Note: the lab machine may have a slightly different version of LLVM than shown in the video. Currently, you should use: clang++-14, llvm-profdata-14, and llvm-cov-14

This lab describes the use of compiler tooling to determine the quality of a test suite with respect to some program under test. The lab video describes a set of commands to the LLVM toolsuite to generate a line coverage report.

For a warmup, create a test suite for the badcalc.cpp demo code (link) that achieves full line coverage. Turn in the test suite and your coverage report.

You are encouraged to apply the coverage reporting technique to your own project code, but how you spend the rest of your lab time after you've completed the coverage report is up to you.

Create a tarball consisting of a single directory with all of your work in it, such as code or written answers. If you have any auxiliary files (READMEs, Makefiles, etc.), be sure to include those as well. Name your work directory l7 and name the tarball l7.tgz. Upload your tarball to the L7 Canvas assignment.

Labs will be graded under one of following criteria:

1. Effort: You will automatically be given full credit for the labwork if the GTA determines that your participation in the lab was meaningful - i.e. you attended the lab session and used the time to make a good-faith attempt to complete the work. It is the sole discretion of the GTA to determine if you put in sufficient effort. Even if you expect an effort-based grade, you should turn in your (possibly incomplete) work.
2. Correctness: If you do not participant meaningfully in lab (i.e. you do not attend the lab session), your grade will be assessed based on the correctness of your lab submission.

The two-criteria grading scheme above is designed to avoid wasting your time. You should not feel obligated to attend the lab, and in fact should only do so if you want help from the GTA on the labwork assignment or whatever project is currently in progress. Here's a handy flowchart for how I suggest you approach lab: