SpiderMonkey now has two new shell flags: --benchmark-mode and --strict-benchmark-mode. These flags consolidate multiple shell options needed for accurate benchmarking into a single command, reducing the risk of running benchmarks with incorrect configurations. The strict variant will abort if the environment is unsuitable for benchmarking (e.g., a debug build). The goal is to standardize benchmarking setup across the team and make it easy for anyone to run benchmarks correctly.
Sort: