Re: [PATCH v2] posix-cpu-timers: Dequeue per-thread timers before exchange_tids()

[Thomas Gleixner <tglx@kernel.org>]

On Mon, Aug 31 2026 at 20:24, Oleg Nesterov wrote:
> On 08/31, Thomas Gleixner wrote:
>>
>> On Wed, Aug 26 2026 at 14:55, Oleg Nesterov wrote:
>> > Can we move posix_cpu_timers_exit() or the whole CONFIG_POSIX_TIMERS
>> > sequence in begin_new_exec() up, right after de_thread() before the
>> > "if (retval)" check?
>>
>> Yes, but that does not buy us anything.
>
> I didn't mean this makes sense, I just tried to understand the Hyunwoo's
> fix, in particular the placement of additional posix_cpu_timers_exit().

Fair enough. But as I was staring at it some more, I thought more about
it. Yes, we can do:

     ret = de_thread()
     if (ret)
     	goto out;

     mop_up_posix_timers();

de_thread() only returns failure _before_ the TID swap, so that's safe
and spares sprinkling this extra invocation into de_thread().

What a mess.