status-im
/
eth2.0-specs
mirror of https://github.com/status-im/eth2.0-specs.git
2
0
Fork 0
Code Issues Projects Releases Wiki Activity

hww feedback for finality rewards fix

This commit is contained in:
Danny Ryan 2020-05-20 10:11:47 -06:00
parent 95c3295eeb
commit 943e51aef1
No known key found for this signature in database
GPG Key ID: 2765A792E42CE07A
3 changed files with 51 additions and 77 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
specs/phase0/beacon-chain.md
View File

@ -1368,7 +1368,7 @@ def get_finality_delay(state: BeaconState) -> uint64:
```python
def in_inactivity_leak(state: BeaconState) -> bool:
def is_in_inactivity_leak(state: BeaconState) -> bool:
return get_finality_delay(state) > MIN_EPOCHS_TO_INACTIVITY_PENALTY
```
@ -1397,8 +1397,9 @@ def get_attestation_component_deltas(state: BeaconState,
for index in get_eligible_validator_indices(state):
if index in unslashed_attesting_indices:
increment = EFFECTIVE_BALANCE_INCREMENT # Factored out from balance totals to avoid uint64 overflow
if in_inactivity_leak(state):
# Full base reward will be cancelled out by inactivity penalty deltas
if is_in_inactivity_leak(state):
# Since full base reward will be canceled out by inactivity penalty deltas,
# optimal participation receives full base reward compensation here.
rewards[index] += get_base_reward(state, index)
else:
reward_numerator = get_base_reward(state, index) * (attesting_balance // increment)
@ -1464,7 +1465,7 @@ def get_inactivity_penalty_deltas(state: BeaconState) -> Tuple[Sequence[Gwei], S
Return inactivity reward/penalty deltas for each validator.
"""
penalties = [Gwei(0) for _ in range(len(state.validators))]
if in_inactivity_leak(state):
if is_in_inactivity_leak(state):
matching_target_attestations = get_matching_target_attestations(state, get_previous_epoch(state))
matching_target_attesting_indices = get_unslashed_attesting_indices(state, matching_target_attestations)
for index in get_eligible_validator_indices(state):

3
tests/core/pyspec/eth2spec/test/helpers/rewards.py
View File

@ -151,7 +151,6 @@ def run_get_inactivity_penalty_deltas(spec, state):
matching_attestations = spec.get_matching_target_attestations(state, spec.get_previous_epoch(state))
matching_attesting_indices = spec.get_unslashed_attesting_indices(state, matching_attestations)
finality_delay = spec.get_previous_epoch(state) - state.finalized_checkpoint.epoch
eligible_indices = spec.get_eligible_validator_indices(state)
for index in range(len(state.validators)):
assert rewards[index] == 0
@ -159,7 +158,7 @@ def run_get_inactivity_penalty_deltas(spec, state):
assert penalties[index] == 0
continue
if finality_delay > spec.MIN_EPOCHS_TO_INACTIVITY_PENALTY:
if spec.is_in_inactivity_leak(state):
base_reward = spec.get_base_reward(state, index)
base_penalty = spec.BASE_REWARDS_PER_EPOCH * base_reward - spec.get_proposer_reward(state, index)
if not has_enough_for_reward(spec, state, index):

116
tests/core/pyspec/eth2spec/test/phase_0/epoch_processing/test_process_rewards_and_penalties.py
View File

@ -63,74 +63,6 @@ def test_genesis_epoch_full_attestations_no_rewards(spec, state):
assert state.balances[index] == pre_state.balances[index]
@with_all_phases
@spec_state_test
def test_full_attestations(spec, state):
attestations = prepare_state_with_attestations(spec, state)
pre_state = state.copy()
yield from run_process_rewards_and_penalties(spec, state)
attesting_indices = spec.get_unslashed_attesting_indices(state, attestations)
assert len(attesting_indices) == len(pre_state.validators)
for index in range(len(pre_state.validators)):
if index in attesting_indices:
assert state.balances[index] > pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
@with_all_phases
@spec_state_test
@leaking()
def test_full_attestations_with_leak(spec, state):
attestations = prepare_state_with_attestations(spec, state)
proposer_indices = [a.proposer_index for a in state.previous_epoch_attestations]
pre_state = state.copy()
yield from run_process_rewards_and_penalties(spec, state)
attesting_indices = spec.get_unslashed_attesting_indices(state, attestations)
assert len(attesting_indices) == len(pre_state.validators)
for index in range(len(pre_state.validators)):
# Proposers can still make money during a leak
if index in proposer_indices:
assert state.balances[index] > pre_state.balances[index]
# If not proposer but participated optimally, should have exactly neutral balance
elif index in attesting_indices:
assert state.balances[index] == pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
@with_all_phases
@spec_state_test
@leaking()
def test_partial_attestations_with_leak(spec, state):
attestations = prepare_state_with_attestations(spec, state)
attestations = attestations[:len(attestations) // 2]
state.previous_epoch_attestations = state.previous_epoch_attestations[:len(attestations)]
proposer_indices = [a.proposer_index for a in state.previous_epoch_attestations]
pre_state = state.copy()
yield from run_process_rewards_and_penalties(spec, state)
attesting_indices = spec.get_unslashed_attesting_indices(state, attestations)
assert len(attesting_indices) < len(pre_state.validators)
for index in range(len(pre_state.validators)):
# Proposers can still make money during a leak
if index in proposer_indices and index in attesting_indices:
assert state.balances[index] > pre_state.balances[index]
# If not proposer but participated optimally, should have exactly neutral balance
elif index in attesting_indices:
assert state.balances[index] == pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
@with_all_phases
@spec_state_test
def test_full_attestations_random_incorrect_fields(spec, state):
@ -224,6 +156,7 @@ def run_with_participation(spec, state, participation_fn):
return att_participants
attestations = prepare_state_with_attestations(spec, state, participation_fn=participation_tracker)
proposer_indices = [a.proposer_index for a in state.previous_epoch_attestations]
pre_state = state.copy()
@ -233,10 +166,20 @@ def run_with_participation(spec, state, participation_fn):
assert len(attesting_indices) == len(participated)
for index in range(len(pre_state.validators)):
if index in participated:
assert state.balances[index] > pre_state.balances[index]
if spec.is_in_inactivity_leak(state):
# Proposers can still make money during a leak
if index in proposer_indices and index in participated:
assert state.balances[index] > pre_state.balances[index]
# If not proposer but participated optimally, should have exactly neutral balance
elif index in attesting_indices:
assert state.balances[index] == pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
if index in participated:
assert state.balances[index] > pre_state.balances[index]
else:
assert state.balances[index] < pre_state.balances[index]
@with_all_phases
@ -246,6 +189,14 @@ def test_almost_empty_attestations(spec, state):
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, 1))
@with_all_phases
@spec_state_test
@leaking()
def test_almost_empty_attestations_with_leak(spec, state):
rng = Random(1234)
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, 1))
@with_all_phases
@spec_state_test
def test_random_fill_attestations(spec, state):
@ -253,6 +204,14 @@ def test_random_fill_attestations(spec, state):
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, len(comm) // 3))
@with_all_phases
@spec_state_test
@leaking()
def test_random_fill_attestations_with_leak(spec, state):
rng = Random(4567)
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, len(comm) // 3))
@with_all_phases
@spec_state_test
def test_almost_full_attestations(spec, state):
@ -260,12 +219,27 @@ def test_almost_full_attestations(spec, state):
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, len(comm) - 1))
@with_all_phases
@spec_state_test
@leaking()
def test_almost_full_attestations_with_leak(spec, state):
rng = Random(8901)
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: rng.sample(comm, len(comm) - 1))
@with_all_phases
@spec_state_test
def test_full_attestation_participation(spec, state):
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: comm)
@with_all_phases
@spec_state_test
@leaking()
def test_full_attestation_participation_with_leak(spec, state):
yield from run_with_participation(spec, state, lambda slot, comm_index, comm: comm)
@with_all_phases
@spec_state_test
def test_duplicate_attestation(spec, state):