Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/jryans/debug-info-metrics-artifact

Artifact for the CC 2024 paper "Accurate Coverage Metrics for Compiler-Generated Debugging Information"
https://github.com/jryans/debug-info-metrics-artifact

coverage debug-info optimisation

Last synced: 3 months ago
JSON representation

Artifact for the CC 2024 paper "Accurate Coverage Metrics for Compiler-Generated Debugging Information"

Host: GitHub
URL: https://github.com/jryans/debug-info-metrics-artifact
Owner: jryans
License: mit
Created: 2024-01-08T12:44:13.000Z (about 1 year ago)
Default Branch: main
Last Pushed: 2024-07-24T12:18:39.000Z (6 months ago)
Last Synced: 2024-10-05T23:38:31.501Z (4 months ago)
Topics: coverage, debug-info, optimisation
Language: Jupyter Notebook
Homepage:
Size: 18 MB
Stars: 1
Watchers: 3
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md
- License: LICENSE

Awesome Lists containing this project

README

        ## Artifact

This is the artifact for the [CC 2024 paper][paper]:

Accurate Coverage Metrics for Compiler-Generated Debugging Information

by J. Ryan Stinnett and Stephen Kell.

[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.10568392.svg)](https://doi.org/10.5281/zenodo.10568392)

## Usage

The `data` directory includes various scripts and notebooks used to produce the

figures included in the paper.

The `figures` directory contains the figures as shown in the published paper.

The basic pipeline to reproduce a figure for a given program under analysis

(e.g. Git which is used in most figures) is:

1. Check out project source at commit in the table below

2. Run `prepare.sh` in the project source directory

3. Run `build.sh` in the project source directory

4. Run `check-quality.sh` in the `data/[project]` directory of this repo

5. Use the notebook in `data/[project]`

Some scripts currently use system-specific paths to locate compilers and other

tools, so some modifications may be needed. Review `vars.sh` and adjust path

values as needed.

## Requirements

Source code for each project under analysis will be needed. For the paper, the

following projects and commits were used:

| Project | Commit |

| ------- | ------ |

| Dash    | b00288fd28c1b39d0f1531b6e6d86de59de4be8d |

| FFmpeg  | c06d3d24047206f9c11bfc5849544b960e5d68eb |

| Git     | bd5df96b7916c1e8a9300ab477aec047b3901cd4 |

| libbz2  | 323678a0d655eb3720618dd7f0dd403eef5e6999 |

| FLINT   | c08c06bf202d1b154742eecc183b8a3d13cddf62 |

| tar     | cf16a2394572301496b618a857e71371020af643 |

Most published figures focus on the Git codebase, so consider starting there.

In addition, some other tools are invoked by scripts here:

| Tools    | Commit |

| -------- | ------ |

| [dbgcov] | 1b89156b7eb3fefad610e432859da21eebbe8046 |

| [debuginfo-quality] | 2a1a0dfc2325b7c210998692d219fcb0bf708e91 |

| [incomplete-debuginfo] | 0df40cb8e1166b4203ecf2dddbb1e304d5772511 |

The following compiler versions are used:

- Clang 12.0.0

- Clang 13.0.0

- Clang 14.0.0

- Clang 15.0.0

- Clang 16.0.0

- Clang 17.0.1

- GCC 10.5.0

- GCC 11.4.0

- GCC 12.3.0

- GCC 13.1.0

Our results were collected on macOS 13.6 for the x86-64 architecture. Other

platforms may give different results, as some compilers may enable different

debug info features by default for each target platform.

[paper]:

  https://doi.org/10.1145/3640537.3641578

[dbgcov]:

  https://github.com/stephenrkell/dbgcov

[debuginfo-quality]:

  https://github.com/jryans/debuginfo-quality/tree/debug-info-metrics

[incomplete-debuginfo]:

  https://github.com/jryans/incomplete-debuginfo/tree/debug-info-metrics