feat: Implement EIP-8024 for Amsterdam #3223
Conversation
| 0xE6 => DUPN => stack_io(0, 1), immediate_size(1); | ||
| 0xE7 => SWAPN => stack_io(0, 0), immediate_size(1); | ||
| 0xE8 => EXCHANGE => stack_io(0, 0), immediate_size(1); |
There was a problem hiding this comment.
The stack_io values are not accurate here, they depend on the immediate, unless we want to put the worst case values here. I don't know what this is used for.
| } | ||
|
|
||
| #[cfg(test)] | ||
| mod tests { |
There was a problem hiding this comment.
The test cases are the same listed in the EIP.
| Self::Internal(InternalResult::InvalidExtDelegateCallTarget) | ||
| } | ||
| InstructionResult::InvalidImmediateEncoding => { | ||
| Self::Halt(HaltReason::OpcodeNotFound.into()) |
There was a problem hiding this comment.
I didn't want to add a new HaltReason for this because IIUC it would be a breaking change for implementers of From<HaltReason>.
CodSpeed Performance ReportMerging #3223 will not alter performanceComparing Summary
|
|
I can't reproduce the nightly test failures. |
Interesting error, jumpdest analysis found that the last opcode is 0x00, but it is jumped over, so it concluded it does not need to append zero bytes. Will see how to do this in an efficient way in |
|
Imo |
|
Unbelievably bullish change |
|
It will take a few months before this becomes active so will move this PR to draft |
3 participants
CFI'd as of last week.