Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
simgrid.git
9 years agoEnsure that MSG_host_self() works from maestro context
Martin Quinson [Mon, 22 Dec 2014 22:03:23 +0000 (23:03 +0100)]
Ensure that MSG_host_self() works from maestro context

9 years ago[mc] Add some FIXMEs for cross-process support
Gabriel Corona [Fri, 19 Dec 2014 10:56:52 +0000 (11:56 +0100)]
[mc] Add some FIXMEs for cross-process support

9 years ago[mc] Cross-process support for MC_ignore
Gabriel Corona [Fri, 19 Dec 2014 08:51:16 +0000 (09:51 +0100)]
[mc] Cross-process support for MC_ignore

9 years ago[mc] Implements privatization support for MC_process_read
Gabriel Corona [Thu, 18 Dec 2014 15:03:03 +0000 (16:03 +0100)]
[mc] Implements privatization support for MC_process_read

This is currently needed for cross-process MC in order to read the
heap state.

9 years ago[mc] Abstract the process and a snapshot types with a address_space superclass
Gabriel Corona [Tue, 16 Dec 2014 12:06:37 +0000 (13:06 +0100)]
[mc] Abstract the process and a snapshot types with a address_space superclass

Add a `address_space`, superclass of `process` and  `snapshot`.

In order to do this, the contract of MC_process_read and
MC_snapshot_read has been uniformized:

  * the order of arguments has been harmonized;

  * a new flag MC_ADDRESS_SPACE_READ_FLAGS_LAZY is used to avoid copy
    when the data is in the current memory;

  * MC_NO_PROCESS_INDEX has been renamed into MC_PROCESS_INDEX_MISSING;

  * MC_ANY_PROCESS_INDEX has been renamed into MC_PROCESS_INDEX_ANY;

  * MC_PROCESS_INDEX_DISABLED is used to access the raw address space
    (without privatisation support);

  * `const void*` is used instead of `void*` when it possible.

Soem cleanup things are still to be done:

 * remove special NULL handling;

 * add support for SMPI privatization in the process object.

9 years ago[mc] More comments for mc_dwarf_execute_expression()
Gabriel Corona [Tue, 16 Dec 2014 11:05:14 +0000 (12:05 +0100)]
[mc] More comments for mc_dwarf_execute_expression()

Give some basic explanation about the DWARF operations.

9 years ago[mc] Add more information about mc_dwarf_register_to_libunwind()
Gabriel Corona [Tue, 16 Dec 2014 10:34:38 +0000 (11:34 +0100)]
[mc] Add more information about mc_dwarf_register_to_libunwind()

9 years agoavoid word being recognized as special by doxygen
Augustin Degomme [Mon, 15 Dec 2014 21:00:27 +0000 (22:00 +0100)]
avoid word being recognized as special by doxygen

9 years agotypos
Augustin Degomme [Mon, 15 Dec 2014 20:52:40 +0000 (21:52 +0100)]
typos

9 years agoupdate doc about privatization .. \nThis was not updated yet. The option was document...
Augustin Degomme [Mon, 15 Dec 2014 20:48:56 +0000 (21:48 +0100)]
update doc about privatization .. \nThis was not updated yet. The option was documented in options.doc, so this my be a duplication of information

9 years agofix tesh
Augustin Degomme [Fri, 12 Dec 2014 16:59:19 +0000 (17:59 +0100)]
fix tesh

9 years agorequalify teshes for vm migration
Augustin Degomme [Fri, 12 Dec 2014 16:57:10 +0000 (17:57 +0100)]
requalify teshes for vm migration

9 years agoupdate changelog
Augustin Degomme [Fri, 12 Dec 2014 16:29:08 +0000 (17:29 +0100)]
update changelog

9 years agodoc update for new appenders
Augustin Degomme [Fri, 12 Dec 2014 16:27:42 +0000 (17:27 +0100)]
doc update for new appenders

9 years agoPatch by F.Chaix : add two "new" log appender methods : split and roll
Augustin Degomme [Fri, 12 Dec 2014 15:46:01 +0000 (16:46 +0100)]
Patch by F.Chaix : add two "new" log appender methods : split and roll
split will create new files when a specified size is reached
roll will overwrite the file when this size is reached
example syntax is : --log=root.appender:splitfile:10000:myfilename_%.txt

The % is a wildcard that will be replaced by the number of the file. If no % is present, it will be at the end

9 years ago[mc] Support for reading heap state from another process
Gabriel Corona [Thu, 11 Dec 2014 12:14:32 +0000 (13:14 +0100)]
[mc] Support for reading heap state from another process

9 years agomore documentation about the simcall mechanism
Martin Quinson [Fri, 12 Dec 2014 12:57:29 +0000 (13:57 +0100)]
more documentation about the simcall mechanism

9 years ago[mc] Fix error handling in MC_process{read,write}
Gabriel Corona [Thu, 11 Dec 2014 13:51:39 +0000 (14:51 +0100)]
[mc] Fix error handling in MC_process{read,write}

9 years ago[mc] Access memory from another process
Gabriel Corona [Tue, 9 Dec 2014 14:29:16 +0000 (15:29 +0100)]
[mc] Access memory from another process

The goal is to be able to move MC in a separate process which should
be more robust and easier to develop:

 * avoid using two heaps (which is combersome);

 * avoid weird interactions bewteen the MC and the application;

 * use optimisation for the whole MC process;

 * avoid the stack-cleaner for the whole MC process.

The functions MC_process_read and MC_process_write are defined to
abstract memory access:

 * when the target process is the current processs, they call
   `memcpy`;

 * otherwise they call `read` or `write` on `/proc/$pid/mem` (on newer
   kernels, `process_vm_readv` and `process_vm_writev`) could be used.

A lot of bits are missing such as:

 * access to `std_heap` is currently not process-aware (the current
   process is used);

 * access to SIMIX layer from MC;

 * communcation/synchronisation between the processes;

 * …

Limitations:

 * for the per-page/chunked snapshot the current implementation uses
   an extra copy (and one syscall per page), we can do better than
   this.

9 years ago(try to) avoid looping forever, temporarily.
Augustin Degomme [Wed, 10 Dec 2014 13:25:01 +0000 (14:25 +0100)]
(try to) avoid looping forever, temporarily.

This value was 10 before, and was set to 3 recently, but weirdly this caused vm migration simulations to loop forever.
It works with an infinite timeout, but the message is lost with a few values (1,2 ,3, 6, 9 were tested and failed, other ones up to 13 at least do succeed), while succeeding with other ones (after a few hundred timeouts, so this seems to be working)
So if an MSG expert sees this and can find the problem here ..

Test should still fail with this patch, but just because teshes were not requalified yet

9 years agoFix windows build (this is now used from the java library)
Augustin Degomme [Wed, 10 Dec 2014 09:30:42 +0000 (10:30 +0100)]
Fix windows build (this is now used from the java library)

9 years agofix dot output with file descriptor checkpoint/restore
Marion Guthmuller [Wed, 10 Dec 2014 12:39:16 +0000 (13:39 +0100)]
fix dot output with file descriptor checkpoint/restore

9 years agoPut a 300 seconds timeout on each test with ctest
Augustin Degomme [Tue, 9 Dec 2014 15:37:00 +0000 (16:37 +0100)]
Put a 300 seconds timeout on each test with ctest
This should really be enough (the default was 1500...)

9 years agoFix build
Augustin Degomme [Tue, 9 Dec 2014 09:46:18 +0000 (10:46 +0100)]
Fix build
Some vm tests are still broken and some are looping, they need to be fixed

9 years ago[mc] Enable the custom mm malloc only in MC
Gabriel Corona [Tue, 9 Dec 2014 12:15:38 +0000 (13:15 +0100)]
[mc] Enable the custom mm malloc only in MC

We can do better in the future: we can avoid using the main mm malloc
in many cases even for MC.

9 years ago[mm] Allow to disable the mm based `malloc` at runtime
Gabriel Corona [Tue, 9 Dec 2014 11:17:18 +0000 (12:17 +0100)]
[mm] Allow to disable the mm based `malloc` at runtime

The goal is to enable the mm based `malloc` only when needed and fall
back to the (more efficient) builtin/next implementation when it is
not needed:

  * run instrospection-less jobs without it;

  * whene the MC and the application are in different processes, the
    MC will be able to run with the standard `malloc` and the
    application will use mm.

As malloc is needed very early in the application initialisation, an
environment variable is used to change the behaviour.

9 years ago[mc] Remove redundant typedefs
Gabriel Corona [Tue, 9 Dec 2014 10:21:49 +0000 (11:21 +0100)]
[mc] Remove redundant typedefs

Clang complains about typedef redefinitions in pre-C11 mode.

9 years ago[mc] Optimise most of XBT
Gabriel Corona [Tue, 9 Dec 2014 08:16:22 +0000 (09:16 +0100)]
[mc] Optimise most of XBT

9 years agobashism -- (fix https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=772329 )
degomme [Mon, 8 Dec 2014 22:37:07 +0000 (23:37 +0100)]
bashism -- (fix https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=772329 )

9 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Mon, 8 Dec 2014 17:14:47 +0000 (18:14 +0100)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

9 years ago[mc] Optimise all the MC compilation units
Gabriel Corona [Mon, 8 Dec 2014 13:48:47 +0000 (14:48 +0100)]
[mc] Optimise all the MC compilation units

9 years ago[mc] Multiple .so support for region snapshots
Gabriel Corona [Fri, 5 Dec 2014 15:05:50 +0000 (16:05 +0100)]
[mc] Multiple .so support for region snapshots

The region snapshoting logic can handle a variable number of .so
files:

 * add more informations to the snapshot regions,

    * the type (heap, library/executable);

    * the correspoding library/executable;

    * the type of storage (dense/flat, chunked/sparse or privatised)
      and the type-specific variables are defined in an enum
      (variant/tagged enum).

 * SMPI privatisation snapshot regions are stored as children of a
   parent snapshot region

Limitation:

 * we might want to use a more modular/extensible approach OO for the
   snapshot region storage type instead of variant-based approach;

 * SMPI can currently only handle privatisation for the local
   variables of the executable so this is only supported in the MC as
   well for this reason but otherwise the MC is ready to support the
   SMPI privatisation of libraries.

9 years agoFixed return values for several MPI_WIN functions
Christian Heinrich [Mon, 8 Dec 2014 12:38:34 +0000 (13:38 +0100)]
Fixed return values for several MPI_WIN functions

(statements had to be wrapped in else-clauses)

9 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Mon, 8 Dec 2014 09:20:23 +0000 (10:20 +0100)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

9 years agoFew temporary debug messages - adrien
Adrien Lebre [Mon, 8 Dec 2014 09:20:12 +0000 (10:20 +0100)]
Few temporary debug messages - adrien

9 years agoAdd a mutex to lock access to the SMPI mailboxes when a message is posted
Augustin Degomme [Sat, 6 Dec 2014 15:05:38 +0000 (16:05 +0100)]
Add a mutex to lock access to the SMPI mailboxes when a message is posted
Should avoid race condition, while keeping isolation from SIMIX

9 years agoRevert "Avoid using simcalls here, as by descheduling the process, we could misplace...
Augustin Degomme [Sat, 6 Dec 2014 02:00:14 +0000 (03:00 +0100)]
Revert "Avoid using simcalls here, as by descheduling the process, we could misplace some messages in mailboxes, and end up deadlocking."

This reverts commit 63ba498484d4fcfc706be1e514d7fedbc6c9f4be.

9 years ago[mc] Multiple .so support in MC_ignore_local_variable()
Gabriel Corona [Fri, 5 Dec 2014 14:50:56 +0000 (15:50 +0100)]
[mc] Multiple .so support in MC_ignore_local_variable()

9 years ago[mc] Basic support for more other libraries than libsimgrid.so
Gabriel Corona [Thu, 4 Dec 2014 15:03:20 +0000 (16:03 +0100)]
[mc] Basic support for more other libraries than libsimgrid.so

9 years agoactivate mpich3 tests for post/wait/start/complete RMA calls
Augustin Degomme [Fri, 5 Dec 2014 13:59:20 +0000 (14:59 +0100)]
activate mpich3 tests for post/wait/start/complete RMA calls

9 years agoAdd MPI_Win_post, MPI_Win_start, MPI_Win_complete, and MPI_Win_wait support.
Augustin Degomme [Fri, 5 Dec 2014 13:51:37 +0000 (14:51 +0100)]
Add MPI_Win_post, MPI_Win_start, MPI_Win_complete, and MPI_Win_wait support.

This is the second (out of 3) of the classic MPI RMA synchronization methods.
This version is naive and may not be what real MPI lib do, as the standard lets the implementer chose the behavior of theses calls.

9 years ago[mc] Move process info in a new s_mc_process_t structure
Gabriel Corona [Thu, 4 Dec 2014 10:02:16 +0000 (11:02 +0100)]
[mc] Move process info in a new s_mc_process_t structure

This is a beginning of the refactoring in order to support MC-ing a
remote process.

9 years ago[mc] Fix distcheck
Gabriel Corona [Thu, 4 Dec 2014 10:41:02 +0000 (11:41 +0100)]
[mc] Fix distcheck

9 years ago[mc] Don't include libunwind.h in non MC builds
Gabriel Corona [Thu, 4 Dec 2014 10:07:40 +0000 (11:07 +0100)]
[mc] Don't include libunwind.h in non MC builds

9 years agoMerge branch 'master'
Gabriel Corona [Thu, 4 Dec 2014 09:45:49 +0000 (10:45 +0100)]
Merge branch 'master'

9 years ago[mc] Don't use unprototyped functions
Gabriel Corona [Thu, 4 Dec 2014 09:43:01 +0000 (10:43 +0100)]
[mc] Don't use unprototyped functions

9 years ago[mc] Remove useless header #includes
Gabriel Corona [Tue, 2 Dec 2014 13:02:26 +0000 (14:02 +0100)]
[mc] Remove useless header #includes

9 years agoforgot to add this include
Augustin Degomme [Tue, 2 Dec 2014 18:17:49 +0000 (19:17 +0100)]
forgot to add this include

9 years agolet's try to please windows
Augustin Degomme [Tue, 2 Dec 2014 17:20:13 +0000 (18:20 +0100)]
let's try to please windows

9 years agoAvoid using simcalls here, as by descheduling the process, we could misplace some...
Augustin Degomme [Tue, 2 Dec 2014 14:50:12 +0000 (15:50 +0100)]
Avoid using simcalls here, as by descheduling the process, we could misplace some messages in mailboxes, and end up deadlocking.

Calling directly SIMIX functions is not really the best, but it may fix a bad heisenbug

9 years agotypos-=2
Augustin Degomme [Tue, 2 Dec 2014 14:44:45 +0000 (15:44 +0100)]
typos-=2

9 years agoavoid problem when freeing pointer with lb!=0
Augustin Degomme [Tue, 2 Dec 2014 14:44:22 +0000 (15:44 +0100)]
avoid problem when freeing pointer with lb!=0

9 years agodo the same thing as before with IB model parameters
Augustin Degomme [Mon, 1 Dec 2014 13:53:19 +0000 (14:53 +0100)]
do the same thing as before with IB model parameters

9 years ago[mc] Modularise header files for MC
Gabriel Corona [Tue, 2 Dec 2014 09:41:38 +0000 (10:41 +0100)]
[mc] Modularise header files for MC

This is a preparation step for the upcoming refactorisation of the MC
code in order to MC an external process.

9 years ago[mc] Define a type for MC object information flags
Gabriel Corona [Tue, 2 Dec 2014 09:00:28 +0000 (10:00 +0100)]
[mc] Define a type for MC object information flags

9 years ago[mc] Remove MC_ignore_global_variable() calls
Gabriel Corona [Mon, 1 Dec 2014 14:31:29 +0000 (15:31 +0100)]
[mc] Remove MC_ignore_global_variable() calls

 - compared_pointer which does not exist;
 - smpi_current_rank does not exist;
 - maestro_stack_start and mastro_stack_end doe not need to be ignored.

9 years ago[mc] Enable MC specific behaviour in replay mode
Gabriel Corona [Mon, 1 Dec 2014 12:47:04 +0000 (13:47 +0100)]
[mc] Enable MC specific behaviour in replay mode

9 years agoRevert "[mc] Enable MC specific behaviour in replay mode"
Gabriel Corona [Mon, 1 Dec 2014 13:01:43 +0000 (14:01 +0100)]
Revert "[mc] Enable MC specific behaviour in replay mode"

This reverts commit 33eca433c4f055cdfcc55e46d125f8708e1848c7.

Build is broken.

9 years ago[mc] Enable MC specific behaviour in replay mode
Gabriel Corona [Mon, 1 Dec 2014 12:47:04 +0000 (13:47 +0100)]
[mc] Enable MC specific behaviour in replay mode

9 years ago[mc] Remove useless condition check
Gabriel Corona [Mon, 1 Dec 2014 12:17:59 +0000 (13:17 +0100)]
[mc] Remove useless condition check

9 years ago[mc] Only enable the umpire test for MC builds
Gabriel Corona [Mon, 1 Dec 2014 11:33:16 +0000 (12:33 +0100)]
[mc] Only enable the umpire test for MC builds

9 years agoUse pthread mutex instead of semaphore in mm
Gabriel Corona [Thu, 27 Nov 2014 10:25:22 +0000 (11:25 +0100)]
Use pthread mutex instead of semaphore in mm

9 years agoFix dist
Gabriel Corona [Mon, 1 Dec 2014 11:16:45 +0000 (12:16 +0100)]
Fix dist

9 years agos/formated/formatted/
Gabriel Corona [Fri, 28 Nov 2014 13:38:31 +0000 (14:38 +0100)]
s/formated/formatted/

9 years ago[mc] Initial support MC record/replay
Gabriel Corona [Thu, 30 Oct 2014 13:39:17 +0000 (14:39 +0100)]
[mc] Initial support MC record/replay

The idea is to record an execution path in MC mode inorder to be able
to replay it outside of the MC (event with a non-MC build). Some very
basic (an unobtrusive) MC code is compiled even when MC is disabled.

9 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Mon, 1 Dec 2014 09:55:12 +0000 (10:55 +0100)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

9 years agoand now, fix the java teshsuite, re-sorry
Martin Quinson [Sat, 29 Nov 2014 13:29:31 +0000 (14:29 +0100)]
and now, fix the java teshsuite, re-sorry

I shouldnt try to hack on simgrid at week-ends :-(

9 years agofix the build of java bundles, sorry
Martin Quinson [Sat, 29 Nov 2014 13:03:13 +0000 (14:03 +0100)]
fix the build of java bundles, sorry

9 years agoDont produce that pdf output that we dont use
Martin Quinson [Sat, 29 Nov 2014 12:31:19 +0000 (13:31 +0100)]
Dont produce that pdf output that we dont use

9 years agoreindent and improve displayed message
Martin Quinson [Sat, 29 Nov 2014 10:56:42 +0000 (11:56 +0100)]
reindent and improve displayed message

9 years agoput together the java-based tests
Martin Quinson [Sat, 29 Nov 2014 10:53:58 +0000 (11:53 +0100)]
put together the java-based tests

9 years agoMerge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid
Adrien Lebre [Sat, 29 Nov 2014 08:59:40 +0000 (09:59 +0100)]
Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/simgrid/simgrid

9 years agomove smpi bandwidth and latency factors out of the ifdef HAVE_SMPI. SMPI and IB netwo...
Augustin Degomme [Fri, 28 Nov 2014 17:04:29 +0000 (18:04 +0100)]
move smpi bandwidth and latency factors out of the ifdef HAVE_SMPI. SMPI and IB network models can be used without using SMPI

9 years agomerge msg_vm.c - adrien (please note that there is one line (destruction of the tx_pr...
Adrien Lebre [Fri, 28 Nov 2014 16:29:16 +0000 (17:29 +0100)]
merge msg_vm.c - adrien (please note that there is one line (destruction of the tx_process) where I don't not remind whether it is usefull or not... this can lead to a crash unfortunately, I will fix it later)

9 years agoRemove warnings in vm
Augustin Degomme [Fri, 28 Nov 2014 09:29:01 +0000 (10:29 +0100)]
Remove warnings in vm

9 years agoFix dist
Augustin Degomme [Fri, 28 Nov 2014 09:20:33 +0000 (10:20 +0100)]
Fix dist

9 years agoremove potential bug / clang warning
Augustin Degomme [Tue, 25 Nov 2014 12:16:19 +0000 (13:16 +0100)]
remove potential bug / clang warning
size_t being undefined, the comparison < 0 was never true

9 years agosupport timeout of migration
Takahiro Hirofuchi [Thu, 27 Nov 2014 11:24:07 +0000 (20:24 +0900)]
support timeout of migration

Fixme: The default timeout value is hard-coded. Modify it and compile
the code if necessary.

9 years agofix indent in migration code
Takahiro Hirofuchi [Thu, 27 Nov 2014 11:14:17 +0000 (20:14 +0900)]
fix indent in migration code

9 years agoremove unnecessary comment out
Takahiro Hirofuchi [Thu, 27 Nov 2014 11:10:55 +0000 (20:10 +0900)]
remove unnecessary comment out

9 years agoremove the unnecessary vm object in migration
Takahiro Hirofuchi [Thu, 27 Nov 2014 10:27:48 +0000 (19:27 +0900)]
remove the unnecessary vm object in migration

9 years agoremove unused code in migration
Takahiro Hirofuchi [Thu, 27 Nov 2014 08:55:02 +0000 (17:55 +0900)]
remove unused code in migration

9 years agoremove trailing space in the migration code
Takahiro Hirofuchi [Thu, 27 Nov 2014 06:49:41 +0000 (15:49 +0900)]
remove trailing space in the migration code

9 years agoremove unused migration code for CPU overheads
Takahiro Hirofuchi [Thu, 27 Nov 2014 06:28:34 +0000 (15:28 +0900)]
remove unused migration code for CPU overheads

This commit should not affect anything.

9 years ago[mc] Test if the stack-cleaner has any effect
Gabriel Corona [Mon, 24 Nov 2014 15:03:10 +0000 (16:03 +0100)]
[mc] Test if the stack-cleaner has any effect

In order to test this:

 * we compile the same test program with and without the stack cleaner
   (`-fstack-cleaner`, `-fno-stack-cleaner`);

 * in this program, we move random bytes in the stack;

 * we expect the stack-cleaner to zero them out.

This test in only used if the configure stack-cleaner is detected to
support the `-fstack-cleaner` CLI option (it is the stack-cleaner
compiler wrapper).

9 years ago[mc] Disable/enable the stack-cleaner from a CLI argument (-f[no-]stack-cleaner)
Gabriel Corona [Mon, 24 Nov 2014 14:33:47 +0000 (15:33 +0100)]
[mc] Disable/enable the stack-cleaner from a CLI argument (-f[no-]stack-cleaner)

9 years ago[mc] Fix the stack cleaner
Gabriel Corona [Mon, 24 Nov 2014 12:19:14 +0000 (13:19 +0100)]
[mc] Fix the stack cleaner

The condition was broken and the %rsp limit was too high.

9 years ago[mc] Fix umpire tests
Gabriel Corona [Mon, 24 Nov 2014 09:19:56 +0000 (10:19 +0100)]
[mc] Fix umpire tests

9 years agoMerge branch 'xp'
Gabriel Corona [Fri, 21 Nov 2014 07:56:14 +0000 (08:56 +0100)]
Merge branch 'xp'

9 years agoRevert "Temporarily disable an option"
Gabriel Corona [Fri, 21 Nov 2014 07:55:38 +0000 (08:55 +0100)]
Revert "Temporarily disable an option"

Back to normal.

9 years agoTemporarily disable an option
gabriel corona [Thu, 20 Nov 2014 15:06:21 +0000 (16:06 +0100)]
Temporarily disable an option

The option somehow changes the results in the MC from previous
experiments. It is disabled temporarily in this commit in order to be
able to reproduce those results with the new commits.

9 years agoInfrastructure for statically defined tracepoints
Gabriel Corona [Fri, 7 Nov 2014 15:17:18 +0000 (16:17 +0100)]
Infrastructure for statically defined tracepoints

3 modes are supported on compilation:

 * normal (no SDT);

 * SDT (systemtap statically defined tracepoint);

 * UST (lttng userspace static tracepoint, compatible with systemtap
   if LTTNG_UST_HAVE_SDT_INTEGRATION).

9 years ago[mc] Remove reference to DW_TAG_mutable_type:
Gabriel Corona [Tue, 18 Nov 2014 09:46:19 +0000 (10:46 +0100)]
[mc] Remove reference to DW_TAG_mutable_type:

Is was in a DWARFv3 draft but was removed from the final version: it
was removed from libdw which breaks compilation.

9 years agotypo -= 2
Martin Quinson [Tue, 18 Nov 2014 08:44:34 +0000 (09:44 +0100)]
typo -= 2

9 years agoprotect these calls to smpi_datatype_size as they are not always relevant
degomme [Tue, 18 Nov 2014 07:00:13 +0000 (08:00 +0100)]
protect these calls to smpi_datatype_size as they are not always relevant

9 years agoFix problem with unknown datatypes in replay/tracing.
degomme [Mon, 17 Nov 2014 22:42:46 +0000 (23:42 +0100)]
Fix problem with unknown datatypes in replay/tracing.
When datatype was unknown to replay, it was replayed as MPI_BYTE.

This modification adds a parameter to encode_datatype, to tell tracing that the datatype size has to be taken into account in the count parameter
This results in the fact that a message of count*datatype_size being replayed as a message of (count*datatype_size)*sizeof(MPI_BYTE), which is the same.

This is not a perfect or elegant solution, but :
- it works.
- it handles manually created datatypes
- it doesn't break previously generated replay files
- it avoids testing each time 50 different datatypes (see encode_datatype function)
- the new parameter avoids doing strcmp with "-1" at each time, performance should not be too bad

9 years agoThere should not be msg datatypes here
degomme [Mon, 17 Nov 2014 22:24:32 +0000 (23:24 +0100)]
There should not be msg datatypes here

9 years ago[mm] Disable HAVE_GNU_LD code in order to get rid of the junkarea
Gabriel Corona [Mon, 17 Nov 2014 14:34:34 +0000 (15:34 +0100)]
[mm] Disable HAVE_GNU_LD code in order to get rid of the junkarea

The HAVE_GNU_LD mode of mmalloc delegates in some cases to standard
malloc()/free() which are resolved with dlsym(). This cause some
bootstrap problems which are only resolved with the junkarea: the
junkarea is regularly broken when adding dependencies because the
junkarea is then too small.

By disabling the HAVE_GNU_LD path, we get rid of the junkarea hack.

9 years agoclean after augustin, as usual
Martin Quinson [Thu, 13 Nov 2014 21:01:35 +0000 (22:01 +0100)]
clean after augustin, as usual

9 years agoRemove the unmodified NAS examples as they are really useless nowadays
Martin Quinson [Thu, 13 Nov 2014 20:59:09 +0000 (21:59 +0100)]
Remove the unmodified NAS examples as they are really useless nowadays

I'm still unsure of what to do with the modified ones. I vote for
removing them if we have enough examples already.