Re: [PATCH v2 1/7] rcutorture: Rewrite RCU priority-boost test around a CPU-hungry reader

From: Sebastian Andrzej Siewior

Date: Thu Nov 19 2026 - 04:14:58 EST


On 2026-11-16 14:44:32 [-0500], Joel Fernandes wrote:
> The old boost test relies on RT throttling being disabled and on grace
> periods taking too long, which is fragile. Instead, each interval one
> CPU runs a normal-priority reader that spins inside an RCU read-side
> critical section until it has consumed test_boost_reader_ms of CPU time
> (measured via sum_exec_runtime, so time spent preempted does not count),
> while a SCHED_FIFO hog on the same CPU tries to starve it. With RT
> throttling enabled, an unboosted reader gets only about 5% of the CPU,
> far less than it needs within test_boost_duration. If the reader
> finishes, boosting worked, otherwise it failed.
…
> Signed-off-by: Joel Fernandes <jfern@xxxxxxxxxx>

This is doing what I though it is doing in the first place :)
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>

I didn't run all the tests for it but did test it initially in my setup.

Sebastian