On 2022-06-13 23:13:49 [+0200], Arnd Bergmann wrote: > Acked-by: Arnd Bergmann <arnd@xxxxxxxx> > > It's probably best to keep this together with the corresponding architecture > specific changes. I was a bit worried about bisection at first, but > then realized > that this is not a problem for mainline since ARCH_SUPPORTS_RT is not > yet enabled on any architecture. So where do I put this patch to? If I remember correctly then arm64 is using this. ARM has its own thing and x86 has this change already. > How are softirqs called on preempt_rt? Does this result in higher stack usage > for the normal task stacks again, or are they run in separate threads anyway? one of: - ksoftirqd thread. - in the force-threaded interrupt after the main handler. - any time after bh-counter hits zero due local_bh_enable(). The last one will cause higher task stacks. > Arnd Sebastian
