CKB Dev Log 2023-07-26

[anipenguin]

♬♪ - @xxuejie

Updates

Features

• Added API implementation for the load_extension syscall (CKB2023) in the ckb-std & ckb-c-stdlib standard libraries
  • Add ckb_load_extension syscall ckb-c-stdlib#47
  • Add load_extension syscall ckb-std#65
• ckb-auth now supports Solana
  • add solana support ckb-auth#1
• ckb-debugger now decodes single RISC-V instructions
  • ckb-debugger support decode instruction ckb-standalone-debugger#94

Improvements

• CurrentCycled syscall now returns the sum of cycles across all virtual machines
  • refactor(scripts): current_cycles returns the sum of cycles of all living machines #4059
• Enhanced CKB to now terminate the miner process correctly once the nonce limit has been reached
  • terminate miner process when nonce limit reached #4075

Fixes

• Fixed an issue where the CKB miner would stop working when handling a large number of instructions
  • Fix(TxPoolService): set started to true before load_persisted_data #4069
• Corrected the inconsistency in the output formats in ckb-cli
  • [Fix] Type of secp256k1_data.out_point.index for ckb util genesis-scripts should be hex encoded ckb-cli#565
• Resolved an issue that could potentially lead to type conversion overflow in the indexer RPC parameters
  • fix: use Uint32 for Indexer search limits ckb-sdk-go#210
• In ckb-vm, we've implemented several fixes to reduce stack consumption, making it more manageable when dealing with spawn syscalls
  • feat(decoder): move cache to heap ckb-vm#357
  • Backport #346 #347 #349 ckb-vm#367
• Furthermore, we've corrected the memory size inconsistency in non-default-sized VM instances
  • Use correct memory pages ckb-vm#360
  • Use correct memory pages and size ckb-vm#363

In Pipeline

• Refactoring the code in ckb-types to separate it into two distinct sections. The first section consists of code auto-generated by molecule, which is a general serialization format. The second section consists of code that is specific to CKB. This segregation is necessary because the CKB-specific code is not directly associated with the molecule serialization format. Instead, it is intrinsically tied to CKB's own logic. This refactoring will enhance code manageability, clarity, and maintainability, aligning more closely with best coding practices.
  • refactor: split ckb-gen-types from ckb-types #4073
• Upon completion of the refactoring process within ckb-types, we intend to proceed with refactoring the CKB Rust SDK. The objective is to better delineate the boundaries between auto-generated code and handwritten code, thereby offering a more structured, cleaner interface.
• Working to optimize the implementation of secp256r1 signatures on CKB-VM. The primary focus is to minimize cycle consumption, thereby increasing the overall efficiency.