validate EL block hash when running consensus block tests

[etan-status]

We currently don't have an easy way to test EL block hash computation. As the EL block hash in consensus-spec-tests is computed correctly, update the test runners that load block from test files to also verify the EL block hash. This increases missing test coverage.

Requires ethereum/consensus-specs#3829

[github-actions]

Unit Test Results

0 tests   0 ✔️  0s ⏱️
0 suites  0 💤
0 files    0 ❌

Results for commit 5acdbf8.

♻️ This comment has been updated with latest results.

[etan-status]

Cross-checked against jtraglia/consensus-specs#2 (with #6718)

[etan-status]

The hash is incorrect when there are empty txes, see status-im/nim-eth#744 (comment)

As such blocks are invalid anyway as of ethereum/consensus-specs#3885, have excluded that condition from block hash consistency check.

[etan-status]

alpha.9 --> used to work but new tests got added after the latest fixes got merged.

• Fix block hash computation for fork transition tests ethereum/consensus-specs#4022

[tersec]

Still in alpha.10:

EF - Electra - Finality  [Preset: minimal] ..... (0.42s)
Invalid `block_hash`: /github-runner/github-runner-node-01/workspace/nimbus-eth2/nimbus-eth2/vendor/nim-eth2-scenarios/tests-v1.5.0-alpha.10/minimal/electra/random/random/pyspec_tests/randomized_6/blocks_0.ssz_snappy
EF - Electra - Random  [Preset: minimal]  (0.00s)

[etan-status]

Yes, they keep adding breakages after the fixes :)

[etan-status]

• Emit correct block hash in random Electra tests ethereum/consensus-specs#4072

[etan-status]

Checked it against alpha.10 + the latest fix PR and nimbus passes with that. As long as no new incorrect block hashes get added by other PRs it should once more be alright (same as we hoped for during the earlier alphas which all worked against nimbus at a time).