Re: [PATCH v2 25/44] mm: Switch vma_merge(), split_vma(), and __split_vma to vma iterator

[Liam Howlett]

* SeongJae Park <sj@xxxxxxxxxx> [230106 12:23]:
> Hello Liam,
> 
> On Thu, 5 Jan 2023 19:15:59 +0000 Liam Howlett <liam.howlett@xxxxxxxxxx> wrote:
> 
> > From: "Liam R. Howlett" <Liam.Howlett@xxxxxxxxxx>
> > 
> > Drop the vmi_* functions and transition all users to use the vma
> > iterator directly.
> > 
> > Signed-off-by: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
> > ---
> >  fs/userfaultfd.c   | 14 ++++----
> >  include/linux/mm.h | 16 +++-------
> >  mm/madvise.c       |  6 ++--
> >  mm/mempolicy.c     |  6 ++--
> >  mm/mlock.c         |  6 ++--
> >  mm/mmap.c          | 79 +++++++++++++---------------------------------
> >  mm/mprotect.c      |  6 ++--
> >  mm/mremap.c        |  2 +-
> >  8 files changed, 47 insertions(+), 88 deletions(-)
> [...]
> > --- a/include/linux/mm.h
> > +++ b/include/linux/mm.h
> > @@ -2830,22 +2830,16 @@ static inline int vma_adjust(struct vm_area_struct *vma, unsigned long start,
> [...]
> > -extern int vmi__split_vma(struct vma_iterator *vmi, struct mm_struct *,
> > -	struct vm_area_struct *, unsigned long addr, int new_below);
> > -extern int split_vma(struct mm_struct *, struct vm_area_struct *,
> > -	unsigned long addr, int new_below);
> > -extern int vmi_split_vma(struct vma_iterator *vmi, struct mm_struct *,
> > -		struct vm_area_struct *, unsigned long addr, int new_below);
> > +extern int __split_vma(struct vma_iterator *vmi, struct vm_area_struct *,
> > +		       unsigned long addr, int new_below);
> > +extern int split_vma(struct vma_iterator *vmi, struct vm_area_struct *,
> > +			 unsigned long addr, int new_below);
> 
> I just found this change for split_vma() is applied to !CONFIG_MMU, which the
> definition of split_vma() is not changed, so cause a build error.  I posted a
> simple fix for that:
> https://lore.kernel.org/linux-mm/20230106171857.149918-1-sj@xxxxxxxxxx/
> 

Thanks.  I think I need revisit the nommu side of things with this
change as well.  I was hoping to avoid that, but it seems to be more
necessary than I had thought.