dynG is alpha: 0.3.0, the latest release, is on PyPI (dyng with the CUDA plugin wheels
dyng-cu12 / dyng-cu13), as are 0.2.0 and 0.1.0, the first. The PyPI package dyng 0.0.1 was
only a name reservation and contains no library code; the release candidates 0.1.0rc1, 0.2.0rc1
and 0.3.0rc1 were published on TestPyPI only. Security fixes go into the latest minor
release (as a new patch release) and into main; older minor releases are not patched.
Problems found in a release candidate are fixed in main before the release.
| Version | Supported |
|---|---|
main |
yes |
| latest minor release (today 0.3.x) | yes |
| 0.3.0rc1 (release candidate, TestPyPI only) | no; please upgrade to 0.3.0 |
| 0.2.x | no; please upgrade to 0.3.0 |
| 0.2.0rc1 (release candidate, TestPyPI only) | no; please upgrade to 0.3.0 |
| 0.1.x | no; please upgrade to 0.3.0 |
| 0.1.0rc1 (release candidate, TestPyPI only) | no; please upgrade to 0.3.0 |
| 0.0.1 (name reservation, no library code) | no code to fix |
| older minor releases | no; please upgrade |
Please do not open a public issue, discussion or pull request for a security problem. Report it privately, by either of these ways:
- GitHub private vulnerability reporting (preferred): open https://github.com/dyng-dev/dyng/security/advisories/new, or go to the repository's Security tab and click Report a vulnerability. The report is visible only to you and the maintainers.
- E-mail to the lead maintainer: sm.shovan@gmail.com, with "dynG security" in the subject.
Please include:
- the affected version or commit, and how you installed dynG (source build, wheel);
- the backend (sequential, OpenMP, CUDA), the operating system, the compiler, and for CUDA the GPU, the driver and the CUDA version;
- a minimal reproducer (input files and the call or command) and what you observed (crash, out-of-bounds access, memory growth, sanitizer report);
- whether the problem is already known to others, and how you would like to be credited.
- Acknowledgement within 7 days. A maintainer confirms that the report was received and who handles it.
- Assessment. We reproduce the problem, decide whether it is a vulnerability (see Scope), and estimate its severity (CVSS). We tell you the outcome and keep you informed at least every 14 days until the report is closed.
- Fix in private. The fix is prepared in a GitHub security advisory's temporary private fork, with a regression test. You are invited to review it.
- Coordinated disclosure. We agree on a disclosure date with you, normally when the fixed
release is available and no later than 90 days after the report (shorter if the problem
is being exploited, longer only with your agreement). On that date we publish the patch
release, the GitHub security advisory (with a CVE requested through GitHub where one is
warranted) and a
Securityentry inCHANGELOG.md. - Credit. Reporters are credited in the advisory unless they ask not to be.
If a report turns out not to be a vulnerability, we say why and, with your agreement, move it to a normal issue.
In scope:
- memory-safety problems (crash, out-of-bounds read or write, use after free, unbounded memory use) in the library, the command-line tools or the Python package, including on malformed or hostile input files (Matrix Market, edge lists, the MOSP text formats, batch files);
- problems in the build, release and CI configuration of this repository that could let someone tamper with published artifacts (for example the GitHub workflows or the PyPI Trusted Publishing setup).
Out of scope:
- results that are merely wrong on valid input: these are bugs, please report them as normal issues (they matter to us just as much);
- resource use that is proportional to a legitimately large input;
- vulnerabilities in third-party software (the CUDA toolkit, the driver, compilers, Python), which should be reported to their vendors; tell us if dynG needs to react.