<component name="ProjectDictionaryState">
  <dictionary name="sergej.jaskiewicz">
    <words>
      <w>lowerings</w>
      <w>unmuted</w>
    </words>
  </dictionary>

  EXPECT_EQ(non_categorized_count, 428);
}

// Just a counter test to see which ops have duplicate lowerings. This isn't a
// correctness test versus a test to easily ensure the op registry is kept
// in sync.
TEST_F(LegalizationOpConfigTest, CountTypesWhichHaveBothMlirAndTf2xlaFallback) {
  int double_lowering_count = 0;

  DialectRegistry dialect_registry;
  dialect_registry.insert<mlir::TF::TensorFlowDialect>();

import org.jetbrains.kotlin.resolve.BindingContext

@OptIn(KaAnalysisNonPublicApi::class)
fun MutableSet<KtFile>.collectReachableInlineDelegatedPropertyAccessors() {
    if (isEmpty()) return

    // One of the compiler lowerings, namely `PropertyReferenceLowering`,
    // optimizes usages of property references in some cases,
    // and if a containing delegated property accessor is inline,
    // it might need this accessor's bytecode.
    //

std::unique_ptr<OperationPass<ModuleOp>> createLegalizeTFPass(
    bool legalize_chlo = true,
    std::optional<StringRef> tf2xla_fallback_device_type = std::nullopt,
    bool prefer_tf2xla = false);

/// Adds the TF to TF lowerings and TF to XLA rewrite patterns to the pattern
/// list.
void PopulateLegalizeTfPatterns(MLIRContext* context,
                                RewritePatternSet* patterns);

def InputLoweringMetricsPass : Pass<"input-lowering-metrics-pass", "mlir::func::FuncOp"> {

  let summary = "Collects various metrics about the input to the lowering "
                "portion of the bridge. This is a logical no-op.";

  let description = [{
    Gathers metrics about the input MLIR to Phase 2 of the TFXLA Bridge, which
    does a strict semantic lowering from Tensorflow ops to XLA HLO.
  }];

}

// BatchMatMulV2 has native as well as fallback lowering patterns available.
// The fallback pattern uses dot_general without broadcast on operands and then
// transposes the output which is faster. However, the fallback pattern doesn't
// support dynamic shaped operands like the native lowering. Verify that
// fallback lowering is preferred for static shaped operands when available.

// CHECK-LABEL: batchmatmulv2

bool IsOpAllowedTf2xlaFallback(const TypeID& type_id);

// Whether this type is Preferred to use a TF2XLA fallback kernel when using
// the MLIR bridge. If this is true, then the TF2XLA fallback kernel will be
// used over the MLIR lowering.
bool IsOpAllowedTf2xlaPreferred(const TypeID& type_id);

}  // namespace mhlo
}  // namespace mlir

struct DecomposeResourceOpsPass
    : public impl::DecomposeResourceOpsPassBase<DecomposeResourceOpsPass> {
  void runOnOperation() override {
    // Add lowering patterns to the list.
    RewritePatternSet patterns(&getContext());
    TF::PopulateDecomposeResourceOpsPatterns(&getContext(), &patterns);

    if (failed(applyPatternsAndFoldGreedily(getOperation(),

using ::tensorflow::thread::ThreadPoolInterface;

using ::tensorflow::tfrt_stub::FallbackTensor;

// -------------------------------------------------------------------------- //
// Run function via the TF->TFRT fallback lowering.
// -------------------------------------------------------------------------- //

namespace {
// Thread pool for running `intra-op` tasks scheduled by the fallback kernels.

// Creates an instance of the VerifyQuantLegalization pass, which verifies all
// quant ops and types are lowered.
std::unique_ptr<OperationPass<func::FuncOp>>
CreateVerifyQuantLegalizationPass();

// Add all passes for lowering TF quant ops and types to MHLO int.
void AddQuantizationLoweringPasses(mlir::OpPassManager &pm);

// Creates an instance of OptimizeIntGraphPass, which optimizes the int graph
// lowered from the quantized graph.