Re: [PATCHv5 3/3] mm/gup_benchemark: add LONGTERM_BENCHMARK test in gup fast path

[Pingfan Liu <kernelfans@gmail.com>]

On Fri, Feb 28, 2020 at 11:43 PM Alexander Duyck
<alexander.duyck@gmail.com> wrote:
>
> On Fri, Feb 28, 2020 at 3:35 AM Pingfan Liu <kernelfans@gmail.com> wrote:
> >
> > Introduce a GUP_LONGTERM_BENCHMARK ioctl to test longterm pin in gup fast
> > path.
>
> The title of the patch has a typo in it. There is only one 'e' in "benchmark".
Yes, it should be GUP_FAST_LONGTERM_BENCHMARK
>
> > Signed-off-by: Pingfan Liu <kernelfans@gmail.com>
> > Cc: Ira Weiny <ira.weiny@intel.com>
> > Cc: Andrew Morton <akpm@linux-foundation.org>
> > Cc: Mike Rapoport <rppt@linux.ibm.com>
> > Cc: Dan Williams <dan.j.williams@intel.com>
> > Cc: Matthew Wilcox <willy@infradead.org>
> > Cc: John Hubbard <jhubbard@nvidia.com>
> > Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>
> > Cc: Keith Busch <keith.busch@intel.com>
> > Cc: Christoph Hellwig <hch@infradead.org>
> > Cc: Shuah Khan <shuah@kernel.org>
> > To: linux-mm@kvack.org
> > Cc: linux-kernel@vger.kernel.org
> > ---
> >  mm/gup_benchmark.c                         | 7 +++++++
> >  tools/testing/selftests/vm/gup_benchmark.c | 6 +++++-
> >  2 files changed, 12 insertions(+), 1 deletion(-)
> >
> > diff --git a/mm/gup_benchmark.c b/mm/gup_benchmark.c
> > index 8dba38e..bf61e7a 100644
> > --- a/mm/gup_benchmark.c
> > +++ b/mm/gup_benchmark.c
> > @@ -8,6 +8,7 @@
> >  #define GUP_FAST_BENCHMARK     _IOWR('g', 1, struct gup_benchmark)
> >  #define GUP_LONGTERM_BENCHMARK _IOWR('g', 2, struct gup_benchmark)
> >  #define GUP_BENCHMARK          _IOWR('g', 3, struct gup_benchmark)
> > +#define GUP_FAST_LONGTERM_BENCHMARK    _IOWR('g', 4, struct gup_benchmark)
> >
> >  struct gup_benchmark {
> >         __u64 get_delta_usec;
> > @@ -57,6 +58,11 @@ static int __gup_benchmark_ioctl(unsigned int cmd,
> >                         nr = get_user_pages_fast(addr, nr, gup->flags,
> >                                                  pages + i);
> >                         break;
> > +               case GUP_FAST_LONGTERM_BENCHMARK:
> > +                       nr = get_user_pages_fast(addr, nr,
> > +                                       (gup->flags & 1) | FOLL_LONGTERM,
> > +                                        pages + i);
> > +                       break;
>
> If I am not mistaken the mask of gup->flags is redundant. It is
> already masked by FOLL_WRITE several lines before this switch
> statement.
Yes, you are right. Thanks for your kind review.

Regards,
Pingfan