# Re: \[PATCH v10 07/14\] unwind\_user/deferred: Make unwind deferral requests NMI-safe

- From: Peter Zijlstra \<peterz@infradead.org\>
- Date: 2025-06-19 09:45:19 
- List: linux-trace-kernel
- Page: https://ratatoskr.run/linux-trace-kernel/2025/06/3098799
- Thread: https://ratatoskr.run/linux-trace-kernel/2025/06/3086227/t.md

```text
On Thu, Jun 19, 2025 at 05:42:31AM -0400, Steven Rostedt wrote:
> 
> 
> On June 19, 2025 5:32:26 AM EDT, Peter Zijlstra [ref] wrote:
> >On Thu, Jun 19, 2025 at 05:07:10AM -0400, Steven Rostedt wrote:
> >
> >> Does #DB make in_nmi() true? If that's the case then we do need to handle that.
> >
> >Yes: #DF, #MC, #BP (int3), #DB and NMI all have in_nmi() true.
> >
> >Ignoring #DF because that's mostly game over, you can get them all
> >nested for up to 4 (you're well aware of the normal NMI recursion
> >crap).
> 
> We probably can implement this with stacked counters.

I would seriously consider dropping support for anything that can't do
cmpxchg at the width you need.

```