[mlir][index] Add boilerplate for the `index` dialect

[Mogball]

Summary

This patch introduces the index dialect and associated boilerplate for
adding ops and enums (comparison predicates).

[Mogball]

update commandline test

[rriddle]

This skeleton LGTM as an initial starting point.

mlir/include/mlir/Dialect/Index/IR/IndexDialect.td

84–86

Can you drop these until needed?

mlir/lib/Dialect/Index/IR/IndexDialect.cpp

13

mlir/lib/Dialect/Index/IR/IndexOps.cpp

14

Same here.

[jpienaar]

mlir/include/mlir/Dialect/Index/IR/IndexDialect.td

25

index type (builtin.index) ? Just to be clear that for here it is the concrete type instance of interest

38

the maximum -> maximum

41

You've switched here to consider what happens for 32 bit targets, but it's a bit abrupt/you need to read to the end before having the context for the start.

mlir/lib/Dialect/Index/IR/IndexAttrs.cpp

2

Same

mlir/lib/Dialect/Index/IR/IndexDialect.cpp

2

No need to specify file type in .cpp file

[nicolasvasilache]

Thanks for pushing on this @Mogball !

[Mogball]

Address reviewer comments

[Mogball]

mlir/include/mlir/Dialect/Index/IR/IndexDialect.td

41

I clarified the paragraph a bit.

[Mogball]

sink changes

[fmayer]

This has broken our sanitizer buildbot: https://lab.llvm.org/buildbot/#/builders/5/builds/28427/steps/9/logs/stdio

[4543/6866] Building IndexOpsDialect.h.inc...
FAILED: tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsDialect.h.inc /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsDialect.h.inc 
cd /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan && /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/bin/mlir-tblgen -gen-dialect-decls -dialect=index -I /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/llvm/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td --write-if-changed -o tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsDialect.h.inc -d tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsDialect.h.inc.d
/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td:25:35: error: Variable not defined: 'NoSideEffect'
    : Op<IndexDialect, mnemonic, [NoSideEffect] # traits>;
                                  ^
[4544/6866] Building IndexOpsTypes.cpp.inc...
FAILED: tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.cpp.inc /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.cpp.inc 
cd /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan && /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/bin/mlir-tblgen -gen-typedef-defs -typedefs-dialect=index -I /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/llvm/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td --write-if-changed -o tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.cpp.inc -d tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.cpp.inc.d
/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td:25:35: error: Variable not defined: 'NoSideEffect'
    : Op<IndexDialect, mnemonic, [NoSideEffect] # traits>;
                                  ^
[4545/6866] Building BufferizationOps.cpp.inc...
[4546/6866] Building BufferizationOps.h.inc...
[4547/6866] Building IndexOpsTypes.h.inc...
FAILED: tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.h.inc /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.h.inc 
cd /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan && /b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/bin/mlir-tblgen -gen-typedef-decls -typedefs-dialect=index -I /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/llvm/include -I/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include -I/b/sanitizer-x86_64-linux-fast/build/llvm_build_msan/tools/mlir/include /b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td --write-if-changed -o tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.h.inc -d tools/mlir/include/mlir/Dialect/Index/IR/IndexOpsTypes.h.inc.d
/b/sanitizer-x86_64-linux-fast/build/llvm-project/mlir/include/mlir/Dialect/Index/IR/IndexOps.td:25:35: error: Variable not defined: 'NoSideEffect'
    : Op<IndexDialect, mnemonic, [NoSideEffect] # traits>;

[Mogball]

Thanks. The fix should've been pushed out already

mlir/include/mlir/Dialect/CMakeLists.txt

 add_subdirectory(AMDGPU)
 add_subdirectory(AMX)
 add_subdirectory(Affine)
 add_subdirectory(Arith)
 add_subdirectory(ArmNeon)
 add_subdirectory(ArmSVE)
 add_subdirectory(Async)
 add_subdirectory(Bufferization)
 add_subdirectory(Complex)
 add_subdirectory(ControlFlow)
 add_subdirectory(DLTI)
 add_subdirectory(EmitC)
 add_subdirectory(Func)
 add_subdirectory(GPU)
+add_subdirectory(Index)
 add_subdirectory(LLVMIR)
 add_subdirectory(Linalg)
 add_subdirectory(MLProgram)
 add_subdirectory(Math)
 add_subdirectory(MemRef)
 add_subdirectory(NVGPU)
 add_subdirectory(OpenACC)
 add_subdirectory(OpenMP)

mlir/include/mlir/Dialect/Index/CMakeLists.txt

+add_subdirectory(IR)

mlir/include/mlir/Dialect/Index/IR/CMakeLists.txt

+set(LLVM_TARGET_DEFINITIONS IndexEnums.td)
+mlir_tablegen(IndexEnums.h.inc -gen-enum-decls)
+mlir_tablegen(IndexEnums.cpp.inc -gen-enum-defs)
+mlir_tablegen(IndexAttrs.h.inc -gen-attrdef-decls -attrdefs-dialect=index)
+mlir_tablegen(IndexAttrs.cpp.inc -gen-attrdef-defs -attrdefs-dialect=index)
+add_mlir_dialect(IndexOps index)
+add_mlir_doc(IndexOps IndexOps Dialects/ -gen-dialect-doc)

mlir/include/mlir/Dialect/Index/IR/IndexAttrs.h

+//===- IndexAttrs.h - Index attribute declarations ----------------*- C++-*-==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef MLIR_DIALECT_INDEX_IR_INDEXATTRS_H
+#define MLIR_DIALECT_INDEX_IR_INDEXATTRS_H
+
+#include "mlir/IR/Attributes.h"
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Declarations
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexEnums.h.inc"
+
+#define GET_ATTRDEF_CLASSES
+#include "mlir/Dialect/Index/IR/IndexAttrs.h.inc"
+
+#endif // MLIR_DIALECT_INDEX_IR_INDEXATTRS_H

mlir/include/mlir/Dialect/Index/IR/IndexDialect.h

+//===- IndexDialect.h - Index dialect declaration -----------------*- C++-*-==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef MLIR_DIALECT_INDEX_IR_INDEXDIALECT_H
+#define MLIR_DIALECT_INDEX_IR_INDEXDIALECT_H
+
+#include "mlir/IR/Dialect.h"
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Declarations
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexOpsDialect.h.inc"
+
+#endif // MLIR_DIALECT_INDEX_IR_INDEXDIALECT_H

mlir/include/mlir/Dialect/Index/IR/IndexDialect.td

+//===- IndexDialect.td - Index dialect definition ----------*- tablegen -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef INDEX_DIALECT
+#define INDEX_DIALECT
+
+include "mlir/IR/DialectBase.td"
+
+//===----------------------------------------------------------------------===//
+// IndexDialect
+//===----------------------------------------------------------------------===//
+
+def IndexDialect : Dialect {
+  let name = "index";
+
+  let summary = "The Index dialect";
+  let description = [{
+    The Index dialect contains operations for manipulating values of the builtin
+    `index` type. The index type models target-specific values of pointer width,
+    like `intptr_t`. Index values are typically used as loop bounds, array
+    subscripts, tensor dimensions, etc.
+
+    The operations in this dialect operate exclusively on scalar index types.
+    The dialect and its operations treat the index type as signless and contains
+    signed and unsigned versions of certain operations where the distinction is
+    meaningful. In particular, the operations and transformations are careful to
+    be aware of the target-independent-ness of the index type, such as when
+    folding.
+
+    The folding semantics of the Index dialect operations ensure that folding
+    produces the same results irrespective of the eventual target pointer width.
+    All index constants are stored in `APInt`s of maximum index bitwidth: 64.
+    Operations are folded using 64-bit integer arithmetic.
+
+    For operations where the values of the upper 32 bits don't impact the values
+    of the lower 32 bits, no additional handling is required because if the
+    target is 32-bit, the truncated folded result will be the same as if the
+    operation were computed with 32-bit arithmetic, and if the target is 64-bit,
+    the fold result is valid by default.
+
+    Consider addition: an overflow in 32-bit is the same as truncating the
+    result computed in 64-bit. For example, `add(0x800000008, 0x800000008)` is
+    `0x1000000010` in 64-bit, which truncates to `0x10`, the same result as
+    truncating the operands first: `add(0x08, 0x08)`. Specifically, an operation
+    `f` can always be folded if it satisfies the following for all 64-bit values
+    of `a` and `b`:
+
+    ```
+    trunc(f(a, b)) = f(trunc(a), trunc(b))
+    ```
+
+    When materializing target-specific code, constants just need to be truncated
+    as appropriate.
+
+    Operations where the values of the upper 32 bits do impact the values of the
+    lower 32 bits are not folded if the results would be different in 32-bit.
+    These are operations that right shift -- division, remainder, etc. These
+    operations are only folded for subsets of `a` and `b` for which the above
+    property is satisfied. This is checked per fold attempt.
+
+    Consider division: the 32-bit computation will differ from 64-bit if the
+    latter results in a high bit shifted into the lower 32 bits. For example,
+    `div(0x100000002, 2)` is `0x80000001` in 64-bit but `0x01` in 32-bit; it
+    cannot be folded. However, `div(0x200000002, 2)` can be folded. The 64-bit
+    result is `0x100000001`, which truncated to 32 bits is `0x01`. The 32-bit
+    result of the operation with truncated operands `div(0x02, 2)` which is
+    `0x01`, the same as truncating the 64-bit result.
+  }];
+
+  let cppNamespace = "::mlir::index";
+
+  let extraClassDeclaration = [{
+    /// Register all dialect attributes.
+    void registerAttributes();
+    /// Register all dialect operations.
+    void registerOperations();
+  }];
+}
+
+#endif // INDEX_DIALECT

mlir/include/mlir/Dialect/Index/IR/IndexEnums.td

+//===- IndexEnums.td - Index enum definitions --------------*- tablegen -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef INDEX_ENUMS
+#define INDEX_ENUMS
+
+include "mlir/IR/EnumAttr.td"
+
+#endif // INDEX_ENUMS

mlir/include/mlir/Dialect/Index/IR/IndexOps.h

+//===- IndexOps.h - Index operation declarations ------------------*- C++-*-==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef MLIR_DIALECT_INDEX_IR_INDEXOPS_H
+#define MLIR_DIALECT_INDEX_IR_INDEXOPS_H
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Declarations
+//===----------------------------------------------------------------------===//
+
+#define GET_OP_CLASSES
+#include "mlir/Dialect/Index/IR/IndexOps.h.inc"
+
+#endif // MLIR_DIALECT_INDEX_IR_INDEXOPS_H

mlir/include/mlir/Dialect/Index/IR/IndexOps.td

+//===- IndexOps.td - Index operation definitions -----------*- tablegen -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef INDEX_OPS
+#define INDEX_OPS
+
+include "mlir/Dialect/Index/IR/IndexDialect.td"
+include "mlir/Interfaces/SideEffectInterfaces.td"
+include "mlir/IR/OpBase.td"
+
+//===----------------------------------------------------------------------===//
+// IndexOp
+//===----------------------------------------------------------------------===//
+
+/// Base class for Index dialect operations.
+class IndexOp<string mnemonic, list<Trait> traits = []>
+    : Op<IndexDialect, mnemonic, [NoSideEffect] # traits>;
+
+#endif // INDEX_OPS

mlir/include/mlir/InitAllDialects.h

 #include "mlir/Dialect/Bufferization/Transforms/FuncBufferizableOpInterfaceImpl.h"
 #include "mlir/Dialect/Complex/IR/Complex.h"
 #include "mlir/Dialect/ControlFlow/IR/ControlFlow.h"
 #include "mlir/Dialect/DLTI/DLTI.h"
 #include "mlir/Dialect/EmitC/IR/EmitC.h"
 #include "mlir/Dialect/Func/IR/FuncOps.h"
 #include "mlir/Dialect/GPU/IR/GPUDialect.h"
 #include "mlir/Dialect/GPU/TransformOps/GPUTransformOps.h"
+#include "mlir/Dialect/Index/IR/IndexDialect.h"
 #include "mlir/Dialect/LLVMIR/LLVMDialect.h"
 #include "mlir/Dialect/LLVMIR/NVVMDialect.h"
 #include "mlir/Dialect/LLVMIR/ROCDLDialect.h"
 #include "mlir/Dialect/Linalg/IR/Linalg.h"
 #include "mlir/Dialect/Linalg/TransformOps/LinalgTransformOps.h"
 #include "mlir/Dialect/Linalg/Transforms/BufferizableOpInterfaceImpl.h"
 #include "mlir/Dialect/Linalg/Transforms/TilingInterfaceImpl.h"
 #include "mlir/Dialect/MLProgram/IR/MLProgram.h"
registry.insert<acc::OpenACCDialect,
                   async::AsyncDialect,
                   bufferization::BufferizationDialect,
                   cf::ControlFlowDialect,
                   complex::ComplexDialect,
                   DLTIDialect,
                   emitc::EmitCDialect,
                   func::FuncDialect,
                   gpu::GPUDialect,
+                  index::IndexDialect,
                   LLVM::LLVMDialect,
                   linalg::LinalgDialect,
                   math::MathDialect,
                   memref::MemRefDialect,
                   ml_program::MLProgramDialect,
                   nvgpu::NVGPUDialect,
                   scf::SCFDialect,
                   omp::OpenMPDialect,

mlir/lib/Dialect/CMakeLists.txt

 add_subdirectory(Affine)
 add_subdirectory(AMDGPU)
 add_subdirectory(Arith)
 add_subdirectory(ArmNeon)
 add_subdirectory(ArmSVE)
 add_subdirectory(Async)
 add_subdirectory(AMX)
 add_subdirectory(Bufferization)
 add_subdirectory(Complex)
 add_subdirectory(ControlFlow)
 add_subdirectory(DLTI)
 add_subdirectory(EmitC)
 add_subdirectory(Func)
 add_subdirectory(GPU)
+add_subdirectory(Index)
 add_subdirectory(Linalg)
 add_subdirectory(LLVMIR)
 add_subdirectory(Math)
 add_subdirectory(MemRef)
 add_subdirectory(MLProgram)
 add_subdirectory(NVGPU)
 add_subdirectory(OpenACC)
 add_subdirectory(OpenMP)

mlir/lib/Dialect/Index/CMakeLists.txt

+add_subdirectory(IR)

mlir/lib/Dialect/Index/IR/CMakeLists.txt

+add_mlir_dialect_library(MLIRIndexDialect
+  IndexAttrs.cpp
+  IndexDialect.cpp
+  IndexOps.cpp
+
+  DEPENDS
+  MLIRIndexOpsIncGen
+
+  LINK_LIBS PUBLIC
+  MLIRDialect
+  MLIRIR
+  )

mlir/lib/Dialect/Index/IR/IndexAttrs.cpp

+//===- IndexAttrs.cpp - Index attribute definitions ------------------------==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexAttrs.h"
+#include "mlir/Dialect/Index/IR/IndexDialect.h"
+
+using namespace mlir;
+using namespace mlir::index;
+
+//===----------------------------------------------------------------------===//
+// IndexDialect
+//===----------------------------------------------------------------------===//
+
+void IndexDialect::registerAttributes() {
+  addAttributes<
+#define GET_ATTRDEF_LIST
+#include "mlir/Dialect/Index/IR/IndexAttrs.cpp.inc"
+      >();
+}
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Declarations
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexEnums.cpp.inc"
+
+#define GET_ATTRDEF_CLASSES
+#include "mlir/Dialect/Index/IR/IndexAttrs.cpp.inc"

mlir/lib/Dialect/Index/IR/IndexDialect.cpp

+//===- IndexDialect.cpp - Index dialect definition -------------------------==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexDialect.h"
+
+using namespace mlir;
+using namespace mlir::index;
+
+//===----------------------------------------------------------------------===//
+// IndexDialect
+//===----------------------------------------------------------------------===//
+
+void IndexDialect::initialize() {
+  registerAttributes();
+  registerOperations();
+}
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Definitions
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexOpsDialect.cpp.inc"

mlir/lib/Dialect/Index/IR/IndexOps.cpp

+//===- IndexOps.cpp - Index operation definitions --------------------------==//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/Index/IR/IndexOps.h"
+#include "mlir/Dialect/Index/IR/IndexDialect.h"
+
+using namespace mlir;
+using namespace mlir::index;
+
+//===----------------------------------------------------------------------===//
+// IndexDialect
+//===----------------------------------------------------------------------===//
+
+void IndexDialect::registerOperations() {
+  addOperations<
+#define GET_OP_LIST
+#include "mlir/Dialect/Index/IR/IndexOps.cpp.inc"
+      >();
+}
+
+//===----------------------------------------------------------------------===//
+// ODS-Generated Definitions
+//===----------------------------------------------------------------------===//
+
+#define GET_OP_CLASSES
+#include "mlir/Dialect/Index/IR/IndexOps.cpp.inc"

mlir/test/mlir-opt/commandline.mlir

 // CHECK-NEXT: bufferization
 // CHECK-NEXT: builtin
 // CHECK-NEXT: cf
 // CHECK-NEXT: complex
 // CHECK-NEXT: dlti
 // CHECK-NEXT: emitc
 // CHECK-NEXT: func
 // CHECK-NEXT: gpu
+// CHECK-NEXT: index
 // CHECK-NEXT: linalg
 // CHECK-NEXT: llvm
 // CHECK-NEXT: math
 // CHECK-NEXT: memref
 // CHECK-NEXT: ml_program
 // CHECK-NEXT: nvgpu
 // CHECK-NEXT: nvvm
 // CHECK-NEXT: omp