feat: WIP start splitting 0.6 and 0.7 logic #634

dancoombs · 2024-03-06T19:28:22Z

THIS IS BEING REPLACED BY: #635

dancoombs · 2024-03-06T19:42:57Z

crates/builder/src/bundle_proposer.rs

@@ -99,7 +98,7 @@ where
    builder_index: u64,
    pool: C,
    simulator: S,
-    entry_point: E,
+    entry_point: Arc<E>,


TODO: figure this out, we should be consistent here, we don't need these Arc as everything inside the entry point class should be clone

dancoombs · 2024-03-06T19:43:15Z

crates/builder/src/bundle_proposer.rs

    SimulationResult, SimulationViolation, Simulator, ViolationError,
 };
 use rundler_types::{
-    chain::ChainSpec, Entity, EntityType, EntityUpdate, EntityUpdateType, GasFees, Timestamp,
-    UserOperation, UserOpsPerAggregator,
+    chain::ChainSpec, Entity, EntityType, EntityUpdate, EntityUpdateType, GasFees, GasOverheads,


TODO: GasOverheads should move to chain config

dancoombs · 2024-03-06T19:45:59Z

crates/builder/src/bundle_proposer.rs

+        + op.total_verification_gas_limit()
+        + op.required_pre_execution_buffer()
+        + op.call_gas_limit()


TODO: verify this change

dancoombs · 2024-03-06T19:47:10Z

crates/builder/src/task.rs

@@ -311,7 +311,7 @@ where
            index,
            self.pool.clone(),
            simulator,
-            ep.clone(),
+            Arc::new(ep.clone()),


TODO: bundle proposer currently will only handle v0.6 UOs. Figure out how to share logic

dancoombs · 2024-03-06T19:48:19Z

crates/pool/src/server/remote/error.rs

@@ -244,13 +244,6 @@ impl From<MempoolError> for ProtoMempoolError {
 impl From<PrecheckViolation> for ProtoPrecheckViolationError {
    fn from(value: PrecheckViolation) -> Self {
        match value {
-            PrecheckViolation::InitCodeTooShort(length) => ProtoPrecheckViolationError {


TODO: removed these, should add check in RPC before sending to pool.

dancoombs · 2024-03-06T19:48:34Z

crates/pool/src/server/remote/protos.rs

+        match op {
+            RundlerUserOperation::V0_6(op) => op.into(),
+            RundlerUserOperation::V0_7(_) => {
+                unimplemented!("V0_7 user operation is not supported")


TODO: implement

dancoombs · 2024-03-06T19:49:00Z

crates/provider/src/ethers/entry_point_v0_6.rs

+const ARBITRUM_NITRO_NODE_INTERFACE_ADDRESS: Address = H160([
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xc8,
+]);
+
+const OPTIMISM_BEDROCK_GAS_ORACLE_ADDRESS: Address = H160([
+    0x42, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x0F,
+]);


TODO: we should be getting this from ChainSpec

dancoombs · 2024-03-06T19:50:28Z

crates/provider/src/ethers/entry_point_v0_6.rs

+// TODO(danc):
+// - Modify this interface to take only `UserOperationV0_6` and remove the `into_v0_6` calls
+// - Places that are already in a V0_6 context can use this directly
+// - Implement a wrapper that takes `UserOperation` and dispatches to the correct entry point version
+//   based on a constructor from ChainSpec.
+// - Use either version depending on the abstraction of the caller.


dancoombs · 2024-03-06T19:51:03Z

crates/provider/src/ethers/entry_point_v0_6.rs

@@ -240,14 +281,119 @@ where
            .context("should compute balances")?;
        Ok(out.balances)
    }
+
+    async fn aggregate_signatures(


Moved these from provider as they are entry point dependent

dancoombs · 2024-03-06T19:52:03Z

crates/provider/src/traits/entry_point.rs

@@ -98,8 +120,38 @@ pub trait EntryPoint: Send + Sync + 'static {
    ) -> Result<ExecutionResult, String>;

    /// Get the deposit info for an address
-    async fn get_deposit_info(&self, address: Address) -> anyhow::Result<DepositInfo>;
+    async fn get_deposit_info(&self, address: Address) -> anyhow::Result<DepositInfoV0_6>;


TODO: this will need to be made generic, can't rely on the 0_6 version.

Similar to ExecutionResult

dancoombs · 2024-03-06T19:53:24Z

crates/rpc/src/debug.rs

@@ -126,7 +127,7 @@ where
            .await
            .map_err(|e| rpc_err(INTERNAL_ERROR_CODE, e.to_string()))?
            .into_iter()
-            .map(|pop| pop.uo.into())
+            .map(|pop| UserOperationV0_6::from(pop.uo).into())


TODO: Should be able to map these to new RPC types that can handle both formats.

dancoombs · 2024-03-06T19:54:06Z

crates/rpc/src/eth/api.rs

@@ -64,7 +64,7 @@ impl Settings {

 #[derive(Debug)]
 struct EntryPointContext<P, E> {
-    gas_estimator: GasEstimatorImpl<P, E>,
+    gas_estimator: GasEstimatorV0_6<P, E>,


TODO: need to handle multiple types of gas estimators here, based on the entry point version. Likely easiest to just use dyn on the trait.

dancoombs · 2024-03-06T19:55:23Z

crates/sim/src/estimation/mod.rs

+/// Gas estimate for a user operation
+#[derive(Debug, Copy, Clone, Deserialize, Serialize)]
+#[serde(rename_all = "camelCase")]
+pub struct GasEstimate {


TODO: This won't work for 0.7 estimation as we need to return more fields. Rethink this interface

0xfourzerofour · 2024-03-06T19:52:08Z

crates/provider/src/ethers/entry_point_v0_6.rs

@@ -87,6 +116,10 @@ where
        user_op: UserOperation,
        max_validation_gas: u64,
    ) -> anyhow::Result<TypedTransaction> {
+        let user_op = user_op
+            .into_v0_6()


I wonder if there is a way we could use some generic magic on the function and implement Into for the different types of user operations. I have not thought about implementation and just spitballing but surely we would be able to use a UO generic an call into() instead of having multiple implementations

Just skimming here so dont have full context yet

We will need multiple implementations regardless because this pulls in the contract interface which is different for v0.6 and v0.7

dancoombs · 2024-03-06T19:56:34Z

crates/sim/src/gas/gas.rs

+        + uo.total_verification_gas_limit()
+        + uo.required_pre_execution_buffer()
+        + uo.call_gas_limit()


TODO: verify the logic changes in this file

dancoombs · 2024-03-06T19:57:13Z

crates/sim/src/precheck.rs

-    /// The init code is too short to contain a factory address.
-    #[display("initCode must start with a 20-byte factory address, but was only {0} bytes")]
-    InitCodeTooShort(usize),


TODO: check these in RPC now.

dancoombs added 2 commits March 5, 2024 18:24

feat(types): add entrypoint 0.7 contract types

eef9d76

feat: WIP start splitting 0.6 and 0.7 logic

7055dcd

dancoombs commented Mar 6, 2024

View reviewed changes

0xfourzerofour reviewed Mar 6, 2024

View reviewed changes

dancoombs commented Mar 6, 2024

View reviewed changes

dancoombs force-pushed the danc/contracts-v0_7 branch 2 times, most recently from 6c4b2f4 to f5d2c6e Compare March 29, 2024 21:12

dancoombs closed this Mar 29, 2024

dancoombs deleted the danc/uo-enum branch November 24, 2024 16:42

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: WIP start splitting 0.6 and 0.7 logic #634

feat: WIP start splitting 0.6 and 0.7 logic #634

dancoombs commented Mar 6, 2024 •

edited

Loading

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

0xfourzerofour Mar 6, 2024

0xfourzerofour Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

dancoombs Mar 6, 2024

feat: WIP start splitting 0.6 and 0.7 logic #634

feat: WIP start splitting 0.6 and 0.7 logic #634

Conversation

dancoombs commented Mar 6, 2024 • edited Loading

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

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

dancoombs commented Mar 6, 2024 •

edited

Loading