Gabriel Corona [Tue, 8 Jul 2014 11:43:29 +0000 (13:43 +0200)]
Merge branch 'mc-fix' into mc-fastsnapshot

Conflicts:
src/mc/mc_diff.c

5 years ago[mc] Fix check on type name
Gabriel Corona [Tue, 8 Jul 2014 10:54:59 +0000 (12:54 +0200)]
[mc] Fix check on type name

The name was changed to include 'struct/class'.

5 years ago[mc] Fix bad parameter passed in mc_diff
Gabriel Corona [Fri, 4 Jul 2014 11:22:00 +0000 (13:22 +0200)]
[mc] Fix bad parameter passed in mc_diff

In the following calls:

    compare_heap_area_with_type(state,
      area1, area2,
      area1_to_compare, area2_to_compare,
      snapshot1, snapshot2,
      previous, type, size, check_ignore,
      pointer_level);

    compare_heap_area_without_type(state,
      area1, area2,
      area1_to_compare, area2_to_compare,
      snapshot1, snapshot2,
      previous, size, check_ignore);

areaX and real_areaX_to_compare do not point the same data in different
address spaces in some cases.

Sometimes real_areaX_to_compare is adjusted to point to the beginning
of the block or fragment:

    area1_to_compare = addr_block1;
    // or
    area1_to_compare = (char *) addr_frag1 + offset1;
      // (when offset1==0)

but areaX is not adjusted accordingly and still point to the original
data: in the called compare_heap_area_with[out]_type(), the two values
are inconsistent.

Moreover in some cases, the type does not correspond:

  * areaX_to_compareX is the beginning of the fragment ;
  * type is the type of areaX,  not the type of the fragment.

5 years agoadd minloc and maxloc for long_long
Augustin Degomme [Mon, 7 Jul 2014 15:22:04 +0000 (17:22 +0200)]
add minloc and maxloc for long_long

5 years agothis should be err_rank, not err_comm
Augustin Degomme [Mon, 7 Jul 2014 15:14:50 +0000 (17:14 +0200)]
this should be err_rank, not err_comm

5 years agoadd a non-existing MPI_2LONG datatype to handle some corner cases in fortran
Augustin Degomme [Mon, 7 Jul 2014 15:09:11 +0000 (17:09 +0200)]
add a non-existing MPI_2LONG datatype to handle some corner cases in fortran

5 years agoXml platform cleaning teshsuite/msg
Paul Bédaride [Mon, 7 Jul 2014 14:51:33 +0000 (16:51 +0200)]
Xml platform cleaning teshsuite/msg

5 years ago[mc] Fix lookup of malloc fragment type
Gabriel Corona [Mon, 7 Jul 2014 08:41:29 +0000 (10:41 +0200)]
[mc] Fix lookup of malloc fragment type

In some cases, the type was looked up in the current heapinfo and not
in the snapshot one:

 * in some cases this is not an issue as state->heapinfo2 is currently
   the same as the current heapinfo but this solution is more robust;

 * if we were supposed to look it from state->heapinfo1 this is wrong.

5 years agoactivate working tests
Augustin Degomme [Mon, 7 Jul 2014 14:07:14 +0000 (16:07 +0200)]
activate working tests
Lower sizes and number of messages for pingping to avoid taking too long