Skip to content

Release v1.5.2

Latest
Latest
Compare
Choose a tag to compare
@github-actions github-actions released this 30 Mar 20:07
· 207 commits to master since this release
v1.5.2
This tag was signed with the committer’s verified signature.
m4rs-mt Marcel Koester
GPG key ID: E68755654F1165F4
Verified
Learn about vigilant mode.
7d55310
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

This new stable release includes bug fixes and internal improvements (get the ILGPU Nuget package and ILGPU Algorithms Nuget package).

Changes

  • Fixed missing stream in F# sample project (#1106).
  • Fixed marshalling ArrayView.Empty as kernel parameter (#1131).
  • Fixed PTX code generation for void functions and out parameters (#1147).
  • Fixed alignment when marshalling Cuda kernel parameters (#1176).
  • Fixed definition of Nvidia LibDevice Modf (#1201).
  • Optimized PTX IntrinsicMath implementation to use LibDevice (#1189).
  • Fixed OpenCL printf newline output (#1182).
  • Updated NVVM to support Cuda SDK v8 (#1187).
  • Added DLL resolver for Cuda on WSL (#1177).

Internal Changes

  • Extended static type information (#1072).
  • Updated version to v1.5.2-beta1 (#1195).
  • Moved OpenCL IntrinsicMath implementations (#1185).
  • Adjusted copyright information of back ported fixes (#1216).
  • Fixed ambiguous method call in T4 transformation (#1224).

CI Changes

  • Disabled assertions on GPU unit tests (#1212).
  • Disabled automatic Cuda testing from CI builds (#1272).
  • Bump Cuda versions (#1111, #1172, #1230, #1314).
  • Downgrade macOS image to build netcoreapp3.1 and net5.0 (#1320).

We would like to thank the entire ILGPU community for providing feedback, submitting issues and feature requests.

Full Changelog: v1.5.1...v1.5.2

Assets 5
Loading