Lisandro Dalcin

I can reproduce the mpi4py failure locally. However, if I try `mpiexec -n {2,3,4,5} python test/test_ulfm.py -v` to narrow down to the previously failing test, now everything runs just fine....

Next I run the following way to EXCLUDE the ULFM tests ``` mpiexec -n 2 python test/main.py -v -e ulfm ``` and then a intercommunicator failed the following different way:...

This may be the same or similar issue being discussed in #12367. @bosilca nailed down the problem https://github.com/open-mpi/ompi/issues/12367#issuecomment-1985804113, but I have no idea if anyone is investigating further or someone...

> If you really want to test each of those fairly, I'm not sure what do you mean by "fair". That way of testing is what everyone else does. Such...

> then terminate must instead cleanly handle every possible failure or unclean termination of an application. I think you may have misunderstood what mpi4py's ULFM tests are about. These tests...

> Let me try to explain again. `mpiexec` is not an MPI application - it does not invoke MPI, it does not even link against `libmpi`. I'm very well aware...

@jsquyres I just found something else: If I run everything the following way ``` mpiexec -n 2 python test/main.py -v ``` then it fails with ``` ... [kw61149:870126] [[11981,1],1] selected...

@bosilca I was talking indeed about "high-stress" spawn tests. I'm in a conundrum here. If I keep spawn tests active, they will keep failing and generating confusion with unrelated changes....

> Perhaps you could disable them, but give us a trivial command line way to enable all the spawn tests Would something as silly as `export MPI_SPAWN=1` to enable these...

Sorry, I have to try again with the full set of configure options I was trying, I'm suspecting of `--disable-static`...