plotchy

Results 26 comments of plotchy

Adding a thought here: If there isn't a post-dominant `return/revert/break/etc` of the `costly-operator`, you can also check if all post-dominance frontier's of the node-id are then post-dominated by a `return/revert/break/etc`....

That's a good suggestion, and I think I could code it. And thx for checking the heuristic, edited it.

This is used in the echidna printer and relies on the `_get_name()` helper function. https://github.com/crytic/slither/blob/db703d82f4876e30f1ca10fb78d85b4dd4c31ec1/slither/printers/guidance/echidna.py#L35-L41 ```python def _get_name(f: Union[Function, Variable]) -> str: # Return the name of the function or...

Ah, @montyly youre totally right. I like the idea of using placeholders. I briefly tested with `*receive()` and `*fallback()`, but I seem to be getting collision issues within the Slither...

@0xalpharush Hmm after looking at that line it seems ambiguous how `functions_entry_points` is designed with regards to `fallback`. IIRC Python has higher precedence on `and` than `or`. I think the...

+1. This is a very helpful RPC for security work. [ETHBMC](https://github.com/RUB-SysSec/EthBMC#ethbmc-a-bounded-model-checker-for-smart-contracts) is an example where it's used for storage access. Love the idea of returning key:val pairs. For massive storage...

I agree Though probably need to separate the arbitrum variants out from each other https://github.com/alloy-rs/chains/blob/0eaf7000dc3d21e47c52bfa57a72a87e28861621/src/named.rs#L352-L357 Arbitrum One is ~.25 s/block https://arbiscan.io/blocks Nova isn't consistent so is probably fine as is...

Likely root issue of both https://github.com/crytic/crytic-compile/issues/269 and https://github.com/crytic/slither/issues/1241 as well

I've spent some more time debugging both of these occurences. Without the full detachment of associationsfilenames (topological sorting occurs on associations, but printing occuring as filenames) there are many edge...

And front loading all non-dependant nodes to the front of the list before a de-duping step helps to alleviate some edge cases where the graph is 'lazily sorted' (a minimal...