rust: Allow for usage of flatbuffers in #![no_std] environment

[MarWit]

Hey,

this PR introduces changes that allow for usage of flatbuffers in #![no_std] environments. The main difference to this PR is that instead of removing thiserror for no_std builds it uses core2 and thiserror_core2 in its place. The core2 crate needed some small changes, so I had to use fork for the time being.

[dbaileychess]

@CasperN Could you check this? This was an internal request.

[CasperN]

LGTM, sorry for the delayed review. Please pull from master, sign the CLA, and I'll merge when tests pass.

[MarWit]

Rebased and CLA sorted out.

[CasperN]

Sorry @MarWit, one more thing, can you bump the minor version number in the flatbuffers/Cargo.toml?

Looks like the CI failure has to do with Java and Android, so it seem irrelevant to this PR.

[MarWit]

Done. Just a little reminder that this change is using our forks of core2 and thiserror-core2 crates -- there is ongoing PR for them, so it could be addressed in subsequent patch, though.

[MarWit]

core2 changes have been merged into the upstream; changed cargo manifest accordingly.

[CasperN]

Rust CI failed with 403, which is ...interesting

I think its fine to merge since the only change from the force push should've been the version number and cargo.toml.

@MarWit are you good to merge?

[MarWit]

Yes. There is still thiserror_core2 that needs to be upstreamed, but this can be done later.

[CasperN]

There is still thiserror_core2 that needs to be upstreamed, but this can be done later.

I'll publish to crates.io once that is upstreamed and updated in our Cargo.toml (they require dependencies to have published versions to publish).

[CasperN]

@MarWit when do you expect the upstreaming to happen?

[MarWit]

I am currently waiting for the feedback from the crate's maintainer, so I sadly do not have any ETA right now. I will let you know when I will have any information.

[CasperN]

Gotcha, note that while we do not publish often, this is blocking publishing; so I'd prefer it gets resolved within a couple weeks - lest we risk slowing down another contributor. I hope that's plenty of time, but let me know if upstreaming gets blocked and we can figure something out.