Closed tmadlener closed 1 year ago
The upstream spack recipe only does setup_dependent_run_environment
, which seems like a bug. You need those env vars to run pythia8 itself, so it should also do setup_run_environment
.
But that it's not currently making it's way into the setup script also seems like a bug, in k4_generate_setup_script. Since we do have dependencies of pythia8 in the stack it should also take into account their run environment.
I was also thinking that, but that change has been introduced ~2 years ago and it has been working for quite some time, so I am not sure why it stopped working now or why it works for Ubuntu but not the rest.
My suspicion is this commit here: https://github.com/spack/spack/commit/72b36ac14446782f47519db8ee5cff6631b2b4c9
My suspicion is this commit here: https://github.com/spack/spack/commit/72b36ac14446782f47519db8ee5cff6631b2b4c9
That might not be it in the end. In my local development environment I created with spack, PYTHIA8
is set correctly. This is also on Ubuntu, so I am not sure whether this is an OS issue?
This isn't happening as of today. That commit lead to some problems related to how setup_dependent_run_environment
works but I think they fixed it after I complained.
The
PYTHIA8
andPYTHIA8DATA
environment variables are not set in the current nightly builds, making the CI for k4SimDelphes fail: https://github.com/key4hep/k4SimDelphes/actions/runs/6731118066/job/18295169237?pr=112#step:5:38In principle these variables are set in the spack recipe here, so I am not sure what is happening in the nightly builds that prevents setting these correctly.