Description
I think we should ensure that the withdrawal indices here are monotonically increasing, or the test may not be valid (we define two withdrawals with index=0
):
execution-spec-tests/tests/shanghai/eip4895_withdrawals/test_withdrawals.py
Lines 611 to 623 in 51d30bd
According to geth's Withdrawal struct implementation, the index is a
monotonically increasing identifier issued by consensus layer
Should a block be able to contain two withdrawals with the same index? This test also performs multiple withdrawals from the same validator index to multiple addresses, which is an unusual test case in itself (there is a 1-1 mapping between validator and withdrawal index and this can not be changed). Surely it's good to test this, but perhaps we should additionally test with more typical values, as found on the beacon chain?