degomme [Fri, 29 Apr 2016 13:28:20 +0000 (15:28 +0200)]
suppression++
degomme [Fri, 29 Apr 2016 13:25:37 +0000 (15:25 +0200)]
leaks --
degomme [Fri, 29 Apr 2016 11:22:04 +0000 (13:22 +0200)]
leak--
degomme [Fri, 29 Apr 2016 09:23:43 +0000 (11:23 +0200)]
bcast on one process is a no-op
degomme [Fri, 29 Apr 2016 09:23:28 +0000 (11:23 +0200)]
leak --
degomme [Fri, 29 Apr 2016 08:55:32 +0000 (10:55 +0200)]
segfaulting is not leaking - exclude this test from memcheck
Gabriel Corona [Fri, 29 Apr 2016 08:47:46 +0000 (10:47 +0200)]
Ensure that the dict subsystem is initialized when creating a dict
degomme [Fri, 29 Apr 2016 00:28:06 +0000 (02:28 +0200)]
this test actually caused an overflow which led to strange things and leaks.
degomme [Thu, 28 Apr 2016 23:44:24 +0000 (01:44 +0200)]
Don't create structures for 0 sized datatypes
Gabriel Corona [Thu, 28 Apr 2016 19:58:16 +0000 (21:58 +0200)]
Remove spurious XBT_PUBLIC
degomme [Thu, 28 Apr 2016 15:01:07 +0000 (17:01 +0200)]
fix warning and build
degomme [Thu, 28 Apr 2016 14:59:18 +0000 (16:59 +0200)]
Avoid allocating 0 bytes (issue with mallocators only?)
degomme [Thu, 28 Apr 2016 14:30:23 +0000 (16:30 +0200)]
Some more cleanups in datatypes end of life
degomme [Thu, 28 Apr 2016 13:37:09 +0000 (15:37 +0200)]
leak-- (duplicate of a predefined datatype can be released)
Gabriel Corona [Thu, 28 Apr 2016 13:36:08 +0000 (15:36 +0200)]
Remove useless boost #includes
Gabriel Corona [Thu, 28 Apr 2016 13:21:06 +0000 (15:21 +0200)]
Use C++ Flag for context factory selection
Gabriel Corona [Thu, 28 Apr 2016 13:20:46 +0000 (15:20 +0200)]
[config] Add a template declareFlag<T>(...)
degomme [Thu, 28 Apr 2016 12:39:57 +0000 (14:39 +0200)]
fix The leak
Gabriel Corona [Thu, 28 Apr 2016 10:38:52 +0000 (12:38 +0200)]
[mc] Fix compilation when MC id disabled
Gabriel Corona [Thu, 28 Apr 2016 09:54:31 +0000 (11:54 +0200)]
[mc] Remove redundant 'mc_mode' global variable
Martin Quinson [Thu, 28 Apr 2016 08:44:08 +0000 (10:44 +0200)]
start the development of 3.14 right away
Martin Quinson [Thu, 28 Apr 2016 08:42:10 +0000 (10:42 +0200)]
note to self
Martin Quinson [Wed, 27 Apr 2016 23:18:29 +0000 (01:18 +0200)]
gosh, that logo is awful
degomme [Thu, 28 Apr 2016 00:15:14 +0000 (02:15 +0200)]
indent to avoid GCC6 warning madness
degomme [Thu, 28 Apr 2016 00:01:08 +0000 (02:01 +0200)]
setup a cleanup routine for mvapich collectives.
We allocate some tables at first use that we never released.
degomme [Wed, 27 Apr 2016 23:59:56 +0000 (01:59 +0200)]
leaks --
Martin Quinson [Wed, 27 Apr 2016 23:04:22 +0000 (01:04 +0200)]
fix a bunch of typos reported by lintian
Lintian is one of the debian quality assurance many robots
Martin Quinson [Wed, 27 Apr 2016 22:53:56 +0000 (00:53 +0200)]
this file was removed
Martin Quinson [Wed, 27 Apr 2016 19:42:19 +0000 (21:42 +0200)]
that was funny, but that's a probable copyright violation
Martin Quinson [Wed, 27 Apr 2016 19:39:16 +0000 (21:39 +0200)]
Fix the detection of backtrace
degomme [Wed, 27 Apr 2016 19:39:46 +0000 (21:39 +0200)]
Ignore leaks that are not ours
degomme [Wed, 27 Apr 2016 19:33:43 +0000 (21:33 +0200)]
Netcards are bound to die, ignore their leaks.
degomme [Wed, 27 Apr 2016 19:10:28 +0000 (21:10 +0200)]
order output of some tests, for windows
Martin Quinson [Wed, 27 Apr 2016 19:26:05 +0000 (21:26 +0200)]
This is v3.13.
Martin Quinson [Wed, 27 Apr 2016 18:52:45 +0000 (20:52 +0200)]
we don't do nsis releases nor separate bindings anymore
Martin Quinson [Wed, 27 Apr 2016 17:20:31 +0000 (19:20 +0200)]
fix the detection of backtrace: not all linuxes have it
In particular, this is not implemented on armhf.
Martin Quinson [Wed, 27 Apr 2016 17:10:22 +0000 (19:10 +0200)]
support for backtraces on windows is gone
Martin Quinson [Wed, 27 Apr 2016 16:19:34 +0000 (18:19 +0200)]
plug a memleak in that example
Martin Quinson [Wed, 27 Apr 2016 09:25:17 +0000 (11:25 +0200)]
these files are gone, don't specify their optim level
degomme [Wed, 27 Apr 2016 15:49:45 +0000 (17:49 +0200)]
Fix awful code.
degomme [Wed, 27 Apr 2016 15:31:09 +0000 (17:31 +0200)]
Send the right size here.
degomme [Wed, 27 Apr 2016 13:21:12 +0000 (15:21 +0200)]
still not our
degomme [Wed, 27 Apr 2016 13:21:00 +0000 (15:21 +0200)]
Add warning, as next Boost release will break boost context factory.
In details, fcontext_t is no longer part of the public API
degomme [Wed, 27 Apr 2016 09:26:11 +0000 (11:26 +0200)]
leak--
degomme [Wed, 27 Apr 2016 09:13:44 +0000 (11:13 +0200)]
silence stupid warning
degomme [Wed, 27 Apr 2016 09:03:14 +0000 (11:03 +0200)]
leak --
Martin Quinson [Wed, 27 Apr 2016 09:12:04 +0000 (11:12 +0200)]
don't test for boost-graph as we don't use it yet
degomme [Wed, 27 Apr 2016 08:54:59 +0000 (10:54 +0200)]
fix some more leaks
smpi_datatype_free does not update refcount, use unuse instead.
degomme [Wed, 27 Apr 2016 08:32:34 +0000 (10:32 +0200)]
not our leak.
degomme [Wed, 27 Apr 2016 08:32:27 +0000 (10:32 +0200)]
remove some bad casts
Gabriel Corona [Wed, 27 Apr 2016 08:34:02 +0000 (10:34 +0200)]
Add missing #include for std::current_exception()
Martin Quinson [Wed, 27 Apr 2016 08:24:56 +0000 (10:24 +0200)]
document signals
Martin Quinson [Wed, 27 Apr 2016 07:59:52 +0000 (09:59 +0200)]
reindent
Frederic Suter [Wed, 27 Apr 2016 07:53:52 +0000 (09:53 +0200)]
edits
Martin Quinson [Wed, 27 Apr 2016 07:35:44 +0000 (09:35 +0200)]
proposal of NEWS and ChangeLog files
Martin Quinson [Wed, 27 Apr 2016 06:50:42 +0000 (08:50 +0200)]
plug some harmless memleaks
degomme [Tue, 26 Apr 2016 23:39:42 +0000 (01:39 +0200)]
leaks --, for a change
Martin Quinson [Tue, 26 Apr 2016 20:43:25 +0000 (22:43 +0200)]
plug the memleaks of the cloud-capping example
Frederic Suter [Tue, 26 Apr 2016 20:17:56 +0000 (22:17 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Conflicts:
tools/cmake/Tests.cmake
Frederic Suter [Tue, 26 Apr 2016 20:04:02 +0000 (22:04 +0200)]
couple cleanups
Martin Quinson [Tue, 26 Apr 2016 19:21:15 +0000 (21:21 +0200)]
energy: plug some memleaks
Martin Quinson [Tue, 26 Apr 2016 19:16:02 +0000 (21:16 +0200)]
plug a memleak in VM
Martin Quinson [Tue, 26 Apr 2016 15:52:43 +0000 (17:52 +0200)]
plug 2 memleaks
They were "still reachable", but annoying nonetheless.
Martin Quinson [Tue, 26 Apr 2016 08:11:54 +0000 (10:11 +0200)]
mbox: kill an unused field
Martin Quinson [Tue, 26 Apr 2016 07:58:49 +0000 (09:58 +0200)]
complete the API of s4u::Mailbox
degomme [Tue, 26 Apr 2016 15:43:32 +0000 (17:43 +0200)]
leak--
degomme [Tue, 26 Apr 2016 14:07:22 +0000 (16:07 +0200)]
If we call them -thread-, might as well use thread factory here.
degomme [Tue, 26 Apr 2016 14:06:57 +0000 (16:06 +0200)]
leaks again
degomme [Tue, 26 Apr 2016 08:56:20 +0000 (10:56 +0200)]
not ours
Gabriel Corona [Tue, 26 Apr 2016 13:56:09 +0000 (15:56 +0200)]
Add a comment/TODO about i386 RawContext
Gabriel Corona [Tue, 26 Apr 2016 09:41:35 +0000 (11:41 +0200)]
RawContext: add comments for x86 raw_swapcontext()
Frederic Suter [Tue, 26 Apr 2016 11:34:55 +0000 (13:34 +0200)]
renaming a bunch of tests
Frederic Suter [Tue, 26 Apr 2016 09:22:55 +0000 (11:22 +0200)]
kill exception example
Frederic Suter [Tue, 26 Apr 2016 09:21:42 +0000 (11:21 +0200)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Frederic Suter [Tue, 26 Apr 2016 09:21:27 +0000 (11:21 +0200)]
renaming properties to platform-properties
Frederic Suter [Tue, 26 Apr 2016 09:20:05 +0000 (11:20 +0200)]
a couple of rewritings
Martin Quinson [Tue, 26 Apr 2016 07:36:24 +0000 (09:36 +0200)]
fix the lua test
Martin Quinson [Tue, 26 Apr 2016 06:49:34 +0000 (08:49 +0200)]
also alias the 3.12 option names
Martin Quinson [Mon, 25 Apr 2016 20:21:02 +0000 (22:21 +0200)]
and now kebab-case the tracing options
degomme [Mon, 25 Apr 2016 23:58:03 +0000 (01:58 +0200)]
leaks --
degomme [Mon, 25 Apr 2016 23:44:17 +0000 (01:44 +0200)]
typo made alias non functional
degomme [Mon, 25 Apr 2016 21:19:20 +0000 (23:19 +0200)]
This one was not ours.
degomme [Mon, 25 Apr 2016 21:19:11 +0000 (23:19 +0200)]
n-th cleanup on datatype refcount handling
degomme [Mon, 25 Apr 2016 18:42:00 +0000 (20:42 +0200)]
Yet another wrapper for main..
Handles the return value of main, and return value of SMPI will be the first non 0 one.
degomme [Mon, 25 Apr 2016 16:19:13 +0000 (18:19 +0200)]
We don't support MPI_Init( 0, 0 ), actually, and crash when this happens.
But these tests are precisely checking an incorrect output, so crashing was a success.
Martin Quinson [Mon, 25 Apr 2016 19:58:38 +0000 (21:58 +0200)]
All options are consistently in kebab-case (but the tracing ones)
Pushing before converting the tracing options as I broke jenkins once
again.
Martin Quinson [Mon, 25 Apr 2016 18:40:58 +0000 (20:40 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Mon, 25 Apr 2016 18:40:04 +0000 (20:40 +0200)]
kebab-case some smpi options
An alias of the old name is setup, so this is transparent for old timers.
degomme [Mon, 25 Apr 2016 13:35:02 +0000 (15:35 +0200)]
leaks -- with smp algos
Gabriel Corona [Mon, 25 Apr 2016 12:52:17 +0000 (14:52 +0200)]
Use our own minimal signal implementation
Gabriel Corona [Mon, 25 Apr 2016 12:50:46 +0000 (14:50 +0200)]
Add #includes in cpu_interface.hpp
Gabriel Corona [Mon, 25 Apr 2016 12:01:52 +0000 (14:01 +0200)]
Fix bad pointer cast in the L07 model
Martin Quinson [Mon, 25 Apr 2016 11:18:50 +0000 (13:18 +0200)]
this file should be compilable in C
Gabriel Corona [Mon, 25 Apr 2016 10:36:51 +0000 (12:36 +0200)]
bg-set-signal.tesh, Don't run the perl interpreter asynchronously
In this tesh:
! expect signal SIGSEGV
& perl segfault.pl
$ cd ..
$ cmake -E remove_directory temp_testdir-bg-set-signal
The segdefault.pl is sometimes deleted (by cmake) before being opened
by the perl interpreter.
Gabriel Corona [Mon, 25 Apr 2016 10:25:13 +0000 (12:25 +0200)]
[mc] Really ignore the output of bugged1_liveness
degomme [Sun, 24 Apr 2016 21:21:38 +0000 (23:21 +0200)]
Fix build
degomme [Sun, 24 Apr 2016 21:15:32 +0000 (23:15 +0200)]
leaks --
Martin Quinson [Sun, 24 Apr 2016 19:49:05 +0000 (21:49 +0200)]
start converting option names to kebab-case
Martin Quinson [Sat, 23 Apr 2016 21:01:36 +0000 (23:01 +0200)]
Revert "We never use the name of the mailbox"
Actually, MC use the name of the mailbox for some reason.
This reverts commit
cd1b6fc186f641dd5e3fb68d793c6253e1b8fcd9.
Martin Quinson [Sat, 23 Apr 2016 20:20:29 +0000 (22:20 +0200)]
populate s4u::mailbox
The evil plan is to use that instead of smx_mailbox_t