llvm-project/mlir/include/mlir-c/ExecutionEngine.h
Shenghang Tsai 7610b13729
[MLIR] Split ExecutionEngine Initialization out of ctor into an explicit method call (#153524)
Retry landing https://github.com/llvm/llvm-project/pull/153373
## Major changes from previous attempt
- remove the test in CAPI because no existing tests in CAPI deal with
sanitizer exemptions
- update `mlir/docs/Dialects/GPU.md` to reflect the new behavior: load
GPU binary in global ctors, instead of loading them at call site.
- skip the test on Aarch64 since we have an issue with initialization there

---------

Co-authored-by: Mehdi Amini <joker.eph@gmail.com>
2025-08-17 23:07:24 +02:00

98 lines
4.1 KiB
C

//===-- mlir-c/ExecutionEngine.h - Execution engine management ---*- 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
//
//===----------------------------------------------------------------------===//
//
// This header provides basic access to the MLIR JIT. This is minimalist and
// experimental at the moment.
//
//===----------------------------------------------------------------------===//
#ifndef MLIR_C_EXECUTIONENGINE_H
#define MLIR_C_EXECUTIONENGINE_H
#include "mlir-c/IR.h"
#include "mlir-c/Support.h"
#ifdef __cplusplus
extern "C" {
#endif
#define DEFINE_C_API_STRUCT(name, storage) \
struct name { \
storage *ptr; \
}; \
typedef struct name name
DEFINE_C_API_STRUCT(MlirExecutionEngine, void);
#undef DEFINE_C_API_STRUCT
/// Creates an ExecutionEngine for the provided ModuleOp. The ModuleOp is
/// expected to be "translatable" to LLVM IR (only contains operations in
/// dialects that implement the `LLVMTranslationDialectInterface`). The module
/// ownership stays with the client and can be destroyed as soon as the call
/// returns. `optLevel` is the optimization level to be used for transformation
/// and code generation. LLVM passes at `optLevel` are run before code
/// generation. The number and array of paths corresponding to shared libraries
/// that will be loaded are specified via `numPaths` and `sharedLibPaths`
/// respectively.
/// TODO: figure out other options.
MLIR_CAPI_EXPORTED MlirExecutionEngine mlirExecutionEngineCreate(
MlirModule op, int optLevel, int numPaths,
const MlirStringRef *sharedLibPaths, bool enableObjectDump);
/// Initialize the ExecutionEngine. Global constructors specified by
/// `llvm.mlir.global_ctors` will be run. One common scenario is that kernel
/// binary compiled from `gpu.module` gets loaded during initialization. Make
/// sure all symbols are resolvable before initialization by calling
/// `mlirExecutionEngineRegisterSymbol` or including shared libraries.
MLIR_CAPI_EXPORTED void mlirExecutionEngineInitialize(MlirExecutionEngine jit);
/// Destroy an ExecutionEngine instance.
MLIR_CAPI_EXPORTED void mlirExecutionEngineDestroy(MlirExecutionEngine jit);
/// Checks whether an execution engine is null.
static inline bool mlirExecutionEngineIsNull(MlirExecutionEngine jit) {
return !jit.ptr;
}
/// Invoke a native function in the execution engine by name with the arguments
/// and result of the invoked function passed as an array of pointers. The
/// function must have been tagged with the `llvm.emit_c_interface` attribute.
/// Returns a failure if the execution fails for any reason (the function name
/// can't be resolved for instance).
MLIR_CAPI_EXPORTED MlirLogicalResult mlirExecutionEngineInvokePacked(
MlirExecutionEngine jit, MlirStringRef name, void **arguments);
/// Lookup the wrapper of the native function in the execution engine with the
/// given name, returns nullptr if the function can't be looked-up.
MLIR_CAPI_EXPORTED void *
mlirExecutionEngineLookupPacked(MlirExecutionEngine jit, MlirStringRef name);
/// Lookup a native function in the execution engine by name, returns nullptr
/// if the name can't be looked-up.
MLIR_CAPI_EXPORTED void *mlirExecutionEngineLookup(MlirExecutionEngine jit,
MlirStringRef name);
/// Register a symbol with the jit: this symbol will be accessible to the jitted
/// code.
MLIR_CAPI_EXPORTED void
mlirExecutionEngineRegisterSymbol(MlirExecutionEngine jit, MlirStringRef name,
void *sym);
/// Dump as an object in `fileName`.
MLIR_CAPI_EXPORTED void
mlirExecutionEngineDumpToObjectFile(MlirExecutionEngine jit,
MlirStringRef fileName);
#ifdef __cplusplus
}
#endif
#endif // MLIR_C_EXECUTIONENGINE_H