Staking Precompiles #358

JoshOrndorff · 2021-04-15T19:49:17Z

This PR adds a precompile to wrap the parachain staking functionality in the EVM. All staking functionality is in a single precompile. This precompile distinguishes what operation to perform by matching a four-byte "function selector" that conforms to solidity conventions. A corresponding solidity interface is provided to enable seamless interaction with users' contracts.

Basically we want to let EVM accounts, including contracts, interact with staking as first-class citizens.

Follow Ups

All follow ups are tracked in Jira Moonbeam-495

used_addresses to share between runtime and node instead of duplicating that code
put NominationDAO.sol in some other demo repo

Checklist

❌ Does it require a purge of the network?
You bumped the runtime version if there are breaking changes in the runtime ?
✔️ Does it require changes in documentation/tutorials ?

…test

…ceed

…spatch not found

runtime/precompiles/src/lib.rs

4meta5

LGTM @JoshOrndorff let me know if there is anything else you'd like to be done and, if not, we can merge

JoshOrndorff

Just a couple more little things. Please also update the description with links to the followup tickets.

runtime/Cargo.toml

runtime/precompiles/src/JoinCandidatesWrapper.sol

runtime/precompiles/src/NominationDAO.sol

JoshOrndorff · 2021-04-29T16:01:06Z

runtime/precompiles/src/lib.rs

+	#[allow(dead_code)]
+	fn used_addresses() -> impl Iterator<Item = R::AccountId> {
+		sp_std::vec![1, 2, 3, 4, 5, 6, 7, 8, 255, 256, 511]
+			.into_iter()
+			.map(|x| hash(x).into())
+	}


Just make sure there is a followup issue for this. Then I'm fine leaving it as it is.

These addresses have been updated in master, by the way. We pushed some out to 1024 and I (somewhat arbitrarily) placed sacrifice at 4095.

runtime/precompiles/src/lib.rs

runtime/precompiles/src/staking.rs

tests/contracts/sources.ts

node/src/chain_spec.rs

runtime/precompiles/src/JoinCandidatesWrapper.sol

runtime/precompiles/src/NominationDAO.sol

runtime/precompiles/src/staking.rs

notlesh

This is looking awesome. I have a bunch of nits -- nothing big. I definitely think the logging should be turned down or removed...

node/src/chain_spec.rs

notlesh · 2021-04-29T17:03:32Z

node/src/chain_spec.rs

+			// TODO Cleanly fetch the addresses from
+			// the runtime/moonbeam precompiles and systematically fill them with code
+			// that will revert if it is called by accident (it shouldn't be because
+			// it is shadowed by the precompile).


I'm not quite following this TODO... is it still relevant / should it be done before merging?

No, we'll do it in a follow-up.

I think a followup makes more sense. There is a design decision to be made. Basically the dilemna is this.

We need to get bogus code under every precompile address. The precompile addresses are codified in the precompiles crate, but the genesis config (which includes the pre-deployed bogus code) is set up here.

So the decision is whether we have the list of addresses in both places (as we do now), or do we call into the precompiles crate to get them somehow. I started sketching that also, but still didn't like it because even the way I sketched it the addresses were still in two places (although at least they were in the same file).

notlesh · 2021-04-29T17:18:17Z

runtime/precompiles/src/StakingInterface.sol

+    // Now the dispatchables
+
+    /// Join the set of collator candidates
+    function join_candidates(uint256 amount) external;


It might be nice to document what this amount is (is it 10^-18 units?)

It looks like it's used below as well, so maybe a comment at the top of the contract...

A better variable name would go a long way. maybe bond_amount. Good news the names don't affect the signatures so we wouldn't have to regenerate them. Agree about the units. Luckily solidity makes this nice with keywords like ether, wei, etc.

notlesh · 2021-04-29T17:20:42Z

runtime/precompiles/src/StakingInterface.sol

+// 	"42a72461": "at_stake(uint256,address)",
+// 	"289b6ba7": "candidate_bond_less(uint256)",
+// 	"c57bd3a8": "candidate_bond_more(uint256)",
+// 	"96b41b5b": "candidate_pool()",
+// 	"b0c0081f": "collator_commission()",
+// 	"427c3c14": "exit_queue()",
+// 	"767e0450": "go_offline()",
+// 	"d2f73ceb": "go_online()",
+// 	"10db2de9": "inflation_config()",
+// 	"8545c833": "is_candidate(address)",
+// 	"8e5080e7": "is_nominator(address)",
+// 	"8f6d27c7": "is_selected_candidate(address)",
+// 	"ad76ed5a": "join_candidates(uint256)",
+// 	"b7694219": "leave_candidates()",
+// 	"e8d68a37": "leave_nominators()",
+// 	"82f2c8df": "nominate(address,uint256)",
+// 	"f6a52569": "nominator_bond_less(address,uint256)",
+// 	"971d44c8": "nominator_bond_more(address,uint256)",
+// 	"9799b4e7": "points(uint256)",
+// 	"4b65c34b": "revoke_nomination(address)",
+// 	"98807d84": "staked(address)",
+// 	"7b46e61b": "total_selected()"


Is the goal to have 1:1 wrapper functions for everything exposed from the pallet?

I'm resisting the urge to suggest we could generate this. Oops...

runtime/precompiles/src/staking.rs

notlesh · 2021-04-29T17:29:08Z

runtime/precompiles/src/staking.rs

+
+		//TODO figure out how much gas it costs to check whether you're a collator candidate.
+		// That function will not naturally be benchmarked because it is not dispatchable
+		// I guess the heavy part will be one storage read.


It could be benchmarked -- benchmarking does support non-dispatchables. If we use that strategy to calculate gas, we'll also want to respect weight -> gas mapping here.

We would only need benchmarking for weight. Then we'll use the weight-gas-mapping at runtime.

notlesh · 2021-04-29T17:33:06Z

tests/util/transactions.ts

@@ -98,6 +98,7 @@ export async function createContract(
  return {
    rawTx,
    contract,
+    contractAddress,


I think you can get the contract address from the contract object:

contract.options.address

@joelamouche This came from your integration test. I would like to get that test passing. I think it should be passing. Maybe you can take a look together with @4meta5 ?

I put that test in jira 495 because it wasn't passing and I was trying to reduce the scope to get this PR to a merge-able state. If you need the test in this PR, feel free to add it back and go for it.

JoshOrndorff · 2021-04-29T20:09:47Z

runtime/precompiles/src/StakingInterface.sol

-// 	"96b41b5b": "candidate_pool()",
-// 	"b0c0081f": "collator_commission()",
-// 	"427c3c14": "exit_queue()",


We shouldn't be removing these. If they aren't implemented, we should have followups. Actually having them in the code is more important than here. The code is what is used to generate these signatures.

We shouldn't be removing these. Actually having them in the code is more important than here. The code is what is used to generate these signatures.

Why do we want to include them if they're not implemented? It is easy to add them back and generate the right signatures. Moreover, removing the dead code does not effect the other signatures because each signature is independently generated.

If they aren't implemented, we should have followups.

All the followups are tracked in jira 495

runtime/tests/integration_test.rs

JoshOrndorff

Looks good to me. I can't approve since I opened this PR :)

Thanks for getting the integration test working.

albertov19 · 2021-05-03T13:13:43Z

In terms of documentation I think we can do a Page under the Staking tab. We'll get the ball rolling once we can allocate some resources to this.

4meta5 and others added 14 commits April 8, 2021 11:12

init

832726b

Merge branch 'master' into amar-int-tests

f179e91

save

737c0fd

Initial pass at Sacrifice precompile

cf4e42d

Add gas-related tests for Sacrifice precompile

9efad64

import pallet instances from runtime and start evm transfer to stake …

69a32b7

…test

transfer to stake without evm transfer

71d5660

rename carl to charlie

1f10d75

push failed test to show how evm call works

97fc054

set gas price and limit to make test pass

688a324

test still fails, push to show that transfer through EVM does not suc…

b3abf42

…ceed

test passes

f154f50

save unfinished pay block authors todo

270ea2d

Sketched idea, mostly right (I think), doesn't compile because of .di…

ad59685

…spatch not found

JoshOrndorff commented Apr 15, 2021

View reviewed changes

runtime/precompiles/src/lib.rs Outdated Show resolved Hide resolved

JoshOrndorff added 5 commits April 15, 2021 20:16

holy shit it compiles!

0d2a184

ditch BigUint

a539aaa

don't need to name phantom field

1e5f8d1

BalanceOf<Runtime>

28a7e50

Don't need temp buffers for inputs

ab7fea5

notlesh reviewed Apr 15, 2021

View reviewed changes

runtime/precompiles/src/lib.rs Outdated Show resolved Hide resolved

no_std

026d4c9

JoshOrndorff commented Apr 15, 2021

View reviewed changes

runtime/precompiles/src/lib.rs Outdated Show resolved Hide resolved

4meta5 and others added 7 commits April 16, 2021 15:22

reward block authors runtime integration test

fbb9852

improve readability

15aae62

increase test inflation config to see test balances go up more

a46d404

move to own file, and rename for pallet-specific precompile

106beec

Try (and fail) to install in Runtime

17ebf9b

Add proper trait bounds so it compiles

93b197c

Better constant use

4d8838e

4meta5 added 4 commits April 28, 2021 17:33

green

b8842b2

Merge branch 'master' into joshy-staking-precompiles

ec9ca76

add is candidate with rust runtime integration test

dd33e71

clean and bump spec version

f2d6443

4meta5 approved these changes Apr 29, 2021

View reviewed changes

4meta5 requested review from crystalin and notlesh April 29, 2021 05:04

JoshOrndorff commented Apr 29, 2021

View reviewed changes

girazoki reviewed Apr 29, 2021

View reviewed changes

node/src/chain_spec.rs Outdated Show resolved Hide resolved

runtime/precompiles/src/JoinCandidatesWrapper.sol Outdated Show resolved Hide resolved

runtime/precompiles/src/NominationDAO.sol Outdated Show resolved Hide resolved

girazoki reviewed Apr 29, 2021

View reviewed changes

runtime/precompiles/src/staking.rs Outdated Show resolved Hide resolved

runtime/precompiles/src/staking.rs Outdated Show resolved Hide resolved

girazoki reviewed Apr 29, 2021

View reviewed changes

runtime/precompiles/src/staking.rs Outdated Show resolved Hide resolved

notlesh suggested changes Apr 29, 2021

View reviewed changes

address some review comments

d116efb

JoshOrndorff commented Apr 29, 2021

View reviewed changes

4meta5 removed their assignment Apr 29, 2021

4meta5 added 2 commits April 29, 2021 23:53

min nomination test commented out but fails

97c9436

add some comments requested in review

cb0addd

4meta5 reviewed Apr 30, 2021

View reviewed changes

runtime/tests/integration_test.rs Outdated Show resolved Hide resolved

4meta5 added 2 commits April 30, 2021 11:26

fix min nomination test as per joshy comment

9557521

fix precompile addresses in dead code

b8f3122

JoshOrndorff commented May 3, 2021

View reviewed changes

JoshOrndorff requested review from notlesh and girazoki May 3, 2021 12:46

notlesh approved these changes May 3, 2021

View reviewed changes

4meta5 merged commit e73891f into master May 3, 2021

4meta5 deleted the joshy-staking-precompiles branch May 3, 2021 14:06

albertov19 mentioned this pull request May 10, 2021

Build a Polkadot Staking dApp with Moonbeam moonbeam-foundation/hackathon-0x-moonbeam#2

Open

JoshOrndorff mentioned this pull request May 12, 2021

Add pre-deploy code under precompile contracts in the spec templates #409

Merged

3 tasks

JoshOrndorff mentioned this pull request Jun 15, 2021

Expand and clean staking precompiles #511

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Staking Precompiles #358

Staking Precompiles #358

JoshOrndorff commented Apr 15, 2021 •

edited by 4meta5

Loading

4meta5 left a comment

JoshOrndorff left a comment

JoshOrndorff Apr 29, 2021

notlesh Apr 29, 2021

notlesh left a comment

notlesh Apr 29, 2021

4meta5 Apr 29, 2021

JoshOrndorff Apr 29, 2021

notlesh Apr 29, 2021

notlesh Apr 29, 2021

JoshOrndorff Apr 29, 2021

notlesh Apr 29, 2021

notlesh Apr 29, 2021

JoshOrndorff Apr 30, 2021

notlesh Apr 29, 2021

JoshOrndorff Apr 29, 2021

4meta5 Apr 30, 2021

JoshOrndorff Apr 29, 2021

4meta5 Apr 29, 2021 •

edited

Loading

JoshOrndorff left a comment

albertov19 commented May 3, 2021

Staking Precompiles #358

Staking Precompiles #358

Conversation

JoshOrndorff commented Apr 15, 2021 • edited by 4meta5 Loading

Follow Ups

Checklist

4meta5 left a comment

Choose a reason for hiding this comment

JoshOrndorff left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

notlesh left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

4meta5 Apr 29, 2021 • edited Loading

Choose a reason for hiding this comment

JoshOrndorff left a comment

Choose a reason for hiding this comment

albertov19 commented May 3, 2021

JoshOrndorff commented Apr 15, 2021 •

edited by 4meta5

Loading

4meta5 Apr 29, 2021 •

edited

Loading