Martin Quinson [Tue, 1 Nov 2016 19:48:26 +0000 (20:48 +0100)]
use std:: instead of fixed size vectors. C is over.
Martin Quinson [Tue, 1 Nov 2016 19:38:55 +0000 (20:38 +0100)]
another try to fix mac OSX (libunwind is not the one we think there)
Martin Quinson [Tue, 1 Nov 2016 18:27:12 +0000 (19:27 +0100)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Tue, 1 Nov 2016 18:13:36 +0000 (19:13 +0100)]
routing: cosmetics and doc improvement
Martin Quinson [Tue, 1 Nov 2016 14:49:40 +0000 (15:49 +0100)]
improve the doc of the routing API
Martin Quinson [Tue, 1 Nov 2016 14:28:08 +0000 (15:28 +0100)]
cosmetics in bypass routing
Martin Quinson [Tue, 1 Nov 2016 14:05:59 +0000 (15:05 +0100)]
cosmetics
Matthieu Volat [Tue, 1 Nov 2016 07:43:22 +0000 (08:43 +0100)]
Use BYTE_ORDER macro, not private __BYTE_ORDER
Matthieu Volat [Tue, 1 Nov 2016 07:24:14 +0000 (08:24 +0100)]
MADV_MERGEABLE is available on linux, guard it.
Matthieu Volat [Mon, 31 Oct 2016 18:03:55 +0000 (19:03 +0100)]
Use MAP_PREFAULT_READ on FreeBSD as en equivalent to MAP_POPULATE.
Matthieu Volat [Mon, 31 Oct 2016 16:49:45 +0000 (17:49 +0100)]
Add -Wno-extern-c-compat when building with clang
Else it triggers an warning due to libunwind.h using empty C structures that
are turned into an error by -Werror.
degomme [Tue, 1 Nov 2016 13:32:22 +0000 (14:32 +0100)]
add suppressions from libunwind.
https://github.com/JuliaLang/julia/issues/4533 reports the same errors we have on ci as libunwind's.
I have some more locally which are more annoying, though.
degomme [Tue, 1 Nov 2016 08:28:35 +0000 (09:28 +0100)]
apparently, this should be in this order
Martin Quinson [Tue, 1 Nov 2016 00:01:39 +0000 (01:01 +0100)]
fix graphviz tests after the host order change in sg_host_list
Martin Quinson [Mon, 31 Oct 2016 23:53:11 +0000 (00:53 +0100)]
cmake: better error message
Martin Quinson [Mon, 31 Oct 2016 23:50:01 +0000 (00:50 +0100)]
cosmetics
Martin Quinson [Mon, 31 Oct 2016 23:45:22 +0000 (00:45 +0100)]
another try to fix Mac: don't guess the unwind component names so widely
Martin Quinson [Mon, 31 Oct 2016 23:30:25 +0000 (00:30 +0100)]
try to fix the windows build (+clang-format)
Martin Quinson [Mon, 31 Oct 2016 23:20:03 +0000 (00:20 +0100)]
fix mac build (another try)
Martin Quinson [Mon, 31 Oct 2016 23:11:26 +0000 (00:11 +0100)]
woops, this code was not unused
I must have mis-grepped :(
Martin Quinson [Mon, 31 Oct 2016 23:02:01 +0000 (00:02 +0100)]
cosmetics
Martin Quinson [Mon, 31 Oct 2016 22:47:15 +0000 (23:47 +0100)]
kill an unused chunk of CMakeLists.txt
Martin Quinson [Mon, 31 Oct 2016 22:43:29 +0000 (23:43 +0100)]
fix Mac OSX build
Martin Quinson [Mon, 31 Oct 2016 22:31:32 +0000 (23:31 +0100)]
this is definitely a backtrace function
Martin Quinson [Mon, 31 Oct 2016 22:18:44 +0000 (23:18 +0100)]
kill empty functions
Martin Quinson [Mon, 31 Oct 2016 22:08:42 +0000 (23:08 +0100)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Mon, 31 Oct 2016 22:05:51 +0000 (23:05 +0100)]
test for libunwind even if not building MC
Martin Quinson [Mon, 31 Oct 2016 21:21:55 +0000 (22:21 +0100)]
plug a memleak
degomme [Mon, 31 Oct 2016 13:39:09 +0000 (14:39 +0100)]
Add a FreeBSD backend for get_memory_map(), allow to enable smpi privatization.
Thanks to M. Volat !
degomme [Mon, 31 Oct 2016 13:35:33 +0000 (14:35 +0100)]
Rework how model checking and mmap get enabled/disabled
Simplify test: model checking relies on linux-only syscalls such as signalfd,
mremap, so it won't work on anything else.
But do not set HAVE_MMAP to 0 if MC cannot be built, other parts could use it
Courtesy of M. Volat
degomme [Mon, 31 Oct 2016 13:34:00 +0000 (14:34 +0100)]
Test for mremap function and store result in internal_config.h.
From M. Volat
degomme [Mon, 31 Oct 2016 13:33:02 +0000 (14:33 +0100)]
Setup and use a LD_LIBRARY variable in build.
libdl is linux-only, so set this variable as the result of a find_library
command here and let it empty on other systems. From M. Volat.
degomme [Mon, 31 Oct 2016 09:18:20 +0000 (10:18 +0100)]
From M. Volat : more portable way of opening shm buffers for privatization
degomme [Mon, 31 Oct 2016 08:31:20 +0000 (09:31 +0100)]
handle MPI_IN_PLACE for this collective
Martin Quinson [Sun, 30 Oct 2016 15:55:47 +0000 (16:55 +0100)]
a platform to test the AS bypass feature
Martin Quinson [Sat, 29 Oct 2016 15:41:18 +0000 (17:41 +0200)]
Spell checking ;)
Martin Quinson [Fri, 28 Oct 2016 18:24:23 +0000 (20:24 +0200)]
fix MC builds
It seems that MC does not love when exceptions are raised.
Martin Quinson [Fri, 28 Oct 2016 17:07:13 +0000 (19:07 +0200)]
Merge branch 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Fri, 28 Oct 2016 17:05:46 +0000 (19:05 +0200)]
fix the last test after the recent change in MSG_hosts_as_dynar order
Martin Quinson [Fri, 28 Oct 2016 17:02:54 +0000 (19:02 +0200)]
still 2 tests to go before it works again
Martin Quinson [Fri, 28 Oct 2016 16:58:23 +0000 (18:58 +0200)]
fix another example
Martin Quinson [Fri, 28 Oct 2016 16:53:47 +0000 (18:53 +0200)]
fix another example after the recent change in msg_hosts_as_dynar
Martin Quinson [Fri, 28 Oct 2016 16:47:27 +0000 (18:47 +0200)]
fix the tesh of many tests after the order change in hosts_as_dynar
Martin Quinson [Fri, 28 Oct 2016 16:46:44 +0000 (18:46 +0200)]
fix an example after the recent change to hosts_as_dynar modifying its order
Martin Quinson [Thu, 27 Oct 2016 08:09:28 +0000 (10:09 +0200)]
Destroy all hosts by the end of the simulation.
This plugs huge memleaks, and make sure that the dtor of the energy
plugin actually makes the expected output.
Martin Quinson [Wed, 26 Oct 2016 17:02:12 +0000 (19:02 +0200)]
tesh: add an option to keep on disk the obtained output
This helps regenerating the tesh files.
Martin Quinson [Tue, 25 Oct 2016 18:40:46 +0000 (20:40 +0200)]
tesh: show the full path to tesh file
Martin Quinson [Tue, 25 Oct 2016 17:26:43 +0000 (19:26 +0200)]
host_list is now a std::map instead of xbt_dict
Martin Quinson [Mon, 24 Oct 2016 20:40:01 +0000 (22:40 +0200)]
move some content out of parsing to the classes
Frederic Suter [Mon, 24 Oct 2016 08:54:13 +0000 (10:54 +0200)]
vacation commit
chase some sonar minor smells
Martin Quinson [Sun, 23 Oct 2016 19:49:57 +0000 (21:49 +0200)]
small cosmetics in the storage
Martin Quinson [Sat, 22 Oct 2016 00:25:15 +0000 (02:25 +0200)]
inline a function called only once
Martin Quinson [Sat, 22 Oct 2016 00:19:42 +0000 (02:19 +0200)]
cosmetics: use C++ type names
Martin Quinson [Sat, 22 Oct 2016 00:14:15 +0000 (02:14 +0200)]
rename MSG_host_(get/set)_params into MSG_vm_(get/set)_params
Martin Quinson [Fri, 21 Oct 2016 21:08:19 +0000 (23:08 +0200)]
Do not throw an exception on assert, but display backtrace+abort
Martin Quinson [Fri, 21 Oct 2016 20:09:51 +0000 (22:09 +0200)]
Merge branches 'master' and 'master' of scm.gforge.inria.fr:/gitroot/simgrid/simgrid
Martin Quinson [Fri, 21 Oct 2016 16:55:50 +0000 (18:55 +0200)]
Create a s4u::VirtualMachine to clean things out.
This is still to be populated, with the content of the msg_vm module.
Martin Quinson [Fri, 21 Oct 2016 16:44:10 +0000 (18:44 +0200)]
rename surf::VirtualMachine to surf::VirtualMachineImpl
This makes room for a s4u::VirtualMachine class that should arrive soon.
Martin Quinson [Thu, 20 Oct 2016 22:11:25 +0000 (00:11 +0200)]
tiny simplification
Martin Quinson [Thu, 20 Oct 2016 15:59:15 +0000 (17:59 +0200)]
continue removing content from HostImpl
degomme [Thu, 20 Oct 2016 13:43:13 +0000 (15:43 +0200)]
Really fix the issue.
degomme [Thu, 20 Oct 2016 13:09:23 +0000 (15:09 +0200)]
Fix bug in SMPI tracing, where messages were sometimes badly matched.
This eparates sent and received messages when trying to match a new one to avoid matching a send with another send.
degomme [Thu, 20 Oct 2016 12:33:05 +0000 (14:33 +0200)]
smpi replayed waitall was not working, as request vector is destroyed inside the waitall called.
Store information before the call to trace them properly at the end of the wait
degomme [Thu, 20 Oct 2016 12:32:02 +0000 (14:32 +0200)]
add some debug messages for smpi tracing
degomme [Wed, 19 Oct 2016 13:24:29 +0000 (15:24 +0200)]
remove now useless file
Martin Quinson [Wed, 19 Oct 2016 20:39:53 +0000 (22:39 +0200)]
cosmetics
Martin Quinson [Tue, 18 Oct 2016 17:36:49 +0000 (19:36 +0200)]
remove a redundent forward declaration
degomme [Wed, 19 Oct 2016 10:14:16 +0000 (12:14 +0200)]
fix case when no fortran compiler is found
degomme [Wed, 19 Oct 2016 09:50:48 +0000 (11:50 +0200)]
cleanup handling of Fortran compiler in cmake (Patch from M. Volat)
degomme [Wed, 19 Oct 2016 08:58:37 +0000 (10:58 +0200)]
add a few fortran (unimplemented) declarations
Yes I know at some point I should add all of them.
degomme [Tue, 18 Oct 2016 15:19:33 +0000 (17:19 +0200)]
add some defines in fortran
degomme [Tue, 18 Oct 2016 15:19:21 +0000 (17:19 +0200)]
Add a check so that process_init exits cleanly when improperly called.
Martin Quinson [Tue, 18 Oct 2016 06:17:43 +0000 (08:17 +0200)]
kill a useless constructor
Martin Quinson [Mon, 17 Oct 2016 20:10:41 +0000 (22:10 +0200)]
HOSTIMPL IS NO LONGER A RESOURCE! At least! \o/
Martin Quinson [Mon, 17 Oct 2016 19:57:00 +0000 (21:57 +0200)]
kill dead code
Martin Quinson [Mon, 17 Oct 2016 19:53:36 +0000 (21:53 +0200)]
Kill HostImpl::cpu_. Use the one of the s4u::Host
Martin Quinson [Mon, 17 Oct 2016 18:55:17 +0000 (20:55 +0200)]
reindent after previous changes
Martin Quinson [Mon, 17 Oct 2016 18:54:54 +0000 (20:54 +0200)]
put the HostImpl as s4u::Host->pimpl instead of as an extension
Martin Quinson [Mon, 17 Oct 2016 15:36:38 +0000 (17:36 +0200)]
hello clang, that's another ForEach macro of ours
Martin Quinson [Mon, 17 Oct 2016 07:33:54 +0000 (09:33 +0200)]
kill a bunch of empty ctor/dtor
Martin Quinson [Mon, 17 Oct 2016 07:10:44 +0000 (09:10 +0200)]
simplification: inline a virtual function that is never derivated
Martin Quinson [Sun, 16 Oct 2016 15:17:56 +0000 (17:17 +0200)]
further ignorable cleanups
Martin Quinson [Sun, 16 Oct 2016 14:54:47 +0000 (16:54 +0200)]
code simplifications around host construction
Storage list is passed as a parameter, not through a global dict of
storages.
Martin Quinson [Sun, 16 Oct 2016 14:14:06 +0000 (16:14 +0200)]
code simplifications in VMs
Martin Quinson [Sun, 16 Oct 2016 13:37:21 +0000 (15:37 +0200)]
use static_cast to please sonar + reindent
Martin Quinson [Sun, 16 Oct 2016 09:57:34 +0000 (11:57 +0200)]
do standard C++
Destroy the elements explicitly instead of having them destroyed when
they are removed from the dict container.
Martin Quinson [Sun, 16 Oct 2016 00:01:22 +0000 (02:01 +0200)]
kill a stupid function. Let's do regular C++
Martin Quinson [Sat, 15 Oct 2016 20:25:09 +0000 (22:25 +0200)]
cleanup in the Link field names
Martin Quinson [Sat, 15 Oct 2016 19:59:39 +0000 (21:59 +0200)]
cleanups in NS3: it's on my way
I need to clean the platform creation stuff. This is blocking for more
interesting changes such as the fact that each cluster has its own
network model.
Martin Quinson [Sat, 15 Oct 2016 19:52:59 +0000 (21:52 +0200)]
NS3.25 is out now
Martin Quinson [Sat, 15 Oct 2016 19:03:05 +0000 (21:03 +0200)]
cleanups in the model initialization
Martin Quinson [Sat, 15 Oct 2016 18:50:07 +0000 (20:50 +0200)]
factorize some code between NS3 cluster creation and other clusters
Martin Quinson [Sat, 15 Oct 2016 15:41:17 +0000 (17:41 +0200)]
cosmetics in the creation of NS3 clusters
Martin Quinson [Sat, 15 Oct 2016 15:31:43 +0000 (17:31 +0200)]
one step further to sort out the spagetti plate of AS creation
Martin Quinson [Sat, 15 Oct 2016 15:21:04 +0000 (17:21 +0200)]
reindent, no real change
Martin Quinson [Sat, 15 Oct 2016 15:18:35 +0000 (17:18 +0200)]
move content from the sg_platf to the object hierarchy
Martin Quinson [Sat, 15 Oct 2016 14:47:44 +0000 (16:47 +0200)]
remove one malloc
Martin Quinson [Sat, 15 Oct 2016 11:19:55 +0000 (13:19 +0200)]
fix NS3 build
Martin Quinson [Sat, 15 Oct 2016 08:52:20 +0000 (10:52 +0200)]
make some field private, to please sonar
Sonar rational is that protected fields are the sign of a logic that
spreads over all sub-classes. I tend to strongly agree with this point.
Martin Quinson [Fri, 14 Oct 2016 22:21:32 +0000 (00:21 +0200)]
cosmetics in clang-format