the project has many workspace crates that contain critical code (like the kmm crate for virtual memory management) but that have no unit tests yet at all. This is mostly due to the ceremony required to set up the machine to run anything at all (like an allocator). Duplicating that across all crates just for running tests would be madness though. Maybe we can spin up a "runtime crate" that encapsulates all that setup logic (including setting up stuff like unwinding, backtracing etc)
the project has many workspace crates that contain critical code (like the
kmm
crate for virtual memory management) but that have no unit tests yet at all. This is mostly due to the ceremony required to set up the machine to run anything at all (like an allocator). Duplicating that across all crates just for running tests would be madness though. Maybe we can spin up a "runtime crate" that encapsulates all that setup logic (including setting up stuff like unwinding, backtracing etc)