NFC: Use the free function variants for dyn_cast/cast/isa/....

loop_nest.cc

@@ -17,6 +17,7 @@
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallString.h"
 #include "mlir/IR/Builders.h"
+#include "mlir/Support/LLVM.h"
 #include "sequence.h"
 #include "util.h"
 
@@ -106,7 +107,7 @@ const IterationSpace &IterationSpaceAnalysis::ComputeIterationSpace(
     loop_names.reserve(num_loops);
 
     for (mlir::Attribute attr : compute_op.Loops()) {
-      LoopAttr loop = attr.cast<LoopAttr>();
+      LoopAttr loop = mlir::cast<LoopAttr>(attr);
       loop_names.push_back(loop.name());
       exprs.push_back(loop.iter());
     }
@@ -265,10 +266,10 @@ mlir::LogicalResult VerifyLoopNestWellFormed(
   int domain_size = shape.Dimensions().size();
 
   for (int i = 0, e = loop_nest.size(); i < e; ++i) {
-    LoopAttr loop = loop_nest[i].dyn_cast<LoopAttr>();
+    LoopAttr loop = mlir::dyn_cast<LoopAttr>(loop_nest[i]);
     // Ensure that symbols are unique in the loop nest.
     for (int j = 0; j < i; ++j) {
-      if (loop.name() == loop_nest[j].cast<LoopAttr>().name()) {
+      if (loop.name() == mlir::cast<LoopAttr>(loop_nest[j]).name()) {
         return mlir::emitError(loc)
                << "name " << loop.name() << " used twice in the same loop nest";
       }
@@ -317,7 +318,7 @@ class LoopNestState {
     int common_prefix_size = 0;
     for (int e = std::min(loop_nest.size(), open_loops_.size());
          common_prefix_size < e; ++common_prefix_size) {
-      LoopAttr loop = loop_nest[common_prefix_size].cast<LoopAttr>();
+      LoopAttr loop = mlir::cast<LoopAttr>(loop_nest[common_prefix_size]);
       if (loop.name() != open_loops_[common_prefix_size]) break;
     }
 
@@ -326,7 +327,7 @@ class LoopNestState {
 
     // Add remaining loops to the current fusion prefix.
     for (mlir::Attribute attribute : loop_nest.drop_front(common_prefix_size)) {
-      LoopAttr loop = attribute.cast<LoopAttr>();
+      LoopAttr loop = mlir::cast<LoopAttr>(attribute);
 
       if (closed_loops_.count(loop.name()) > 0) {
         return op.EmitError() << "occurrences of loop " << loop.name()
@@ -500,7 +501,7 @@ static mlir::LogicalResult VerifyLoopRanges(
     const LoopFusionAnalysis &fusion_analysis,
     const SequenceAnalysis &sequence_analysis) {
   for (mlir::Attribute attr : loop_nest) {
-    LoopAttr loop = attr.cast<LoopAttr>();
+    LoopAttr loop = mlir::cast<LoopAttr>(attr);
     const LoopFusionClass &fusion_class = fusion_analysis.GetClass(loop.name());
     for (const auto &dimension : fusion_class.getDomain()) {
       if (sequence_analysis.IsBefore(op, dimension.value.defining_op())) {
@@ -670,7 +671,7 @@ mlir::LogicalResult LoopFusionAnalysis::Init(
 // Returns the unroll factor of the `pos`-th loop in the given compute op.
 // Expects the op to have a well-formed loop nest attribute.
 static unsigned ExtractUnrollFactor(const ComputeOpInstance &op, unsigned pos) {
-  auto loop = op.Loops()[pos].cast<LoopAttr>();
+  auto loop = mlir::cast<LoopAttr>(op.Loops()[pos]);
   if (mlir::IntegerAttr unroll_factor = loop.unroll()) {
     return unroll_factor.getInt();
   }
@@ -686,7 +687,7 @@ mlir::LogicalResult LoopFusionAnalysis::RegisterLoop(
   loop_names.reserve(loop_pos);
   iter_exprs.reserve(loop_pos);
   for (int i = 0; i < loop_pos; ++i) {
-    LoopAttr loop = op.Loops()[i].cast<LoopAttr>();
+    LoopAttr loop = mlir::cast<LoopAttr>(op.Loops()[i]);
     loop_names.push_back(loop.name());
     iter_exprs.push_back(loop.iter());
   }
@@ -698,7 +699,7 @@ mlir::LogicalResult LoopFusionAnalysis::RegisterLoop(
   auto loop_nest_mapping =
       MappingAttr::get(op.context(), op.domain_size(), iter_exprs);
 
-  LoopAttr loop = op.Loops()[loop_pos].cast<LoopAttr>();
+  LoopAttr loop = mlir::cast<LoopAttr>(op.Loops()[loop_pos]);
   auto [it, was_inserted] =
       fusion_classes_.try_emplace(loop.name(), loop.name(), op, loop_nest);
   LoopFusionClass &fusion_class = it->second;

sair_op_interfaces.h

@@ -26,6 +26,7 @@
 #include "mlir/IR/Operation.h"
 #include "mlir/IR/Value.h"
 #include "mlir/Interfaces/SideEffectInterfaces.h"
+#include "mlir/Support/LLVM.h"
 #include "mlir/Support/LogicalResult.h"
 #include "sair_attributes.h"
 #include "sair_types.h"
@@ -67,7 +68,7 @@ class ValueOperand {
 
   // Returns the type of the value referenced by the operand.
   ValueType GetType() const {
-    return operand_->get().getType().cast<ValueType>();
+    return mlir::cast<ValueType>(operand_->get().getType());
   }
 
   // Returns the operation owning the operand.
@@ -510,14 +511,14 @@ class OperandInstance {
 inline auto OpInstance::getDomain() const {
   return llvm::map_range(GetDomainValues(), [](mlir::Value v) {
     OpInstance dim_op = OpInstance(llvm::cast<SairOp>(v.getDefiningOp()));
-    return ResultInstance(dim_op, v.cast<OpResult>().getResultNumber());
+    return ResultInstance(dim_op, mlir::cast<OpResult>(v).getResultNumber());
   });
 }
 
 inline auto OpInstance::DomainWithDependencies() const {
   DomainShapeAttr shape = GetShape();
   return llvm::map_range(llvm::enumerate(GetDomainValues()), [=](auto p) {
-    auto value = p.value().template cast<mlir::OpResult>();
+    auto value = mlir::cast<mlir::OpResult>(p.value());
     OpInstance dim_op = OpInstance(llvm::cast<SairOp>(value.getOwner()));
     ValueAccessInstance access = {
         .value = ResultInstance(dim_op, value.getResultNumber()),

sequence.cc

@@ -20,6 +20,7 @@
 
 #include "llvm/ADT/iterator_range.h"
 #include "llvm/Support/Debug.h"
+#include "mlir/Support/LLVM.h"
 #include "loop_nest.h"
 #include "sair_op_interfaces.h"
 #include "sair_ops.h"
@@ -454,7 +455,7 @@ ProgramPoint SequenceAnalysis::FindInsertionPoint(
     llvm::ArrayRef<mlir::Attribute> new_loops = new_op.Loops();
     num_common_loops = std::min<int>(new_loops.size(), num_common_loops);
     for (; num_common_loops > 0; --num_common_loops) {
-      auto loop = new_loops[num_common_loops - 1].cast<LoopAttr>();
+      auto loop = mlir::cast<LoopAttr>(new_loops[num_common_loops - 1]);
       if (loop.name() == start_loop_nest[num_common_loops - 1]) break;
     }
     if (num_common_loops <= num_loops) break;

test/passes.cc

@@ -15,6 +15,7 @@
 #include "test/passes.h"
 
 #include "mlir/IR/Builders.h"
+#include "mlir/Support/LLVM.h"
 #include "sair_attributes.h"
 #include "sair_dialect.h"
 
@@ -34,7 +35,7 @@ static llvm::SmallVector<T, 4> GetAttrVector(llvm::StringRef name,
   llvm::SmallVector<T, 4> vector;
   vector.reserve(array.size());
   for (mlir::Attribute element : array.getValue()) {
-    vector.push_back(element.cast<T>());
+    vector.push_back(mlir::cast<T>(element));
   }
   return vector;
 }

transforms/lower_proj_any.cc

@@ -15,6 +15,7 @@
 #include "mlir/Dialect/Func/IR/FuncOps.h"
 #include "mlir/Dialect/SCF/IR/SCF.h"
 #include "mlir/Pass/Pass.h"
+#include "mlir/Support/LLVM.h"
 #include "loop_nest.h"
 #include "sair_dialect.h"
 #include "sair_ops.h"
@@ -86,7 +87,7 @@ class LowerProjAny : public impl::LowerProjAnyPassBase<LowerProjAny> {
         // guaranteed to be identity by verifiers.
         for (MappingExpr expr :
              mapping.Dimensions().drop_front(num_common_loops)) {
-          if (!expr.isa<MappingNoneExpr>()) {
+          if (!mlir::isa<MappingNoneExpr>(expr)) {
             return op.emitError()
                    << "cannot lower operation to proj_last on scalars";
           }

util.h

@@ -16,6 +16,7 @@
 #define THIRD_PARTY_SAIR_TRANSFORMS_UTIL_H_
 
 #include "mlir/IR/Builders.h"
+#include "mlir/Support/LLVM.h"
 #include "sair_attributes.h"
 #include "sair_op_interfaces.h"
 
@@ -95,7 +96,7 @@ std::function<mlir::ArrayAttr(mlir::ArrayAttr)> MkArrayAttrMapper(
     llvm::SmallVector<mlir::Attribute> output;
     output.reserve(array.size());
     for (mlir::Attribute attr : array.getValue()) {
-      output.push_back(scalar_fn(attr.cast<T>()));
+      output.push_back(scalar_fn(mlir::cast<T>(attr)));
     }
     return mlir::ArrayAttr::get(array.getContext(), output);
   };