Should --cost items be audited
Nobody has claimed this yet.
- Dominant language
- C
- Stars
- 13.6k
- Forks
- 2.6k
- PR merge metrics
- No merged PRs in 30d
Description
As I have found out, ,doing a comprehensive check of gpg, the -cost items do not always go from quicker to slower. Here is a grid showing speeds with a mix of all cost2/cost3 values The vertical grid is hash type, and the horizontal is encryption type. Both are in order by 'cost' numbers, but as can be seen,
| . | idea | 3des | cast5 | bf | a128 | a192 | a256 | 2f | c128 | c192 | c256 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| md5 | 52.1 | 24.6 | 49.4 | 40.8 | 51.5 | 25.5 | 25.4 | 21.8 | 53.3 | 25.7 | 25.6 |
| sha1 | 41.6 | 20.7 | 41.7 | 33.2 | 40.2 | 20.8 | 20.5 | 19.9 | 40.7 | 20.8 | 20.7 |
| ripmd | 19.7 | 10.0 | 20.0 | 18.5 | 19.3 | 8.8 | 9.3 | 8.5 | 18.2 | 9.5 | 10.0 |
| s256 | 20.7 | 19.8 | 20.8 | 18.3 | 20.9 | 20.2 | 20.7 | 19.8 | 20.4 | 18.4 | 20.5 |
| s384 | 27.2 | 27.4 | 27.6 | 24.0 | 27.8 | 28.4 | 28.8 | 27.6 | 26.4 | 26.9 | 27.3 |
| s512 | 27.5 | 28.5 | 28.7 | 24.5 | 27.6 | 27.8 | 28.4 | 27.3 | 28.5 | 27.1 | 26.5 |
| s224 | 17.9 | 19.7 | 19.6 | 17.1 | 19.8 | 19.6 | 10.1 | 9.9 | 20.1 | 20.2 | 9.8 |
Is there anything we can (or should) do to help normalize how cost works ? I am sure there are other formats which the --cost values do not line up with throughput.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No source file, test, or entry point is named. Begin by reproducing the reported gpg benchmark grid and examining how cost2 and cost3 values are assigned across hash and encryption types; done requires an agreed way to audit or normalize cost ordering without losing the reported throughput data.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- c
- Domain
- cryptography, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100