Fine-tune scheduling of parallel reductions

hercules_cg/src/fork_tree.rs

hercules_cg/src/gpu.rs

hercules_opt/src/fork_guard_elim.rs

hercules_opt/src/gcm.rs

hercules_samples/dot/build.rs

hercules_samples/dot/src/main.rs

hercules_samples/matmul/build.rs

hercules_samples/matmul/src/main.rs

hercules_test/hercules_interpreter/src/interpreter.rs

hercules_test/hercules_tests/tests/loop_tests.rs

juno_frontend/src/semant.rs

juno_samples/cava/src/main.rs

juno_samples/concat/src/main.rs

juno_samples/edge_detection/src/main.rs

juno_samples/fork_join_tests/src/fork_join_tests.jn

juno_samples/fork_join_tests/src/gpu.sch

juno_samples/fork_join_tests/src/main.rs

juno_samples/matmul/src/main.rs

juno_samples/nested_ccp/src/main.rs

juno_samples/patterns/src/main.rs