All pull requests

Pull requests

• [Clang][SYCL] Detect Intel GPUs for --offload-arch=native on a SPIR-V target

  #229938·srividya-sundaram opened Oct 7, 2026Contributor·

  ·

• [Clang][SYCL] Reject an NVIDIA offload target for SYCL

  #229937·srividya-sundaram opened Oct 7, 2026Contributor·

  ·

• [MLIR][XeGPU] Rename Xe3 uArch to Xe3p and add 2D block memory restrictions

  mlir:gpumlir
  #229936·silee2 opened Oct 7, 2026Contributor·

  ·
  2

• [InlineCost][Draft] Add MemorySSA to InlineCost duplicate load detection

  #229934·EthanLuisMcDonough opened Oct 7, 2026Member·3 tasks·

  ·

• [RuntimeDyld] Implement more relocations for 32-bit PowerPC

  #229933·mattst88 opened Oct 7, 2026Contributor·

  ·

• [clang-tidy][docs] Fix ExplicitConstructorCheck example links

  clang-tidyclang-tools-extra
  #229932·joker-eph opened Oct 7, 2026Contributor·

  ·
  1

• [Darwin][TSan][Test-only] Make flaky norace-objcxx-run-time.mm test diagnosable

  compiler-rt:sanitizercompiler-rt:tsancompiler-rt
  #229931·DanBlackwell opened Oct 7, 2026Contributor·

  ·
  1

• [lldb] Use StringRef in formatters::ExtractIndexFromString

  lldb
  #229930·bulbazord opened Oct 7, 2026Member·

  ·
  1

• [VPlan] Don't merge allocas across lanes, parts or recipes.

  llvm:transformsvectorizers
  #229929·fhahn opened Oct 7, 2026Contributor·

  ·
  1

  2

• [AArch64][GlobalISel] More test coverage for vector immediates. NFC

  backend:AArch64skip-precommit-approval
  #229928·davemgreen opened Oct 7, 2026Contributor·

  ·
  1

• [LV] Have ExtractValue match the shape of the operand.

  llvm:transformsvectorizers
  #229927·fhahn opened Oct 7, 2026Contributor·

  ·
  2

• [AMDGPU] Analyze subword loads in redundant AND deletion

  backend:AMDGPUllvm:globalisel
  #229926·LU-JOHN opened Oct 7, 2026Contributor·

  ·
  1

• [VPlan] Re-run makeScalarizationDecisions after call widening.

  llvm:transformsvectorizers
  #229925·fhahn opened Oct 7, 2026Contributor·

  ·
  4

• AMDGPU: Stop relying on -amdgpu-scalarize-global-loads=false in ALU tests

  #229924·arsenm opened Oct 7, 2026Contributor·

  ·
  2

• [lldb] Update formatter_bytecode.py to ABI version 2

  lldb
  #229923·kastiglione opened Oct 7, 2026Contributor·

  ·
  2

• [AMDGPU] Add a static performance simulator for GFX1250

  backend:AMDGPU
  #229922·vangthao95 opened Oct 7, 2026Contributor·

  ·
  7

• [VPlan] Model the connected main-loop CFG and epilogue resume values.

  llvm:transformsvectorizers
  #229921·fhahn opened Oct 7, 2026Contributor·

  ·
  1

• [NVPTX] Generalize register-or-immediate operands for integer instructions

  backend:NVPTX
  #229920·AlexMaclean opened Oct 7, 2026Member·

  ·
  1

  

• [MemProf] Match call sites with line 0 (C++ non-virtual thunks)

  llvm:transformsPGO
  #229919·david-xl opened Oct 7, 2026Contributor·

  ·
  12

• [AMDGPU] Require wave ID masks to be constants

  backend:AMDGPUllvm:analysis
  #229917·krzysz00 opened Oct 7, 2026Contributor·2/2·

  ·
  1

• [AMDGPU][NFC] Pre-commit tests to not match non-constant wave ID masks

  backend:AMDGPUllvm:analysis
  #229916·krzysz00 opened Oct 7, 2026Contributor·1/2·

  ·
  2

• [CIR] Put try_throw's normal destination in the throw's region

  clangClangIR
  #229915·mmha opened Oct 7, 2026Contributor·

  ·
  2

• [flang][DOCONCURRENT] Fix device mapping of several live-ins kinds

  flang:fir-hlfirflang
  #229912·afanfa opened Oct 7, 2026·

  ·
  4

• [mlir][bytecode] Allow custom encodings for mutable types

  mlir:coremlir
  #229910·bcardosolopes opened Oct 7, 2026Member·

  ·
  1

• github-upload-release: Update script to work with draft releases

  #229909·tstellar opened Oct 7, 2026Contributor·

  ·
  1