}                                                                        \
  }
DEFINE_GETATTR(TF_OperationGetAttrInt, int64_t, int64_t, i);
DEFINE_GETATTR(TF_OperationGetAttrFloat, float, float, f);
DEFINE_GETATTR(TF_OperationGetAttrBool, unsigned char, bool, b);
DEFINE_GETATTR(TF_OperationGetAttrType, TF_DataType, DataType, type);
#undef DEFINE_GETATTR

void TF_OperationGetAttrShape(TF_Operation* oper, const char* attr_name,

      "/job:localhost/replica:0/task:0/device:CPU:0",
      "/job:localhost/replica:0/task:0/device:CPU:1"};
  ParallelDevice parallel_device(devices);
  const FunctionDef assert_and_collective = FunctionDefHelper::Define(
      // Name
      "AssertAndCollective",
      // Args
      {"x: float", "condition: bool"},
      // Return values
      {"y: float"},
      // Attr def
      {},
      // Nodes
      {

  if (op_reg_data->shape_inference_fn == nullptr) {
    status->status =
        InvalidArgument("No shape inference function exists for op '",
                        node_def.op(), "', did you forget to define it?");
    return;
  }

  status->status = c.Run(op_reg_data->shape_inference_fn);
  if (!status->status.ok()) return;

  // Set output_shapes.
  TF_ShapeAndTypeList* output_shapes_result =

  EXPECT_NE(TF_OK, TF_GetCode(status)) << TF_Message(status);

  // Even after the prevoiusly failed cluster update, another update and op
  // execution should work fine as long as the provided server_def is valid.
  TFE_ContextUpdateServerDef(ctx, 0, serialized.data(), serialized.size(),
                             status);
  EXPECT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);

#include "tensorflow/c/tf_status_helper.h"
#include "tensorflow/core/platform/path.h"
#include "tensorflow/core/platform/stacktrace_handler.h"
#include "tensorflow/core/platform/test.h"

#define ASSERT_TF_OK(x) ASSERT_EQ(TF_OK, TF_GetCode(x)) << TF_Message(x)
#define EXPECT_TF_OK(x) EXPECT_EQ(TF_OK, TF_GetCode(x)) << TF_Message(x)

static const char* content = "abcdefghijklmnopqrstuvwxyz1234567890";
// We will work with content_view instead of content.

  }

 private:
  bool first_call_ = true;
};

// After the graph pass is registered, it takes effect globally and can affect
// other test cases. Define a static variable to switch it on and off.
bool GraphErrorInjectionPass::enabled_ = false;

// Test to ensure that a registered graph optimization pass is only executed

  // re-read the same value, expect another lookup
  cache1.LookupOrCompute("a", &value, compute_func, &status);
  TF_EXPECT_OK(status.status);
  EXPECT_EQ(value, 1);
  EXPECT_EQ(num_compute_calls, 2);

  // Define a new cache with max_age > 0 and verify correct behavior.
  tf_gcs_filesystem::ExpiringLRUCache<int> cache2(2, 4);
  num_compute_calls = 0;
  value = -1;

  // Read our first value

    forward_op_.outputs.push_back(retvals[i]);
  }
  // TODO(b/166669239): This is needed to support AttrBuilder::Get for string
  // attributes. Number type attrs and DataType attrs work fine without this.
  // Consider getting rid of this and making the behavior between number types
  // and string consistent.
  forward_op_.attrs.BuildNodeDef();
  // TODO(b/170307493): Populate skip_input_indices here.

  };
  tf_gcs_filesystem::RamFileBlockCache cache(block_size, block_size, 0,
                                             fetcher);
  std::vector<char> out;
  // Reading the first 16 bytes should be fine.
  TF_EXPECT_OK(ReadCache(&cache, "", 0, block_size, &out));
  EXPECT_TRUE(first_block);
  EXPECT_EQ(out.size(), block_size);
  // Reading at offset file_size + 4 will read the second block (since the read

}

// REGISTER_OP for CApiAttributesTest test cases.
// Registers two ops, each with a single attribute called 'v'.
// The attribute in one op will have a type 'type', the other
// will have list(type).
#define ATTR_TEST_REGISTER_OP(type)                           \
  REGISTER_OP("CApiAttributesTestOp" #type)                   \
      .Attr("v: " #type)                                      \