single_version_override with patch = […] has no effect
I’m trying to use the single_version_override
feature of MODULES.bazel
in order to apply a local patch to an existing public repository in the Bazel Common Registry. (Specifically google_benchmark
doesn’t build on an arm64 windows system and I want to both locally unblock playing with it and eventually work on upstreaming a fix if they’re interested in taking it).
Fetch hash of Bazel output artifact with minimal building and fetching?
I would like to determine the hash of a Bazel target artifact, with the minimal amount of building and fetching.
How to build a Python Spark bundle with Bazel?
Spark accepts a variety of Python bundle formats:
How can I pass a path/directory to a rule in Bazel
Working with directories in Bazel is discouraged since Bazel will not accurately track the directory contents, however, I often find myself
running into the same problem where the input to rule just has to be a directory. Most commonly, I am setting something like PATH
, PYTHONPATH
, RESOURCES
, or TOOL_PATH
in a rule’s env dictionary and these environment variables have to be either a single directory or column separated paths/directories.
bazel query ignore all **/.cache/**/BUILD files
I want to execute bazel query //...
but I have many .cache
folders in my repo and they may contain BUILD files in their subfolders that may cause bazel errors. I want bazel to exclude all BUILD files that matches the glob pattern **/.cache/**/BUILD
. What should I do?
Is there any way to override bazelrc location without changing command line?
I’m trying to integrate bazel into a Linux distro packaging system.
Why won’t Bazel creating symlinks during the build process?
From the directory with the WORKSPACE
file I’m building a java target in bazel that should leave all the symlinks:
coreapp bootstrap is failing after main is moved to bazel
When running bootstrap the job fails for below
Is the remote cache uploaded during the build process?
When multiple targets of a project are built together, does it upload the remote cache during the build process, or wait until all builds are completed before uploading?
How to build a c++ binary with bazel and toolchains, where the compilers are in docker images
I have docker images containing different versions of gcc compiler. How would I build a simple c++ hello-world application multiple times, using each docker image as the execution-platform. There doesn’t appear to be any clear examples of how to construct this using toolchains, and its unclear if you are expected to inherit aspects of existing cc rules or if you need to build rules and toolchains from scratch.