- Feb 08, 2016
-
-
Felix Gruber authored
(cherry picked from commit 60459f97) Signed-off-by:
Steffen Müthing <muething@dune-project.org>
-
- Sep 12, 2015
-
-
Markus Blatt authored
(cherry picked from commit 9588afdf) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
- Sep 01, 2015
-
-
Markus Blatt authored
(cherry picked from commit 3bef4991) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
- Jul 01, 2015
-
-
Dominic Kempf authored
With the recent changes to bigunsignedint, the index cannot be initialized with a negative value anymore. We change this to 1 to overcome the problem. This commit together with the recent bugfixes by Ansgar fixes the AMG code on master. (cherry picked from commit bf023947) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
Ansgar Burchardt authored
(cherry picked from commit 6100b778) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
Ansgar Burchardt authored
Since 0c697d8d messages of size zero are no longer sent. This can reduce the number of requests we have to wait for. (cherry picked from commit f35437e9) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
Ansgar Burchardt authored
The storage of bigunsignedint was changed to std::uint16_t in commit 541b3dfc. This implements the same change for the MPITraits<bigunsignedint<k>> Note that this assumes that MPITraits<std::uint16_t> is defined. This is the case as long as uint16_t maps to one of the standard integer types (unsigned char/short/int/long). Thanks to Markus for pointing this out. Reference: https://dune-project.org/flyspray/index.php?do=details&task_id=1657#comment6393 (cherry picked from commit 586375b4) Signed-off-by:
Carsten Gräser <graeser@dune-project.org>
-
- Apr 20, 2015
-
-
(cherry picked from commit 7aa683f1)
-
-
Markus Blatt authored
Before this commit communcation would be initiated even on empty interfaces. While this does not lead to overwriting data or producing segmentation faults, it does lead to accessing the first component of empty vectors to get a pointer to pass to MPI functions. This produces error output with valgrind and other memory debuggers. Therefore with this commit we will first check whether the interface contains values and do nothing if it doesn't. We also augmented the test to catch the case wher only some processes have an empty interface.
-
- Apr 10, 2015
-
-
Steffen Müthing authored
This check was added to improve the error message returned when using MPI without correctly initializing it (e.g. by creating a parallel grid), see FS#1612 for further information.
-
- Feb 27, 2015
-
-
Christian Engwer authored
@Markus: too much CMake?! ;-)
-
Markus Blatt authored
There is not need to send message that contain no data. It might even produce problems on some implementation. Therfore this patch always checks the size and simply sets the request to inactive (aka MPI_REQUEST_NULL) if no data would be send or received. The patch is to be cherry-picked to master if Christian approves.
-
- Jan 11, 2015
-
-
Christoph Grüninger authored
Like we do for Autotools. This is only supported for CMake 3.0 or newer. Oder versions ignore the property and report the test as failed.
-
- Nov 08, 2014
-
-
Jö Fahlke authored
"Makefile.in", "*.o", etc. need only be listed in the toplevel .gitignore, the rules are applied recursively. Test programs etc. should be listed as "/program" in their directory's .gitignore, so they are not accidentially ignored in a lower level directory.
-
- Oct 17, 2014
-
-
Oliver Sander authored
-
- Oct 06, 2014
-
-
Ansgar Burchardt authored
-
- Oct 01, 2014
-
-
Tobias Malkmus authored
This very old gcc bug https://gcc.gnu.org/bugzilla/show_bug.cgi?id=10200 is fixed in clang but not in icc or gcc, see FS#1505.
-
- Sep 30, 2014
-
-
Steffen Müthing authored
In order to simplify fallback handling between code with and without MPI support, this patch makes the specialization of CollectiveCommunication for MPI communicators default-constructible by defaulting the underlying MPI communicator to MPI_COMM_WORLD, which should be reasonable for most scenarios. As the non-specialized version for sequential code is already default-constructible, this allows grid constructors to accept a CollectiveCommunication object regardless of whether MPI support is enabled or not and to provide a default value for this parameter. This patch resolves FS#1497.
-
- Sep 29, 2014
-
-
Dominic Kempf authored
CollectiveCommunication serves as a fallback implementation, when code that is written for parallel computations is executed in sequential. The template parameters BinaryFunctions are no-op then, but they are still there to define the interface. However, in a generic (parallel) implementation these will be given. Including binaryfunctions.hh here avoids error messages due to no-op template parameters whose names could not be found by the compiler.
-
- May 14, 2014
-
-
Oliver Sander authored
-
- Apr 09, 2014
-
-
Christoph Grüninger authored
-
- Apr 07, 2014
-
-
Benjamin Bykowski authored
-
- Mar 22, 2014
-
-
Christoph Grüninger authored
-
- Mar 18, 2014
-
-
Oliver Sander authored
-
- Mar 06, 2014
-
-
Christian Engwer authored
-
Christian Engwer authored
-
- Feb 14, 2014
-
-
Christoph Grüninger authored
-
Christoph Grüninger authored
These headers were moved in Dune 2.3.
-
- Jan 17, 2014
-
-
Christian Engwer authored
when we try to send data, it might happen, that for some reason we don't want to any data. This happens, when the dynamic size of our CommPolicy is 0. In that case we must not add a MessageInformation, otherwise we get an MPI error.
-
Christian Engwer authored
-
- Jan 16, 2014
-
-
Markus Blatt authored
-
- Jan 14, 2014
-
-
Robert Kloefkorn authored
-
- Jan 10, 2014
-
-
Martin Nolte authored
Somehow, the vector version of CollectiveCommunication< MPI_Comm >::prod passed std::plus to allreduce (instead of std::multiplies). This patch corrects this flaw.
-
- Dec 18, 2013
-
-
Markus Blatt authored
VariableSizeCommunicator now appears in doxyen and the parallel index sets are a subsection of the parallel section.
-
- Dec 17, 2013
-
-
Markus Blatt authored
There entries in doxyerror.log about undocumented function parameters and parameters that had a wrong name in the documentation. This patch fixes this.
-
Markus Blatt authored
This patch introduces a new macro DUNE_UNUSED_PARAMETER(parm) for marking unused parameters that are there for a purpose (e.g. the prescribed interfac The macro simply performs a cast to void. In addition all unused parameter w that I found were removed.
-
- Dec 13, 2013
-
-
Markus Blatt authored
-
Markus Blatt authored
-
Markus Blatt authored
With this patch VariableSizeCommunicator now really supports a varying number of data entries per index during communication.
-