Evaluate grsec's performance hit

Something we’ll need to evaluate is the performance hit (the test suite runs I see take 35% longer than those for feature/8415-overlayfs-stretch, which is quite bigger than I would have expected; but I’ve not done any measurement on bare metal yet). If it’s too big then we may need to look closer and perhaps disable some grsec features whose cost/benefit ratio is not interesting enough.

Parent Task: #7649 (closed)

Original created by @intrigeri on 12110 (Redmine)