remove hard dependency on modules for mathtoxevm

Author: ianayl

mlir/include/mlir/Conversion/Passes.td

@@ -800,7 +800,7 @@ def ConvertMathToSPIRVPass : Pass<"convert-math-to-spirv"> {
 // MathToXeVM
 //===----------------------------------------------------------------------===//
 
-def ConvertMathToXeVM : Pass<"convert-math-to-xevm", "ModuleOp"> {
+def ConvertMathToXeVM : Pass<"convert-math-to-xevm"> {
   let summary =
       "Convert (fast) math operations to native XeVM/SPIRV equivalents";
   let description = [{

mlir/test/Conversion/MathToXeVM/check-module-behavior.mlir

@@ -0,0 +1,42 @@
+// RUN: mlir-opt --pass-pipeline="builtin.module(convert-math-to-xevm)" %s \
+// RUN:   | FileCheck %s -check-prefixes='CHECK,CHECK-ENTIRE-MODULE'
+// RUN: mlir-opt --pass-pipeline="builtin.module(gpu.module(convert-math-to-xevm))" %s \
+// RUN:   | FileCheck %s -check-prefixes='CHECK,CHECK-ONLY-GPU'
+//
+// Check that MathToXeVM handles nested modules while respecting pass manager.
+
+// CHECK-LABEL: @test_module
+module @test_module {
+  // CHECK-ENTIRE-MODULE: llvm.func @_Z22__spirv_ocl_native_expf(f32) -> f32
+  // CHECK-ONLY-GPU-NOT: llvm.func @_Z22__spirv_ocl_native_expf(f32) -> f32
+
+  // CHECK-LABEL: @test_gpu
+  gpu.module @test_gpu {
+    // CHECK: llvm.func @_Z22__spirv_ocl_native_expf(f32) -> f32
+    gpu.func @exp_gpu() {
+      %c1_f32 = arith.constant 1. : f32
+
+      // CHECK: math.exp
+      %exp_normal_f32 = math.exp %c1_f32 : f32
+
+      // CHECK: llvm.call @_Z22__spirv_ocl_native_expf(%{{.*}}) {fastmathFlags = #llvm.fastmath<afn>} : (f32) -> f32
+      %exp_fast_f32 = math.exp %c1_f32 fastmath<afn> : f32
+
+      gpu.return
+    }
+  }
+
+  // CHECK-LABEL: @exp_func
+  func.func @exp_func() {
+    %c1_f32 = arith.constant 1. : f32
+
+    // CHECK: math.exp
+    %exp_normal_f32 = math.exp %c1_f32 : f32
+
+    // CHECK-ENTIRE-MODULE: llvm.call @_Z22__spirv_ocl_native_expf(%{{.*}}) {fastmathFlags = #llvm.fastmath<afn>} : (f32) -> f32
+    // CHECK-ONLY-GPU: math.exp
+    %exp_fast_f32 = math.exp %c1_f32 fastmath<afn> : f32
+
+    return
+  }
+}