feat: blob_hash #489
feat: blob_hash #489
Conversation
|
|
||
| // If index is within bounds, get the hash at that index | ||
| // Otherwise return zero bytes | ||
| let (high, low) = split_felt(blob_hashes.value.len); |
There was a problem hiding this comment.
i guess it's fine to do this because we don't want to over analyze the evm
but blob_hashes.value.len right now is between 2 and 5, in the future it'll be at most 200 or maybe 1000
| if (out_of_bounds.value == 0) { | ||
| tempvar blob_hash = Bytes32(new Bytes32Struct(0, 0)); | ||
| } else { | ||
| tempvar blob_hash = blob_hashes.value.data[index.value.low]; |
There was a problem hiding this comment.
if we're only using index.value.low, we should probably assert index.value.high == 0, as well as blob_hashes.value.len.high ==0,
then just compare the two low felt values
There was a problem hiding this comment.
we should probably assert index.value.high == 0,
no, this would be a deviation to spec. MAX_U256 would be a valid blob_hash index query and would return zero. We must not assume anything on the validity range of this input value.
for len(blob_hashes), we can assume (and assert) that this number is way below 2**128.
so either:
- just do u256 comparison, which is done now and is correct
- assert len(blob_hashes) < 2128, and return zero if input >= 2128 or if input < len
| // If index is within bounds, get the hash at that index | ||
| // Otherwise return zero bytes | ||
| let (high, low) = split_felt(blob_hashes.value.len); | ||
| let out_of_bounds = U256_lt(index, U256(new U256Struct(low, high))); |
There was a problem hiding this comment.
in_bounds ?
| let out_of_bounds = U256_lt(index, U256(new U256Struct(low, high))); | |
| let in_bounds = U256_lt(index, U256(new U256Struct(low, high))); |
| if (out_of_bounds.value == 0) { | ||
| tempvar blob_hash = Bytes32(new Bytes32Struct(0, 0)); | ||
| } else { | ||
| tempvar blob_hash = blob_hashes.value.data[index.value.low]; |
There was a problem hiding this comment.
we should probably assert index.value.high == 0,
no, this would be a deviation to spec. MAX_U256 would be a valid blob_hash index query and would return zero. We must not assume anything on the validity range of this input value.
for len(blob_hashes), we can assume (and assert) that this number is way below 2**128.
so either:
- just do u256 comparison, which is done now and is correct
- assert len(blob_hashes) < 2128, and return zero if input >= 2128 or if input < len
Close #479