- `util/print_env.sh` for debugging pipelines.

- This can be a tag or a branch (uses `git clone -b X`)

Update the CI scripts for the DIS stub to be consistent with the more generic DVMP scripts. Also did some more restructuring of the benchmark to make it easier to split off a generic benchmark library in the future.

Fix issue where caching leads to DVMP samples to not be re-generated when the requested number of events changes.

Updated benchmark output file to always write to a 'tests' key, making it easier to detect if a json file contains test results

Add benchmark/exception headers as symlink to the benchmarks for now, in the future these should be moved to the util library

disabled debug logging for now, will re-enable after having established that running many jobs works

Got rid of fname script for dvmp, and split up generator artifacts from reconstruction artifacts (input versus results directories

 - We should limit the number of new environment variables. They
 increase complexity which should be mostly absorbed by using containers.
 - Any new environment variables should be set via the ci yml file
 or defined the UI. https://docs.gitlab.com/ee/ci/variables/README.html#custom-environment-variables
 - Removed dummy scripts
 - Removed dvmp dummy jobs.
 - Fixed dvmp output data. Note the out artifacts should go in
 `results/dvmp` if they are in the `dvmp` directory. This will just make
 things easier later on for collecting and summarizing results.
 - updated  dvmp/scripts/run_generator_instance.sh
 - Added dvmp.sh to run geant4/reconstruction. Note `dvmp/dvmp.sh` could be named something more specific, like `dvmp/dvmp_barrel_decays.sh` depending on the input, analysis, etc...