All pull requests

Pull requests

• fix: correct cuTile source locations for generated constants

  #1385·Arthur031221 opened Oct 3, 2026·2 of 5 tasks·

  ·
  1

• fix(codegen): compare llc against the IR producer stated at build time

  #1384·sachinsharma3191 opened Oct 3, 2026Contributor·

  ·
  1

  1

• ci: run CodeQL on an 8-core runner

  #1377·roivanov opened Oct 2, 2026Collaborator·2 of 6 tasks·

  ·
  1

• Umbrella readme

  #1376·cryos opened Oct 2, 2026Collaborator·

  ·
  1

• fix: validate cluster launch dimensions

  #1374·efegokdemir opened Oct 1, 2026·2 tasks done·

  ·
  1

  4

• fix(cuda-macros,codegen): keep all-generic library .oxart via #72 anchor

  #1372·dundysm opened Oct 1, 2026·3 of 6 tasks·

  ·
  1

  1

• feat(finalizer): fail a link that drops an expected kernel

  #1367·lucifer1004 opened Sep 30, 2026Contributor·

  ·
  1

• fix(mir-importer): preserve emitted diverging device calls

  #1364·YunusMutlu opened Sep 29, 2026Contributor·5 tasks done·

  ·
  1

• fix(mir-importer): preserve non-terminating drops

  #1361·YunusMutlu opened Sep 29, 2026Contributor·4 tasks done·

  ·
  1

• fix(mir-importer): lower is_val_statically_known to false

  #1359·Arthur031221 opened Sep 29, 2026·2 of 5 tasks·

  ·
  1

  1

• fix(mir-lower): cap enum alignment anchors

  #1357·YunusMutlu opened Sep 28, 2026Contributor·4 tasks done·

  ·
  1

• feat(intrinsics): admit standard-metadata m16n8k16 bf16 sparse MMA

  #1355·YeonwooSung opened Sep 28, 2026·

  ·

• feat(intrinsics): admit bf16x2 to FP8x2 pack conversions

  #1354·YeonwooSung opened Sep 28, 2026·

  ·

• fix(cuda-macros): correct gpu_printf default float and unsigned formatting

  #1353·YeonwooSung opened Sep 28, 2026·

  ·

• feat(wgmma): add f16 m64n128k16 async MMA

  #1352·YeonwooSung opened Sep 28, 2026·

  ·

• fix(cuda-device): select tcgen05 alloc warp by linear thread id

  #1351·YeonwooSung opened Sep 28, 2026·2 tasks done·

  ·

• fix(ptx-schedule): refuse decisions-json paths that overwrite PTX

  #1350·YeonwooSung opened Sep 28, 2026·

  ·
  1

• fix(cuda-device): run mbarrier init on one linear thread

  #1349·YeonwooSung opened Sep 28, 2026·1 of 2 tasks·

  ·

• fix(codegen): warn when llc major differs from rustc LLVM version

  #1348·amansahani opened Sep 27, 2026·4 tasks done·

  ·

• feat(unroll): recognize range for loops

  #1346·midagedev opened Sep 27, 2026Contributor·5 tasks done·

  ·

• fix(shared): enforce typed shared-memory alignment

  #1345·YunusMutlu opened Sep 26, 2026Contributor·3 of 5 tasks·

  ·
  1

• fix(mir-importer): lower from-end ConstantIndex after Subslice

  #1343·YunusMutlu opened Sep 26, 2026Contributor·4 of 5 tasks·

  ·
  1

• feat(cuda-device): add SharedPtr, a CTA-shared pointer typed addrspace(3)

  #1341·lucifer1004 opened Sep 26, 2026Contributor·

  ·

• test(codegen): guard iterator-consumed local-array scalarization

  #1340·0z5a opened Sep 26, 2026·

  ·

• fix(mir-lower): unify equal-width shift representations

  #1338·0z5a opened Sep 26, 2026·

  ·
  1