Skip to main content

This version of GitHub Enterprise was discontinued on 2023-01-18. No patch releases will be made, even for critical security issues. For better performance, improved security, and new features, upgrade to the latest version of GitHub Enterprise. For help with the upgrade, contact GitHub Enterprise support.

Troubleshooting CodeQL runner in your CI system

If you're having problems with the CodeQL runner, you can troubleshoot by using these tips.

Code scanning is available for organization-owned repositories in GitHub Enterprise Server. This feature requires a license for GitHub Advanced Security. For more information, see "About GitHub Advanced Security."

Note: The CodeQL runner is being deprecated. On GitHub Enterprise Server 3.0 and greater, you can install CodeQL CLI version 2.6.3 to replace CodeQL runner.

For more information, see the CodeQL runner deprecation. For information on migrating to CodeQL CLI, see "Migrating from the CodeQL runner to CodeQL CLI."

The init command takes too long

Before the CodeQL runner can build and analyze code, it needs access to the CodeQL bundle, which contains the CodeQL CLI and the CodeQL libraries.

When you use the CodeQL runner for the first time on your machine, the init command downloads the CodeQL bundle to your machine. This download can take a few minutes. The CodeQL bundle is cached between runs, so if you use the CodeQL runner again on the same machine, it won't download the CodeQL bundle again.

To avoid this automatic download, you can manually download the CodeQL bundle to your machine and specify the path using the --codeql-path flag of the init command.

No code found during the build

If the analyze command for the CodeQL runner fails with an error No source code was seen during the build, this indicates that CodeQL was unable to monitor your code. Several reasons can explain such a failure.

  1. Automatic language detection identified a supported language, but there is no analyzable code of that language in the repository. A typical example is when our language detection service finds a file associated with a particular programming language like a .h, or .gyp file, but no corresponding executable code is present in the repository. To solve the problem, you can manually define the languages you want to analyze by using the --languages flag of the init command. For more information, see "Configuring CodeQL runner in your CI system."

  2. You're analyzing a compiled language without using the autobuild command and you run the build steps yourself after the init step. For the build to work, you must set up the environment such that the CodeQL runner can monitor the build process. The init command generates instructions for how to export the required environment variables, so you can copy and run the script after you've run the init command.

    • On macOS and Linux:
      $ . codeql-runner/codeql-env.sh
    • On Windows, using the Command shell (cmd) or a batch file (.bat):
      > call codeql-runner\codeql-env.bat
    • On Windows, using PowerShell:
      > cat codeql-runner\codeql-env.sh | Invoke-Expression

    The environment variables are also stored in the file codeql-runner/codeql-env.json. This file contains a single JSON object which maps environment variable keys to values. If you can't run the script generated by the init command, then you can use the data in JSON format instead.

    Note:� If you used the --temp-dir flag of the init command to specify a custom directory for temporary files, the path to the codeql-env files might be different.

  3. You're analyzing a compiled language on macOS without using the autobuild command and you run the build steps yourself after the init step. If SIP (System Integrity Protection) is enabled, which is the default on recent versions of OSX, analysis might fail. To fix this, prefix the build command with the $CODEQL_RUNNER environment variable. For example, if your build command is cmd arg1 arg2, you should run $CODEQL_RUNNER cmd arg1 arg2.

  4. The code is built in a container or on a separate machine. If you use a containerized build or if you outsource the build to another machine, make sure to run the CodeQL runner in the container or on the machine where your build task takes place. For more information, see "Running CodeQL code scanning in a container."