V2.2.2 eigenda #6
Conversation
| // record preimage data | ||
| log.Info("Recording preimage data for EigenDA") | ||
| shaDataHash := sha256.New() | ||
| shaDataHash.Write(sequencerMsg) |
There was a problem hiding this comment.
why are we recording the hash of the sequencer message and not the underlying data?
There was a problem hiding this comment.
Oh actually we are recording the underlying data, see line 211. here we are calculating the key of it :D
There was a problem hiding this comment.
I see, but it seems like sha256(data) != dataHash on line 211. Right now the dataHash variable is based on the hash of the sequencer message, which itself is a reference the EigenDA blob, not the blob itself.
To put it formally: data != sequencerMsg, which breaks the invariant of the preimage cache necessary for fraud proofs to work as expected.
There was a problem hiding this comment.
Yes you are right, here's the map working as the preimage cache, the key of the map I put here is sha256(reference) and value is data itself, and later when i query from the map again (file cmd/replay/main.go), i also hashed data reference, so i can get correct data. But i got you, maybe you prefer using hash(data) as key? i can make the change : D
There was a problem hiding this comment.
Oh but i double check PreimageEigenDAReader it has to implement interface EigenDAReader, so the parameter need to be blob reference lol
There was a problem hiding this comment.
Maybe I misunderstand. Isn't the preimage cache for reading data that isn't in the inbox? The blob reference is in the inbox.
renlulu
force-pushed
the
v2.2.2-eigenda
branch
from
5145ef3 to
836195f
Compare
* refactor: clean doc * doc: add todo for replay * doc: add alt license header * doc: change some comment for replay
For review purpose, don't merge!