This way goto source doesn't end up in a generated file for those parts and they can be edited directly.
I have way too often accidentally edited the generated file.
The arcu crate is a more general implementation of the Rcu I implemented in here in scryer. It contains some bug-fixes regarding race-conditions in the Rcu update function, which could cause leaks and uses after free.
Source of the problem was the Relaxed load/strore/update of the reference count in side the Arc not being properly ordered with other load/stores.
there are technically still two locks
- the Mutex to serialize consurrent updates to the AtomTable
- the RwLock as part of GLOBAL_ATOM_TABLE
the former is the age lock of RESIZE ATOM TABLE in <https://arxiv.org/pdf/1608.00989.pdf>, though we use it for the whole update as we don't match the whole structure and can't reserve an atom slot as is, so we need to also lock concurrent updates without resizing
the later should be uncontendet as we only write AtomTable::new iff the current value is a dangling Weak
* makes most pub things in src/ pub(crate) as not to expose things accidentally
* only those things needed by src/bin/scryer-prolog.rs and tests/scryer.rs
should be pub
* split src/main.rs into src/lib.rs and src/bin/scryer-prolog.rs
* add tests folder and run most of the files in src/tests with cargo test
added bytes method to Stream in src/machine/streams.rs to check if stdout is as expected
