Date:   Wed, 05 Jun 2019 14:57:49 -0700
From:   Dan Williams <dan.j.williams@...el.com>
To:     akpm@...ux-foundation.org
Cc:     David Hildenbrand <david@...hat.com>,
        Mike Rapoport <rppt@...ux.ibm.com>,
        Jane Chu <jane.chu@...cle.com>,
        Michael Ellerman <mpe@...erman.id.au>,
        Pavel Tatashin <pasha.tatashin@...een.com>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Jonathan Corbet <corbet@....net>,
        Logan Gunthorpe <logang@...tatee.com>,
        Paul Mackerras <paulus@...ba.org>,
        Toshi Kani <toshi.kani@....com>,
        Oscar Salvador <osalvador@...e.de>,
        Jeff Moyer <jmoyer@...hat.com>, Michal Hocko <mhocko@...e.com>,
        Vlastimil Babka <vbabka@...e.cz>, stable@...r.kernel.org,
        Jérôme Glisse <jglisse@...hat.com>,
        linux-mm@...ck.org, linux-nvdimm@...ts.01.org,
        linux-kernel@...r.kernel.org, osalvador@...e.de, mhocko@...e.com
Subject: [PATCH v9 00/12] mm: Sub-section memory hotplug support

Changes since v8 [1]:
- Rebase on next-20190604 to incorporate the removal of the
  MHP_MEMBLOCK_API flag and other cleanups from David.

- Move definition of subsection_mask_set() earlier into "mm/sparsemem:
  Add helpers track active portions of a section at boot" (Oscar)

- Cleanup unnecessary IS_ENABLED(CONFIG_SPARSEMEM_VMEMMAP) in
  section_deactivate() in response to a request (declined) to split the
  pure CONFIG_SPARSEMEM bits from section_{de,}activate(). I submit that
  the maintenance is less error prone, especially when modifying common
  logic, if the implementations remain unified. (Oscar)

- Cleanup sparse_add_section() vs sparse_index_init() return code.
  (Oscar)

- Document ZONE_DEVICE and subsection semantics relative to
  CONFIG_SPARSEMEM_VMEMMAP in Documentation/vm/memory-model.rst. (Mike)

[1]: https://lore.kernel.org/lkml/155718596657.130019.17139634728875079809.stgit@dwillia2-desk3.amr.corp.intel.com/

---

The memory hotplug section is an arbitrary / convenient unit for memory
hotplug. 'Section-size' units have bled into the user interface
('memblock' sysfs) and can not be changed without breaking existing
userspace. The section-size constraint, while mostly benign for typical
memory hotplug, has and continues to wreak havoc with 'device-memory'
use cases, persistent memory (pmem) in particular. Recall that pmem uses
devm_memremap_pages(), and subsequently arch_add_memory(), to allocate a
'struct page' memmap for pmem. However, it does not use the 'bottom
half' of memory hotplug, i.e. never marks pmem pages online and never
exposes the userspace memblock interface for pmem. This leaves an
opening to redress the section-size constraint.

To date, the libnvdimm subsystem has attempted to inject padding to
satisfy the internal constraints of arch_add_memory(). Beyond
complicating the code, leading to bugs [2], wasting memory, and limiting
configuration flexibility, the padding hack is broken when the platform
changes this physical memory alignment of pmem from one boot to the
next. Device failure (intermittent or permanent) and physical
reconfiguration are events that can cause the platform firmware to
change the physical placement of pmem on a subsequent boot, and device
failure is an everyday event in a data-center.

It turns out that sections are only a hard requirement of the
user-facing interface for memory hotplug and with a bit more
infrastructure sub-section arch_add_memory() support can be added for
kernel internal usages like devm_memremap_pages(). Here is an analysis
of the current design assumptions in the current code and how they are
addressed in the new implementation:

Current design assumptions:

- Sections that describe boot memory (early sections) are never
  unplugged / removed.

- pfn_valid(), in the CONFIG_SPARSEMEM_VMEMMAP=y, case devolves to a
  valid_section() check

- __add_pages() and helper routines assume all operations occur in
  PAGES_PER_SECTION units.

- The memblock sysfs interface only comprehends full sections

New design assumptions:

- Sections are instrumented with a sub-section bitmask to track (on x86)
  individual 2MB sub-divisions of a 128MB section.

- Partially populated early sections can be extended with additional
  sub-sections, and those sub-sections can be removed with
  arch_remove_memory(). With this in place we no longer lose usable memory
  capacity to padding.

- pfn_valid() is updated to look deeper than valid_section() to also check the
  active-sub-section mask. This indication is in the same cacheline as
  the valid_section() so the performance impact is expected to be
  negligible. So far the lkp robot has not reported any regressions.

- Outside of the core vmemmap population routines which are replaced,
  other helper routines like shrink_{zone,pgdat}_span() are updated to
  handle the smaller granularity. Core memory hotplug routines that deal
  with online memory are not touched.

- The existing memblock sysfs user api guarantees / assumptions are
  not touched since this capability is limited to !online
  !memblock-sysfs-accessible sections.

Meanwhile the issue reports continue to roll in from users that do not
understand when and how the 128MB constraint will bite them. The current
implementation relied on being able to support at least one misaligned
namespace, but that immediately falls over on any moderately complex
namespace creation attempt. Beyond the initial problem of 'System RAM'
colliding with pmem, and the unsolvable problem of physical alignment
changes, Linux is now being exposed to platforms that collide pmem
ranges with other pmem ranges by default [3]. In short,
devm_memremap_pages() has pushed the venerable section-size constraint
past the breaking point, and the simplicity of section-aligned
arch_add_memory() is no longer tenable.

These patches are exposed to the kbuild robot on my libnvdimm-pending
branch [4], and a preview of the unit test for this functionality is
available on the 'subsection-pending' branch of ndctl [5].

[2]: https://lore.kernel.org/r/155000671719.348031.2347363160141119237.stgit@dwillia2-desk3.amr.corp.intel.com
[3]: https://github.com/pmem/ndctl/issues/76
[4]: https://git.kernel.org/pub/scm/linux/kernel/git/djbw/nvdimm.git/log/?h=libnvdimm-pending
[5]: https://github.com/pmem/ndctl/commit/7c59b4867e1c


---

Dan Williams (12):
      mm/sparsemem: Introduce struct mem_section_usage
      mm/sparsemem: Add helpers track active portions of a section at boot
      mm/hotplug: Prepare shrink_{zone,pgdat}_span for sub-section removal
      mm/sparsemem: Convert kmalloc_section_memmap() to populate_section_memmap()
      mm/hotplug: Kill is_dev_zone() usage in __remove_pages()
      mm: Kill is_dev_zone() helper
      mm/sparsemem: Prepare for sub-section ranges
      mm/sparsemem: Support sub-section hotplug
      mm: Document ZONE_DEVICE memory-model implications
      mm/devm_memremap_pages: Enable sub-section remap
      libnvdimm/pfn: Fix fsdax-mode namespace info-block zero-fields
      libnvdimm/pfn: Stop padding pmem namespaces to section alignment


 Documentation/vm/memory-model.rst    |   39 ++++
 arch/powerpc/include/asm/sparsemem.h |    3 
 arch/x86/mm/init_64.c                |    4 
 drivers/nvdimm/dax_devs.c            |    2 
 drivers/nvdimm/pfn.h                 |   15 -
 drivers/nvdimm/pfn_devs.c            |   95 +++------
 include/linux/memory_hotplug.h       |    7 -
 include/linux/mm.h                   |    4 
 include/linux/mmzone.h               |   92 +++++++--
 kernel/memremap.c                    |   61 ++----
 mm/memory_hotplug.c                  |  171 +++++++++-------
 mm/page_alloc.c                      |   10 +
 mm/sparse-vmemmap.c                  |   21 +-
 mm/sparse.c                          |  359 +++++++++++++++++++++++-----------
 14 files changed, 534 insertions(+), 349 deletions(-)

Powered by blists - more mailing lists