Demo of TrustInSoft Analyzer integration with Jenkins for CI

The purpose of this repository is to demonstrate the possible ways of integrating TrustInSoft Analyzer in Jenkins CI.

The code used for this demo is the Intel tinycrypt library, supplemented with a test driver of 10 tests used for the TrustInSoft analysis.

Setup

You must:

1. Set up a Jenkins Server instance.
   (The objective of this demo is not to explain how to install Jenkins, but this is rather simple, and there exists plenty of documentation and tutorials available freely on the Internet. This is not a formal recommendation, but you may look at How To Install Jenkins on Ubuntu 20.04 on DigitalOcean.)
2. Set up a Jenkins Agent that runs a version of Ubuntu supported by the TrustInSoft Analyzer (18.04, 20.04, or 22.04 as of Dec 2022). Note, that the Jenkins server is by default a Jenkins agent, too. So, at least for small scale deployment, you may not need to do more than installing the Jenkins server.
3. Install the tis-analyzer package on the Jenkins agent and configure access to a valid license (FlexNet or Node locked).

Jenkins job configuration

There are different ways of defining Jenkins jobs:

• The most modern and flexible way is through pipeline jobs based on Jenkinsfile. The advantage of this approach is that the job configuration itself is code, hence all configuration changes can be tracked as any other code changes (exactly like the C source code). This is an implementation of the Infrastructure as Code principle (IaC) - configuration is directly stored in the code repository.
• Another solution is to use Jenkins' classic freestyle jobs. This may also be OK, although it has the drawback that the Job configuration is stored in the Jenkins instance itself, and therefore provides little traceability of job configuration changes.

Note: In both job examples, we will leverage the run_all.sh script that is a convenient way to run all the 10 analyses defined in this repository. This is a convenience only and absolutely not mandatory for the jobs. You may just as well run the tis-analyzer commands that you want directly in the job.

Jenkins job based on pipeline

As per the advantage of Infrastructure as Code, most of the configuration is defined in the Jenkinsfile. The only steps to be taken in Jenkins are:

• Create a Job of kind Pipeline or Multibranch pipeline.
  (The main difference is that a multi-branch pipeline will allow to conveniently analyze different branches of your code, whereas a pipeline is for one branch only.)

Jenkins jobs based on Freestyle jobs

In this type of job, the configuration is not textual. It's configured in the Jenkins' GUI itself. (Jenkins may then store in database and/or XML files, but's that internal) Configuration is rather self-explanatory. See Jenkins screenshots below:

Source Code Management

1. Set Git->Repositories->Repository URL to https://github.com/TrustInSoft/demo-jenkins.git
2. Click Add Repository

Build Triggers

1. Choose Build periodically
2. Set Schedule to H * * * *

Build Environment

1. Choose Delete workspace before build starts
2. Choose Inject environment variables to the build process
   1. Set Properties Content to TIS_VERSION=1.44 or whatever version is installed on your Jenkins agent (Can be 1.44.1, 1.45 or higher in the future). If several versions are installed, then chose the one you want to use for your pipeline (if no specific need, it's very recommended to use the latest)

Build Steps

1. Insert all the Build Steps and Post-Build Actions Step 1 is obviously mandatory to run the analysis. Step 2 is optional yet very recommended to produce the analysis HTML report Step 3 is optionel, and to run only if you wish to produce a MISRA report

   1. Add step Execute shell (1)

      /home/tis/${TIS_VERSION}/bin/tis-setenv.sh
      source ~/.tis.conf
      # The -n <i> parameter below pertains to the number of analyses you want to run in parallel
      # Benefits: The more analyses you run in parallel, the faster your pipeline
      # Constraints:
      # - You need at least as many token licenses as the number of analyses to run in parallel
      # - You need enough CPU power (cores) and RAM to run the number of parallel analyses yuo have chosen
      trustinsoft/run_all.sh -n 1
      

   2. Add step Execute shell (2)

      # This step is to be run to produce the HTML tis-report
      # The parameter passed to tis-report is ALWAYS "_report" unless you changed the default configuration of tis-analyzer 
      /home/tis/${TIS_VERSION}/bin/tis-setenv.sh
      source ~/.tis.conf
      tis-report _results
      

   3. Add step Execute shell (3)

      # This step is to be run only if you want a MISRA compliance report
      # The parameter passed to tis-misra is the root directory of the source code ("lib" in te current sample project)
      /home/tis/${TIS_VERSION}/bin/tis-setenv.sh
      source ~/.tis.conf
      tis-misra lib
      

Post-build Actions

1. Add step Archive the artifacts
   1. Set Files to archive to
      tis_report.html, _results/**.json, _results/**.csv, tis_misra_report/**