Thomas Quinn
Thomas Quinn
The "Order array broadcast delivery with send epoch" seems to be causing memory corruption problems in ChaNGa. On of the errors looks like this: `Processor 375 Exiting: Called CmiAbort ------------...
ChaNGa, using the HAPI interface to CUDA, fails at exit because cudaFreeHost() reports an invalid argument when called in the releasePool() method. From code inspection, it is apparent that, if...
This is with Charm version 7.0.0 and ChaNGa. I am running on frontera with the mpi build of charm on 40 nodes. I am occasionally getting hangs during the use...
Running ChaNGa under valgrind reports that CkArray::springCleaning() is accessing freed memory. This is with ChaNGa version 3.5 commit v3.5-11-gc7ba57c0 and charm version v7.1.0-devel-321-g606459e74 This is built on an AMD/infiniband machine...
*Original issue: https://charm.cs.illinois.edu/redmine/issues/551* --- The code documentation for Ck::IO::Options describes "basePE", "skipPEs", and "activePEs" to specify which processors are to participate in an IO session. However, the session array is...
Note that no simulation code is changed, just the code for testing.
Makefile.in always sets "-m64" in compile flags which is only valid for certain architectures (e.g. Intel) and not others (e.g., ARM).
This is @harshithamenon 's code to replicate treepieces to improve performance by distributing cache requests over several processors.
There are a few `#ifdef CUDA`s around the cleanup of registeredTreepieces---a vector of treepieces on each node. This information could be useful elsewhere (e.g. the TreePieceReplica code) so all these...
testcollapse/compare.py is updated to python3.