Well, something is broken in the test suite it looks like...

Is the block_gas_limit value being respected? Looking through the Frontier code, it seems like there’s a check to see if a single transaction exceeds the block_gas_limit, but I don’t see any part that verifies whether the total gas consumption of the transactions in a block is less than the block_gas_limit.

Yes, this is also a problem

@MOZGIII

According to the next:

Each transaction receipt stores already cumulative_gas_used instead of gas_used. If this behavior is intended, the original code is written correctly. Unlike the return type of eth_getTransactionReceipt RPC call contain both cumulative_gas_used and gas_used, Receipt has only one gas_used field. I need to look into further what Ethereum protocol specifies.

See #602

Right, but then, should we rename the struct field accordingly to avoid further confusion?

Is the block_gas_limit value being respected? Looking through the Frontier code, it seems like there’s a check to see if a single transaction exceeds the block_gas_limit, but I don’t see any part that verifies whether the total gas consumption of the transactions in a block is less than the block_gas_limit.

No check for the block_gas_limit; the block_gas_limit is a fixed value derived from the block_max_weight. In fact, we convert the transaction gas to the transaction weight and ensure all transaction weights are less than the block max weight.