On Wed, Aug 03, 2022 at 04:42:43PM -0700, Linus Torvalds wrote: > On Wed, Aug 3, 2022 at 4:24 PM Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > > > > On Wed, Aug 03, 2022 at 06:59:36PM -0400, Steven Rostedt wrote: > > > > > > preempt_disable_inlock() ? > > > > preempt_disable_locked()? > > Heh. Shed painting in full glory. > > Let's try just "preempt_enable_under_spinlock()" and see. > > It's a bit long, but it's still shorter than the existing usage pattern. > > And we don't have "inlock" anywhere else, and while "locked" is a real > pattern we have, it tends to be about other things (ie "I hold the > lock that you need, so don't take it"). > > And this is _explicitly_ only about spinning locks, because sleeping > locks don't do the preemption disable even without RT. > > So let's make it verbose and clear and unambiguous. It's not like I > expect to see a _lot_ of those. Knock wood. Should we have it take a spinlock_t pointer? We could have lockdep check it is actually held.
