Added Intrinsics and some performance optimisations #58
Conversation
- Added Intrinsics fast path - Added Thread Static cache for GetMacDataRfc8439 - Added ability for benchmarks to run in .NET48 (for a pre/post intrinsics comparison)
|
Commit 28d47c9 added |
|
Hello macaba, Glad to hear from you again. 😀 Allow me some time to check it in the different platforms since the CI didn't kick on Azure DevOps. |
There was a problem hiding this comment.
@macaba noticed why the CI failed, could you kindly address this little issue?
| <TargetFrameworks>netstandard1.6;netstandard2.0;netstandard2.1;netcoreapp3.1;netcoreapp5.0</TargetFrameworks> | ||
| <TargetFrameworks Condition="'$(OS)' != 'Unix'">netstandard1.6;netstandard2.0;netstandard2.1;netcoreapp3.1;netcoreapp5.0;net45;net48</TargetFrameworks> |
There was a problem hiding this comment.
Could you remove .NET 5 since it's still in preview, so the CI doesn't complain?
It will surely be added once .NET 5 is released by the end of the year.
There was a problem hiding this comment.
I just pushed an update to support .NET 5 now that is RTM.
|
@macaba I did a quick test under Windows and the Wycheproof test vectors from ChaCha20Poly1305 and XChaCha20Poly1305 are failing. Could you check those out, please? Apparently, the changes on |
|
Hello @macaba I wonder if you got any progress on this? I'm planning on releasing the updated version on NuGet tagged as 2.0.0 since in the current state (prior to the PR) we got some breaking changes. Would be nice to have yours included. |
daviddesmet
added
the
major
daviddesmet
changed the title
|
This issue has been automatically marked as stale because it has not had recent activity. It will be closed in one week if no further activity occurs. |
Hello!
I finally got round to pulling in all your latest updates. Fantastic work on the Span & BitUtils support.
A quick bit of benchmarking.
Before adding my pure intrinsics processing (i.e. using BitUtils which I suspect emits some intrinsics under the hood):
A decent improvement vs. Framework.
After adding intrinsics:
I'm particularly impressed with the memory allocated when using the Span Encrypt method. When I did a memory trace, the
GetMacDataRfc8439was the only method significantly allocating memory. It would be ideal to come up with some sort of caching or memory context scheme for this.