Joseph Schuchart
Joseph Schuchart
1. Implement global-to-global `dash::copy`, needed for 2. `dash::transform` that honors the distribution of the output range for local input ranges, which 3. now also works with std iterators, .e.g, coming...
While working on #374 I realized that some algorithms in DASH synchronize units at the end (e.g., `dash::for_each`) and some don't (e.g., `dash::generate`). This should be unified. @fuchsto I assume...
The current implementation of `dash::accumulate` has a flaw: only unit 0 (randomly chosen?) actually receives a valid result. This is neither documented nor does the user have a chance to...
Consider the following code snippet: ``` struct point_t { int x, y = 1; }; std::vector vec(N); assert(vec[0].y == 1); // succeeds dash::Array arr(N); assert(vec[0].y == 1); // may or...
@fuchsto: Please have a look at this. Your test `MatrixTest.DelayedAlloc` succeeds while the new test `MatrixTest.VariadicDelayedAlloc` fails while checking the validity of the pattern arguments, which are the same as...
In my humble try to compile DASH with all compilers we support, I also tried to use the latest Cray CCE to compile the DASH part. While this has not...
Right now, `dash::transform` can be called with a local input range. The user is responsible for blocking the output range, i.e., he has to call `dash::transform` for every target unit...
TUD has a Jenkins CI server running that we could use to run our tests on multiple nodes of their cluster. This will cover a whole class of problems that...
I am seeing issues on my local machine when running the `ThreadsafetyTest`s under OpenMPI. The used compiler is GCC 5.3.0, both in Debug and Release mode. __OpenMPI 1.10.5__ I am...
Access to C11 _Atomic variables incurs sequential memory ordering, which compiles to atomic operations and in many cases is not desired (initialization, non-multi-threaded runs). This change makes GCC atomics the...