[MLIR] Mark LLVM::FMAOp as legal (#144671)

Mark LLVM::FMAOp as legal in configureGpuToNVVMConversionLegality, since
we can handle intrinsic lowering in the NVPTX backend and emit
fma.rn.f32.

Author: chelini

mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp

@@ -429,10 +429,10 @@ void mlir::configureGpuToNVVMConversionLegality(ConversionTarget &target) {
   target.addLegalDialect<::mlir::NVVM::NVVMDialect>();
   target.addIllegalDialect<gpu::GPUDialect>();
   target.addIllegalOp<LLVM::CopySignOp, LLVM::CosOp, LLVM::ExpOp, LLVM::Exp2Op,
-                      LLVM::FAbsOp, LLVM::FCeilOp, LLVM::FFloorOp, LLVM::FMAOp,
-                      LLVM::FRemOp, LLVM::LogOp, LLVM::Log10Op, LLVM::Log2Op,
-                      LLVM::PowOp, LLVM::RoundEvenOp, LLVM::RoundOp,
-                      LLVM::SinOp, LLVM::SqrtOp>();
+                      LLVM::FAbsOp, LLVM::FCeilOp, LLVM::FFloorOp, LLVM::FRemOp,
+                      LLVM::LogOp, LLVM::Log10Op, LLVM::Log2Op, LLVM::PowOp,
+                      LLVM::RoundEvenOp, LLVM::RoundOp, LLVM::SinOp,
+                      LLVM::SqrtOp>();
 
   // TODO: Remove once we support replacing non-root ops.
   target.addLegalOp<gpu::YieldOp, gpu::GPUModuleOp>();

mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir

@@ -1027,7 +1027,7 @@ module attributes {transform.with_named_sequence} {
       legal_ops = ["func.func", "gpu.module", "gpu.yield"],
       illegal_dialects = ["gpu"],
       illegal_ops = ["llvm.copysign", "llvm.cos", "llvm.exp", "llvm.exp2", "llvm.fabs", "llvm.fceil",
-                    "llvm.ffloor", "llvm.fma", "llvm.frem", "llvm.log", "llvm.log10", "llvm.log2", "llvm.pow",
+                    "llvm.ffloor", "llvm.frem", "llvm.log", "llvm.log10", "llvm.log2", "llvm.pow",
                     "llvm.roundeven", "llvm.round", "llvm.sin", "llvm.sqrt"],
       partial_conversion
     } : !transform.any_op

mlir/test/Integration/GPU/CUDA/dump-ptx.mlir

@@ -2,7 +2,7 @@
 // RUN:  | mlir-opt -gpu-lower-to-nvvm-pipeline -debug-only=serialize-to-isa \
 // RUN:  2>&1 | FileCheck %s
 
-// CHECK: Generated by LLVM NVPTX Back-End
+// CHECK-LABEL: Generated by LLVM NVPTX Back-End
 // CHECK: .visible .func kernel_a()
 // CHECK: ret;
 gpu.module @bar {
@@ -11,3 +11,15 @@ gpu.module @bar {
     llvm.return
   }
 }
+
+// CHECK-LABEL: Generated by LLVM NVPTX Back-End
+// CHECK: .visible .func  ({{.+}}) fma(
+// CHECK: fma.rn.f32
+
+gpu.module @foo {
+  llvm.func @fma(%arg0: f32, %arg1: f32) -> f32
+    attributes { gpu.kernel } {
+    %res = llvm.intr.fma (%arg0, %arg1, %arg1) : (f32, f32, f32) -> f32
+    llvm.return %res : f32
+  }
+}