feat!: serialise RAM and ROM opcodes

[guipublic]

This PR serialise RAM and ROM opcodes from ACIR into BB acir format.

BEGIN_COMMIT_OVERRIDE
feat!: Add serialization logic for RAM and ROM opcodes (#153)
END_COMMIT_OVERRIDE

[TomAFrench]

Can we add a couple of test cases similar to this for a simple block? Ideally we'd have an test for each constraint type so we don't rely on noir integration tests to check we're handling them correctly here.

[guipublic]

I added a unit test, it is failing because it cannot locate barretenberg so I cannot test it, but anyways the relevant barretenberg PR is not merge yet.

[TomAFrench]

Looks good. We should be good to merge once the conflicts are resolved.

Huh, looks like the upstream BB PR hasn't been merged yet. Strange that this PR's test is passing in that case.

[guipublic]

Huh, looks like the upstream BB PR hasn't been merged yet. Strange that this PR's test is passing in that case.

I suppose it is because I added the new data at the end when serialising, so if you read less, you still get valid data.

[TomAFrench]

I've bumped the BBerg version to include the relevant PRs so the test should be actually testing stuff now.

[TomAFrench]

It would be good to see a commit which breaks the blocks test case (e.g. use a witness which doesn't satisfy the block constraints) to show that this test is meaningful and then revert it.

[vezenovm]

It would be good to see a commit which breaks the blocks test case (e.g. use a witness which doesn't satisfy the block constraints) to show that this test is meaningful and then revert it.

Or just leave it in and show verify false

[vezenovm]

Just a small nit

[guipublic]

It would be good to see a commit which breaks the blocks test case (e.g. use a witness which doesn't satisfy the block constraints) to show that this test is meaningful and then revert it.

Or just leave it in and show verify false

I have added a failing test