Re: [PATCH v3 -next 00/15] sysctl: move sysctls from vm_table into its own files

On Thu, Oct 24, 2024 at 04:07:10PM +0800, yukaixiong wrote:
...
> 
> 
> >>   mm/swap.c                          |  16 ++-
> >>   mm/swap.h                          |   1 +
> >>   mm/util.c                          |  67 +++++++--
> >>   mm/vmscan.c                        |  23 +++
> >>   mm/vmstat.c                        |  44 +++++-
> >>   net/sunrpc/auth.c                  |   2 +-
> >>   security/min_addr.c                |  11 ++
> >>   23 files changed, 330 insertions(+), 312 deletions(-)
> >>
> >> -- 
> >> 2.34.1
> >>
> > General comment for the patchset in general. I would consider making the
> > new sysctl tables const. There is an effort for doing this and it has
> > already lanted in linux-next. So if you base your patch from a recent
> > next release, then it should just work. If you *do* decide to add a
> > const qualifier, then note that you will create a dependency with the
> > sysctl patchset currently in next and that will have to go in before.
> >
> > Best
> >
> 
> Sorry,  I don't understand what is the meaning of "create a dependency 
> with the sysctl patchset".
The patches in the sysctl subsys that allow you to qualify the ctl_table
as const are not in mainline yet. They are in linux-next. This means
that if these patches go into the next kernel release before the
sysctl-next branch, it will have compilation errors. Therefore the
sysctl-next branch needs to be pulled in to the new kernel release
before this patchest. This also means that for this to build properly it
has to be based on a linux-next release.

> 
> Do you just want me to change all "static struct ctl_table" type table 
> into "static const struct ctl_table" type in my patchset?
You should const qualify them if the maintainer that is pulling in these
patches is ok with it. You should *not* const qualify them if the
maintainer prefers otherwise.

Please get back to me if I did not address your questions.

Best

-- 

Joel Granados