curtismuntz / bazel_compilers Goto Github PK

Building binaries for foreign architectures is only half the battle. Development cycles are much faster if you can test the cross compiled code locally. Add qemu scripts to run / test the cross compiled binaries.

There was an issue in c230931 . This should be similar to OpenLightingProject/ola#1125 .

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Update actions/cache action to v4
• Update actions/checkout action to v4
• Update bazelbuild/setup-bazelisk action to v3
• Update dependency ubuntu to v22
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

First, thank you very much for making your project, it was very useful in my case!

I was trying to build Tensorflow Lite C library for ARM but I got this error message:

/home/user/Documents/Bazel/tensorflow/lite/c/BUILD:22:1: Linking of rule '//tensorflow/lite/c:libtensorflowlite_c.so' failed (Exit 1)
external/org_linaro_components_toolchain_gcc_armv7/arm-linux-gnueabihf/libc/usr/lib/Scrt1.o: In function _start': /home/tcwg-buildslave/workspace/tcwg-make-release_0/snapshots/glibc.git~release~2.25~master/csu/../sysdeps/arm/start.S:128: undefined reference to main'
collect2: error: ld returned 1 exit status
Target //tensorflow/lite/c:tensorflowlite_c failed to build
Use --verbose_failures to see the command lines of failed build steps.

I solved it manually cloning your repository, copying it into my local copy of tensorflow project, referencing it using the local_repository command instead of http_archive (in WORKSPACE file):

local_repository(
name = "murtis_bazel_compilers",
path = "//home/user/Documents/Bazel/murtis_bazel_compilers/"
)

load("@murtis_bazel_compilers//compilers:dependencies.bzl", "cross_compiler_dependencies")

cross_compiler_dependencies()

so I could finally change the file cc_toolchain_config.bzl. I commented the line 648 (the -pie flag):

#"-pie",

and then it worked.

I am sure there must be a better way to fix this so it would work for both executables and dynamic libraries like Tensorflow Lite C library. Thank you very much for a fix in advance - for all others who might also get this error in similar cases.