V8 introduces two new WebAssembly optimizations in Chrome M137: speculative call_indirect inlining and deoptimization support. These optimizations use runtime feedback to make assumptions and generate faster machine code, particularly benefiting WasmGC programs. The combination delivers 50%+ speedups on Dart microbenchmarks and
•20m read time• From v8.dev
Table of contents
Background #Motivation #Speculative Inlining #Deoptimization #Results #Conclusion and Outlook #Sort: