Question

我想使用bazel交叉编译TensorFlow的C ++代码并获得arm可执行版本，在Arm架构机器上运行它。

关注维基https://github.com/bazelbuild/bazel/wiki/Building-with-a-custom-toolchain，我修改了一些代码，然后将代码推送到mygithub:bazel_toolchain。

然后我将bazel_toolchain目录放在TensorFlow github存储库的克隆文件tensorflow / bazel_toolchain中

现在树是/ root / tensorflow_master / tensorflow / bazel_toochain

我添加hello.cc代码如下：

hello.cc代码：

 #include "tensorflow/cc/client/client_session.h"
 #include "tensorflow/cc/ops/standard_ops.h"
 #include "tensorflow/core/framework/tensor.h"

 int main() {
   using namespace tensorflow;
   using namespace tensorflow::ops;
   Scope root = Scope::NewRootScope();
   // Matrix A = [3 2; -1 0]
   auto A = Const(root, { {3.f, 2.f}, {-1.f, 0.f} });
   // Vector b = [3 5]
   auto b = Const(root, { {3.f, 5.f} });
   // v = Ab^T
   auto v = MatMul(root.WithOpName("v"), A, b, MatMul::TransposeB(true));
   std::vector<Tensor> outputs;
   ClientSession session(root);
   // Run and fetch v
   TF_CHECK_OK(session.Run({v}, &outputs));
   // Expect outputs[0] == [19; -3]
   LOG(INFO) << outputs[0].matrix<float>();
   return 0;
}

构建文件如下：

filegroup(
 name = "srcs",
 srcs = glob(["**"]),
 visibility = ["//tensorflow/bazel_toolchain:__pkg__"],
)

cc_binary(
 name = "hello",
 srcs = ["hello.cc"],
 deps = [
    "//tensorflow/cc:cc_ops",
    "//tensorflow/cc:client_session",
    "//tensorflow/core:tensorflow",
 ],
 )

当我在/ root / tensorflow_master /目录下执行命令时，

 bazel build --crosstool_top=//tensorflow/bazel_toolchain/tools/arm_compiler:toolchain --cpu=armeabi-v7a 
 //tensorflow/bazel_toolchain:hello

它注意到了

tensorflow-master/tensorflow/bazel_toolchain/tools/arm_compiler/BUILD:46:1: no such package '@org_linaro_components_toolchain_gcc_5_3_1//': The repository could not be resolved and referenced by '//tensorflow/bazel_toolchain/tools/arm_compiler:linaro_linux_linker_files'.

错误：分析目标＆＃39; // tensorflow / bazel_toolchain：你好＆＃39;失败;建立中止

我该怎么办？非常感谢！

Answer 1

The build file you shared is not correct, there is ) after srcs that effectively closes cc_binary, so deps are not an attribute on cc_binary but a declaration of a variable in the BUILD file (I think). Move deps into cc_binary.

The error message is however unrelated to above. Can you share the command line you use to build hello-world? Do you use --crosstool_top to use your custom crosstool? Can you edit your question to add content of your WORKSPACE file?

如何构建tensorflow c ++代码使用bazel的自定义工具链？

1 个答案: