user0/rust
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update documentation

Browse source
This commit is contained in:
Jakub Beránek 2024-12-29 19:20:39 +01:00 • committed by Jakub Beránek
parent 0cc11f8c48
commit e4b6ccef43
1 changed files with 15 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

29
src/ci/docker/README.md
View file

@ -1,29 +1,30 @@
# Docker images for CI
This folder contains a bunch of docker images used by the continuous integration
(CI) of Rust. An script is accompanied (`run.sh`) with these images to actually
execute them. To test out an image execute:
(CI) of Rust. A script is accompanied (`run.sh`) with these images to actually
execute them.
Note that a single Docker image can be used by multiple CI jobs, so the job name
is the important thing that you should know. You can examine the existing CI jobs in
the [`jobs.yml`](../github-actions/jobs.yml) file.
To run a specific CI job locally, you can use the following script:
```
./src/ci/docker/run.sh $image_name
python3 ./src/ci/github-actions/ci.py run-local <job-name>
```
for example:
For example, to run the `x86_64-gnu-llvm-18-1` job:
```
./src/ci/docker/run.sh x86_64-gnu
python3 ./src/ci/github-actions/ci.py run-local x86_64-gnu-llvm-18-1
```
Images will output artifacts in an `obj/$image_name` dir at the root of a repository. Note
that the script will overwrite the contents of this directory.
To match conditions in rusts CI, also set the environment variable `DEPLOY=1`, e.g.:
```
DEPLOY=1 ./src/ci/docker/run.sh x86_64-gnu
```
The job will output artifacts in an `obj/<image-name>` dir at the root of a repository. Note
that the script will overwrite the contents of this directory. `<image-name>` is set based on the
Docker image executed in the given CI job.
**NOTE**: In CI, the script outputs the artifacts to the `obj` directory,
while locally, to the `obj/$image_name` directory. This is primarily to prevent
while locally, to the `obj/<image-name>` directory. This is primarily to prevent
strange linker errors when using multiple Docker images.
For some Linux workflows (for example `x86_64-gnu-llvm-18-N`), the process is more involved. You will need to see which script is executed for the given workflow inside the [`jobs.yml`](../github-actions/jobs.yml) file and pass it through the `DOCKER_SCRIPT` environment variable. For example, to reproduce the `x86_64-gnu-llvm-18-3` workflow, you can run the following script: