Two optimization changes were merged in the same commit, and the regression suite now fails. Why does this make diagnosis harder than it would be with separate commits?
Select an answer to reveal the explanation.
Short Explanation
One failing test, two suspects. When you bundle two optimizations into a single commit, the failure points at the commit, not one change, so you have to investigate both together.
Full Explanation
The correct choice is the one saying that the failing test cannot indicate which of the two changes caused it, so both must be investigated together instead of isolating one. With both optimizations in one commit, the regression failure implicates the commit as a whole, not either change individually, so isolating the cause takes extra work that a smaller, separately reviewed change would have avoided.
The answer claiming that merged commits cannot be reverted is incorrect because merged commits can be reverted like any other; that is not what makes diagnosis harder.
The answer claiming that the regression suite only reports failures for the most recent file changed in a commit is incorrect because regression suites report failures against the code state, not against the most recently touched file.
The answer claiming that combining changes in one commit prevents the build from producing a stack trace is incorrect because stack traces are a property of the failing code path, unrelated to how many changes were bundled into the commit.
For the exam, know that combining changes is what makes attribution difficult, which is the practical reason optimization is applied in small increments.