Mac下Bazel构建Linux Python二进制文件时CC工具链缺失报错

问题根源 #

Python构建依赖CC工具链（比如Python C扩展、Bazel Python规则底层依赖），而Go规则自带交叉编译工具链，所以Go的类似配置能正常运行，Python则需要额外配置Linux x86_64的CC交叉工具链。

解决步骤 #

1. 校验平台定义正确性

先确认//parser:linux_x86的平台配置是否包含正确的约束值，示例如下：

# parser/BUILD中的平台定义
platform(
    name = "linux_x86",
    constraint_values = [
        "@platforms//os:linux",
        "@platforms//cpu:x86_64",
    ],
)

2. 安装Linux x86_64交叉编译工具链

在Mac上通过Homebrew安装对应工具链：

brew install x86_64-linux-gnu-gcc

3. 配置Bazel CC工具链

在项目中新建toolchain/BUILD文件，定义Linux x86的CC工具链：

load("@rules_cc//cc:defs.bzl", "cc_toolchain", "cc_toolchain_config")

# 工具链文件组，指向Homebrew安装的交叉编译工具路径
filegroup(
    name = "linux_x86_toolchain_files",
    srcs = glob([
        "/usr/local/bin/x86_64-linux-gnu-*",
        "/usr/local/x86_64-linux-gnu/lib/**",
    ]),
    visibility = ["//visibility:public"],
)

filegroup(name = "empty")

# 简化的工具链配置
cc_toolchain_config(
    name = "linux_x86_toolchain_config",
    target_cpu = "x86_64",
    compiler = "gcc",
    host_system_name = "darwin",
    target_system_name = "linux",
    target_libc = "glibc",
    abi_version = "gcc",
    abi_libc_version = "glibc_2.19",
)

# 定义CC工具链
cc_toolchain(
    name = "linux_x86_cc_toolchain",
    toolchain_identifier = "linux_x86_64",
    target_system_name = "linux",
    target_cpu = "x86_64",
    target_libc = "glibc",
    compiler = "gcc",
    toolchain_config = ":linux_x86_toolchain_config",
    all_files = ":linux_x86_toolchain_files",
    compiler_files = ":linux_x86_toolchain_files",
    dwp_files = ":empty",
    linker_files = ":linux_x86_toolchain_files",
    objcopy_files = ":empty",
    strip_files = ":empty",
    supports_param_files = 0,
)

4. 注册工具链并引入rules_cc

在WORKSPACE文件中添加：

# 引入rules_cc规则集
http_archive(
    name = "rules_cc",
    urls = ["https://github.com/bazelbuild/rules_cc/releases/download/0.0.9/rules_cc-0.0.9.tar.gz"],
    sha256 = "cd06d15dd8bb59926e4d65f9003bfc20f9da4b2519985c27e190cddc8b7a7806",
)

# 注册自定义的Linux x86 CC工具链
register_toolchains("//toolchain:linux_x86_cc_toolchain")

5. 验证Python二进制配置

确保parser/BUILD中的py_binary指定了目标平台：

load("@rules_python//python:defs.bzl", "py_binary")

py_binary(
    name = "parser_bin",
    srcs = ["main.py"],
    platform = "//parser:linux_x86",
    deps = [...], # 你的依赖
)

6. 重新构建并调试

运行构建命令，带调试参数确认工具链匹配：

bazel build --platforms=//parser:linux_x86 //parser:parser_bin --toolchain_resolution_debug

如果输出中显示匹配到//toolchain:linux_x86_cc_toolchain，则配置生效。

替代方案：直接在Linux环境构建 #

如果交叉编译配置麻烦，可使用Docker在Linux容器中直接构建：

docker run -v $(pwd):/workspace -w /workspace l.gcr.io/google/bazel:latest bazel build //parser:parser_bin

内容的提问来源于stack exchange，提问作者JackDaniels