Re: [PATCH] mm/damon/core: modified control->repeat loop at the kdamond_call()

[SeongJae Park <sj@kernel.org>]

On Wed, 31 Dec 2025 15:23:49 +0900 JaeJoon Jung <rgbi3307@gmail.com> wrote:

> On Wed, 31 Dec 2025 at 10:28, SeongJae Park <sj@kernel.org> wrote:
> >
> > On Mon, 29 Dec 2025 16:57:15 -0800 SeongJae Park <sj@kernel.org> wrote:
> >
> > > On Tue, 30 Dec 2025 09:14:37 +0900 JaeJoon Jung <rgbi3307@gmail.com> wrote:
> > >
> > > > I will reflect the above in patch v2.
> > > > Would you like me to resend patch v2 ?
> > >
> > > I just found my fix [1] of the bug that you reported has a race condition.  And
> > > to fix that, I think kdamond_call() will anyway need to be modified.  I will
> > > shortly send more details as a reply to the fix [1].  Can you please hold this
> > > refactoring until all discussions about the bug fix is done?  Sorry for your
> > > inconvenience.
> > >
> > > [1] https://lore.kernel.org/20251228183105.289441-1-sj@kernel.org
> >
> > I ended up sending the new version of the fix [1] without changes on
> > kdamond_call().  Unless you have concerns on the fix, please feel free to
> > proceed to v2 of this patch.  Thank you for patiently waiting.
> >
> > [1] https://lore.kernel.org/20251231012315.75835-1-sj@kernel.org
> 
> My patch for the kdamond_call() function further improves control->list
> to behave as a FIFO.  This also improves stability against the control->list
> collision issue encountered when damon_call() is called.
> If there are no further objections, I'll send you Patch v2.

If you want to say something about damon_call() uaf issue, let's make a
conclusion about it first, rather than mixing topics.


Thanks,
SJ

[...]