Skip to content

rabbi1155/MATRIXAI

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

WEEKLY UPDATES on MATRIX SMARTCONTRACT PROJECT

11th June - 15th June, 2018

Semantic Computing
  • Discuss and research on how to achieve auto extraction of contract semantics

  • Propose a high-level design on auto extraction of contract semantics based on contract CFG (Control Flow Graph)

  • Employ EVM bytecode parsing framework to generate CFG and automatically extract eigenvectors, as well as execute simple tests

  • Finalized a Semantic Computing blueprint which is based on contract eigenvectors, including algorithm design, third-party libraries selection, the building of developing enviromnets, etc. Semantic Computing is currently in dev phase

Upgradable and Safe VM
  • Discuss and research on how to address VM architecture and implementation details; go through go-ethereum VM code structures and dev interfaces;

  • Finalize the high-level safe upgrade design for the VM end

  • Based on the VM implementations of go-ethereum1.8.11, we completed the extensions to core module of interpreter, as well as JUMP_TABLE mapping

  • Introduced a "Patched Contract" support to the VM, and added a —patchfile command line option;

  • Prepared use cases for regression testing on new features;

  • Completed a demo version of "Patched Contract" on VM side against BEC/SMT overflow vulnerbility.

18th June - 22th June, 2018

Semantic Computing
  • Achieve the goal of automatic extraction of contract semantics characteristics as regards the EVM bytecode instructions; tests also done;

  • Completed the training process to the GAlib-based semantic computing model, and performed the fitting training of simple data set for the model parameters;

  • Completed framework codes for contract semantics clustering; finalized the use of third party libraries and the format of interactive data;

  • Carrying out the analysis work on contract semantics template ; omnichannel preparations for demo use; as well as result analysis.

Upgradable and Safe VM
  • Go on with the development of "Patched Contract", and go through code structures based on step-by-step debugging;

  • Completed the development work of Transaction Function Recognition, and extended on the map data structure in statedb;

  • Determined the implementation plan of patch permission authentication mechanism, focusing on the authentication scheme based on account address;

  • Conducted research on contract upgrade, existing technology roadmap of patches, as well as coding design patterns;

  • Built up a private chain for testing purpose, and completed the application of contract patches on the private chain

25th June - 29th June, 2018

Semantic Computing
  • Completed the automatic framework of Semantic Computing, including characteristics extraction, model training and clustering;

  • Complated the crawling and preparation of datasets;

  • Optimized the training process to the GAlib-based semantic computing model, and completed the fitting training of contract code dataset for the model parameters;

  • Completed the code similarity calculation algorithm based on Bipartie Matching algorithm;

  • Completed the premilinary real contract data semantics clustering with not so ideal results; result analysis and extended experiments are in progress;

  • Discussed together on the plan of framework optimization

Upgradable and Safe VM
  • Research and discuss on contract upgrade plans and come out with a report; discuss on limitations and application scenarios of each upgrade plan;
  • Go on with the development of "Patched Contract", refactor the code implementations under Core package; go through code structures based on step-by-step debugging;
  • Implement refactoring and optimizations on VM codes deployed and upgraded on private chain;
  • Read through go-ethereum API documetns and principles, as well as web3 related documents;
  • Get down with technical proposal design on upgradable contract and VM

3rd July - 7th July, 2018

Semantic Computing
  • Completed the semantics analysis on the current token based or walled based contract, and summarized several typical application scenarios;

  • Completed design on automatically generated interaction scenarios by smart contracts;

  • Completed the framework implementation of cluster based smartcontract semantics computing tools;

  • Achieved the optimized extraction of semantic information, and completed the automatic training of weighting parameter based on GAlib;

Upgradable and Safe VM
  • Discuss and research on the design scheme of Fault Tolerance system, analysing the practicability from the perspectives of technology and application of this proposal

  • Go on with the development of "Patched Contract" supporting VM, and completed the optimized implementation of the PC address redirecting;

  • Regarding the security vulnerability prediction engine related work, we focused on the taint analysis technology, and analyzed the feasibility of this technology at the virtual machine level;

  • Completed the merge of open-source taint analysis and patching;

  • Performed testing and debugging on security vulnerability prediction functions in private test-net.

9rd July - 13rd July, 2018

Semantic Computing
  • Completed the design refination of the interactive scenarios generated automatically by smartcontracts, and finalized the prototype in form of web service, through different discussions;

  • Frontend & background design and development in progress as regards smartcontract auto-generated prototype application scenarios

  • Further research on existing contracts, and refine the feature classification and induction of application scenarios

  • Completed the semantic extraction optimization plan of automatic contract semantics clustering tools, inlucding the optimization of original semantic vectors, computing models, etc;

  • Completed the extention of dataset for evaluation and preparation of dataset; fine-tuned the parameters of automatic semantic clustering model;

  • Evaluating the effects of clustering based on full datasets.

Upgradable and Safe VM

About

Matrix's AI research and related contents

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C 77.4%
  • Python 7.4%
  • Verilog 6.9%
  • VHDL 3.5%
  • Tcl 1.1%
  • HTML 0.9%
  • Other 2.8%