Analysis for Code Size Opportunities in RISC V
Talkat RISC-V Forum: Developer Tools and Tool Chains 2021 · on Wednesday, 2021-06-02· at 16:35· Virtual·
summary
Describes a practical evaluation of the Zce extension using an objdump-parsed Python script that builds instruction records, simulates replacing sequences (tablejal, compressed multiply, push/pop), and accounts for table size and encoding-space cost. Uses a multi-compiler, multi-benchmark suite including V8 and common embedded suites and reports up to 15–20% code-size savings for table-jump and push/pop cases while checking overlap, encoding-bit tradeoffs, and implementation/performance risk.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 0:14Why code size is important
- 1:47How to achieve code size density
- 3:50Python script
- 4:50Optimizations
- 6:09Verification
- 8:07Extensions
- 9:18Table Job
- 11:16Push Pop
- 13:18Encoding Space
- 14:00Benchmarking
- 14:59Benchmarks
- 16:05Results
- 19:04What we did
- 19:50Outro
related talks
17mRecent Developments in Optimizing Compilers for RISC-V
23mCode Size of RISC V versus ARM using the Embench™ 0 5 Benchmark Suite
23mCode Density Improvements Beyond The C Standard Extension
19mCode Size Reduction Deep Dive: Zc Extensions and Dictionary Compression Custom Instruction
15mCompared Analysis of GCC Codegen for AArch64 and RISC-V
19mGCC Compiler Code Size Density
01h 10mWhy RISC-V Through ISA Customization Demonstration
10mRISC-V Instruction Set Explorer (RISE)