[lkml]   [2020]   [Sep]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Subject[PATCH v5 00/17] device-dax: support sub-dividing soft-reserved ranges
Changes since v4 [1]:
- Rebased on
in -mm [2]. I.e. patches that did not need fixups from v4 are not

- Folded all fixes

- Replaced "device-dax: kill dax_kmem_res" with:

device-dax/kmem: introduce dax_kmem_range()
device-dax/kmem: move resource name tracking to drvdata
device-dax/kmem: replace release_resource() with release_mem_region() address David's request to make those cleanups easier to review.
Note that I dropped changes to how IORESOURCE_BUSY is manipulated since
David and I are still debating the best way forward there.

- Broke out some of dax-bus reworks in "device-dax: introduce 'seed'
devices" to a new "device-dax: introduce 'struct dev_dax' typed-driver

- Added a conversion of xen_alloc_unallocated_pages() from pgmap.res to
pgmap.range. I found it odd that there is no corresponding
memunmap_pages() triggered by xen_free_unallocated_pages()?

- Not included, a conversion of virtio_fs to use pgmap.range for its new
usage of devm_memremap_pages(). It appears the virtio_fs changes are
merged after -mm? My mental model of -mm was that it applies on top of
linux-next? In any event, Vivek, you will need to coordinate a
conversion to pgmap.range for the virtio_fs dax-support merge. Maybe
that should go through Andrew as well?

- Lowercase all the subject lines per akpm's preference

- Received a 0day robot build-success notification over 122 configs

- Thanks to Joao for looking after this set while I was out.



Andrew, this series replaces



dax-hmem-introduce-dax_hmemregion_idle-parameter.patch your stack.

Let me know if there is a different / preferred way to refresh a bulk of
patches in your queue when only a subset need updates.


The device-dax facility allows an address range to be directly mapped
through a chardev, or optionally hotplugged to the core kernel page
allocator as System-RAM. It is the mechanism for converting persistent
memory (pmem) to be used as another volatile memory pool i.e. the
current Memory Tiering hot topic on linux-mm.

In the case of pmem the nvdimm-namespace-label mechanism can sub-divide
it, but that labeling mechanism is not available / applicable to
soft-reserved ("EFI specific purpose") memory [3]. This series provides
a sysfs-mechanism for the daxctl utility to enable provisioning of
volatile-soft-reserved memory ranges.

The motivations for this facility are:

1/ Allow performance differentiated memory ranges to be split between
kernel-managed and directly-accessed use cases.

2/ Allow physical memory to be provisioned along performance relevant
address boundaries. For example, divide a memory-side cache [4] along
cache-color boundaries.

3/ Parcel out soft-reserved memory to VMs using device-dax as a security
/ permissions boundary [5]. Specifically I have seen people (ab)using
memmap=nn!ss (mark System-RAM as Persistent Memory) just to get the
device-dax interface on custom address ranges. A follow-on for the VM
use case is to teach device-dax to dynamically allocate 'struct page' at
runtime to reduce the duplication of 'struct page' space in both the
guest and the host kernel for the same physical pages.



Dan Williams (14):
device-dax: make pgmap optional for instance creation
device-dax/kmem: introduce dax_kmem_range()
device-dax/kmem: move resource name tracking to drvdata
device-dax/kmem: replace release_resource() with release_mem_region()
device-dax: add an allocation interface for device-dax instances
device-dax: introduce 'struct dev_dax' typed-driver operations
device-dax: introduce 'seed' devices
drivers/base: make device_find_child_by_name() compatible with sysfs inputs
device-dax: add resize support
mm/memremap_pages: convert to 'struct range'
mm/memremap_pages: support multiple ranges per invocation
device-dax: add dis-contiguous resource support
device-dax: introduce 'mapping' devices
device-dax: add an 'align' attribute

Joao Martins (3):
device-dax: make align a per-device property
dax/hmem: introduce dax_hmem.region_idle parameter
device-dax: add a range mapping allocation attribute

arch/powerpc/kvm/book3s_hv_uvmem.c | 14
drivers/base/core.c | 2
drivers/dax/bus.c | 1039 ++++++++++++++++++++++++++++++--
drivers/dax/bus.h | 11
drivers/dax/dax-private.h | 58 ++
drivers/dax/device.c | 112 ++-
drivers/dax/hmem/hmem.c | 17 -
drivers/dax/kmem.c | 178 +++--
drivers/dax/pmem/compat.c | 2
drivers/dax/pmem/core.c | 14
drivers/gpu/drm/nouveau/nouveau_dmem.c | 15
drivers/nvdimm/badrange.c | 26 -
drivers/nvdimm/claim.c | 13
drivers/nvdimm/nd.h | 3
drivers/nvdimm/pfn_devs.c | 13
drivers/nvdimm/pmem.c | 27 -
drivers/nvdimm/region.c | 21 -
drivers/pci/p2pdma.c | 12
drivers/xen/unpopulated-alloc.c | 45 +
include/linux/memremap.h | 11
include/linux/range.h | 6
lib/test_hmm.c | 15
mm/memremap.c | 299 +++++----
tools/testing/nvdimm/dax-dev.c | 22 -
tools/testing/nvdimm/test/iomap.c | 2
25 files changed, 1557 insertions(+), 420 deletions(-)

base-commit: 6764736525f27a411ba2c0c430aaa2df7375f3ac

 \ /
  Last update: 2020-09-25 22:46    [W:0.205 / U:1.720 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site