Skip to content

Add README badges, Status and License sections; assert copyright holder in LICENSE #3

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jrconradt merged 1 commit into from
Jun 22, 2026
Merged

Add README badges, Status and License sections; assert copyright holder in LICENSE #3

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion LICENSE
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -175,7 +175,7 @@

END OF TERMS AND CONDITIONS

Copyright 2026 Infalligence Labs
Copyright 2026 Infalligence Labs LLC

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
Expand Down
10 changes: 10 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,7 @@
# LinearPipe.Net

[![CI](https://github.com/jrconradt/LinearPipe.Net/actions/workflows/ci.yml/badge.svg)](https://github.com/jrconradt/LinearPipe.Net/actions/workflows/ci.yml) ![.NET](https://img.shields.io/badge/.NET-10-512BD4) ![Arch](https://img.shields.io/badge/Linux_x86--64-AVX--512F-orange) [![License](https://img.shields.io/badge/License-Apache_2.0-blue)](LICENSE)

A memory-mapped, chunk-disjoint linear data-plane pipeline. Source and sink are mmap'd; the chunk space is partitioned into disjoint spans across worker threads; each 64-byte chunk is loaded, run through a per-chunk SIMD transform, and written to the sink with a non-temporal store. Wait-freedom is structural — workers never share a chunk, so no locks or atomics are needed on the hot path.

`new LinearPipeline(sourcePath, sinkPath, transform, consts, options)` then `Flow()`. The transform is an unmanaged function pointer — `delegate* unmanaged[Cdecl, SuppressGCTransition]<ulong*, byte*, byte*, void>` — called once per 64-byte chunk with (output, input, consts) pointers, so there is no managed dispatch on the hot path. `consts` points at baked constants the transform reads (may be null). `PipelineOptions` carries the worker count and per-worker core affinity — either a `AffinityBase`/`AffinityStride` arithmetic map or an explicit `AffinityCores` array for NUMA-local placement.
Expand All @@ -16,3 +18,11 @@ Requires Linux on x86-64 with AVX512F. The sink uses `vmovntdq` non-temporal sto
dotnet build LinearPipe.Net.slnx
dotnet test LinearPipe.Net.slnx
```

## Status

Active development; API unstable. Used as the streaming substrate behind [FWHT.Net](https://github.com/jrconradt/FWHT.Net)'s benchmark. Linux on x86-64 with AVX-512F is required — there is no fallback store path.

## License

Apache-2.0. Copyright 2026 Infalligence Labs LLC — see [LICENSE](LICENSE).
Loading