Repository navigation

Cythonize `cuda.core` more - #1070

Merged

leofang merged 20 commits into

NVIDIA:mainfrom

leofang:cythonize_more

Oct 7, 2025

Merged

Cythonize `cuda.core` more#1070
leofang merged 20 commits into
NVIDIA:mainfrom
leofang:cythonize_more

Commits on Oct 3, 2025
- release gil
  leofang
  committed
  1f84a5a
- release gil for record
  leofang
  committed
  2c86825
- use __dealloc__ in event/stream
  leofang
  committed
  efe7fe0
- nit: remove print
  leofang
  committed
  fa88b28
- fix linter error
  leofang
  committed
  1083d90
- reduce further the number of Python objects held by Stream
  leofang
  committed
  c06dc9d
- replace a few more __del__ by __dealloc__
  leofang
  committed
  b713c5c
- improve __cuda_stream__ performance
  leofang
  committed
  d764f5b
- cythonize Buffer & DMR (WIP - not working!)
  leofang
  committed
  3684644
- minor fixes - still failing
  leofang
  committed
  2c9d5ec
- fix casting and avoid repetive assignment
  leofang
  committed
  638bf59
Commits on Oct 4, 2025
- ensure we have C access for DeviceMemoryResource
  leofang
  committed
  0c64e8e
- restore the contracts for now (the deallocation stream should be fixed)
  leofang
  committed
  4c4bb17
- fully cythonize Stream
  leofang
  committed
  c593ac9
- make linter happy
  leofang
  committed
  fae5e7f
- Merge branch 'main' into cythonize_more
  leofang
  authored
  bf712bf
Commits on Oct 6, 2025
- address review comments
  leofang
  committed
  2dd9138
- Merge branch 'main' into cythonize_more
  leofang
  committed
  a5e6bcf
Commits on Oct 7, 2025
- give our ABC class (MemoryResource) fast access in C/Cython
  leofang
  committed
  3809a33
- make it clear we need two CUDA contexts for retriving the stream's device
  leofang
  committed
  22b0e2e