First three patches of the patch series "mm/damon: fix misc bugs in
DAMON modules" [1] was trying to fix boot time DAMON sample modules
enabling issues by avoiding starting DAMON before the module
initialization phase.  However, probably by a mistake during a merge,
only half of the change is merged, and the part for avoiding the
starting of DAMON before the module initialized is missed.  So the
problem is not solved.  Fix those.

Note that the broken commits are merged into 6.17-rc1, but also
backported to relevant stable kernels.  So this series also need to be
merged into the stable kernels.  Hence Cc-ing stable@.

[1] https://lore.kernel.org/20250706193207.39810-1-sj@kernel.org

SeongJae Park (3):
  samples/damon/wsse: avoid starting DAMON before initialization
  samples/damon/prcl: avoid starting DAMON before initialization
  samples/damon/mtier: avoid starting DAMON before initialization

 samples/damon/mtier.c | 3 +++
 samples/damon/prcl.c  | 3 +++
 samples/damon/wsse.c  | 3 +++
 3 files changed, 9 insertions(+)


base-commit: 186951910f4e44e20738d85c0421032634ddb298
-- 
2.39.5

On Mon,  8 Sep 2025 19:22:35 -0700 SeongJae Park <sj@kernel.org> wrote:

> First three patches of the patch series "mm/damon: fix misc bugs in
> DAMON modules" [1] was trying to fix boot time DAMON sample modules
> enabling issues by avoiding starting DAMON before the module
> initialization phase.  However, probably by a mistake during a merge,
> only half of the change is merged, and the part for avoiding the
> starting of DAMON before the module initialized is missed.  So the
> problem is not solved.  Fix those.
> 
> Note that the broken commits are merged into 6.17-rc1, but also
> backported to relevant stable kernels.  So this series also need to be
> merged into the stable kernels.  Hence Cc-ing stable@.

That's unfortunate, but the about doesn't actually tell us what this
series does.  

> [1] https://lore.kernel.org/20250706193207.39810-1-sj@kernel.org

Presumably it's in there somewhere?

On Mon, 8 Sep 2025 19:35:48 -0700 Andrew Morton <akpm@linux-foundation.org> wrote:

> On Mon,  8 Sep 2025 19:22:35 -0700 SeongJae Park <sj@kernel.org> wrote:
> 
> > First three patches of the patch series "mm/damon: fix misc bugs in
> > DAMON modules" [1] was trying to fix boot time DAMON sample modules
> > enabling issues by avoiding starting DAMON before the module
> > initialization phase.  However, probably by a mistake during a merge,
> > only half of the change is merged, and the part for avoiding the
> > starting of DAMON before the module initialized is missed.  So the
> > problem is not solved.  Fix those.
> > 
> > Note that the broken commits are merged into 6.17-rc1, but also
> > backported to relevant stable kernels.  So this series also need to be
> > merged into the stable kernels.  Hence Cc-ing stable@.
> 
> That's unfortunate, but the about doesn't actually tell us what this
> series does.  

Good point.  The issue is that the sample modules can crash if those are
enabled at boot time before DAMON is initialized, via kernel command line.

Would you prefer me sending another version of this patch series with an
elaborated cover letter?

> 
> > [1] https://lore.kernel.org/20250706193207.39810-1-sj@kernel.org
> 
> Presumably it's in there somewhere?

You're right.  Both the cover letter and the individual fix patches (first
three of the series) describes the issue and origin broken commit.

Please let me know if there is anything I can help for this patch series from
my side :)


Thanks,
SJ

On Mon,  8 Sep 2025 20:51:41 -0700 SeongJae Park <sj@kernel.org> wrote:

> > > Note that the broken commits are merged into 6.17-rc1, but also
> > > backported to relevant stable kernels.  So this series also need to be
> > > merged into the stable kernels.  Hence Cc-ing stable@.
> > 
> > That's unfortunate, but the about doesn't actually tell us what this
> > series does.  
> 
> Good point.  The issue is that the sample modules can crash if those are
> enabled at boot time before DAMON is initialized, via kernel command line.
> 
> Would you prefer me sending another version of this patch series with an
> elaborated cover letter?

Please just send out the appropriate words and I'll paste it in.

On Mon, 8 Sep 2025 21:18:45 -0700 Andrew Morton <akpm@linux-foundation.org> wrote:

> On Mon,  8 Sep 2025 20:51:41 -0700 SeongJae Park <sj@kernel.org> wrote:
> 
> > > > Note that the broken commits are merged into 6.17-rc1, but also
> > > > backported to relevant stable kernels.  So this series also need to be
> > > > merged into the stable kernels.  Hence Cc-ing stable@.
> > > 
> > > That's unfortunate, but the about doesn't actually tell us what this
> > > series does.  
> > 
> > Good point.  The issue is that the sample modules can crash if those are
> > enabled at boot time before DAMON is initialized, via kernel command line.
> > 
> > Would you prefer me sending another version of this patch series with an
> > elaborated cover letter?
> 
> Please just send out the appropriate words and I'll paste it in.

Thank you for the guidance, Andrew.  How about below?

"""
First three patches of the patch series "mm/damon: fix misc bugs in DAMON
modules" [1] was trying to fix boot time DAMON sample modules enabling issues.
The issues are the modules can crash if those are enabled before DAMON is
enabled, like using boot time parameter options. The three patches were fixing
the issues by avoiding starting DAMON before the module initialization phase.

However, probably by a mistake during a merge, only half of the change is
merged, and the part for avoiding the starting of DAMON before the module
initialized is missed.  So the problem is not solved and thus the modules can
still crash if enabled before DAMON is initialized.  Fix those by applying the
unmerged parts again.

Note that the broken commits are merged into 6.17-rc1, but also backported to
relevant stable kernels.  So this series also needs to be merged into the
stable kernels.  Hence Cc-ing stable@.
"""


Thanks,
SJ