Re: [PATCH v2 1/1] psi: Fix uaf issue when psi trigger is destroyed while being polled

[Linus Torvalds]

On Tue, Jan 11, 2022 at 11:27 AM Suren Baghdasaryan <surenb@xxxxxxxxxx> wrote:
>
> Thanks for the explanation!
> So, it sounds like the best (semantically correct) option I have here
> is smp_store_release() to set the pointer, and then smp_load_acquire()
> to read it. Is my understanding correct?

Yeah, that's the clearest one from a memory ordering standpoint, and
is generally also cheap.

                Linus