Re: [PATCH] rqspinlock: Fix order in raw_res_spin_(un)lock_irq to allow schedule

[Peter Zijlstra]

On Tue, Jun 09, 2026 at 01:22:35PM +0200, Arnd Bergmann wrote:
> On Tue, Jun 9, 2026, at 11:49, Gabriele Monaco wrote:
> > raw_res_spin_unlock_irqrestore() calls raw_res_spin_unlock() and then
> > restores interrupts, this means preemption is enabled when interrupts
> > are still disabled (as part of raw_res_spin_unlock()) so this cannot
> > trigger an actual preemption.
> > This is inconsistent with other spinlock implementations
> > (raw_spin_unlock_irqrestore() and bpf_res_spin_unlock_irqrestore()
> > itself).
> >
> > Adjust the macro to ensure interrupts are enabled before enabling
> > preemption, allowing to schedule at that point. Make the same
> > modification in the error path of raw_res_spin_lock_irqsave().
> >
> > Fixes: 101acd2e78b1 ("rqspinlock: Add macros for rqspinlock usage")

Yeah, this is right. spinlocks always get one preempt_disable, in
addition they might also get irq or bh disable.