EthereumRuntime: Use EXTCODECOPY to retrieve bytecode
#51
Conversation
Codecov Report
@@ Coverage Diff @@
## master #51 +/- ##
=========================================
+ Coverage 64.62% 65.5% +0.87%
=========================================
Files 9 9
Lines 212 200 -12
Branches 58 55 -3
=========================================
- Hits 137 131 -6
+ Misses 41 37 -4
+ Partials 34 32 -2
Continue to review full report at Codecov.
|
contracts/Verifier.sol
Outdated
|
|
||
| if ((dispute.state & END_OF_EXECUTION) != 0) { | ||
| if (uint8(code[_executionState.pc]) != OP_REVERT && uint8(code[_executionState.pc]) != OP_RETURN) { | ||
| if (opcode != OP_REVERT && opcode != OP_RETURN) { |
There was a problem hiding this comment.
This question is unrelated to the PR.
Now I think more about this, I wonder if the operator in plasma chain would submit reverted transactions to main net?
Is this will open a kind of DOS attack?
There was a problem hiding this comment.
This can be challenged. This just makes sure that a challenger can proof that the solvers solution was too 'long'
There was a problem hiding this comment.
I see, this is just for the case solver didn't stop after hitting REVERT. My question was totally unrelated.
But given that, then stateHash is already enough to cover this case, isn't it?
Because if opcode is REVERT, EVMRuntime will return some error?
There was a problem hiding this comment.
Also there is opcode invalid. Should it be included here?
There was a problem hiding this comment.
Yes, stateHash is enough because they have to proof the execution and also agree on prioer execution steps.
Also there is opcode invalid. Should it be included here?
Hmm, no. Every valid execution ends with either RETURN or REVERT.
Or most of the time: Out Of Gas xD
There was a problem hiding this comment.
I think so too. When something invalid happen it will REVERT.
But here: https://solidity.readthedocs.io/en/v0.5.2/assembly.html#opcodes
there is an invalid which also said to end execution with invalid instruction
Do I miss anything...
There was a problem hiding this comment.
I think assert will end with invalid but haven't found something conclusive
There was a problem hiding this comment.
@peara
If we want to allow this, we might also include SELFDESTRUCT. But that wasn't at the table.
I think we don't want to allow that for PLASMA @johannbarbie ?
There was a problem hiding this comment.
@peara But still pointer, STOP is missing 😎
|
Yay or Nay ? 🙂 |
|
Yay 😄 |
Introduces the EVMCode helper library
Related #18