Re: [akpm-mm:mm-stable 219/240] mm/memory.c:5410:41: error: implicit declaration of function 'vma_is_tcp'; did you mean 'vma_is_dax'?

[Suren Baghdasaryan <surenb@google.com>]

On Mon, Aug 14, 2023 at 1:09 PM Suren Baghdasaryan <surenb@google.com> wrote:
>
> On Mon, Aug 14, 2023 at 1:06 PM Andrew Morton <akpm@linux-foundation.org> wrote:
> >
> > On Sun, 13 Aug 2023 16:17:21 +0800 kernel test robot <lkp@intel.com> wrote:
> >
> > > tree:   https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-stable
> > > head:   5fb2ea3111f4ecc6dc4891ce5b00f0217aae9a04
> > > commit: 4aaa60dad4d1c96151dec51098aed866bb6e867d [219/240] mm: allow per-VMA locks on file-backed VMAs
> > > config: x86_64-defconfig (https://download.01.org/0day-ci/archive/20230813/202308131610.jF4ncWp6-lkp@intel.com/config)
> > > compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
> > > reproduce: (https://download.01.org/0day-ci/archive/20230813/202308131610.jF4ncWp6-lkp@intel.com/reproduce)
> > >
> > > If you fix the issue in a separate patch/commit (i.e. not just a new version of
> > > the same patch/commit), kindly add following tags
> > > | Reported-by: kernel test robot <lkp@intel.com>
> > > | Closes: https://lore.kernel.org/oe-kbuild-all/202308131610.jF4ncWp6-lkp@intel.com/
> > >
> > > All errors (new ones prefixed by >>):
> > >
> > >    In file included from include/linux/build_bug.h:5,
> > >                     from include/linux/container_of.h:5,
> > >                     from include/linux/list.h:5,
> > >                     from include/linux/smp.h:12,
> > >                     from include/linux/kernel_stat.h:5,
> > >                     from mm/memory.c:42:
> > >    mm/memory.c: In function 'lock_vma_under_rcu':
> > > >> mm/memory.c:5410:41: error: implicit declaration of function 'vma_is_tcp'; did you mean 'vma_is_dax'? [-Werror=implicit-function-declaration]
> > >     5410 |         if (unlikely(!vma->anon_vma && !vma_is_tcp(vma)))
> > >          |                                         ^~~~~~~~~~
> > >    include/linux/compiler.h:77:45: note: in definition of macro 'unlikely'
> > >       77 | # define unlikely(x)    __builtin_expect(!!(x), 0)
> > >          |                                             ^
> > >    cc1: some warnings being treated as errors
> > >
> >
> > This later gets accidentally fixed by Suren's "mm: handle userfaults
> > under VMA lock".

Andrew, is "mm: handle userfaults under VMA lock" merged into
mm-stable? I could not find it and in fact the whole "Per-VMA lock
support for swap and userfaults" patchset
(https://lore.kernel.org/all/CAJuCfpHSFikZ=h34yS980BmUP5M=+j6rB4_b-q7MCc10Xs24+w@mail.gmail.com/)
seems to be missing in mm-stable. That's problematic because Matthew's
"Handle most file-backed faults under the VMA lock" patchset
(https://lore.kernel.org/all/20230724185410.1124082-1-willy@infradead.org/)
requires at least one patch from my patchset to work correctly, this
one: https://lore.kernel.org/all/20230630211957.1341547-4-surenb@google.com/.

An additional note,
https://lore.kernel.org/all/20230812002033.1002367-1-willy@infradead.org/
is fixing a known issue in "Handle most file-backed faults under the
VMA lock" patchset and it's missing from mm-stable too. As Matthew
mentioned in that patch, ideally it should be placed before "mm:
handle faults that merely update the accessed bit under the VMA lock"

Thanks,
Suren.

> >
> > If Matthew can suggest a fix for this I can queue it close to "mm:
> > allow per-VMA locks on file-backed VMAs" to minimize the size of the
> > bisection hole?
>
> I thought Matthew's suggestion here
> https://lore.kernel.org/all/ZNerqcNS4EBJA%2F2v@casper.infradead.org/
> would fix this. Dos it not?