Martin Quinson [Tue, 18 Sep 2018 12:40:21 +0000 (14:40 +0200)]
dockerfiles: install our files under /source/ and refresh images
Martin Quinson [Tue, 18 Sep 2018 10:58:27 +0000 (12:58 +0200)]
tuto smpi: finish (for now) the platform section; draft the install section
Martin Quinson [Tue, 18 Sep 2018 07:50:32 +0000 (09:50 +0200)]
cosmetics
Martin Quinson [Tue, 18 Sep 2018 07:41:57 +0000 (09:41 +0200)]
cosmetics in graphical representations of cluster descriptions
Martin Quinson [Tue, 18 Sep 2018 06:39:59 +0000 (08:39 +0200)]
hopefully fix the NS3 test
Martin Quinson [Tue, 18 Sep 2018 05:39:18 +0000 (07:39 +0200)]
fix make distcheck, as usual :(
Martin Quinson [Tue, 18 Sep 2018 00:03:38 +0000 (02:03 +0200)]
ignore a directory generated by sphinx
Martin Quinson [Tue, 18 Sep 2018 00:02:56 +0000 (02:02 +0200)]
docs: sphinx 1.8.0 was released, so use it
Martin Quinson [Mon, 17 Sep 2018 23:58:32 +0000 (01:58 +0200)]
Merge branch 'master' of github.com:simgrid/simgrid
Martin Quinson [Mon, 17 Sep 2018 22:47:16 +0000 (00:47 +0200)]
Rename cluster.xml to cluster_backbone.xml
also, fix the make dist and some cosmetics.
Martin Quinson [Mon, 17 Sep 2018 22:30:18 +0000 (00:30 +0200)]
cleanups in the cluster platform files
Martin Quinson [Mon, 17 Sep 2018 22:16:46 +0000 (00:16 +0200)]
docs: prefer svg to png, and inclusion to copy/paste
Martin Quinson [Mon, 17 Sep 2018 21:46:10 +0000 (23:46 +0200)]
cosmetics on the graphical TOC
Martin Quinson [Mon, 17 Sep 2018 07:54:46 +0000 (09:54 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Fri, 14 Sep 2018 20:54:04 +0000 (22:54 +0200)]
fix sectionning and one typo
Arnaud Legrand [Fri, 14 Sep 2018 09:17:09 +0000 (11:17 +0200)]
Graphical representation of example platforms
Martin Quinson [Thu, 13 Sep 2018 22:31:12 +0000 (00:31 +0200)]
Merge pull request #292 from kovin/master
Cover with a test Mailbox::ready() method introduced in commit
1ed0e64dc40
Martin Quinson [Thu, 13 Sep 2018 20:00:17 +0000 (22:00 +0200)]
Merge branch 'master' into master
Martin Quinson [Tue, 11 Sep 2018 23:53:17 +0000 (01:53 +0200)]
SMPI tuto: Start stealing content from SMPI courseware
Martin Quinson [Tue, 11 Sep 2018 23:17:36 +0000 (01:17 +0200)]
tuto smpi: add a picture explaining how it works
Martin Quinson [Tue, 11 Sep 2018 23:16:33 +0000 (01:16 +0200)]
allow to have hidden/shown code blocks in the doc
Arnaud Giersch [Tue, 11 Sep 2018 20:35:20 +0000 (22:35 +0200)]
Add an assert/fixme around Actor::set_auto_restart.
Arnaud Giersch [Tue, 11 Sep 2018 20:27:20 +0000 (22:27 +0200)]
Use a std::vector for actors_at_boot_.
Several actors may use the same name (e.g. app-masterworker-multicore).
Also fixes a memory leak.
Martin Quinson [Tue, 11 Sep 2018 16:37:58 +0000 (18:37 +0200)]
start the SMPI tuto
Arnaud Giersch [Fri, 31 Aug 2018 11:19:34 +0000 (13:19 +0200)]
Typo.
Martin Quinson [Mon, 10 Sep 2018 21:30:38 +0000 (23:30 +0200)]
tuto: don't speak of s4u processes (but actors)
Martin Quinson [Mon, 10 Sep 2018 21:17:30 +0000 (23:17 +0200)]
docs: simplify and document that file
Martin Quinson [Mon, 10 Sep 2018 21:01:01 +0000 (23:01 +0200)]
killing trailing whitespaces on png files is not cleaver
Martin Quinson [Mon, 10 Sep 2018 20:33:39 +0000 (22:33 +0200)]
DTD: remove the last occurence of <gpu>
Martin Quinson [Mon, 10 Sep 2018 20:30:49 +0000 (22:30 +0200)]
tesh: informative message for another error condition
Martin Quinson [Mon, 10 Sep 2018 19:58:04 +0000 (21:58 +0200)]
Fix the DTD to not allow to mix internal node content with leaf content in a given zone
Fix https://github.com/simgrid/simgrid/issues/296
Martin Quinson [Mon, 10 Sep 2018 14:19:17 +0000 (16:19 +0200)]
fix the SMPI tests that mandate smpi/wtime == 0
Martin Quinson [Mon, 10 Sep 2018 13:03:49 +0000 (15:03 +0200)]
align doc and code on a more sensible value
Martin Quinson [Mon, 10 Sep 2018 12:42:52 +0000 (14:42 +0200)]
Merge branch 'master' of framagit.org:simgrid/simgrid
Martin Quinson [Mon, 10 Sep 2018 12:39:55 +0000 (14:39 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Mon, 10 Sep 2018 12:35:57 +0000 (14:35 +0200)]
Improve option smpi/wtime
- Set default value to 1ms instead of 0. This default settings may
lead to slower simulation, but it works in more situations.
- Also apply this delay in gettimeofday() and clock_gettime()
- Improve the documentation.
Augustin Degomme [Mon, 10 Sep 2018 11:39:29 +0000 (13:39 +0200)]
Allow insertion of time inside gettimeofday and clock_gettime
Done with --cfg=smpi/wtime, which was previously only for MPI_Wtime.
This should avoid some infinite loops. Keep 0 as default for now.
Martin Quinson [Mon, 10 Sep 2018 11:02:22 +0000 (13:02 +0200)]
move smpi_mpi_wtime near to the other time-related functions
Martin Quinson [Thu, 6 Sep 2018 19:39:26 +0000 (21:39 +0200)]
dont use send/receive on mailboxes, but put/get
FREDERIC SUTER [Wed, 5 Sep 2018 10:56:09 +0000 (12:56 +0200)]
Update app_s4u.rst
FREDERIC SUTER [Wed, 5 Sep 2018 10:17:03 +0000 (12:17 +0200)]
Update application.rst
Martin Quinson [Mon, 3 Sep 2018 19:41:38 +0000 (21:41 +0200)]
try to fix windows builds
ContextJava uses ContextThread as a superclass now, but they are not
in the same lib, so ContextThread must be exported as public.
FREDERIC SUTER [Mon, 3 Sep 2018 12:37:59 +0000 (14:37 +0200)]
Update intro_yours.rst
Augustin Degomme [Wed, 29 Aug 2018 12:31:17 +0000 (14:31 +0200)]
Multiply memset size by size of element in umpire.
FREDERIC SUTER [Mon, 3 Sep 2018 12:00:55 +0000 (14:00 +0200)]
Update intro_install.rst
FREDERIC SUTER [Mon, 3 Sep 2018 11:17:08 +0000 (13:17 +0200)]
Update intro_concepts.rst
Martin Quinson [Mon, 3 Sep 2018 07:34:38 +0000 (09:34 +0200)]
fix make distcheck
Martin Quinson [Mon, 3 Sep 2018 07:20:56 +0000 (09:20 +0200)]
Somehow fix the killing of actors in Java
Things are somehow fixed, as all tests seem to pass, but the situation
is still very messy after this commit. Contents:
- Reimplement ContextJava as subclass of ContextThread to reduce duplication.
- Don't send the StopRequest exception on host failure if we are in
Java because *some* of the actors don't catch it well, resulting in
simulation failure.
- Forcefully kill the process ("exit(0)" in C) after MSG_run() because
dead actors are sometimes not completely killed, preventing the
simulation from ending.
See the comment in ActorImpl for a better understanding of this mess
and how to fix it in the future.
Martin Quinson [Sun, 2 Sep 2018 19:35:09 +0000 (21:35 +0200)]
cosmetics while debuging backtraces
Martin Quinson [Sun, 2 Sep 2018 00:17:06 +0000 (02:17 +0200)]
java: obey our coding standard
Martin Quinson [Sun, 2 Sep 2018 00:09:27 +0000 (02:09 +0200)]
don't catch an exception that is never thrown
xbt_os_thread_create() asserts that it succeeds, it does not throw
anything. So put the documentation in the doc instead of displaying it
when that non-existent exception is received.
Martin Quinson [Sun, 2 Sep 2018 00:02:21 +0000 (02:02 +0200)]
java: cosmetics
Martin Quinson [Sat, 1 Sep 2018 23:11:54 +0000 (01:11 +0200)]
that was converted to sphinx
Martin Quinson [Sat, 1 Sep 2018 20:56:32 +0000 (22:56 +0200)]
Remove the deprecated 'state' attribute from the doc
This fixes https://github.com/simgrid/simgrid/issues/295
Martin Quinson [Sat, 1 Sep 2018 20:53:51 +0000 (22:53 +0200)]
docs: write the overall section of 'Applications'
Martin Quinson [Fri, 31 Aug 2018 15:58:58 +0000 (17:58 +0200)]
sphinx: one warning less
Martin Quinson [Thu, 30 Aug 2018 09:37:40 +0000 (11:37 +0200)]
Bummer. Really fix out of tree builds (I hope)
Martin Quinson [Thu, 30 Aug 2018 07:38:36 +0000 (09:38 +0200)]
fix out of tree builds
Martin Quinson [Wed, 29 Aug 2018 21:11:37 +0000 (23:11 +0200)]
fix maestro-set
Martin Quinson [Wed, 29 Aug 2018 20:50:07 +0000 (22:50 +0200)]
disable the platform-failure tests for now, sorry
I fail to debug such complex tests, I need smaller ones such as the
activity-lifecycle that I'm currently growing.
But broken tests in the git prevents everybody from working, including
me. I broke msg-maestro-set-thread at some point and did not even
notice :(
Sorry for breaking the failure platform tests in the first place.
Martin Quinson [Wed, 29 Aug 2018 20:31:09 +0000 (22:31 +0200)]
kill a superseeded sub-test, and fix another one
Processes on failing host are killed right away, so it cannot report
that the host failed as expected.
This whole test should be converted to activity-lifecycle.
Martin Quinson [Wed, 29 Aug 2018 20:13:26 +0000 (22:13 +0200)]
fix make dist
Martin Quinson [Wed, 29 Aug 2018 20:11:31 +0000 (22:11 +0200)]
this test is superseeded by activity-lifecycle
Martin Quinson [Wed, 29 Aug 2018 20:04:11 +0000 (22:04 +0200)]
simplify the actor finalization a tiny bit by using a callback
This is part of the removal of all trace-related pimpl all over the
code of MSG (my goal is to kill MSG_process_cleanup_from_SIMIX() all
together).
Note that I changed from Container::by_name() to
Container::by_name_or_null. It seems that not all actors have a
container by their name, not sure why.
Martin Quinson [Wed, 29 Aug 2018 19:24:26 +0000 (21:24 +0200)]
Convert all xbt_ex(network_error) throwing locations
Martin Quinson [Wed, 29 Aug 2018 19:19:40 +0000 (21:19 +0200)]
typo
Martin Quinson [Wed, 29 Aug 2018 13:19:17 +0000 (15:19 +0200)]
sonar
Martin Quinson [Wed, 29 Aug 2018 13:18:47 +0000 (15:18 +0200)]
woops
Martin Quinson [Wed, 29 Aug 2018 12:17:35 +0000 (14:17 +0200)]
fix 32b builds
Martin Quinson [Wed, 29 Aug 2018 11:14:01 +0000 (13:14 +0200)]
please sonar on rethrow
Martin Quinson [Wed, 29 Aug 2018 09:35:10 +0000 (11:35 +0200)]
Display a msg when contexts are killed by uncatched exceptions
and when I want to really kill an actor (eg when its host is turned
off), I launch an uncatchable kernel::Context::StopRequest instead of
a catchable simgrid::HostFailureException (which will be used in case
of remote exec and similar)
Maybe there should be a config flag to decide if we want to kill the
simulation when an actor fails. The current setting forces the user to
add try/catch (simgrid::Exception) around their main functions. That's
not a bad thing either, not sure.
Martin Quinson [Wed, 29 Aug 2018 00:10:12 +0000 (02:10 +0200)]
Let's exhaustively test the activity lifecycle
This test is not complete yet. It aims at being as exhaustive and
paranoid as possible, just like cloud-sharing even if I didn't find a
good DSL to specify the tests this time.
Martin Quinson [Wed, 29 Aug 2018 00:08:18 +0000 (02:08 +0200)]
improve debug messages and error reporting
Martin Quinson [Tue, 28 Aug 2018 23:59:17 +0000 (01:59 +0200)]
Properly kill the context on HostFailureException
Before, simix was kinda thinking that the actor was dead, but the
context was still running, leading to a Holy Big Mess!
Augustin Degomme [Tue, 28 Aug 2018 15:46:08 +0000 (17:46 +0200)]
update doc
Augustin Degomme [Tue, 28 Aug 2018 15:39:33 +0000 (17:39 +0200)]
Switch to ompi for umpire tests.
MPICH changes brought SMP-aware algorithm, which MC does not really like.
I guess the init_smp is the culprit here, as it uses badly various collectives.
Augustin Degomme [Tue, 28 Aug 2018 15:37:50 +0000 (17:37 +0200)]
update ompi selector as well with "recent" version
Augustin Degomme [Tue, 28 Aug 2018 14:32:41 +0000 (16:32 +0200)]
Requalify automatic tesh, as another algorithm is used in init_smp now.
Augustin Degomme [Tue, 28 Aug 2018 14:29:22 +0000 (16:29 +0200)]
update doc with new algo
Augustin Degomme [Tue, 28 Aug 2018 14:23:24 +0000 (16:23 +0200)]
Upgrade MPICH collective selector to 3.3.
Add SMP variants of some algorithms, and protect against side effects.
Martin Quinson [Sun, 26 Aug 2018 23:49:51 +0000 (01:49 +0200)]
circleci: do not optimise builds, you're supposed to be as fast as hell
Augustin Degomme [Tue, 28 Aug 2018 08:19:07 +0000 (10:19 +0200)]
Fix https://github.com/simgrid/simgrid/issues/294
Martin Quinson [Sun, 26 Aug 2018 23:42:53 +0000 (01:42 +0200)]
Not sure of why it helps now
Martin Quinson [Sun, 26 Aug 2018 23:42:32 +0000 (01:42 +0200)]
fix travis builds
Martin Quinson [Sun, 26 Aug 2018 22:37:52 +0000 (00:37 +0200)]
strenghten this test
Martin Quinson [Sun, 26 Aug 2018 20:50:16 +0000 (22:50 +0200)]
MSG_process_sleep should intercept HostFailureException and report it accordingly
Don't ask me how it could have worked before, but they were a
C++ try/catch in teshsuite/msg/host_on_off_processes. In a MSG code!!
Martin Quinson [Sun, 26 Aug 2018 20:14:26 +0000 (22:14 +0200)]
When the host dies, the actor need an exception even if it's not blocked on an activity
Martin Quinson [Sun, 26 Aug 2018 11:30:15 +0000 (13:30 +0200)]
some compilers cannot see that this value is initialized in all cases
Martin Quinson [Sun, 26 Aug 2018 09:49:40 +0000 (11:49 +0200)]
Do not throw exception in maestro when host->is_off + sleep()
Also, use ActorImpl::throw_exception() instead of messing with its
wannabe private exception_ field.
Martin Quinson [Sun, 26 Aug 2018 00:56:22 +0000 (02:56 +0200)]
convert some xbt_ex(tracing_error) into xbt_assert
Martin Quinson [Sun, 26 Aug 2018 00:09:00 +0000 (02:09 +0200)]
implement THROW_IMPOSSIBLE and THROW_UNIMPLEMENTED with std::runtime_error directly
Martin Quinson [Sat, 25 Aug 2018 23:31:24 +0000 (01:31 +0200)]
convert a xbt_ex(arg_error) into a std::invalid_argument
Martin Quinson [Sat, 25 Aug 2018 23:14:32 +0000 (01:14 +0200)]
deprecate SIMIX_process_throw for ActorImpl::throw_exception
Martin Quinson [Sat, 25 Aug 2018 21:38:15 +0000 (23:38 +0200)]
convert xbt_ex(timeout_error) catching locations to TimeoutError
Martin Quinson [Sat, 25 Aug 2018 22:51:00 +0000 (00:51 +0200)]
Do not swallow exceptions I don't know
Martin Quinson [Sat, 25 Aug 2018 22:36:16 +0000 (00:36 +0200)]
Do not convert TimeoutError to xbt_ex(timeout) in case they were a wait_any
If there is an issue while dealing with a test_any or a wait_any, the
caller must be told which activity failed. I'm not sure of how to
cleanly do so. For now, we use exception.value to store the rank of
that activity in the container.
To modify the exception, C++ leaves us no way but to rethrow it and
recatch it, change its value field, and re-store it in the
issuer->exception. But then, the exception become of the catching
type. Wicked! Vicious! It means that since we were catching (xbt_ex&
e), we actually converted the simgrid::TimeoutException into a xbt_ex.
And this conversion was done in any case, even if the value was set
only if the simcall was actually a wait_any or test_any...
With this commit, we catch, extend and rethrow any TimeoutException,
and if it's not such an xbt_ex, we do the same for a xbt_ex.
A proper version could involve a WaitAnyException (with failing_rank
and cause fields), or maybe the TimeoutException could contain a
pointer to the timeouted activity so that the caller can find its rank
by itself.
Martin Quinson [Sat, 25 Aug 2018 20:02:07 +0000 (22:02 +0200)]
convert all xbt_ex(timeout_error) throwing locations to simgrid::TimeoutError
Martin Quinson [Sat, 25 Aug 2018 19:51:19 +0000 (21:51 +0200)]
convert all xbt_ex(host_error) catching locations to simgrid::HostFailureException
Martin Quinson [Sat, 25 Aug 2018 19:35:15 +0000 (21:35 +0200)]
Replace all xbt_ex(host_error) throwing location with HostFailureException
Martin Quinson [Sat, 25 Aug 2018 13:57:02 +0000 (15:57 +0200)]
convert some catch locations to simgrid::HostFailureException