diff options
| author | Krzysztof Drewniak <Krzysztof.Drewniak@amd.com> | 2022-07-07 17:32:06 +0000 |
|---|---|---|
| committer | Krzysztof Drewniak <Krzysztof.Drewniak@amd.com> | 2022-07-12 15:21:22 +0000 |
| commit | d6ef3d20b4e3768dc30fb229dfa938d8059fffef (patch) | |
| tree | 3684eeb9f0db1257543bad691fc0b3e99f24b4cc /clang/docs | |
| parent | 6626f6fec3d37b78b628b858bdadbbb8301e1a2f (diff) | |
| download | llvm-d6ef3d20b4e3768dc30fb229dfa938d8059fffef.zip llvm-d6ef3d20b4e3768dc30fb229dfa938d8059fffef.tar.gz llvm-d6ef3d20b4e3768dc30fb229dfa938d8059fffef.tar.bz2 | |
[mlir] Remove VectorToROCDL
Between issues such as
https://github.com/llvm/llvm-project/issues/56323, the fact that this
lowering (unlike the code in amdgpu-to-rocdl) does not correctly set
up bounds checks (and thus will cause page faults on reads that might
need to be padded instead), and that fixing these problems would,
essentially, involve replicating amdgpu-to-rocdl, remove
--vector-to-rocdl for being broken. In addition, the lowering does not
support many aspects of transfer_{read,write}, like supervectors, and
may not work correctly in their presence.
We (the MLIR-based convolution generator at AMD) do not use this
conversion pass, nor are we aware of any other clients.
Migration strategies:
- Use VectorToLLVM
- If buffer ops are particularly needed in your application, use
amdgpu.raw_buffer_{load,store}
A VectorToAMDGPU pass may be introduced in the future.
Reviewed By: ThomasRaoux
Differential Revision: https://reviews.llvm.org/D129308
Diffstat (limited to 'clang/docs')
| -rw-r--r-- | clang/docs/ClangFormattedStatus.rst | 10 | ||||
| -rw-r--r-- | clang/docs/tools/clang-formatted-files.txt | 2 |
2 files changed, 0 insertions, 12 deletions
diff --git a/clang/docs/ClangFormattedStatus.rst b/clang/docs/ClangFormattedStatus.rst index 7e714f6..6f62c97 100644 --- a/clang/docs/ClangFormattedStatus.rst +++ b/clang/docs/ClangFormattedStatus.rst @@ -6829,11 +6829,6 @@ tree in terms of conformance to :doc:`ClangFormat` as of: March 06, 2022 17:32:2 - `1` - `0` - :good:`100%` - * - mlir/include/mlir/Conversion/VectorToROCDL - - `1` - - `1` - - `0` - - :good:`100%` * - mlir/include/mlir/Conversion/VectorToSCF - `1` - `1` @@ -7609,11 +7604,6 @@ tree in terms of conformance to :doc:`ClangFormat` as of: March 06, 2022 17:32:2 - `2` - `0` - :good:`100%` - * - mlir/lib/Conversion/VectorToROCDL - - `1` - - `1` - - `0` - - :good:`100%` * - mlir/lib/Conversion/VectorToSCF - `1` - `1` diff --git a/clang/docs/tools/clang-formatted-files.txt b/clang/docs/tools/clang-formatted-files.txt index 5e1c6f1..cc55f9b4 100644 --- a/clang/docs/tools/clang-formatted-files.txt +++ b/clang/docs/tools/clang-formatted-files.txt @@ -7627,7 +7627,6 @@ mlir/include/mlir/Conversion/TosaToSCF/TosaToSCF.h mlir/include/mlir/Conversion/TosaToStandard/TosaToStandard.h mlir/include/mlir/Conversion/VectorToGPU/VectorToGPU.h mlir/include/mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h -mlir/include/mlir/Conversion/VectorToROCDL/VectorToROCDL.h mlir/include/mlir/Conversion/VectorToSCF/VectorToSCF.h mlir/include/mlir/Conversion/VectorToSPIRV/VectorToSPIRV.h mlir/include/mlir/Conversion/VectorToSPIRV/VectorToSPIRVPass.h @@ -8068,7 +8067,6 @@ mlir/lib/Conversion/TosaToStandard/TosaToStandard.cpp mlir/lib/Conversion/TosaToStandard/TosaToStandardPass.cpp mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp -mlir/lib/Conversion/VectorToROCDL/VectorToROCDL.cpp mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp mlir/lib/Conversion/VectorToSPIRV/VectorToSPIRVPass.cpp mlir/lib/Dialect/Traits.cpp |