All pull requests

Pull requests

• fix(codegen): keep the artifact objects of rlibs out of the LTO input

  #1496·fxeqxmulfx opened Oct 11, 2026·4 of 5 tasks·

  ·
  1

  1

• docs(cuda-oxide): state what sync_threads orders at compile time

  #1493·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

  1

• fix(cuda-core): return an error on interior NUL in module loading

  #1492·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

• docs(cuda-oxide): link the book and crate READMEs to NVIDIA/cuda-rust

  #1491·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

• fix(cuda-oxide): point repository metadata and the ICE report URL at NVIDIA/cuda-rust

  #1490·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

• fix(cargo-oxide): scaffold and clone from NVIDIA/cuda-rust

  #1489·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

• fix(cargo-oxide): take the scaffold's host crates from cuda-host's git source

  #1488·alejandro-soto-franco opened Oct 11, 2026Contributor·

  ·
  1

• fix: return jit errors for malformed cutile dimensions

  #1486·uurl opened Oct 11, 2026Contributor·

  ·
  1

  1

• fix(intrinsics): admit FP16 m16n8k8 MMA on sm_75

  #1484·praveenperera opened Oct 11, 2026·9 tasks done·

  ·
  1

  1

• fix(cutile): attribute generated entry setup to the user signature

  #1480·elibol opened Oct 10, 2026Collaborator·

  ·
  1

• build: add CUDA 13.4 and 13.x versions to default candidates

  #1474·temotskipa opened Oct 10, 2026·2 of 4 tasks·

  ·
  1

• fix: standardize non-negative enum types for windows msvc

  #1473·temotskipa opened Oct 10, 2026·2 tasks done·

  ·
  1

  1

• fix(cuda-macros): obey #[inline] on generic device functions

  #1472·fxeqxmulfx opened Oct 10, 2026·4 of 5 tasks·

  ·
  1

  1

• fix(cutile): preserve boolean and unsigned literal values

  #1470·filipkunc opened Oct 10, 2026·2 tasks done·

  ·
  1

  1

• fix: keep symbolic shape on reduce results

  #1456·Arthur031221 opened Oct 8, 2026·

  ·
  1

  1

• fix(mir-importer): normalize pointer address spaces in tuple construction

  cuda-oxide
  #1454·uurl opened Oct 8, 2026Contributor·

  ·
  1

  1

• fix(cutile): validate full tensor shapes before allocation

  #1452·uurl opened Oct 8, 2026Contributor·

  ·
  1

  1

• fix: validate runtime tensor strides before kernel launch

  #1450·uurl opened Oct 8, 2026Contributor·

  ·
  1

  1

• fix(cutile-rs): reject overflowing section lengths and varints in bytecode decoder

  #1448·dawnop opened Oct 8, 2026·

  ·
  1

  1

• feat(cuda-core): add borrowed DeviceBuffer subviews

  runtime
  #1447·uurl opened Oct 8, 2026Contributor·4 tasks done·

  ·
  1

• feat(oxide-artifacts): add strong anchor-only host objects, bump to 0.2.2

  cuda-oxide
  #1446·dundysm opened Oct 8, 2026·8 tasks done·

  ·
  1

• feat(cuda-core): add CudaContext::mem_info for the (free, total) device memory query

  #1445·midagedev opened Oct 8, 2026Contributor·2 tasks done·

  ·
  1

  1

• cuda-async: non-blocking drop for in-flight device futures (reaper)

  cutile-rspost-migration-pr
  #1417·elibol opened Oct 8, 2026Collaborator·

  ·
  1

• docs: describe the CUDA graph replay-safety contract

  cutile-rspost-migration-pr
  #1416·elibol opened Oct 8, 2026Collaborator·

  ·

• fix(simt): soundness and correctness fixes for the cuda-oxide bridge and VMM

  cutile-rspost-migration-pr
  #1415·elibol opened Oct 8, 2026Collaborator·

  ·