"Software Performance Tuning with the Apple CHUD Tools"

Abstract

In recent years, many tools have been created to allow software engineers to analyze the execution of their code. While tools such as gprof often work well, most are not integrated very well with each other or the rest of the development environment, and interpreting the data that they provide can be a challenge. Because Apple's MacOS X is based on UNIX, most open source performance analysis tools can be used. However, we have also integrated several key performance tools together and added graphical data visualization to produce the CHUD toolset (Available for at http://developer.apple.com/tools/download/).With the CHUD tools, programmers can examine the performance of their code using a set of integrated tools that can perform most common performance-measurement tasks, including:--Traces of function call behavior (like gprof) --Sampled measurements of program execution timing --Traces of software events, such as system calls --Hardware event counter measurements Moreover, instead of just presenting a few key figures from these measurements in a brief report, the CHUD tools present their results in several textual *and* graphical formats, with integrated hyperlinks to related assembly and source code, so that programmers can easily examine both how their programs work on a large-scale level or zoom in and look at individual program phases in several different ways.This tutorial is targeted primarily at students and software engineers who work on UNIXbased systems and want to expand the repertoire of tools that they can use to analyze and improve the performance of their code. However, the material should also be useful to educators who teach performance-oriented programming techniques, as the graphical nature of Shark's output makes it easy to demonstrate program behaviors in an eye-catching manner.

Visualizing massively multithreaded applications with ThreadScope

“…However, the utility of Dtrace event logs is limited because Dtrace cannot detect individual memory accesses. Examples of graphs based on Dtrace output are Full instruction tracers, such as Apple's libamber [7] trace generator or Sandia's cycle-accurate Structural Simulation Toolkit [8], record every instruction, and thus can track every memory operation. This thorough data collection has a relatively high cost.…”

“…ThreadScope's current set of event-collection tools are based on one of three data collection methods: instrumenting the threading library, system-level runtime tracing, and full instruction traces. In particular, there are implementations for the qthread library [4], Dtrace [6], Apple's libamber [7], and Sandia's SST simulator [8]. Because ThreadScope presents trace-based parallel application structure, it works best when that structure does not depend on the input.…”

“…Full instruction tracers, such as Apple's libamber [7] trace generator or Sandia's cycle-accurate Structural Simulation Toolkit [8], record every instruction, and thus can track every memory operation. This thorough data collection has a relatively high cost.…”

“…Tracing the program in Figure 1 produces the event description in Figure 2. ThreadScope includes several tools for generating these event descriptions from several tracing tools, including Dtrace [6], Apple's libamber [7], and the SST simulator [8]. Other tracing tools could be used to produce similar event descriptions; the basic requirement is the ability to detect thread and synchronization operations.…”