Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
simgrid.git
14 months agoMerge branch 'simgrid-udpor-integration' into 'master'
Arnaud Giersch [Thu, 23 Feb 2023 17:00:25 +0000 (18:00 +0100)]
Merge branch 'simgrid-udpor-integration' into 'master'

Phase 1 of Integration of UDPOR: Adding UDPOR Constructs

See merge request simgrid/simgrid!132

14 months agoUseless casts to (void*).
Arnaud Giersch [Thu, 23 Feb 2023 16:14:51 +0000 (17:14 +0100)]
Useless casts to (void*).

14 months agoComment is not relevant anymore.
Arnaud Giersch [Thu, 23 Feb 2023 15:48:28 +0000 (16:48 +0100)]
Comment is not relevant anymore.

[ci-skip]

14 months agoUse type aid_t for an actor's PID.
Arnaud Giersch [Thu, 23 Feb 2023 14:58:28 +0000 (15:58 +0100)]
Use type aid_t for an actor's PID.

14 months agoMore Sonar issues.
Arnaud Giersch [Thu, 23 Feb 2023 14:53:26 +0000 (15:53 +0100)]
More Sonar issues.

14 months agoUse normal strings instead of f-strings (Sonar).
Arnaud Giersch [Thu, 23 Feb 2023 14:37:40 +0000 (15:37 +0100)]
Use normal strings instead of f-strings (Sonar).

14 months agoUse C++17's std::scoped_lock where appropriate.
Arnaud Giersch [Thu, 23 Feb 2023 13:00:47 +0000 (14:00 +0100)]
Use C++17's std::scoped_lock where appropriate.

Stick to C++14's std::lock_guard in public headers.

14 months agoDon't access variable ouside of the critical section.
Arnaud Giersch [Thu, 23 Feb 2023 12:53:15 +0000 (13:53 +0100)]
Don't access variable ouside of the critical section.

TSan warned about a data race there.
Also use a std::scoped_lock to handle the mutexi (Sonar).

14 months agoVarious cleanups.
Arnaud Giersch [Tue, 21 Feb 2023 10:59:44 +0000 (11:59 +0100)]
Various cleanups.

Includes:
* prefer simpler overload for channel send/receive
* use direct-initialization for variables
* rename variable and avoid it to be shadowed later
* remove unnecessary casts

14 months agoPartially revert 2ad536e710c5936ff8e525e4bbb5e7046f292aac (Simplify the library initi...
Martin Quinson [Wed, 22 Feb 2023 15:55:44 +0000 (16:55 +0100)]
Partially revert 2ad536e710c5936ff8e525e4bbb5e7046f292aac (Simplify the library initialization) to fix a memleak

14 months agoImplement the semaphore functions in sthread
Martin Quinson [Wed, 22 Feb 2023 15:54:18 +0000 (16:54 +0100)]
Implement the semaphore functions in sthread

14 months agoAdd tests for adding events to configurations
Maxwell Pirtle [Wed, 22 Feb 2023 12:40:08 +0000 (13:40 +0100)]
Add tests for adding events to configurations

14 months agoAdd preliminary tests for Configuration
Maxwell Pirtle [Wed, 22 Feb 2023 12:09:50 +0000 (13:09 +0100)]
Add preliminary tests for Configuration

14 months agoprotect the dict initialization against multithreading
Martin Quinson [Wed, 22 Feb 2023 11:37:36 +0000 (12:37 +0100)]
protect the dict initialization against multithreading

14 months agoImplement adding events to configurations
Maxwell Pirtle [Wed, 22 Feb 2023 10:21:06 +0000 (11:21 +0100)]
Implement adding events to configurations

When adding an event to a configuration, we
assert that the configuration actually contains
all of the dependencies of that event within it;
otherwise, we would violate the property that a
configuration contains all of its dependencies.

Note that we do not assert whether or not the
configuration is actually conflict-free: we leave
that to the correctness of UDPOR and other tests which
will assert UDPOR is doing the right thing. But having
this assertion will certainly help catch bugs with the
UDPOR implementation more quickly (termination would
happen at the moment that the algorithm encounters
a spot where it chose an event incorrectly).

14 months agoAdd extensive tests for checking valid configurations
Maxwell Pirtle [Wed, 22 Feb 2023 09:36:33 +0000 (10:36 +0100)]
Add extensive tests for checking valid configurations

An example event structure was given and all
possible subsets of its events were tested for
completeness. Each subset was tested for whether
it was a maximal subset of events (i.e. one for
which no element causes any of the others) and
for whether the set of events qualified for being
a configuration.

An additional observation was made concerning the determination of
whether or not a set of events is maximal. To compute
the maximal events of a given set of events, it suffices
to search the entire history of the events and eliminate
events from the original set as they appear. Note that only
events in the set itself can ever be maximal, as all other
events discovered in the search must have come from one
of these event's dependencies, or one of their dependencies'
dependencies, and so on.

14 months agoCosmetics
Martin Quinson [Tue, 21 Feb 2023 21:08:45 +0000 (22:08 +0100)]
Cosmetics

- Rename xbt_main.cpp -> xbt_misc.cpp to reflect its content
- Stick to our naming schema (Engine::pimpl -> Engine::pimpl_)
- Don't say the loop iterator is const to then const_cast it
- Kill the leftover xbt::binary_name and xbt::cmdline local vars

14 months agologs: protect finalization against multiple calls
Martin Quinson [Tue, 21 Feb 2023 17:21:57 +0000 (18:21 +0100)]
logs: protect finalization against multiple calls

14 months agoLet the xbt_log module register its finalizer itself
Martin Quinson [Tue, 21 Feb 2023 15:00:06 +0000 (16:00 +0100)]
Let the xbt_log module register its finalizer itself

And kill an internal header file that is now empty

14 months agoAdd method to check if an EventSet is a config
Maxwell Pirtle [Tue, 21 Feb 2023 14:53:53 +0000 (15:53 +0100)]
Add method to check if an EventSet is a config

A method was added that determines whether or not
some set of events (i.e. an `EventSet`) are actually
a valid configuration or not. The question of choosing
valid configurations is up to UDPOR, so while enforcing
that each and every event added to a configuration
be checked to ensure that the configuration remains
as such, it is not strictly necessary. However, it
will help us identify bugs with configurations if
and when they arise.

The key idea behind checking whether a set of events
forms a configuration is to again leverage the power
of History. If while searching through the history
of a set of events you stumble upon an event that is
not part of the original group, then you know that
said group cannot be a configuration since at least
one of the member's dependencies are not part of the
group.

14 months agoKill simgrid::xbt::cmdline and simgrid::xbt::binary_name
Martin Quinson [Tue, 21 Feb 2023 14:53:00 +0000 (15:53 +0100)]
Kill simgrid::xbt::cmdline and simgrid::xbt::binary_name

14 months agoSimplify the library initialization + deprecate 2 XBT functions
Martin Quinson [Mon, 20 Feb 2023 14:46:02 +0000 (15:46 +0100)]
Simplify the library initialization + deprecate 2 XBT functions

Work toward making EngineImpl::initialize() the only entry point of
the library initialization. It's now impossible to initialize xbt
separately of EngineImpl.

It should help reducing the mess in that code, which results from the
many ways of initializing the library. simgrid-mc, unit-tests, smpi
and others each have their own way of initializing everything,
resulting in a spagetthi and fragile code. It's a bit of a waste to
initialize an EngineImpl even when you don't want to actually run a
simulation, but easier code is always better.

This commit also includes another one because I git amended locally by
error, sorry for the mess. The other commit was about the deprecation
of xbt_procname and xbt_getpid that were xbt functions relying on S4U.

14 months agoRemove a useless lib constructor (simpler is better)
Martin Quinson [Mon, 20 Feb 2023 09:00:22 +0000 (10:00 +0100)]
Remove a useless lib constructor (simpler is better)

14 months agoInitialize the dict module on need, w/o relying on lib constructors
Martin Quinson [Mon, 20 Feb 2023 08:51:18 +0000 (09:51 +0100)]
Initialize the dict module on need, w/o relying on lib constructors

14 months agoAdd first batch of tests for History class
Maxwell Pirtle [Tue, 21 Feb 2023 13:47:40 +0000 (14:47 +0100)]
Add first batch of tests for History class

14 months agoAdd first implementations of History class
Maxwell Pirtle [Tue, 21 Feb 2023 13:08:54 +0000 (14:08 +0100)]
Add first implementations of History class

14 months agoIntroduce extrinsic equivalence between events
Maxwell Pirtle [Tue, 21 Feb 2023 12:19:39 +0000 (13:19 +0100)]
Introduce extrinsic equivalence between events

Extrinsic equivalence between two UnfoldingEvents
simply means that we consider to events to be
"the same". This equivalence is used when UDPOR
creates a new event to determine if in fact that
event has already been created. This requires determining
the equivalence between the transitions associated with
those events as well as the dependencies of each of
the events.

14 months agoAdd tests for the Unfolding object
Maxwell Pirtle [Tue, 21 Feb 2023 09:35:37 +0000 (10:35 +0100)]
Add tests for the Unfolding object

The Unfolding object currently doesn't have much functionality,
so there is not much that needs to be tested with it. It is
effectively a wrapper around a collection of std::unique_ptr.
Eventually, more functionality may be added when more of the
implementation of UDPOR is added, although its use will still
be quite limited even later

14 months agoAdd set operator tests for EventSet
Maxwell Pirtle [Tue, 21 Feb 2023 09:22:42 +0000 (10:22 +0100)]
Add set operator tests for EventSet

This commit introduces test cases for the
common set operators, viz. set union, set
difference, and testing subsets.

14 months agoAdd tests for EventSet equality
Maxwell Pirtle [Mon, 20 Feb 2023 15:31:58 +0000 (16:31 +0100)]
Add tests for EventSet equality

14 months agoAdd History object skeleton
Maxwell Pirtle [Mon, 20 Feb 2023 13:58:51 +0000 (14:58 +0100)]
Add History object skeleton

The History class represents the entire
history of a collection of events "lazily",
i.e. without needing to compute the entire
history if it is not actually needed

14 months agoRemove StateManager
Maxwell Pirtle [Mon, 20 Feb 2023 12:54:05 +0000 (13:54 +0100)]
Remove StateManager

The StateManager class was created to wrap
some of the functionality used in tiny_simgrid,
but it turns out that this functionality isn't
actually needed. States are managed entirely
by the UdporChecker/UDPOR in what is
effectively a stack, and it doesn't make sense
for events to talk about states when states
are associated with configurations

14 months agoAdd first batch of tests for UDPOR
Maxwell Pirtle [Mon, 20 Feb 2023 12:38:02 +0000 (13:38 +0100)]
Add first batch of tests for UDPOR

This commit adds tests for the UDPOR constructs
that will be used when UDPOR is finally implemented.
UDPOR will rely heavily on the correctness of
the methods herein and hereafter (in subsequent
commits) tested, so it is important that each function
be tested to the maximum extent possible

14 months ago"Finalize" interface to UnfoldingEvent
Maxwell Pirtle [Fri, 17 Feb 2023 12:59:59 +0000 (13:59 +0100)]
"Finalize" interface to UnfoldingEvent

The interface to the UnfoldingEvent was carried over
from the previous implementation of UDPOR in the
"tiny_simgrid" repository. However, the implementation
needed to be adjusted to better suit how SimGrid is
going to implement UDPOR. Instead of holding a string
tag identifying a transition, each UnfoldingEvent now
directly holds a reference to the transition from the
state prior.

This caused some changes to need to be made in the
UdporChecker.cpp. These changes have moved the pseudocode
of UDPOR one large step closer to the actual implementation
that will be brought to life over the next few weeks

14 months agoAllow actor transitions to be shared
Maxwell Pirtle [Fri, 17 Feb 2023 10:47:50 +0000 (11:47 +0100)]
Allow actor transitions to be shared

Transitions that were stored in `ActorState`
prior to this commit were stored as `std::unique_ptr`
instances. But UDPOR events need to hold
references to the transitions that they are
"associated with". While UDPOR will discard events
at the appropriate time, we want to make sure the
reference to the transition in some `ActorState`
(ultimately in some `State` instance) remains
valid event after UDPOR discard the State object.
This could e.g. be helpful while debugging

14 months agoAdd Unfolding data structure
Maxwell Pirtle [Thu, 16 Feb 2023 09:32:42 +0000 (10:32 +0100)]
Add Unfolding data structure

The Unfolding data structure is used to
house all of the events UDPOR needs to maintain
as it searches through the unfolding. The
Unfolding may eventually also be used to store
and maintain state, although this technically
only needs to be delegated to the checker itself
as it performs its exploration

14 months agoAdd implementations for EventSet + Configurations
Maxwell Pirtle [Thu, 16 Feb 2023 09:10:01 +0000 (10:10 +0100)]
Add implementations for EventSet + Configurations

14 months agoAdd xbt_assert() for state handle overflow
Maxwell Pirtle [Thu, 16 Feb 2023 08:10:23 +0000 (09:10 +0100)]
Add xbt_assert() for state handle overflow

With the StateManager, the current implementation
vends out integer handles that are referred to by
each UnfoldingEvent to map it to the appropriate
State.

A better API, however, may be to have a each event
manage its own state. We should verify first that
each UnfoldingEvent object corresponds to a unique
State object, but this may be the best option

14 months agoUse std::unordered_set instead of std::set for EventSet
Maxwell Pirtle [Tue, 14 Feb 2023 14:04:40 +0000 (15:04 +0100)]
Use std::unordered_set instead of std::set for EventSet

The ordering of the events in the set does not
matter in an EventSet. The std::unordered_set construct
is thus a better fit (std::set orders its elements).
While this disallows the use of the nice functions
in the standard library for forming unions etc.,
it is trivial to implement the functions `make_union()`
etc.

14 months agoMove UDPOR constructs out of udpor_global.cpp
Maxwell Pirtle [Tue, 14 Feb 2023 13:53:44 +0000 (14:53 +0100)]
Move UDPOR constructs out of udpor_global.cpp

The udpor_global.hpp/cpp have been removed in
favor of seperate files for the primary
constructs that will be used in UDPOR. The source
files have been adjusted as necessary to allow
for a complete compilation

14 months agoAdd `udpor` directory under `mc/explo`
Maxwell Pirtle [Tue, 14 Feb 2023 13:20:10 +0000 (14:20 +0100)]
Add `udpor` directory under `mc/explo`

The `udpor` directory was created to better
house/separate all of the functionality that will
eventually be required by the UDPOR algorithm.
Currently, the files are not used as part of the
compilation process. Subsequent commits will remove
the `udpor_global.cpp` file and adjust the compilation
as necessary

14 months agoRemove currently unused methods from udpor_globals
Maxwell Pirtle [Tue, 14 Feb 2023 12:57:40 +0000 (13:57 +0100)]
Remove currently unused methods from udpor_globals

The udpor_globals.cpp file contained many currently
unused functions that were remove with this commit.
Subsequent commits will move the globals into more
relevant files under `explo`

14 months agoAdd minor clean ups
Maxwell Pirtle [Tue, 14 Feb 2023 12:21:43 +0000 (13:21 +0100)]
Add minor clean ups

14 months agoAdd implementation for state management
Maxwell Pirtle [Tue, 7 Feb 2023 14:04:09 +0000 (15:04 +0100)]
Add implementation for state management

14 months agoAdd incomplete implementations of udpor_globals.cpp
Maxwell Pirtle [Tue, 7 Feb 2023 12:47:05 +0000 (13:47 +0100)]
Add incomplete implementations of udpor_globals.cpp

The classes EventSet, Configuration, and StateManager
have implementations added in order to allow SimGrid
to link properly. Future commits will provide actual
implementations for many of the TODOs that remain

14 months agoAdd `udpor` namespace
Maxwell Pirtle [Mon, 6 Feb 2023 15:13:31 +0000 (16:13 +0100)]
Add `udpor` namespace

14 months agoBegin porting implementation from tiny_simgrid
Maxwell Pirtle [Mon, 6 Feb 2023 14:30:44 +0000 (15:30 +0100)]
Begin porting implementation from tiny_simgrid

More of the structure of the current implementation
of the UDPOR algorithm was imported into SimGrid
with minor alterations to better fit the EventSet
class paradigm, along with some other cleaning up.

The next phase will be two-fold:

1. Importantly, state management will need to be
added into SimGrid. This shouldn't be too difficult,
assuming that we can follow a similar strategy as
The Pham had in tiny_simgrid of mapping State
objects to state id's and using those ids to later
manipulate the state. In SimGrid, the state will be
be used to manipulate the remote process to which
the Exploration has access, instead of manipulating
what's effectively a dummy object

2. Implementing most of the TODO functions
in EventSet etc. This functionality will come
in future commits.

Eventually, we'll want to make an iterative version
of UDPOR to prevent stack overflows, as it is likely
that the unfoldings of programs of any significant size
can get quite deep. For now, though, this is not
strictly necessary to get a first draft implemented

14 months agoAdd methods between EventSet and Configuration
Maxwell Pirtle [Mon, 6 Feb 2023 10:54:40 +0000 (11:54 +0100)]
Add methods between EventSet and Configuration

The distinction between a Configuration and an
EventSet is a subtle one. Indeed, a configuration
in UDPOR is merely a set. There may be the opportunity
perhaps even to replace the Configuration class
with a typedef, although this may not be possible or
improve readability.

14 months agoConvert EventSet into class from typedef
Maxwell Pirtle [Mon, 6 Feb 2023 08:37:54 +0000 (09:37 +0100)]
Convert EventSet into class from typedef

The EventSet object was simply typedef-ed to
a std::deque<UnfoldingEvent *>. While this
worked, it caused a lot of the code in tiny-simgrid
to be a bit more difficult to read since it
required the use of a more C-like idiom of passing
`this` as an explicit argument in EvtSetTools.

Effectively, the idea to absorb all of the EvtSetTools
methods into the EventSet class directly. This will
give a more natural C++ (instead of C) look to the
resulting code in the UDPOR algorithm

14 months agoRevert "Upgrade embedded xxhash (-> v0.8.1)."
Arnaud Giersch [Mon, 20 Feb 2023 09:37:04 +0000 (10:37 +0100)]
Revert "Upgrade embedded xxhash (-> v0.8.1)."

It doesn't compile on arm64

This reverts commit 4127fe802ff334591a6ba5b315190523fc2a660e.

14 months agoMerge branch 'fix-probe' into 'master'
Arnaud Giersch [Mon, 20 Feb 2023 09:25:44 +0000 (09:25 +0000)]
Merge branch 'fix-probe' into 'master'

Send actor action probes one at a time

See merge request simgrid/simgrid!131

14 months agoSend actor action probes one at a time
Maxwell Pirtle [Mon, 20 Feb 2023 08:41:16 +0000 (09:41 +0100)]
Send actor action probes one at a time

To avoid the scenario whereby the AppSide
delivers a message larger than the maximum
datagram side allowed by the application,
we instead send a single message for each actor.
This prevents us from needing to complicate the
sending logic in Channel.cpp, which would have
been made more complicated should we have wanted
to send variable-sized messages between the
checker and the application side

14 months agoRemove more occurences of 'surf' with uppercases
Martin Quinson [Sun, 19 Feb 2023 21:34:31 +0000 (22:34 +0100)]
Remove more occurences of 'surf' with uppercases

14 months agoDocument that surf is gone
Martin Quinson [Sun, 19 Feb 2023 15:09:01 +0000 (16:09 +0100)]
Document that surf is gone

14 months agoUpgrade embedded xxhash (-> v0.8.1).
Arnaud Giersch [Sun, 19 Feb 2023 14:48:56 +0000 (15:48 +0100)]
Upgrade embedded xxhash (-> v0.8.1).

Taken from
https://github.com/RedSpah/xxhash_cpp/commit/d21e37c34c4ad4ad562e4ec01ec036e342c422e3

14 months agoUpgrade embedded catch2 (-> v2.13.10).
Arnaud Giersch [Sun, 19 Feb 2023 14:37:48 +0000 (15:37 +0100)]
Upgrade embedded catch2 (-> v2.13.10).

Taken from
https://github.com/catchorg/Catch2/releases/tag/v2.13.10

14 months agoFinally kill surf_interface.hpp watched_host was actually useless
Martin Quinson [Sun, 19 Feb 2023 14:45:51 +0000 (15:45 +0100)]
Finally kill surf_interface.hpp watched_host was actually useless

14 months agorename another file with had 'surf' in its name
Martin Quinson [Sun, 19 Feb 2023 11:53:23 +0000 (12:53 +0100)]
rename another file with had 'surf' in its name

14 months agoKill a file with 'surf' in its name, and further empty surf_interface.cpp
Martin Quinson [Sun, 19 Feb 2023 11:49:24 +0000 (12:49 +0100)]
Kill a file with 'surf' in its name, and further empty surf_interface.cpp

14 months agoRework a comment (to remove 'surf')
Martin Quinson [Sun, 19 Feb 2023 11:21:55 +0000 (12:21 +0100)]
Rework a comment (to remove 'surf')

14 months agoRework compile optimizations for MC.
Arnaud Giersch [Sun, 19 Feb 2023 08:37:22 +0000 (09:37 +0100)]
Rework compile optimizations for MC.

14 months agoMissing includes.
Arnaud Giersch [Sun, 19 Feb 2023 06:27:24 +0000 (07:27 +0100)]
Missing includes.

14 months agoMove -Wno-unused-local-typedefs to common flags.
Arnaud Giersch [Sun, 19 Feb 2023 06:20:34 +0000 (07:20 +0100)]
Move -Wno-unused-local-typedefs to common flags.

It's supported by Clang too, and needed with at least ns-3.31.

14 months agokill src/simix
Martin Quinson [Sat, 18 Feb 2023 21:48:39 +0000 (22:48 +0100)]
kill src/simix

14 months agoMove the non-deprecated bits of simix to kernel::actor::Simcall
Martin Quinson [Sat, 18 Feb 2023 21:44:30 +0000 (22:44 +0100)]
Move the non-deprecated bits of simix to kernel::actor::Simcall

14 months agoKill src/include
Martin Quinson [Sat, 18 Feb 2023 20:49:57 +0000 (21:49 +0100)]
Kill src/include

The -Isrc/include makes it difficult to recognize in the source which
\#include takes a public header and which one is a private header.

Plus, move the third-party headers to a specific directory.

14 months agoreduce the amount of header includes in headers, to speedup recompilation
Martin Quinson [Sat, 18 Feb 2023 20:21:56 +0000 (21:21 +0100)]
reduce the amount of header includes in headers, to speedup recompilation

14 months agoMerge branch 'master' of framagit.org:simgrid/simgrid
Martin Quinson [Sat, 18 Feb 2023 19:10:28 +0000 (20:10 +0100)]
Merge branch 'master' of framagit.org:simgrid/simgrid

14 months agoRemove an obsolete cmake warning
Martin Quinson [Sat, 18 Feb 2023 18:37:22 +0000 (19:37 +0100)]
Remove an obsolete cmake warning

14 months agoMove some content of surf_interface to a new math_utils.h
Martin Quinson [Sat, 18 Feb 2023 18:28:37 +0000 (19:28 +0100)]
Move some content of surf_interface to a new math_utils.h

14 months agoUpdate path.
Arnaud Giersch [Sat, 18 Feb 2023 18:04:03 +0000 (19:04 +0100)]
Update path.

Error also seen with clang 11 and ns3.31.

14 months agoMove the files related to the platform parsing to kernel/xml
Martin Quinson [Sat, 18 Feb 2023 17:45:08 +0000 (18:45 +0100)]
Move the files related to the platform parsing to kernel/xml

14 months agomove a symbol declaration out of surf_interface.hpp
Martin Quinson [Sat, 18 Feb 2023 17:32:11 +0000 (18:32 +0100)]
move a symbol declaration out of surf_interface.hpp

14 months agoCMake requires ns3 v3.28+, so don't work around a bug in v3.27
Martin Quinson [Sat, 18 Feb 2023 17:25:55 +0000 (18:25 +0100)]
CMake requires ns3 v3.28+, so don't work around a bug in v3.27

14 months agoMove the ns3 helper files to kernel/models/ns3
Martin Quinson [Sat, 18 Feb 2023 17:23:50 +0000 (18:23 +0100)]
Move the ns3 helper files to kernel/models/ns3

14 months agoMove the surf tesh tests with the other models' ones
Martin Quinson [Sat, 18 Feb 2023 16:59:53 +0000 (17:59 +0100)]
Move the surf tesh tests with the other models' ones

14 months agoRemove some more occurence of 'surf' in comments and such
Martin Quinson [Sat, 18 Feb 2023 16:31:27 +0000 (17:31 +0100)]
Remove some more occurence of 'surf' in comments and such

14 months agoMisc Sonar issues.
Arnaud Giersch [Sat, 18 Feb 2023 16:42:44 +0000 (17:42 +0100)]
Misc Sonar issues.

14 months agoVarious cleanups.
Arnaud Giersch [Sat, 18 Feb 2023 16:39:33 +0000 (17:39 +0100)]
Various cleanups.

14 months agoFurther rename the identifiers of flexml to simgrid_parse_*
Martin Quinson [Sat, 18 Feb 2023 16:21:27 +0000 (17:21 +0100)]
Further rename the identifiers of flexml to simgrid_parse_*

14 months agoChange the namespace of flexml-generated parser from surfxml to simgrid_parse
Martin Quinson [Sat, 18 Feb 2023 10:48:52 +0000 (11:48 +0100)]
Change the namespace of flexml-generated parser from surfxml to simgrid_parse

14 months agoTwo more references of the surf/precision option name
Martin Quinson [Sat, 18 Feb 2023 10:32:21 +0000 (11:32 +0100)]
Two more references of the surf/precision option name

14 months agoRemove surf from the content from the examples of simulated filesystems
Martin Quinson [Sat, 18 Feb 2023 10:07:33 +0000 (11:07 +0100)]
Remove surf from the content from the examples of simulated filesystems

14 months agoRename ActivityImpl::surf_action_ to ActivityImpl::model_action_
Martin Quinson [Sat, 18 Feb 2023 09:38:38 +0000 (10:38 +0100)]
Rename ActivityImpl::surf_action_ to ActivityImpl::model_action_

14 months agoMove HostImpl to kernel/resource with its friends
Martin Quinson [Fri, 17 Feb 2023 22:02:54 +0000 (23:02 +0100)]
Move HostImpl to kernel/resource with its friends

14 months agomove the models to the right directory (empty src/surf a bit)
Martin Quinson [Fri, 17 Feb 2023 21:31:02 +0000 (22:31 +0100)]
move the models to the right directory (empty src/surf a bit)

The remaining parts of src/surf need to be cleaned before being moved

Also, remove almost all #include "surf_interface.hpp"

14 months agoAllow precision/bmf as a synonym to bmf/precision
Martin Quinson [Thu, 16 Feb 2023 23:05:18 +0000 (00:05 +0100)]
Allow precision/bmf as a synonym to bmf/precision

14 months agoRename option maxmin/precision to precision/work-amount for clarity
Martin Quinson [Thu, 16 Feb 2023 23:01:45 +0000 (00:01 +0100)]
Rename option maxmin/precision to precision/work-amount for clarity

14 months agoRename option "surf/precision" to "precision/timing" for clarity.
Martin Quinson [Thu, 16 Feb 2023 22:44:37 +0000 (23:44 +0100)]
Rename option "surf/precision" to "precision/timing" for clarity.

14 months agoRename sg_surf_precision to sg_precision_timing for clarity
Martin Quinson [Thu, 16 Feb 2023 22:39:54 +0000 (23:39 +0100)]
Rename sg_surf_precision to sg_precision_timing for clarity

14 months agoUse surf_precision as a value for PMPI_Wtick
Martin Quinson [Thu, 16 Feb 2023 22:31:53 +0000 (23:31 +0100)]
Use surf_precision as a value for PMPI_Wtick

the unit of maxmin_precision is in flops or bytes (that's an amount of
work) while the unit of surf_precision is in seconds (that's a time).

14 months agodoc: add a little graph about the host energy plugin
Martin Quinson [Thu, 16 Feb 2023 16:48:10 +0000 (17:48 +0100)]
doc: add a little graph about the host energy plugin

14 months agoMerge branch 'fix-mc-tests' into 'master'
Martin Quinson [Thu, 16 Feb 2023 21:13:41 +0000 (21:13 +0000)]
Merge branch 'fix-mc-tests' into 'master'

Fix root node edge case for `State` instance

See merge request simgrid/simgrid!129

14 months agoInitialize uninitialized transition for root state
Maxwell Pirtle [Thu, 16 Feb 2023 14:51:37 +0000 (15:51 +0100)]
Initialize uninitialized transition for root state

The transition in the root state was uninitialized
and this appears to have caused the last of the remaining
issues with the test suite. Of course, reading from the
uninitialized pointer `transition_` is bound to cause
a bunch of problems...

14 months agoFix root-node edge cases for states
Maxwell Pirtle [Thu, 16 Feb 2023 12:11:07 +0000 (13:11 +0100)]
Fix root-node edge cases for states

14 months agofix MC+gcc builds
Martin Quinson [Thu, 16 Feb 2023 10:05:59 +0000 (11:05 +0100)]
fix MC+gcc builds

14 months agoCI: build the model-checker on framagit, to test the MRs
Martin Quinson [Thu, 16 Feb 2023 09:52:21 +0000 (10:52 +0100)]
CI: build the model-checker on framagit, to test the MRs

14 months agofix the build (but not all tests) of MC
Martin Quinson [Thu, 16 Feb 2023 08:45:47 +0000 (09:45 +0100)]
fix the build (but not all tests) of MC

14 months agoMerge branch 'actor-comms' into 'master'
Martin Quinson [Wed, 15 Feb 2023 23:59:31 +0000 (23:59 +0000)]
Merge branch 'actor-comms' into 'master'

Serialize pending transitions when responding to `ACTOR_STATUS` messages on the application side

See merge request simgrid/simgrid!127

14 months agoproperly deal with network/optim (end of the ModuleGroup cleanup)
Martin Quinson [Wed, 15 Feb 2023 23:49:30 +0000 (00:49 +0100)]
properly deal with network/optim (end of the ModuleGroup cleanup)

The network/optim was abusing the model description mechanism, because
it was introduced before the options could be restricted to a list of
values. That's a pretty old cruft :)

Moreover, the config mechanism (probably) allowed network/optim:TI but then
ignored it silently. Not nice for the users.