[all-commits] [llvm/llvm-project] 3d04da: [NVPTX] Add support for Shared Cluster Memory addr...

[modiking via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 3d04da5bc022269b3238b2a84347dce7a7afd51b
      https://github.com/llvm/llvm-project/commit/3d04da5bc022269b3238b2a84347dce7a7afd51b
  Author: modiking <mmo at nvidia.com>
  Date:   2025-04-22 (Tue, 22 Apr 2025)

  Changed paths:
    M clang/test/CodeGenCUDA/builtins-sm90.cu
    M llvm/docs/NVPTXUsage.rst
    M llvm/include/llvm/IR/IntrinsicsNVVM.td
    M llvm/lib/IR/AutoUpgrade.cpp
    M llvm/test/Assembler/auto_upgrade_nvvm_intrinsics.ll
    M llvm/test/CodeGen/NVPTX/cp-async-bulk-tensor-g2s.ll
    M llvm/test/CodeGen/NVPTX/cp-async-bulk.ll
    M llvm/test/CodeGen/NVPTX/distributed-shared-cluster.ll
    M mlir/include/mlir/Dialect/LLVMIR/NVVMDialect.h
    M mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td
    M mlir/test/Target/LLVMIR/nvvm/tma_bulk_copy.mlir

  Log Message:
  -----------
  [NVPTX] Add support for Shared Cluster Memory address space [2/2] (#136768)

Adds support for new Shared Cluster Memory Address Space
(SHARED_CLUSTER, addrspace 7). See
https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#distributed-shared-memory
for details.

Follow-up to https://github.com/llvm/llvm-project/pull/135444

1. Update existing codegen/intrinsics in LLVM and MLIR that now use this
address space
2. Auto-upgrade previous intrinsics that used SMEM (addrspace 3) but
were really taking in a shared cluster pointer to the new address space



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications