Tag Archives: rcu

Given that it is much more comfortable chasing down race conditions reported by rcutorture than those reported from the field, it would be good to make race conditions more probable during rcutorture runs than in production. A number of tricks … Continue reading →

The rcutorture test suite has quite a few options, including locktorture, rcuscale, refscale, and scftorture in addition to rcutorture itself. These tests can be run with the assistance of either KASAN or KCSAN. Given that RCU contains kernel modules, there … Continue reading →

Although Linux-kernel RCU gets most of the attention, without rcutorture, RCU would not be what it is today. To see this, note that the old saying “If it ain’t tested, it don’t work!” is if anything more valid today than … Continue reading →

This installment of the rcutorture series takes a high-level look at its design. At the highest level, rcutorture is a stress test with a few unit-test components thrown in for good measure. It also includes scripts to handle both single-system … Continue reading →

I recently took a close look at rcutorture‘s console output and noticed the following string: rtbf: 0 rtb: 0. The good news is that there were no rcutorture priority-boosting failures (rtbf: 0). The bad news is that this was only because there was no … Continue reading →

An earlier post discussed ways of making rcutorture more aggressive, but even with these techniques, rcutorture‘s level of aggression is limited by build time on the one hand and the confines of a single system on the other. This post … Continue reading →

Continuing further into the Linux-kernel Documentation/RCU/Design/Requirements/Requirements.rst file uncovers RCU’s final two fundamental guarantees:   The common-case RCU primitives are unconditional, and RCU users can perform a guaranteed read-to-write upgrade. The first guarantee is trivially verified by inspection of the RCU … Continue reading →

Even more reading of the Linux-kernel Documentation/RCU/Design/Requirements/Requirements.rst file encounters RCU’s memory-barrier guarantees. These guarantees are a bit ornate, but roughly speaking guarantee that RCU read-side critical sections lapping over one end of a given grace period are fully ordered with … Continue reading →