Augustin Degomme [Tue, 9 Oct 2012 16:51:14 +0000 (18:51 +0200)]
forgot to handle MPI_ANY_SOURCE as a valid source, they were caught as invalid
Augustin Degomme [Tue, 9 Oct 2012 16:44:14 +0000 (18:44 +0200)]
add a few tests for smpi
jean-noel quintin [Tue, 9 Oct 2012 17:40:32 +0000 (18:40 +0100)]
[examples,smpi] remove warnings and resolve a bug (I hope I amn't sure)
Martin Quinson [Tue, 9 Oct 2012 15:27:59 +0000 (17:27 +0200)]
improve a comment
navarro [Tue, 9 Oct 2012 09:50:10 +0000 (11:50 +0200)]
Add a tesh file for smpi vector
jean-noel quintin [Tue, 9 Oct 2012 08:12:51 +0000 (09:12 +0100)]
Merge remote-tracking branch 'origin/master'
jean-noel quintin [Tue, 9 Oct 2012 08:12:28 +0000 (09:12 +0100)]
[smpi,example] an example of matrix multiplication with non contignous memory
I hope this example is relevant and could increase the coverage
Marion Guthmuller [Tue, 9 Oct 2012 01:54:29 +0000 (03:54 +0200)]
set size to -1 for free fragments in new fragmented block
Marion Guthmuller [Tue, 9 Oct 2012 00:39:02 +0000 (02:39 +0200)]
set value -1 in unsigned short cannot work fine ...
Marion Guthmuller [Mon, 8 Oct 2012 23:53:15 +0000 (01:53 +0200)]
set size_used negative for free fragment/block
Martin Quinson [Mon, 8 Oct 2012 23:24:21 +0000 (01:24 +0200)]
try harder to break mmalloc wrt double-free
jean-noel quintin [Mon, 8 Oct 2012 23:23:04 +0000 (00:23 +0100)]
Merge remote-tracking branch 'origin/master'
jean-noel quintin [Mon, 8 Oct 2012 23:22:39 +0000 (00:22 +0100)]
[smpi] correction bug for detached comm
-all the test passed bug on real exemple it doesn't work
I have to found why and where to free these requests
Sorry for the introduction of this bug.
Marion Guthmuller [Mon, 8 Oct 2012 23:15:23 +0000 (01:15 +0200)]
UNLOCK forgotten before THROWF
Martin Quinson [Mon, 8 Oct 2012 23:01:16 +0000 (01:01 +0200)]
checks that double free are correctly detected
Marion Guthmuller [Mon, 8 Oct 2012 23:00:44 +0000 (01:00 +0200)]
check if free on fragment already free
Martin Quinson [Mon, 8 Oct 2012 22:42:20 +0000 (00:42 +0200)]
add the new mmalloc test to the testsuite
Martin Quinson [Mon, 8 Oct 2012 22:37:46 +0000 (00:37 +0200)]
better integration of this test in our framework (like, use XBT for the logs and such)
Martin Quinson [Mon, 8 Oct 2012 22:28:30 +0000 (00:28 +0200)]
Bug fix in mrealloc where the previous content could get lost
When reallocating a large block, mrealloc tries to be cleaver and
first frees the area before requesting it again to avoid a copy if
there is some free adjacent memory.
This trick obviously don't mix well with the mmalloc memseting the
allocated memory to 0 before returning it.
So, split the mmalloc in two, and use the non-memsetting one from
mrealloc.
Martin Quinson [Mon, 8 Oct 2012 22:02:35 +0000 (00:02 +0200)]
we all love debug informations
Martin Quinson [Mon, 8 Oct 2012 21:14:02 +0000 (23:14 +0200)]
add a basic tesh test about mmalloc
Martin Quinson [Mon, 8 Oct 2012 20:56:30 +0000 (22:56 +0200)]
kill a very old TODO file. We don't care of running mmalloc between several (system) processes
Martin Quinson [Mon, 8 Oct 2012 19:46:49 +0000 (21:46 +0200)]
we need the simgrid library too. True story.
Martin Quinson [Mon, 8 Oct 2012 19:44:15 +0000 (21:44 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Mon, 8 Oct 2012 19:43:13 +0000 (21:43 +0200)]
make it compile in a recent tree
Marion Guthmuller [Mon, 8 Oct 2012 19:42:25 +0000 (21:42 +0200)]
add executable chord_before_dsend in CMakeLists
Marion Guthmuller [Mon, 8 Oct 2012 19:25:21 +0000 (21:25 +0200)]
version of chord before use MSG_task_dsend
Marion Guthmuller [Mon, 8 Oct 2012 12:35:58 +0000 (14:35 +0200)]
model-checker : restore struct smx_ctx_sysv in smx_context_sysv.c
Arnaud Giersch [Mon, 8 Oct 2012 07:10:24 +0000 (09:10 +0200)]
Don't turn warnings about variables possibly clobbered by longjmp into errors.
Gcc is currently not smart enough and sometimes spits spurious warnings.
Arnaud Giersch [Mon, 8 Oct 2012 07:10:24 +0000 (09:10 +0200)]
Declare the function as static, and make the compiler happy.
Also revert "add a declaration to remove a warning"
There's no need to make this function public.
This reverts commit
ce0d4e3a7a5045051980d849700960a78ec78cfb.
Arnaud Giersch [Mon, 8 Oct 2012 07:10:24 +0000 (09:10 +0200)]
Use binary dir for bindir, and cd to home dir.
degomme [Mon, 8 Oct 2012 08:23:04 +0000 (10:23 +0200)]
add a free method for datatypes
Marion Guthmuller [Sun, 7 Oct 2012 15:21:22 +0000 (17:21 +0200)]
model-checker : forget include ucontext.h for struct smx_ctx_sysv
Marion Guthmuller [Sun, 7 Oct 2012 12:47:46 +0000 (14:47 +0200)]
model-checker : test if variable is initialized (warning with optimizations)
Marion Guthmuller [Sat, 6 Oct 2012 18:14:09 +0000 (20:14 +0200)]
model-checker : ignore differences in data segment of binary due to .plt section
Marion Guthmuller [Sat, 6 Oct 2012 17:01:04 +0000 (19:01 +0200)]
model-checker : remove maestro stack from stack comparison
Marion Guthmuller [Sat, 6 Oct 2012 09:26:11 +0000 (11:26 +0200)]
model-checker : cleanup in mc examples
Marion Guthmuller [Fri, 5 Oct 2012 17:26:25 +0000 (19:26 +0200)]
model-checker : forget to remove unused variables
Marion Guthmuller [Fri, 5 Oct 2012 17:12:11 +0000 (19:12 +0200)]
Get local variables in binary and libsimgrid.so and their values for each system state and compare them for state equality detection (to complete if optimizations are enabled)
Marion Guthmuller [Fri, 5 Oct 2012 11:27:35 +0000 (13:27 +0200)]
model-checker : remove unused variable with debug mode disabled
Marion Guthmuller [Fri, 5 Oct 2012 11:20:34 +0000 (13:20 +0200)]
model-checker : extend MC_ignore mechanism for global variables in libsimgrid
Marion Guthmuller [Fri, 5 Oct 2012 09:59:33 +0000 (11:59 +0200)]
model-checker : display information about variables for comparison of local variables
Marion Guthmuller [Mon, 1 Oct 2012 15:23:03 +0000 (17:23 +0200)]
cosmetics
Marion Guthmuller [Mon, 1 Oct 2012 15:08:01 +0000 (17:08 +0200)]
model-checker : check if a difference detected in stack comparison is a pointer on block or fragment previously free
Marion Guthmuller [Sun, 30 Sep 2012 11:53:59 +0000 (13:53 +0200)]
model-checker : functions only used in mc_global declared as static and remove unnecessary strdup
Marion Guthmuller [Sun, 30 Sep 2012 11:50:38 +0000 (13:50 +0200)]
model-checker : store local variables in raw heap
Marion Guthmuller [Sun, 30 Sep 2012 01:14:49 +0000 (03:14 +0200)]
model-checker : memory free
Marion Guthmuller [Sat, 29 Sep 2012 21:39:50 +0000 (23:39 +0200)]
model-checker : double free memory fixed
Marion Guthmuller [Sat, 29 Sep 2012 20:32:21 +0000 (22:32 +0200)]
model-checker : cosmetics reindent
Marion Guthmuller [Sat, 29 Sep 2012 20:31:04 +0000 (22:31 +0200)]
model-checker : move functions about snapshot comparison in a separate file mc_compare.c
Marion Guthmuller [Sat, 29 Sep 2012 20:23:53 +0000 (22:23 +0200)]
model-checker : take snapshot of stacks (local variables values and stack pointer)
Marion Guthmuller [Sat, 29 Sep 2012 20:20:47 +0000 (22:20 +0200)]
model-checker : store equality detected in heap comparison if fragment or block numbers are different
Marion Guthmuller [Sat, 29 Sep 2012 19:58:20 +0000 (21:58 +0200)]
model-checker : move struct s_smx_ctx_sysv in smx_private.h, needed for stacks comparison
Marion Guthmuller [Thu, 27 Sep 2012 16:33:31 +0000 (18:33 +0200)]
model-checker : detect if binary program has been compiled with -g and with optimizations
Marion Guthmuller [Thu, 27 Sep 2012 16:28:14 +0000 (18:28 +0200)]
model-checker : ignore stack areas in heap comparison algorithm for the comparison of local variables with dwarf
Marion Guthmuller [Thu, 27 Sep 2012 16:08:28 +0000 (18:08 +0200)]
model-checker : add libunwind in simgrid dependencies
Marion Guthmuller [Thu, 20 Sep 2012 08:53:54 +0000 (10:53 +0200)]
model-checker : previous algorithm for heap comparison, used to compare the results with the last one
Marion Guthmuller [Thu, 20 Sep 2012 08:50:22 +0000 (10:50 +0200)]
model-checker : print local variables only with debug mode enabled
Marion Guthmuller [Thu, 20 Sep 2012 08:43:58 +0000 (10:43 +0200)]
model-checker : memset forgotten in mmalloc
degomme [Fri, 5 Oct 2012 15:51:53 +0000 (17:51 +0200)]
handle sends of several structs in the same message
degomme [Fri, 5 Oct 2012 10:05:14 +0000 (12:05 +0200)]
bug with MPI_Type_contiguous
degomme [Fri, 5 Oct 2012 09:31:39 +0000 (11:31 +0200)]
add a declaration to remove a warning
degomme [Fri, 5 Oct 2012 09:14:21 +0000 (11:14 +0200)]
add more checking to inputs of SMPI functions
warns about too complex datatypes
avoid creating a complex datatype if it is actually contiguous
navarro [Fri, 5 Oct 2012 13:58:49 +0000 (15:58 +0200)]
Protect from SMPI
suter [Fri, 5 Oct 2012 07:12:58 +0000 (09:12 +0200)]
just bragging :)
jean-noel quintin [Thu, 4 Oct 2012 15:09:16 +0000 (16:09 +0100)]
Merge remote-tracking branch 'origin/master'
jean-noel quintin [Thu, 4 Oct 2012 15:08:14 +0000 (16:08 +0100)]
[smpi] reduce the amount of memory used with the detached communication and non-contiguous memory
jean-noel quintin [Thu, 4 Oct 2012 15:07:23 +0000 (16:07 +0100)]
[smpi] remove a memory leak linked to the detached communication
degomme [Thu, 4 Oct 2012 11:35:32 +0000 (13:35 +0200)]
generalize non contiguous send method to other MPI types (hvector, indexed, hindexed, and structs, and add tests for those)
suter [Thu, 4 Oct 2012 09:57:31 +0000 (11:57 +0200)]
slight changes in the paje trace
suter [Thu, 4 Oct 2012 09:15:55 +0000 (11:15 +0200)]
small improvements to make ctest happy
suter [Thu, 4 Oct 2012 08:52:47 +0000 (10:52 +0200)]
tell tesh that this test returns -1 and not 0
suter [Thu, 4 Oct 2012 08:43:44 +0000 (10:43 +0200)]
add a new simdag/dax test to the memcheck too
suter [Thu, 4 Oct 2012 08:42:21 +0000 (10:42 +0200)]
add a new simdag/dax test
suter [Thu, 4 Oct 2012 08:31:29 +0000 (10:31 +0200)]
add a tesh file for the example of dax loading that works fine
suter [Thu, 4 Oct 2012 08:29:34 +0000 (10:29 +0200)]
modify the test to not die poorly anymore. Just gently exit with an
error message.
suter [Thu, 4 Oct 2012 08:04:44 +0000 (10:04 +0200)]
add a little dax to make a real test with dax_test
Martin Quinson [Thu, 4 Oct 2012 07:55:05 +0000 (09:55 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Thu, 4 Oct 2012 06:03:22 +0000 (08:03 +0200)]
Thank the Jean Noel dream team in the Changelog
Arnaud Giersch [Wed, 3 Oct 2012 21:18:31 +0000 (23:18 +0200)]
Fix errors about uninitialized variables.
Not sure if it is correct but it cannot be worse and, at least,
it compiles..
Arnaud Giersch [Wed, 3 Oct 2012 14:02:26 +0000 (16:02 +0200)]
Use xbt_abort instead of abort.
Maybe this will improve the code coverage analysis.
Augustin Degomme [Wed, 3 Oct 2012 17:16:32 +0000 (19:16 +0200)]
avoid breaking the shiny new and working functionality from jean-noel
Augustin Degomme [Wed, 3 Oct 2012 16:52:01 +0000 (18:52 +0200)]
remove some warnings
Augustin Degomme [Wed, 3 Oct 2012 16:24:16 +0000 (18:24 +0200)]
merge with jean noel's commit
jean-noel quintin [Wed, 3 Oct 2012 15:28:45 +0000 (16:28 +0100)]
Merge remote-tracking branch 'origin/master'
Conflicts:
src/smpi/smpi_base.c
jean-noel quintin [Wed, 3 Oct 2012 15:24:39 +0000 (16:24 +0100)]
[smpi] add a gestion of non-contignous data
Augustin Degomme [Tue, 2 Oct 2012 16:06:29 +0000 (18:06 +0200)]
add support of MPI_PROC_NULL, correct behavior of waitall and waitsome for some special cases (when some requests are NULL), in order to conform a little more to the MPI standard
Augustin Degomme [Tue, 2 Oct 2012 11:06:54 +0000 (13:06 +0200)]
provide Test and Wait functions with a better MPI_STATUS_IGNORE and MPI_REQUEST_NULL handling, as well as returning correct empty statuses when needed
Also add MPI_Testsome support
Arnaud Giersch [Tue, 2 Oct 2012 12:37:30 +0000 (14:37 +0200)]
Call __gcov_flush() on abort so that gcda files are correctly written.
This is obviously only activated with enable_coverage=ON.
Arnaud Giersch [Tue, 2 Oct 2012 09:30:46 +0000 (11:30 +0200)]
Mark as volatile variables that "might be clobbered by ‘longjmp’ or ‘vfork’".
Arnaud Giersch [Tue, 2 Oct 2012 08:45:18 +0000 (10:45 +0200)]
Use atexit to call xbt_postexit.
Martin Quinson [Tue, 2 Oct 2012 09:51:38 +0000 (11:51 +0200)]
Do not kill all processes when MSG shuts down: simix will cleanup anyway
Plus, it broke Java since this function is now called using atexit(),
and it seems that in Java, the thread running these functions is not
maestro. So our mechanism was trying to fire a simcall from this
thread, which is a Java system thread that is unknown from SimGrid.
As a result, the JVM was segfaulting when exiting.
Martin Quinson [Tue, 2 Oct 2012 08:37:15 +0000 (10:37 +0200)]
Deprecate MSG_clean
Martin Quinson [Tue, 2 Oct 2012 07:32:36 +0000 (09:32 +0200)]
Ensures that SIMIX_clean is called automatically
Arnaud Giersch [Thu, 27 Sep 2012 14:17:51 +0000 (16:17 +0200)]
Fix paths for test smpi-replay in out-of-source builds.
Arnaud Giersch [Thu, 27 Sep 2012 14:17:27 +0000 (16:17 +0200)]
Fix paths for test simdag-test-fail in out-of-source builds.
Augustin Degomme [Mon, 1 Oct 2012 10:51:56 +0000 (12:51 +0200)]
fix bug in smpi tracing for wait
Augustin Degomme [Fri, 28 Sep 2012 09:17:51 +0000 (11:17 +0200)]
update the smpi replay tesh file again, as the problems with bandwidth and latency factors are now fixed
Augustin Degomme [Fri, 28 Sep 2012 08:05:21 +0000 (10:05 +0200)]
changes the way smpi tracing handles process ranks, to be able to use subgroups
Augustin Degomme [Fri, 28 Sep 2012 08:04:38 +0000 (10:04 +0200)]
various fixes for probing/testing