Use lexical ordering for test-suite tests LIT's default order is "smart", which on subsequent runs orders tests based on their execution time (negated for failure). Unfortunately, while great for llvm's regression suite, that ordering is not great for benchmark stability, where we want things to be as consistent as possible from run to run. Thanks to Ahmad Yasin for spotting this. rdar://185192119
LNT is an infrastructure for performance testing. The software itself consists of two main parts, a web application for accessing and visualizing performance data, and command line utilities to allow users to generate and submit test results to the server.
The package was originally written for use in testing LLVM compiler technologies, but is designed to be usable for the performance testing of any software.
The official LNT documentation is available online at: https://llvm.org/docs/lnt
The LNT source is available in the llvm-lnt repository: https://github.com/llvm/llvm-lnt
To get started, please look at the Developer Guide in the official documentation.