James Dinan
James Dinan
Add a shmemx unit test for fork/spawn compatibility. I started drafting this test, but am waiting on feedback before generating a PR. Ref: https://github.com/openshmem-org/specification/issues/223
Is it possible to generate the C++ bindings using template metaprogramming instead of the current generators? If so, is that better than what we do now?
Add a new unit test where threads call collectives at the same time, but using different psync/pwrk arrays.
The C11/C++ bindings checkers are compiled using the library flags, which can cause them to be linked with communication libraries resulting in compile-time issues.
Missed #680 apparently because of a gap in testing coverage over large message reductions. Is this case covered or should we add a test that sweeps a range of message...
Add a common "utilities" header for tests for common error checking/reporting things. * Replace usage of ```assert``` with a ```TEST_CHECK_ERROR()``` routine that exits with ```shmem_global_exit``` on error. * Perform scalar...
Wrinkle is that we don't copy the string into a library buffer. Is it ok to lowercase the string returned by getenv?
This is apparently a toxic combination. If you remove ```--enable-remote-virtual-addressing``` all is well. Not sure why this breaks or whether it can actually be fixed... ``` $ mpiexec -np 2...
In transport_ofi.c, is the following TODO still needed? ``` hints.tx_attr = &tx_attr; /* TODO: fill tx_attr */ ``` If (true) create an issue for it. Else, please remove the TODO.