feat: merge-train/avm

yarn-project/simulator/README.md

@@ -12,7 +12,7 @@ It's able to simulate three different types of functions:
 
 Private functions are simulated and proved client-side, and verified client-side in the private kernel circuit.
 
-The public inputs of private functions is defined [here](../stdlib/src/structs/private_circuit_public_inputs.ts).
+The public inputs of private functions are defined [here](../stdlib/src/structs/private_circuit_public_inputs.ts).
 
 They are run with the assistance of a DB oracle that provides any private data requested by the function.
 
@@ -22,17 +22,17 @@ Private functions can call another private function, and can request to call a p
 
 Public functions are simulated and proved on the sequencer side, and verified by the public kernel circuit.
 
-The public inputs of public functions is defined [here](../stdlib/src/structs/avm/avm_circuit_public_inputs.ts).
+The public inputs of public functions are defined [here](../stdlib/src/structs/avm/avm_circuit_public_inputs.ts).
 
 They are run with the assistance of an oracle that provides any value read from the public state tree.
 
-Public functions can call other public function, but no private functions.
+Public functions can call other public functions, but cannot call private functions.
 
 See the specifications of the [Aztec Virtual Machine (AVM) for public execution](./docs/avm/index.md).
 
 ### Unconstrained Functions
 
-Unconstrained functions are useful to extract useful data for users that could produce very large execution traces - such as the summed balance of all a users notes
+Unconstrained functions are useful to extract data for users that could produce very large execution traces - such as the summed balance of all of a user's notes.
 They are not proved, and are simulated client-side.
 
 They are run with the assistance of a DB oracle that provides any private data requested by the function.

yarn-project/simulator/docs/avm/README.md

@@ -39,7 +39,7 @@ public execution requests.
 The AVM:
 * Executes specified public bytecode, instruction by instruction, given some arguments.
 * Meters execution by tracking gas costs per-executed-instruction.
-* Tracks both "mana" (aka L2 gas) and "data availability" gas.
+* Tracks both L2 gas (computation) and DA gas (data availability).
 * Supports nested contract calls and conditional error recovery.
 * Manages access to public state, L1↔L2 messages, public logs, and some limited private state.
 * Finalizes state updates initiated during private execution.

yarn-project/simulator/docs/avm/gas.md

@@ -1,10 +1,10 @@
 # Gas Metering
 
-The AVM tracks gas consumption across two dimensions: **L2 gas** (execution costs, also called **mana**) and **DA gas** (data availability costs).
+The AVM tracks gas consumption across two dimensions: **L2 gas** (execution costs) and **DA gas** (data availability costs). Note that L2 gas is _not_ the same as mana: mana is the higher-level unit of account that incorporates L2 gas, DA gas, and L1 gas costs together.
 
 ## Gas Dimensions
 
-* **L2 Gas** (mana): roughly represents the computational cost of executing (and especially proving) operations.
+* **L2 Gas**: roughly represents the computational cost of executing (and especially proving) operations.
 * **DA Gas**: represents the cost of publishing data to Layer 1 for data availability. This includes:
   - State updates that must be published to L1
   - Emitting logs