# MLIR:         %[[conv:.*]] = "tfl.conv_2d"(%[[ARG_0]], %[[weight]], %[[bias]]) <{dilation_h_factor = 1 : i32, dilation_w_factor = 1 : i32, fused_activation_function = "NONE", padding = "SAME", stride_h = 1 : i32, stride_w = 1 : i32}
# MLIR:         %[[reshape:.*]] = "tfl.reshape"(%[[conv]], %[[shape]]) : (tensor<1x1x1x186x!quant.uniform<i8:f32, 0.09363494573854933:22>>, tensor<3xi32>)

# MLIR:         %[[conv:.*]] = "tfl.conv_2d"(%[[ARG_0]], %[[weight]], %[[bias]]) <{dilation_h_factor = 1 : i32, dilation_w_factor = 1 : i32, fused_activation_function = "NONE", padding = "SAME", stride_h = 1 : i32, stride_w = 1 : i32}
# MLIR:         %[[reshape:.*]] = "tfl.reshape"(%[[conv]], %[[shape]]) : (tensor<1x1x1x186x!quant.uniform<i8:f32, 0.09363494573854933:22>>, tensor<3xi32>)

// issue 7205

var i9 interface{}

func f9() bool {
	g8()
	x := i9
	y := interface{}(g18()) // ERROR "live at call to convT: x.data$" "live at call to g18: x.data$" "stack object .autotmp_[0-9]+ \[2\]string$"
	i9 = y                  // make y escape so the line above has to call convT
	return x != y
}

// liveness formerly confused by UNDEF followed by RET,
// leading to "live at entry to f10: ~r1" (unnamed result).

// issue 7205

var i9 interface{}

func f9() bool {
	g8()
	x := i9
	y := interface{}(g18()) // ERROR "live at call to convT: x.data$" "live at call to g18: x.data$" "stack object .autotmp_[0-9]+ \[2\]string$"
	i9 = y                  // make y escape so the line above has to call convT
	return x != y
}

// liveness formerly confused by UNDEF followed by RET,
// leading to "live at entry to f10: ~r1" (unnamed result).

// RUN: tf-quant-opt %s -split-input-file -quant-insert-quantized-functions -quant-quantize-composite-functions | FileCheck %s

module {
  func.func @conv(%arg0: tensor<1x2x2x3xf32>) -> (tensor<*xf32>, tensor<*xf32>) {

  ASSERT_THAT(new_config.specs().specs(), SizeIs(1));

  const QuantizationSpec& spec = new_config.specs().specs(0);
  EXPECT_THAT(spec.matcher().function_name().regex(),
              StrEq("^.*(conv|dot_general).*"));
  EXPECT_TRUE(spec.method().has_weight_only_ptq());

  const WeightOnlyPtq& weight_only_ptq_spec = spec.method().weight_only_ptq();

  EXPECT_THAT(weight_only_ptq_spec.input_quantized_types(),

    const auto module_op = op->getParentOfType<ModuleOp>();
    const SymbolTable symbol_table(module_op);
    func::FuncOp func = symbol_table.lookup<func::FuncOp>(function_name);

    if (function_name.contains("conv")) {
      return (*(func.getOps<mlir::stablehlo::ConvolutionOp>().begin()))
          .getDimensionNumbers()
          .getKernelOutputFeatureDimension();
    } else if (function_name.contains("dot_general")) {

}

func mustGetLabels(t test.Failer, obj object.K8sObject, path string) map[string]string {
	t.Helper()
	got := mustGetPath(t, obj, path)
	conv, ok := got.(map[string]any)
	if !ok {
		t.Fatalf("could not convert %v", got)
	}
	ret := map[string]string{}
	for k, v := range conv {
		sv, ok := v.(string)
		if !ok {
			t.Fatalf("could not convert to string %v", v)
		}
		ret[k] = sv
	}
	return ret
}

// CHECK-NEXT:  %[[conv:.*]] = "tfl.conv_2d"(%arg0, %[[q_cst_0]], %[[q_cst_1]]) <{dilation_h_factor = 1 : i32, dilation_w_factor = 1 : i32, fused_activation_function = "NONE", padding = "SAME", stride_h = 2 : i32, stride_w = 2 : i32}>
// CHECK-NEXT:  %[[reshape:.*]] = "tfl.reshape"(%[[conv]], %[[cst]]) : (tensor<1x112x112x32x!quant.uniform<u8:f32, 0.023528476789885875>>, tensor<2xi32>)

// or `std::nullopt` if the given op is not per-channel quantizable.
std::optional<int64_t> GetDotGeneralQuantizationDim(
    ::mlir::stablehlo::DotGeneralOp dot_general_op);

// Checks if a `StringRef` contains 'conv' or 'dot_general'.
bool ContainsConvOrDot(StringRef str);

}  // namespace mlir::quant