On Fri, 8 Mar 2024, Michael Schmitz wrote:
IMHO, Gueunter would do better to instrument the spinlock tests on the
assumption that the locking in arch/m68k really is correct.
I've come to agree - maybe logging any run queue locks taken by the init
task with IRQs enabled might help?
If I understood how the spinlock test algorithm worked, I'd be adding
sanity checks to it to test the assumptions underlying that algorithm.
I would not be surprised if it uses logic that sees irqs_disabled() and
then assumes that all hard irqs are disabled, or something similar.
I don't think I'd be adding printk() because the BUG is intermittent and
printk() messes with timing.
