[0/3] userfaultfd remap option

Message

Suren Baghdasaryan Sept. 14, 2023, 3:26 p.m. UTC

This patch series introduces UFFDIO_REMAP feature to userfaultfd, which
has long been implemented and maintained by Andrea in his local tree [1],
but was not upstreamed due to lack of use cases where this approach would
be better than allocating a new page and copying the contents.

UFFDIO_COPY performs ~20% better than UFFDIO_REMAP when the application
needs pages to be allocated [2]. However, with UFFDIO_REMAP, if pages are
available (in userspace) for recycling, as is usually the case in heap
compaction algorithms, then we can avoid the page allocation and memcpy
(done by UFFDIO_COPY). Also, since the pages are recycled in the
userspace, we avoid the need to release (via madvise) the pages back to
the kernel [3].
We see over 40% reduction (on a Google pixel 6 device) in the compacting
thread’s completion time by using UFFDIO_REMAP vs. UFFDIO_COPY. This was
measured using a benchmark that emulates a heap compaction implementation
using userfaultfd (to allow concurrent accesses by application threads).
More details of the usecase are explained in [3].

Furthermore, UFFDIO_REMAP enables remapping swapped-out pages without
touching them within the same vma. Today, it can only be done by mremap,
however it forces splitting the vma.

Main changes since Andrea's last version [1]:
1. Trivial translations from page to folio, mmap_sem to mmap_lock
2. Replace pmd_trans_unstable() with pte_offset_map_nolock() and handle its
possible failure
3. Move pte mapping into remap_pages_pte to allow for retries when source
page or anon_vma is contended. Since pte_offset_map_nolock() start RCU
read section, we can't block anymore after mapping a pte, so have to unmap
the ptesm do the locking and retry.
4. Add and use anon_vma_trylock_write()  to avoid blocking while in RCU
read section.
5. Accommodate changes in mmu_notifier_range_init() API, switch to
mmu_notifier_invalidate_range_start_nonblock() to avoid blocking while in
RCU read section.
6. Open-code now removed __swp_swapcount()
7. Replace pmd_read_atomic() with pmdp_get_lockless()
8. Add new selftest for UFFDIO_REMAP

[1] https://gitlab.com/aarcange/aa/-/commit/2aec7aea56b10438a3881a20a411aa4b1fc19e92
[2] https://lore.kernel.org/all/[e-mail]/
[3] https://lore.kernel.org/linux-mm/[e-mail]/

Andrea Arcangeli (2):
  userfaultfd: UFFDIO_REMAP: rmap preparation
  userfaultfd: UFFDIO_REMAP uABI

Suren Baghdasaryan (1):
  selftests/mm: add UFFDIO_REMAP ioctl test

 fs/userfaultfd.c                             |  49 ++
 include/linux/rmap.h                         |   5 +
 include/linux/userfaultfd_k.h                |  17 +
 include/uapi/linux/userfaultfd.h             |  22 +
 mm/huge_memory.c                             | 118 ++++
 mm/khugepaged.c                              |   3 +
 mm/rmap.c                                    |  13 +
 mm/userfaultfd.c                             | 586 +++++++++++++++++++
 tools/testing/selftests/mm/uffd-common.c     |  34 +-
 tools/testing/selftests/mm/uffd-common.h     |   1 +
 tools/testing/selftests/mm/uffd-unit-tests.c |  62 ++
 11 files changed, 908 insertions(+), 2 deletions(-)