Add dummy implementation of FinalizedState::history_tree

[conradoplg]

Motivation

#2135 adds support for the ZIP-221 history tree in the non-finalized state. However, it needs to read the tree from the finalized state, and a todo! method was added to it. However, that makes zebrad crash.

Specifications

N/A

Designs

N/A

Solution

This adds a dummy implementation that will prevent it from crashing (but it won't give the correct result)

Review

Not sure if this is really needed (can we just wait for #2135?), so it's a draft. I also haven't tested it yet.

@teor2345 might want to take a look. This is critical only if stopping zebrad from panicking is critical right now.

Reviewer Checklist

• Code implements Specs and Designs
• Tests for Expected Behaviour
• Tests for Errors

Follow Up Work

#2135 will properly fix this.

[teor2345]

Not sure if this is really needed (can we just wait for #2135?), so it's a draft. I also haven't tested it yet.

We need to calculate the note commitment trees in the non-finalized and finalized state before we do #2135. Leaving the non-finalized tests disabled until all that work is done risks introducing more bugs.

@teor2345 might want to take a look. This is critical only if stopping zebrad from panicking is critical right now.

This is critical - we don't want to merge any more code until the non-finalized state is covered by tests, and they're passing.

[teor2345]

This PR breaks the out-of-order block commit test - it never finishes:

test service::tests::state_behaves_when_blocks_are_committed_out_of_order has been running for over 60 seconds
Error: The operation was canceled.

https://github.com/ZcashFoundation/zebra/pull/2461/checks?check_run_id=3013159372#step:10:1750

[teor2345]

This history tree gets passed to check::block_commitment_is_valid_for_chain_history, which will fail because the history tree doesn't have the right blocks.

So we'll also need to disable check::block_commitment_is_valid_for_chain_history.

And at that point we might as well revert the PR. 😢

[conradoplg]

Yep 😢 I'll close this PR since it doesn't make sense anymore.

[conradoplg]

Closing this because it seems we can't avoid doing everything bottom-up, so we need to implement everything that is required by #2135 beforing merging its PR.