All pull requests

Pull requests

• Honour CMAKE_INSTALL_INCLUDEDIR when installing public headers

  #14457·tombriden opened Nov 11, 2026·

  ·
  11

  

• Validate CUDA SIMPLE tests with NVRTC in slang-test by default

  pr: non-breaking
  #14455·jkwak-work opened Nov 11, 2026Collaborator·

  ·
  5

  

• Lower integer mad to integer multiply-add instead of float Fma

  pr: non-breaking
  #14452·nv-slang-bot[bot] opened Nov 11, 2026Bot·

  ·
  1

  4

  

• Catch exceptions at getLayout, findAndCheckEntryPoint and getTargetHostCallable

  pr: non-breaking
  #14450·nv-slang-bot[bot] opened Nov 11, 2026Bot·

  ·
  1

  4

  

• Serialize slang-llvm loading and LLVM global initialization for parallel codegen

  pr: non-breaking
  #14446·nv-slang-bot[bot] opened Nov 10, 2026Bot·

  ·
  1

  3

  

• tests: declare diag= exit codes and update the #14344 fixture in docs/generated

  pr: non-breaking
  #14442·nv-slang-bot[bot] opened Nov 10, 2026Bot·

  ·
  4

  

• Skip impossible load/store redundancy searches

  pr: non-breaking
  #14435·tangent-vector opened Nov 10, 2026Contributor·

  ·
  6

  

• SPIR-V/GLSL: do floating-point atomic compare-exchange on integers

  #14414·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  6

  

• render-test: accept indented //TEST_INPUT: lines

  #14413·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  2

  

• SPIR-V/GLSL: diagnose floating-point atomic compare-exchange; fix RWByteAddressBuffer FloatBitwise CAS

  pr: non-breaking
  #14412·nv-slang-bot[bot] opened Nov 8, 2026Bot·

  ·
  1

  4

  

• CUDA: read pointer values through ImmutablePtr with __ldg

  #14406·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  2

  

• CUDA: keep -fp-mode precise arithmetic from FMA contraction

  #14405·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  9

  

• Metal: retry a weak compare-exchange that fails spuriously

  #14404·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  2

  

• WGSL: negate signed vectors with unary minus

  #14403·albertrodriguezclo opened Nov 8, 2026Contributor·

  ·
  1

  2

  

• Add generic shift operators for ILogical

  pr: non-breaking
  #14402·nv-slang-bot[bot] opened Nov 8, 2026Bot·

  ·
  1

  4

  

• CUDA: emit _rn arithmetic under -fp-mode precise

  pr: non-breaking
  #14396·nv-slang-bot[bot] opened Nov 7, 2026Bot·

  ·
  4

  

• Check module-scope varyings against the entry points of their module

  pr: breaking change
  #14395·nv-slang-bot[bot] opened Nov 7, 2026Bot·

  ·
  4

  

• Add literate commenting guidelines to coding conventions

  pr: non-breaking
  #14393·jhelferty-nv opened Nov 7, 2026Contributor·

  ·
  5

  

• Metal: emit read-only buffers as pointers to const

  #14392·albertrodriguezclo opened Nov 6, 2026Contributor·

  ·
  1

  2

  

• WGSL: read a groupshared barrier condition with workgroupUniformLoad

  #14391·albertrodriguezclo opened Nov 6, 2026Contributor·

  ·
  1

  6

  

• Add an invariant modifier for vertex position outputs

  #14389·TareHimself opened Nov 6, 2026·

  ·
  22

  

• Allow static const globals initialized from resource shader parameters

  #14379·minco3 opened Nov 6, 2026Contributor·

  ·
  1

  27

  

• Resolve binary-module dependencies relative to the module source

  pr: non-breaking
  #14362·nv-slang-bot[bot] opened Nov 5, 2026Bot·

  ·
  1

  4

  

• Keep loads through pointers ordered before calls with side effects

  pr: non-breaking
  #14346·albertrodriguezclo opened Nov 5, 2026Contributor·

  ·
  1

  20

  

• Add eliminateUndefinedOpaqueHandles pass for SPIR-V

  pr: non-breaking
  #14342·pdeayton-nv opened Nov 5, 2026Contributor·

  ·
  1

  9