Feature/stored nodes minimal #65
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
Stored nodes have encoded values with at least uncompressed commitments and children hashes. All (little-endian) bytes are also tail truncated to compress their representation. Furthermore, null nodes are represented by empty bytes (0x80 in RLP encoding). The commit also includes a major refactoring of location attribute for nodes. Previously, leaf location was their full key, which was a little awkward. Now, node location is truly its location in the trie, no matter what kind of node it is. Among other things, this means that we store a StemNode's depth along its stem, so we can retrieve its location. Coming up is the possibility to store commitments in compressed form, which would save a lot of space (50% on commitments). Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
|
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
matkt
reviewed
Jul 10, 2024
src/main/java/org/hyperledger/besu/ethereum/trie/verkle/node/BranchNode.java
Outdated
Show resolved
Hide resolved
| if (scalars.get(i) == Bytes32.ZERO) { | ||
| children.add(new NullNode<V>()); | ||
| } else { | ||
| children.add(new StoredNode<V>(this, nextLoc, scalars.get(i))); |
There was a problem hiding this comment.
why not just doing
Bytes childrenLocationPrefix = Bytes.concatenate(location,Bytes.of(0));
children.add(new StoredNode(this, Bytes.concatenate(childrenLocationPrefix,Bytes.of(i)), scalars.get(i)));
There was a problem hiding this comment.
I think this was a somewhat lame attempt to avoid concatenates, as you mentioned it takes a lot of the execution time.
But I am pretty sure this is not better, and less readable. So reverting is good.
| for (int i = 0; i < nChild; i++) { | ||
| children.add(new StoredNode<>(this, Bytes.concatenate(stem, Bytes.of(i)))); | ||
| nextBuffer.set(location.size(), Bytes.of(i)); | ||
| Bytes nextLoc = nextBuffer.copy(); |
There was a problem hiding this comment.
same here, seems to be hard to understand like that
There was a problem hiding this comment.
same as above
src/main/java/org/hyperledger/besu/ethereum/trie/verkle/node/InternalNode.java
Show resolved
Hide resolved
| * @param value A commitment value | ||
| * @return The encoded commitment | ||
| */ | ||
| public static Bytes encodeCommitment(Bytes value) { |
There was a problem hiding this comment.
why this two methods are static ?
There was a problem hiding this comment.
because they do not use any info from an instance, but are rather only utility methods for the class.
But I am not particularly attached to it, so if you think that some other way is better or more consistent with besu, no problem
src/main/java/org/hyperledger/besu/ethereum/trie/verkle/node/Node.java
Outdated
Show resolved
Hide resolved
src/main/java/org/hyperledger/besu/ethereum/trie/verkle/node/StoredNode.java
Show resolved
Hide resolved
src/test/java/org/hyperledger/besu/ethereum/trie/verkle/SimpleBatchedVerkleTrieTest.java
Outdated
Show resolved
Hide resolved
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
Signed-off-by: Karim Taam <karim.t2am@gmail.com>
fix tests and spotless
matkt
reviewed
Aug 12, 2024
src/main/java/org/hyperledger/besu/ethereum/trie/verkle/visitor/CommitVisitor.java
Outdated
Show resolved
Hide resolved
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com>
matkt
added a commit
that referenced
this pull request
Oct 1, 2024
This is a continuation of the previous PR #65 It appeared that the previous PR had some bugs, and meanwhile it was decided to add stem extensions into the DB in order to be able to get stemNodes by stem without requiring a nearest request. In this commit, InternalNodes include stem extensions into its encodedValues, useful to create the children Stored Nodes including their appropriate keys (location or stem). The key is then used to load the StoredNode. StoredNode are also separated into Internal and Stem Nodes as well to reflect this difference. --------- Signed-off-by: Thomas Zamojski <thomas.zamojski@quadratic-labs.com> Co-authored-by: Karim Taam <karim.t2am@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR description
Storing Nodes' info with RLP encodings, list containing at least uncompressed commitments and children hashes.
All (little-endian) bytes are tail truncated to compress their representation.
Furthermore, null commitments/hashes are represented by empty bytes (0x80 in RLP encoding).
The commit also includes a major refactoring of location attribute for nodes.
Previously, leaf location was their full key, which was a little awkward.
Now, node location is truly its location in the trie, no matter what kind of node it is.
Among other things, this means that we store a StemNode's depth along its stem, so we can retrieve its location.
Fixed Issue(s)
Issue #60