Skip to content

jfriel-oqc/mlir-hello

 
 

Repository files navigation

MLIR Hello Dialect nightly build status

This is the minimal example to look into the way to implement the hello-world kind of program with MLIR. The basic code structure is borrowed from standalone and Toy language in LLVM project.

We constantly check the compatibility with the latest LLVM/MLIR in the nightly build. The status of the build is shown in the badge above.

Prerequisites

We need to build our own MLIR in the local machine in advance. Please follow the build instruction for MLIR here.

Building

Please make sure to build LLVM project first according to the instruction.

mkdir build && cd build
cmake -G Ninja .. -DLLVM_DIR=/path/to/llvm-project/build/lib/cmake/llvm \
  -DMLIR_DIR=/path/to/llvm-project/build/lib/cmake/mlir

cmake --build . --target hello-opt

To run the test, check-hello target will be usable.

To build the documentation from the TableGen description of the dialect operations, run

cmake --build . --target mlir-doc

Execution

hello-opt will lower the MLIR into the bytecode of LLVM.

# Lower MLIR to LLVM IR
$ ./build/bin/hello-opt ./test/Hello/print.mlir > /path/to/print.ll

# Execute the code with LLVM interpreter
$ lli /path/to/print.ll 

1.000000 2.000000 3.000000
4.000000 5.000000 6.000000

Operations

hello.constant (::hello::ConstantOp)

constant

Constant operation turns a literal into an SSA value. The data is attached to the operation as an attribute. For example:

  %0 = "hello.constant"()
  { value = dense<[[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]]> : tensor<2x3xf64> }
  : () -> tensor<2x3xf64>

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable, NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Attributes:

Attribute MLIR Type Description
value ::mlir::DenseElementsAttr 64-bit float elements attribute

Results:

Result Description
«unnamed» tensor of 64-bit float values

hello.print (::hello::PrintOp)

print operation

Syntax:

operation ::= `hello.print` $input attr-dict `:` type($input)

The "print" builtin operation prints a given input tensor, and produces no results.

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable, NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Operands:

Operand Description
input tensor of 64-bit float values or memref of 64-bit float values

hello.world (::hello::WorldOp)

print Hello, World

The "world" operation prints "Hello, World", and produces no results.

Traits: AlwaysSpeculatableImplTrait

Interfaces: ConditionallySpeculatable, NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

About

MLIR Sample dialect

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published