projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 960ab52) | patch
[mlir][nvvm] Add `cp.async.bulk.tensor.shared.cluster.global`
authorGuray Ozen <guray.ozen@gmail.com>
Mon, 17 Jul 2023 12:42:30 +0000 (14:42 +0200)
committerGuray Ozen <guray.ozen@gmail.com>
Mon, 17 Jul 2023 15:10:39 +0000 (17:10 +0200)
commit28555793b1e58ee2e147ea62de42a9583fadbc5a
tree91479399c8a4e504ffdf24594150ac9c0936f19ftree | snapshot
parent960ab5225bea8f58e87eab8fd46d55ce9562b33dcommit | diff
[mlir][nvvm] Add `cp.async.bulk.tensor.shared.cluster.global`

This work introduce `cp.async.bulk.tensor.shared.cluster.global` in NVVM dialect that executes load using TMA.

Depends on D155056

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D155060
mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td diff | blob | history
mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp diff | blob | history
mlir/test/Conversion/NVVMToLLVM/nvvm-to-llvm.mlir diff | blob | history
Domain: System / Toolchain;