SN Systems Tech Blog

Filter:

Understanding the performance of code using LLVM's (llvm-mca)

ARTICLE BY: Andrea DiBiagio
POSTED: Dec 13, 2018
TAGS: Compiler, LLVM, LLVM Developers Meeting, Toolchain

LLVM-mca is a LLVM based tool that uses information available in LLVM’s scheduling models to statically measure the performance of machine code in a specific CPU. The goal of this tool is not just to predict the performance of the code when run on the target, but also to help with diagnosing potential performance issues. In this talk we, will discuss how llvm-mca works and walk the audience through example uses of this tool.

READ THE WHOLE POST 

Debugging of optimized code: Extending the lifetime of local variables

ARTICLE BY: Wolfgang Pieb
POSTED: Nov 30, 2017
TAGS: Compiler, Debugger, LLVM, LLVM Developers Meeting, Toolchain

When developing programs, local variables and function parameters are often optimized away by the backend of the compiler to gain the best runtime performance. As a result, these variables and parameters may not be visible when a developer is attempting to debug their optimized programs. For the PS4, game developers commonly write their code with optimizations enabled so they can maintain graphical frame rates (30-60 fps). As such, they regularly encounter loss of debug information, which impacts their productivity. We have introduced an option that forces the various optimization passes to keep local variables and parameters around. This LLVM talk addresses the implementation, effectiveness and performance impact of this feature in front of the LLVM compiler community.

READ THE WHOLE POST 

Demo of a repository for statically compiled programs

ARTICLE BY: Paul Bowen-Huggett
POSTED: Nov 17, 2016
TAGS: LLVM, LLVM Developers Meeting, Toolchain

Paul presents a proof of concept of an approach which improves compile and link times by replacing the conventional use of object files with an incrementally updated repository without requiring change to existing build infrastructure. It aims to present the idea at a high-level using a live demo of some trivial tools and initiate discussion of a real implementation within the LLVM framework.

READ THE WHOLE POST