Re: [PATCH] fs/drop_caches: move drop_caches sysctls into its own file

[Luis Chamberlain]

On Tue, Mar 21, 2023 at 03:28:36PM +0100, Christian Brauner wrote:
> On Tue, Mar 21, 2023 at 09:09:07PM +0800, Yangtao Li wrote:
> > This moves the fs/drop_caches.c respective sysctls to its own file.
> > 
> > Signed-off-by: Yangtao Li <frank.li@xxxxxxxx>
> > ---
> >  fs/drop_caches.c   | 25 ++++++++++++++++++++++---
> >  include/linux/mm.h |  6 ------
> >  kernel/sysctl.c    |  9 ---------
> >  3 files changed, 22 insertions(+), 18 deletions(-)
> > 
> > diff --git a/fs/drop_caches.c b/fs/drop_caches.c
> > index e619c31b6bd9..3032b83ce6f2 100644
> > --- a/fs/drop_caches.c
> > +++ b/fs/drop_caches.c
> > @@ -12,8 +12,7 @@
> >  #include <linux/gfp.h>
> >  #include "internal.h"
> >  
> > -/* A global variable is a bit ugly, but it keeps the code simple */
> > -int sysctl_drop_caches;
> > +static int sysctl_drop_caches;
> >  
> >  static void drop_pagecache_sb(struct super_block *sb, void *unused)
> >  {
> > @@ -47,7 +46,7 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused)
> >  	iput(toput_inode);
> >  }
> >  
> > -int drop_caches_sysctl_handler(struct ctl_table *table, int write,
> > +static int drop_caches_sysctl_handler(struct ctl_table *table, int write,
> >  		void *buffer, size_t *length, loff_t *ppos)
> >  {
> >  	int ret;
> > @@ -75,3 +74,23 @@ int drop_caches_sysctl_handler(struct ctl_table *table, int write,
> >  	}
> >  	return 0;
> >  }
> > +
> > +static struct ctl_table drop_caches_table[] = {
> > +	{
> > +		.procname	= "drop_caches",
> > +		.data		= &sysctl_drop_caches,
> > +		.maxlen		= sizeof(int),
> > +		.mode		= 0200,
> > +		.proc_handler	= drop_caches_sysctl_handler,
> > +		.extra1		= SYSCTL_ONE,
> > +		.extra2		= SYSCTL_FOUR,
> > +	},
> > +	{}
> > +};
> > +
> > +static int __init drop_cache_init(void)
> > +{
> > +	register_sysctl_init("vm", drop_caches_table);
> 
> Does this belong under mm/ or fs/?

To not break old userspace it must be kept under "vm" because the
patch author is moving it from the kernel/sysctl.c table which used
the "vm" table.

Moving it to "fs" would be a highly functional change which should
require review from maintainers it would not break existing userspace
expecations.

> And is it intended to be moved into a completely separate file?

What do you mean by this?

> Feels abit wasteful for 20 lines of code...

Not sure what you mean by this either. The commit log sucks, please
review got log kernel/sysclt.c for much better commit logs for the
rationale of moving sysctls out out kernel/sysctl.c to their own
respective places.

  Luis