Gas Optimizations

Total: 17 contexts over 8 issues

Gas Optimizations

The IR-based code generator was introduced with an aim to not only allow code generation to be more transparent and auditable but also to enable more powerful optimization passes that span across functions.

You can enable it on the command line using --via-ir or with the option {"viaIR": true}.

This will take longer to compile, but you can just simple test it before deploying and if you got a better benchmark then you can add --via-ir to your deploy command

We can use assembly to emit events efficiently by utilizing scratch space and the free memory pointer. This will allow us to potentially avoid memory expansion costs.
Note: In order to do this optimization safely, we will need to cache and restore the free memory pointer.

For example, for a generic emit event for eventSentAmountExample:

We can use the following assembly emit events:

Proof Of Concept

Saves ~13 gas per instance

Proof Of Concept

If a function modifier or require such as onlyOwner/onlyX is used, the function will revert if a normal user tries to pay the function. Marking the function as payable will lower the gas cost for legitimate callers because the compiler will not include checks for whether a payment was provided. The extra opcodes avoided are CALLVALUE(2), DUP1(3), ISZERO(3), PUSH2(3), JUMPI(10), PUSH1(3), DUP1(3), REVERT(0), JUMPDEST(1), POP(2) which costs an average of about 21 gas per call to the function, in addition to the extra deployment cost.

Proof Of Concept

Contracts most called functions could simply save gas by function ordering via Method ID. Calling a function at runtime will be cheaper if the function is positioned earlier in the order (has a relatively lower Method ID) because 22 gas are added to the cost of a function for every position that came before it. The caller can save on gas if you prioritize most called functions.

Given 4 variables a, b, c and d represented as such:

0 0 0 0 0 1 1 0 <- a
0 1 1 0 0 1 1 0 <- b
0 0 0 0 0 0 0 0 <- c
1 1 1 1 1 1 1 1 <- d

To have a == b means that every 0 and 1 match on both variables. Meaning that a XOR (operator ^) would evaluate to 0 ((a ^ b) == 0), as it excludes by definition any equalities.Now, if a != b, this means that there’s at least somewhere a 1 and a 0 not matching between a and b, making (a ^ b) != 0.Both formulas are logically equivalent and using the XOR bitwise operator costs actually the same amount of gas.However, it is much cheaper to use the bitwise OR operator (|) than comparing the truthy or falsy values.These are logically equivalent too, as the OR bitwise operator (|) would result in a 1 somewhere if any value is not 0 between the XOR (^) statements, meaning if any XOR (^) statement verifies that its arguments are different.

Proof Of Concept