Bug#830708: RFP: coz-profiler -- Finding Code that Counts with Causal Profiling
The Coz viewer uses a few javascript libraries that are currently missing in Debian. I've submitted WNPP requests for both, d3-tip and science.js: d3-tip: https://bugs.debian.org/843964 science.js: https://bugs.debian.org/843963 -- Happy hacking Petter Reinholdtsen
Bug#830708: RFP: coz-profiler -- Finding Code that Counts with Causal Profiling
[Petter Reinholdtsen] > The package depend on libelfin, > https://github.com/ccurtsinger/libelfin >, to work, and will > fail to build without it. I've requested a package for libelfin in https://github.com/aclements/libelfin >. I'm working on a draft package of coz, the build rules are in the debian branch available from https://github.com/petterreinholdtsen/coz >. -- Happy hacking Petter Reinholdtsen
Bug#830708: RFP: coz-profiler -- Finding Code that Counts with Causal Profiling
The package depend on libelfin, https://github.com/ccurtsinger/libelfin >, to work, and will fail to build without it. Surprisingly enough, the build will clone the git repository and make the source available, so the problem is not too abvious when building the code for the first time. In addition, the build will download and build several example programs, but I suspect those failures will not be a fatal problem in the build process. -- Happy hacking Petter Reinholdtsen
Bug#830708: RFP: coz-profiler -- Finding Code that Counts with Causal Profiling
Package: wnpp Severity: wishlist * Package name: coz-profiler Version : n/a, git repo without tags Upstream Author : Charlie Curtsinger and Emery Berger, University of Massachusetts Amherst * URL : https://github.com/plasma-umass/coz * License : BSD 2-clause Programming Lang: C++ Description : Finding Code that Counts with Causal Profiling Coz is a new kind of profiler that unlocks optimization opportunities missed by traditional profilers. Coz employs a novel technique we call causal profiling that measures optimization potential. This measurement matches developers' assumptions about profilers: that optimizing highly-ranked code will have the greatest impact on performance. Causal profiling measures optimization potential for serial, parallel, and asynchronous programs without instrumentation of special handling for library calls and concurrency primitives. Instead, a causal profiler uses performance experiments to predict the effect of optimizations. This allows the profiler to establish causality: "optimizing function X will have effect Y," exactly the measurement developers had assumed they were getting all along. -- Happy hacking Petter Reinholdtsen