Re: [PATCH v4 2/3] mm/swap: use swap_ops to register swap device's methods

[Chris Li <chrisl@kernel.org>]

On Tue, Apr 21, 2026 at 6:48 PM Baoquan He <baoquan.he@linux.dev> wrote:
>
> On 04/17/26 at 07:30pm, Chris Li wrote:
> > On Thu, Apr 16, 2026 at 8:40 PM Baoquan He <baoquan.he@linux.dev> wrote:
> ...snip...
> > > +int init_swap_ops(struct swap_info_struct *sis)
> > > +{
> > > +       /*
> > > +        * ->flags can be updated non-atomically, but that will
> > > +        * never affect SWP_FS_OPS, so the data_race is safe.
> > > +        */
> > > +       if (data_race(sis->flags & SWP_FS_OPS))
> > > +               sis->ops = &bdev_fs_swap_ops;
> > > +       /*
> > > +        * ->flags can be updated non-atomically, but that will
> > > +        * never affect SWP_SYNCHRONOUS_IO, so the data_race is safe.
> > > +        */
> > > +       else if (data_race(sis->flags & SWP_SYNCHRONOUS_IO))
> > > +               sis->ops = &bdev_sync_swap_ops;
> > > +       else
> > > +               sis->ops = &bdev_async_swap_ops;
> > > +
> > > +       if (!sis->ops || !sis->ops->read_folio || !sis->ops->write_folio)
> > > +               return -1;
> > Nitpick:
> >
> > For int type error return, you should use  -EINVAL or some thing with
> > error code. If you don't care about error code, change the return type
> > to bool instead.
>
> Thanks for careful reviewing, returning -EINVAL looks better, I will
> change like that and repost.

You don't have to repost it as a new series. You can post one small
incremental update for the -EINVAL change and ask Andrew to fold it
in.

Hi Andrew,

I think this swap ops series can be merge to mm-unstable with the
review. Keep in mind that Baoquan might have a very minor follow up
patch relate to -EINAL error code.

Chris