Replacing Blake2s compression function by Blake2s hash function #190
Conversation
| size_t to_pad = input_size % BLAKE2s_block_size; | ||
| for (size_t i = 0; i < BLAKE2s_block_size - to_pad; i++) { | ||
| padded_input.push_back(FieldT("0")); | ||
| } |
There was a problem hiding this comment.
Seems like here you will still pad with 0's even if input_size % BLAKE2s_block_size == 0. In fact the result of your mod is stored in to_pad, and to_pad is not used inside a conditional to pad only if to_pad != 0 as suggested in your comment above. If to_pad is 0 here - assuming input_size = BLAKE2s_block_size, then your padded_input vector will be of length 2*BLAKE2s_block_size.
|
@rrtoledo please make sure to rebase your branch on top of develop since the build has been repaired. |
dtebbs
force-pushed
the
blake-hash
branch
2 times, most recently
from
45e00a1
to
5037687
Compare
libzeth/circuits/blake2s/blake2s.hpp
Outdated
| @@ -165,6 +165,7 @@ template<typename FieldT> class BLAKE2s_256 : public libsnark::gadget<FieldT> | |||
| }; | |||
|
|
|||
| } // namespace libzeth | |||
|
|
|||
There was a problem hiding this comment.
Not sure this is consistent with the other hpp/tcc files where we wrote the include on the line following the } closing the libzeh namespace
There was a problem hiding this comment.
Ah, I didn't realise it had been switched to that.
It seems slightly unnatural since in all other cases we have a line before #include blocks, and line between } // namespace ... and following code (hence it stood out).
Anyway changed for now to be consistent.
Replacing Blake2s compression function by Blake2s hash function
Tackles #158