Aggressive Optimizations in Clear Linux: Avoiding Regression and Code Bloat

How does Clear Linux, applying such aggressive code optimizations during compilation, avoid regression and code bloat?

To answer the performance (regression) part of your question: For every release we measure a few dozen tests (each with subtests) and we plot the results in graphs to make sure we can see clearly that we don’t go backwards – and if we do, we investigate.

as for code bloat: if it’s space on disk, we accept a bit of that. For example we ship with basically complete debuginfo in almost all our binaries … this costs about 20% diskspace for a binary, but makes debugging and other developer tasks a LOT easier so we allow for that tradeoff.
Code bloat in terms of it costing performance: that really falls in the earlier answer – we care if and when it’s real … but not in the general abstract sense.

2 Likes