Re: [PATCHv2 1/2] rcu/tree: handle VM stoppage in stall detection

[Sergey Senozhatsky]

On (21/05/23 20:46), Paul E. McKenney wrote:
> 
> In theory, yes, sort of, anyway.  In practice, highly unlikely.
> The most plausible way for this to happen is for this code path to be
> delayed for a long time on a 32-bit system, so that jiffies+ULONG_MAX/2
> actually arrives.  But in that case, all sorts of other complaints
> would happen first.

I see.

> But I could make this a cmpxchg(), if that is what you are getting at.

No, it's good. I was just curious what scenario I was missing.