From 202ab273e6eca134b69882f100c666fcd3affbcf Mon Sep 17 00:00:00 2001
From: Julian Gross <julian.gross@dfki.de>
Date: Thu, 9 Jan 2020 09:52:50 +0100
Subject: [PATCH] [mlir] Added missing GPU lowering ops.

Summary:
This diff adds missing GPU lowering ops to MLIR.

Reviewers: herhut, pifon2a, ftynse

Tags: #pre-merge_beta_testing, #llvm

Differential Revision: https://reviews.llvm.org/D72439
---
 .../Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp  |  9 ++++-
 .../GPUToROCDL/LowerGpuOpsToROCDLOps.cpp           |  9 ++++-
 mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir    | 45 ++++++++++++++++++++++
 mlir/test/Conversion/GPUToROCDL/gpu-to-rocdl.mlir  | 45 ++++++++++++++++++++++
 4 files changed, 106 insertions(+), 2 deletions(-)

diff --git a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
index c2493f7..00f89d3 100644
--- a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
+++ b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
@@ -712,7 +712,8 @@ public:
     populateGpuToNVVMConversionPatterns(converter, patterns);
     ConversionTarget target(getContext());
     target.addIllegalDialect<gpu::GPUDialect>();
-    target.addIllegalOp<LLVM::ExpOp>();
+    target.addIllegalOp<LLVM::FAbsOp, LLVM::FCeilOp, LLVM::CosOp,
+                        LLVM::ExpOP>();
     target.addIllegalOp<FuncOp>();
     target.addLegalDialect<LLVM::LLVMDialect>();
     target.addLegalDialect<NVVM::NVVMDialect>();
@@ -739,6 +740,12 @@ void mlir::populateGpuToNVVMConversionPatterns(
                                           NVVM::GridDimYOp, NVVM::GridDimZOp>,
               GPUAllReduceOpLowering, GPUShuffleOpLowering, GPUFuncOpLowering,
               GPUReturnOpLowering>(converter);
+  patterns.insert<OpToFuncCallLowering<AbsFOp>>(converter, "__nv_fabsf",
+                                               "__nv_fabs");
+  patterns.insert<OpToFuncCallLowering<CeilFOp>>(converter, "__nv_ceilf",
+                                               "__nv_ceil");
+  patterns.insert<OpToFuncCallLowering<CosOp>>(converter, "__nv_cosf",
+                                               "__nv_cos");
   patterns.insert<OpToFuncCallLowering<ExpOp>>(converter, "__nv_expf",
                                                "__nv_exp");
 }
diff --git a/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp b/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp
index 8377064..119479d 100644
--- a/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp
+++ b/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp
@@ -51,12 +51,19 @@ public:
         GPUIndexIntrinsicOpLowering<gpu::GridDimOp, ROCDL::GridDimXOp,
                                     ROCDL::GridDimYOp, ROCDL::GridDimZOp>>(
         converter);
+    patterns.insert<OpToFuncCallLowering<AbsFOp>>(converter, "_ocml_fabs_f32",
+                                                 "_ocml_fabs_f64");
+    patterns.insert<OpToFuncCallLowering<CeilFOp>>(converter, "_ocml_ceil_f32",
+                                                 "_ocml_ceil_f64");
+    patterns.insert<OpToFuncCallLowering<CosOp>>(converter, "_ocml_cos_f32",
+                                                 "_ocml_cos_f64");
     patterns.insert<OpToFuncCallLowering<ExpOp>>(converter, "_ocml_exp_f32",
                                                  "_ocml_exp_f64");
 
     ConversionTarget target(getContext());
     target.addLegalDialect<LLVM::LLVMDialect, ROCDL::ROCDLDialect>();
-    target.addIllegalOp<LLVM::ExpOp>();
+    target.addIllegalOp<LLVM::FAbsOp, LLVM::FCeilOp, LLVM::CosOP,
+                        LLVM::ExpOp>();
     target.addDynamicallyLegalOp<FuncOp>(
         [&](FuncOp op) { return converter.isSignatureLegal(op.getType()); });
     if (failed(applyPartialConversion(m, target, patterns, &converter)))
diff --git a/mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir b/mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir
index b1820cb..24bf565 100644
--- a/mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir
+++ b/mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir
@@ -112,6 +112,51 @@ module attributes {gpu.kernel_module} {
 // -----
 
 module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @__nv_fabsf(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @__nv_fabs(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_fabs
+  func @gpu_fabs(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.absf %arg_f32 : f32
+    // CHECK: llvm.call @__nv_fabsf(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.absf %arg_f64 : f64
+    // CHECK: llvm.call @__nv_fabs(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @__nv_ceilf(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @__nv_ceil(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_ceil
+  func @gpu_ceil(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.ceilf %arg_f32 : f32
+    // CHECK: llvm.call @__nv_ceilf(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.ceilf %arg_f64 : f64
+    // CHECK: llvm.call @__nv_ceil(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @__nv_cosf(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @__nv_cos(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_cos
+  func @gpu_cos(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.cos %arg_f32 : f32
+    // CHECK: llvm.call @__nv_cosf(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.cos %arg_f64 : f64
+    // CHECK: llvm.call @__nv_cos(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
   // CHECK: llvm.func @__nv_expf(!llvm.float) -> !llvm.float
   // CHECK: llvm.func @__nv_exp(!llvm.double) -> !llvm.double
   // CHECK-LABEL: func @gpu_exp
diff --git a/mlir/test/Conversion/GPUToROCDL/gpu-to-rocdl.mlir b/mlir/test/Conversion/GPUToROCDL/gpu-to-rocdl.mlir
index f803266..372fe27 100644
--- a/mlir/test/Conversion/GPUToROCDL/gpu-to-rocdl.mlir
+++ b/mlir/test/Conversion/GPUToROCDL/gpu-to-rocdl.mlir
@@ -39,6 +39,51 @@ module attributes {gpu.kernel_module} {
 // -----
 
 module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @_ocml_fabs_f32(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @_ocml_fabs_f64(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_fabs
+  func @gpu_fabs(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.absf %arg_f32 : f32
+    // CHECK: llvm.call @_ocml_fabs_f32(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.absf %arg_f64 : f64
+    // CHECK: llvm.call @_ocml_fabs_f64(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @_ocml_ceil_f32(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @_ocml_ceil_f64(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_ceil
+  func @gpu_ceil(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.ceilf %arg_f32 : f32
+    // CHECK: llvm.call @_ocml_ceil_f32(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.ceilf %arg_f64 : f64
+    // CHECK: llvm.call @_ocml_ceil_f64(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
+  // CHECK: llvm.func @_ocml_cos_f32(!llvm.float) -> !llvm.float
+  // CHECK: llvm.func @_ocml_cos_f64(!llvm.double) -> !llvm.double
+  // CHECK-LABEL: func @gpu_cos
+  func @gpu_cos(%arg_f32 : f32, %arg_f64 : f64) {
+    %result32 = std.cos %arg_f32 : f32
+    // CHECK: llvm.call @_ocml_cos_f32(%{{.*}}) : (!llvm.float) -> !llvm.float
+    %result64 = std.cos %arg_f64 : f64
+    // CHECK: llvm.call @_ocml_cos_f64(%{{.*}}) : (!llvm.double) -> !llvm.double
+    std.return
+  }
+}
+
+// -----
+
+module attributes {gpu.kernel_module} {
   // CHECK: llvm.func @_ocml_exp_f32(!llvm.float) -> !llvm.float
   // CHECK: llvm.func @_ocml_exp_f64(!llvm.double) -> !llvm.double
   // CHECK-LABEL: func @gpu_exp
-- 
2.7.4