Search Results

Take this code snippet: function powOfTen(uint256 x) internal pure returns (int256 result) { assembly { switch x case 1 { result := -18 } } } Which doesn't compile: ParserErro …

An alternative to @alberto's solution is to use the sub instruction: function powOfTen(uint256 x) internal pure returns (int256 result) { assembly { switch x case 1 { result := sub …

Take the following code snippet: pragma solidity ^0.8.0; contract MyContract { function foo() public returns (uint256 result) { // do unsafe work } function bar() public returns …

Take the following function: solidity ^0.8.0; function foo(uint256 x) external pure returns (uint256) { uint256 y = x << 255; return y; } If you set x = 2, the result is 2**256, which is abo …

Say you have a contract defined like this: pragam solidity >=0.8.0; contract Foo { uint256 public constant MY_CONSTANT = 6174; } If I just import Foo in another contract, without inheriting from …

I have an array which is made up of user-defined value types: type UD60x18 is uint256; UD60x18 memory foo[] = new UD60x18[](2); Now, I would like to turn this into uint256. Can I cast this array to u …

We know that: Since Solidity v0.8, the compiler has checks for {over,under}flow by default for all arithmetic operations. Division is an arithmetic operation that rounds down to the integer part of t …

Yes, there is at least one situation where overflow can occur. I don't know about underflows though. When you divide the minimum of a signed type by -1, you get the mirror image of that number in the …

I would like to define a function like this: function convert(ufixed x) external pure returns (uint256 result) { result = x * 1e18; } Unfortunately, the code above does not compile: TypeError: t …

There's now a nicer way to do this in Solidity v0.8.19, thanks to the addition of user-defined operators. I wrote a value type UC that makes unchecked for loops more concise and more readable: // SPDX …

It's designed to be that way. The shift operation is not classified as an arithmetic operation. As per the discussion I had with @chriseth in the Solidity gitter channel:

I would like to write a for loop like this: for (uint256 i = 0; i < length; unchecked { i += 1 }) { // ... } But unfortunately this does not compile (fails with an error "expected primary express …

Note: as of Solidity v0.8.22, this optimization is no longer needed. See this and this. You can do this by putting the post-iteration increment operation at the end of the loop: for (uint256 i = 0; i …

Yes, modulo is cheaper in unchecked arithmetic! Take the following code: pragma solidity ^0.8.0; contract Modulos { function foo(uint256 x, uint256 y) external view returns (uint256 result, uint2 …

It's cheaper to use unchecked blocks and check for overflows and underflows manually than let Solidity do it. Refer to Is the modulo operation cheaper in unchecked arithmetic? for more details.