automatic code-reformat

src/PIM/Conversion/ONNXToSpatial/Patterns/Tensor/Concat.cpp

@@ -23,8 +23,6 @@ struct Concat : public OpConversionPattern<ONNXConcatOp> {
   }
 };
 
-void populateConcatPatterns(RewritePatternSet& patterns, MLIRContext* ctx) {
-  patterns.insert<Concat>(ctx);
-}
+void populateConcatPatterns(RewritePatternSet& patterns, MLIRContext* ctx) { patterns.insert<Concat>(ctx); }
 
 } // namespace onnx_mlir

src/PIM/Conversion/ONNXToSpatial/Patterns/Tensor/Gather.cpp

@@ -19,11 +19,8 @@ static int64_t normalizeAxis(int64_t axis, int64_t rank) { return axis >= 0 ? ax
 
 static int64_t normalizeIndex(int64_t index, int64_t dimSize) { return index >= 0 ? index : dimSize + index; }
 
-static Value extractSliceAt(Value input,
-                            int64_t axis,
-                            int64_t offset,
-                            ConversionPatternRewriter& rewriter,
-                            Location loc) {
+static Value
+extractSliceAt(Value input, int64_t axis, int64_t offset, ConversionPatternRewriter& rewriter, Location loc) {
   auto inputType = cast<RankedTensorType>(input.getType());
   SmallVector<OpFoldResult> offsets(inputType.getRank(), rewriter.getIndexAttr(0));
   SmallVector<OpFoldResult> sizes;
@@ -110,34 +107,42 @@ struct Gather : OpConversionPattern<ONNXGatherOp> {
     SmallVector<int64_t> flatIndices(indicesAttr.getValues<int64_t>().begin(), indicesAttr.getValues<int64_t>().end());
     Location loc = gatherOp.getLoc();
 
-    auto computeOp = createSpatCompute<1>(
-      rewriter, loc, TypeRange {gatherOp.getResult().getType()}, {}, adaptor.getData(), [&](Value data) -> LogicalResult {
-        Value result;
-        if (indicesType.getRank() == 1) {
-          result = concatGatherSlices(data, axis, flatIndices, axisDim, rewriter, loc);
-        } else if (indicesType.getRank() == 2) {
-          int64_t rowCount = indicesType.getShape()[0];
-          int64_t rowWidth = indicesType.getShape()[1];
-          SmallVector<Value> rows;
-          rows.reserve(rowCount);
-          for (int64_t row = 0; row < rowCount; ++row) {
-            ArrayRef<int64_t> rowIndices(flatIndices.data() + row * rowWidth, rowWidth);
-            Value gatheredRow = concatGatherSlices(data, axis, rowIndices, axisDim, rewriter, loc);
-            if (!gatheredRow)
-              return failure();
-            rows.push_back(addLeadingGatherDim(gatheredRow, axis, rewriter, loc));
-          }
-          result =
-            rows.size() == 1 ? rows.front() : tensor::ConcatOp::create(rewriter, loc, /*axis=*/axis, rows).getResult();
-        } else {
-          return failure();
-        }
+    auto computeOp =
+      createSpatCompute<1>(rewriter,
+                           loc,
+                           TypeRange {gatherOp.getResult().getType()},
+                           {},
+                           adaptor.getData(),
+                           [&](Value data) -> LogicalResult {
+                             Value result;
+                             if (indicesType.getRank() == 1) {
+                               result = concatGatherSlices(data, axis, flatIndices, axisDim, rewriter, loc);
+                             }
+                             else if (indicesType.getRank() == 2) {
+                               int64_t rowCount = indicesType.getShape()[0];
+                               int64_t rowWidth = indicesType.getShape()[1];
+                               SmallVector<Value> rows;
+                               rows.reserve(rowCount);
+                               for (int64_t row = 0; row < rowCount; ++row) {
+                                 ArrayRef<int64_t> rowIndices(flatIndices.data() + row * rowWidth, rowWidth);
+                                 Value gatheredRow = concatGatherSlices(data, axis, rowIndices, axisDim, rewriter, loc);
+                                 if (!gatheredRow)
+                                   return failure();
+                                 rows.push_back(addLeadingGatherDim(gatheredRow, axis, rewriter, loc));
+                               }
+                               result = rows.size() == 1
+                                        ? rows.front()
+                                        : tensor::ConcatOp::create(rewriter, loc, /*axis=*/axis, rows).getResult();
+                             }
+                             else {
+                               return failure();
+                             }
 
-        if (!result)
-          return failure();
-        spatial::SpatYieldOp::create(rewriter, loc, result);
-        return success();
-      });
+                             if (!result)
+                               return failure();
+                             spatial::SpatYieldOp::create(rewriter, loc, result);
+                             return success();
+                           });
     if (failed(computeOp))
       return failure();
     rewriter.replaceOp(gatherOp, computeOp->getResults());

src/PIM/Conversion/ONNXToSpatial/Patterns/Tensor/Resize.cpp

@@ -15,11 +15,8 @@ using namespace mlir;
 namespace onnx_mlir {
 namespace {
 
-static Value extractSliceAt(Value input,
-                            int64_t axis,
-                            int64_t offset,
-                            ConversionPatternRewriter& rewriter,
-                            Location loc) {
+static Value
+extractSliceAt(Value input, int64_t axis, int64_t offset, ConversionPatternRewriter& rewriter, Location loc) {
   auto inputType = cast<RankedTensorType>(input.getType());
   SmallVector<OpFoldResult> offsets(inputType.getRank(), rewriter.getIndexAttr(0));
   SmallVector<OpFoldResult> sizes;
@@ -67,8 +64,7 @@ struct Resize : OpConversionPattern<ONNXResizeOp> {
     if (!inputType || !resultType || !inputType.hasStaticShape() || !resultType.hasStaticShape())
       return failure();
 
-    if (resizeOp.getMode() != "nearest"
-        || resizeOp.getCoordinateTransformationMode() != "asymmetric"
+    if (resizeOp.getMode() != "nearest" || resizeOp.getCoordinateTransformationMode() != "asymmetric"
         || resizeOp.getNearestMode() != "floor")
       return failure();
 
@@ -76,9 +72,10 @@ struct Resize : OpConversionPattern<ONNXResizeOp> {
         || llvm::any_of(resultType.getShape(), [](int64_t dim) { return dim <= 0; }))
       return failure();
 
-    auto computeOp = createSpatCompute<1>(
-      rewriter, resizeOp.getLoc(), TypeRange {resultType}, {}, adaptor.getX(), [&](Value x) {
-        Value result = buildNearestResize(x, inputType.getShape(), resultType.getShape(), /*axis=*/0, rewriter, resizeOp.getLoc());
+    auto computeOp =
+      createSpatCompute<1>(rewriter, resizeOp.getLoc(), TypeRange {resultType}, {}, adaptor.getX(), [&](Value x) {
+        Value result =
+          buildNearestResize(x, inputType.getShape(), resultType.getShape(), /*axis=*/0, rewriter, resizeOp.getLoc());
         spatial::SpatYieldOp::create(rewriter, resizeOp.getLoc(), result);
       });
     rewriter.replaceOp(resizeOp, computeOp.getResults());

src/PIM/Conversion/ONNXToSpatial/Patterns/Tensor/Split.cpp

@@ -12,12 +12,8 @@ namespace {
 
 static int64_t normalizeAxis(int64_t axis, int64_t rank) { return axis >= 0 ? axis : rank + axis; }
 
-static Value extractSliceAt(Value input,
-                            int64_t axis,
-                            int64_t offset,
-                            int64_t size,
-                            ConversionPatternRewriter& rewriter,
-                            Location loc) {
+static Value extractSliceAt(
+  Value input, int64_t axis, int64_t offset, int64_t size, ConversionPatternRewriter& rewriter, Location loc) {
   auto inputType = cast<RankedTensorType>(input.getType());
   SmallVector<OpFoldResult> offsets(inputType.getRank(), rewriter.getIndexAttr(0));
   SmallVector<OpFoldResult> sizes;
@@ -33,9 +29,8 @@ static Value extractSliceAt(Value input,
 struct Split : OpConversionPattern<ONNXSplitOp> {
   using OpConversionPattern::OpConversionPattern;
 
-  LogicalResult matchAndRewrite(ONNXSplitOp splitOp,
-                                ONNXSplitOpAdaptor adaptor,
-                                ConversionPatternRewriter& rewriter) const override {
+  LogicalResult
+  matchAndRewrite(ONNXSplitOp splitOp, ONNXSplitOpAdaptor adaptor, ConversionPatternRewriter& rewriter) const override {
     auto inputType = dyn_cast<RankedTensorType>(adaptor.getInput().getType());
     if (!inputType || !inputType.hasStaticShape())
       return failure();