[Mlir-commits] [mlir] [MLIR] Add XeGPU dialect for Intel GPU (PR #78483)

[Mehdi Amini]

================
@@ -0,0 +1,46 @@
+//===- XeGPUDialect.td - XeGPU 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 MLIR_DIALECT_XEGPU_IR_XEGPUDIALECT_TD
+#define MLIR_DIALECT_XEGPU_IR_XEGPUDIALECT_TD
+
+include "mlir/IR/OpBase.td"
+include "mlir/IR/OpAsmInterface.td"
+include "mlir/IR/AttrTypeBase.td"
+include "mlir/IR/BuiltinTypes.td"
+include "mlir/IR/BuiltinTypeInterfaces.td"
+include "mlir/Interfaces/SideEffectInterfaces.td"
+include "mlir/Interfaces/ViewLikeInterface.td"
+include "mlir/Interfaces/CastInterfaces.td"
+include "mlir/Interfaces/ControlFlowInterfaces.td"
+include "mlir/Interfaces/CopyOpInterface.td"
+include "mlir/Interfaces/InferTypeOpInterface.td"
+include "mlir/Interfaces/ShapedOpInterfaces.td"
+
+def XeGPU_Dialect : Dialect {
+    let name = "xegpu";
+    let cppNamespace = "::mlir::xegpu";
+    let summary = "The XeGPU dialect that models Intel GPU's ISA";
+    let description = [{
+      The XeGPU dialect models Intel Xe ISA semantics but works at vector and
+      TensorDesc data type. It provides 1:1 mappings to match Xe instructions 
+      like DPAS and 2D block load. The matrix size being processed at this level
+      exactly matches the hardware instructions or the intrinsic supported by
+      the lower-level GPU compiler.
+    }];
+
+    let dependentDialects = [
+      "arith::ArithDialect",
+      "memref::MemRefDialect"
----------------
joker-eph wrote:

Dependencies are used when you must load the other dialect to use this one.
For example that can come from folding, constant materialization, or canonicalization patterns.

https://github.com/llvm/llvm-project/pull/78483