renegade-project/linux-next
2
0
Fork 0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-11-16 06:35:39 +08:00
Code

hmm: heterogeneous memory management documentation

Browse Source 
Patch series "HMM (Heterogeneous Memory Management)", v25.

Heterogeneous Memory Management (HMM) (description and justification)

Today device driver expose dedicated memory allocation API through their
device file, often relying on a combination of IOCTL and mmap calls.
The device can only access and use memory allocated through this API.
This effectively split the program address space into object allocated
for the device and useable by the device and other regular memory
(malloc, mmap of a file, share memory, â) only accessible by
CPU (or in a very limited way by a device by pinning memory).

Allowing different isolated component of a program to use a device thus
require duplication of the input data structure using device memory
allocator.  This is reasonable for simple data structure (array, grid,
image, â) but this get extremely complex with advance data
structure (list, tree, graph, â) that rely on a web of memory
pointers.  This is becoming a serious limitation on the kind of work
load that can be offloaded to device like GPU.

New industry standard like C++, OpenCL or CUDA are pushing to remove
this barrier.  This require a shared address space between GPU device
and CPU so that GPU can access any memory of a process (while still
obeying memory protection like read only).  This kind of feature is also
appearing in various other operating systems.

HMM is a set of helpers to facilitate several aspects of address space
sharing and device memory management.  Unlike existing sharing mechanism
that rely on pining pages use by a device, HMM relies on mmu_notifier to
propagate CPU page table update to device page table.

Duplicating CPU page table is only one aspect necessary for efficiently
using device like GPU.  GPU local memory have bandwidth in the TeraBytes/
second range but they are connected to main memory through a system bus
like PCIE that is limited to 32GigaBytes/second (PCIE 4.0 16x).  Thus it
is necessary to allow migration of process memory from main system memory
to device memory.  Issue is that on platform that only have PCIE the
device memory is not accessible by the CPU with the same properties as
main memory (cache coherency, atomic operations, ...).

To allow migration from main memory to device memory HMM provides a set of
helper to hotplug device memory as a new type of ZONE_DEVICE memory which
is un-addressable by CPU but still has struct page representing it.  This
allow most of the core kernel logic that deals with a process memory to
stay oblivious of the peculiarity of device memory.

When page backing an address of a process is migrated to device memory the
CPU page table entry is set to a new specific swap entry.  CPU access to
such address triggers a migration back to system memory, just like if the
page was swap on disk.  HMM also blocks any one from pinning a ZONE_DEVICE
page so that it can always be migrated back to system memory if CPU access
it.  Conversely HMM does not migrate to device memory any page that is pin
in system memory.

To allow efficient migration between device memory and main memory a new
migrate_vma() helpers is added with this patchset.  It allows to leverage
device DMA engine to perform the copy operation.

This feature will be use by upstream driver like nouveau mlx5 and probably
other in the future (amdgpu is next suspect in line).  We are actively
working on nouveau and mlx5 support.  To test this patchset we also worked
with NVidia close source driver team, they have more resources than us to
test this kind of infrastructure and also a bigger and better userspace
eco-system with various real industry workload they can be use to test and
profile HMM.

The expected workload is a program builds a data set on the CPU (from
disk, from network, from sensors, â).  Program uses GPU API (OpenCL,
CUDA, ...) to give hint on memory placement for the input data and also
for the output buffer.  Program call GPU API to schedule a GPU job, this
happens using device driver specific ioctl.  All this is hidden from
programmer point of view in case of C++ compiler that transparently
offload some part of a program to GPU.  Program can keep doing other stuff
on the CPU while the GPU is crunching numbers.

It is expected that CPU will not access the same data set as the GPU while
GPU is working on it, but this is not mandatory.  In fact we expect some
small memory object to be actively access by both GPU and CPU concurrently
as synchronization channel and/or for monitoring purposes.  Such object
will stay in system memory and should not be bottlenecked by system bus
bandwidth (rare write and read access from both CPU and GPU).

As we are relying on device driver API, HMM does not introduce any new
syscall nor does it modify any existing ones.  It does not change any
POSIX semantics or behaviors.  For instance the child after a fork of a
process that is using HMM will not be impacted in anyway, nor is there any
data hazard between child COW or parent COW of memory that was migrated to
device prior to fork.

HMM assume a numbers of hardware features.  Device must allow device page
table to be updated at any time (ie device job must be preemptable).
Device page table must provides memory protection such as read only.
Device must track write access (dirty bit).  Device must have a minimum
granularity that match PAGE_SIZE (ie 4k).

Reviewer (just hint):
Patch 1  HMM documentation
Patch 2  introduce core infrastructure and definition of HMM, pretty
         small patch and easy to review
Patch 3  introduce the mirror functionality of HMM, it relies on
         mmu_notifier and thus someone familiar with that part would be
         in better position to review
Patch 4  is an helper to snapshot CPU page table while synchronizing with
         concurrent page table update. Understanding mmu_notifier makes
         review easier.
Patch 5  is mostly a wrapper around handle_mm_fault()
Patch 6  add new add_pages() helper to avoid modifying each arch memory
         hot plug function
Patch 7  add a new memory type for ZONE_DEVICE and also add all the logic
         in various core mm to support this new type. Dan Williams and
         any core mm contributor are best people to review each half of
         this patchset
Patch 8  special case HMM ZONE_DEVICE pages inside put_page() Kirill and
         Dan Williams are best person to review this
Patch 9  allow to uncharge a page from memory group without using the lru
         list field of struct page (best reviewer: Johannes Weiner or
         Vladimir Davydov or Michal Hocko)
Patch 10 Add support to uncharge ZONE_DEVICE page from a memory cgroup (best
         reviewer: Johannes Weiner or Vladimir Davydov or Michal Hocko)
Patch 11 add helper to hotplug un-addressable device memory as new type
         of ZONE_DEVICE memory (new type introducted in patch 3 of this
         serie). This is boiler plate code around memory hotplug and it
         also pick a free range of physical address for the device memory.
         Note that the physical address do not point to anything (at least
         as far as the kernel knows).
Patch 12 introduce a new hmm_device class as an helper for device driver
         that want to expose multiple device memory under a common fake
         device driver. This is usefull for multi-gpu configuration.
         Anyone familiar with device driver infrastructure can review
         this. Boiler plate code really.
Patch 13 add a new migrate mode. Any one familiar with page migration is
         welcome to review.
Patch 14 introduce a new migration helper (migrate_vma()) that allow to
         migrate a range of virtual address of a process using device DMA
         engine to perform the copy. It is not limited to do copy from and
         to device but can also do copy between any kind of source and
         destination memory. Again anyone familiar with migration code
         should be able to verify the logic.
Patch 15 optimize the new migrate_vma() by unmapping pages while we are
         collecting them. This can be review by any mm folks.
Patch 16 add unaddressable memory migration to helper introduced in patch
         7, this can be review by anyone familiar with migration code
Patch 17 add a feature that allow device to allocate non-present page on
         the GPU when migrating a range of address to device memory. This
         is an helper for device driver to avoid having to first allocate
         system memory before migration to device memory
Patch 18 add a new kind of ZONE_DEVICE memory for cache coherent device
         memory (CDM)
Patch 19 add an helper to hotplug CDM memory

Previous patchset posting :
v1 http://lwn.net/Articles/597289/
v2 https://lkml.org/lkml/2014/6/12/559
v3 https://lkml.org/lkml/2014/6/13/633
v4 https://lkml.org/lkml/2014/8/29/423
v5 https://lkml.org/lkml/2014/11/3/759
v6 http://lwn.net/Articles/619737/
v7 http://lwn.net/Articles/627316/
v8 https://lwn.net/Articles/645515/
v9 https://lwn.net/Articles/651553/
v10 https://lwn.net/Articles/654430/
v11 http://www.gossamer-threads.com/lists/linux/kernel/2286424
v12 http://www.kernelhub.org/?msg=972982&p=2
v13 https://lwn.net/Articles/706856/
v14 https://lkml.org/lkml/2016/12/8/344
v15 http://www.mail-archive.com/linux-kernel@xxxxxxxxxxxxxxx/msg1304107.html
v16 http://www.spinics.net/lists/linux-mm/msg119814.html
v17 https://lkml.org/lkml/2017/1/27/847
v18 https://lkml.org/lkml/2017/3/16/596
v19 https://lkml.org/lkml/2017/4/5/831
v20 https://lwn.net/Articles/720715/
v21 https://lkml.org/lkml/2017/4/24/747
v22 http://lkml.iu.edu/hypermail/linux/kernel/1705.2/05176.html
v23 https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg1404788.html
v24 https://lwn.net/Articles/726691/

This patch (of 19):

This adds documentation for HMM (Heterogeneous Memory Management).  It
presents the motivation behind it, the features necessary for it to be
useful and and gives an overview of how this is implemented.

Link: http://lkml.kernel.org/r/20170817000548.32038-2-jglisse@redhat.com
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: David Nellans <dnellans@nvidia.com>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: Aneesh Kumar <aneesh.kumar@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Evgeny Baskakov <ebaskakov@nvidia.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Mark Hairgrove <mhairgrove@nvidia.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Cc: Sherry Cheung <SCheung@nvidia.com>
Cc: Subhash Gutti <sgutti@nvidia.com>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Bob Liu <liubo95@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Jérôme Glisse 2017-09-08 16:11:19 -07:00 committed by Linus Torvalds
parent 8135d8926c
commit bffc33ec53
2 changed files with 391 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

384
Documentation/vm/hmm.txt Normal file
View File

@ -0,0 +1,384 @@
Heterogeneous Memory Management (HMM)
Transparently allow any component of a program to use any memory region of said
program with a device without using device specific memory allocator. This is
becoming a requirement to simplify the use of advance heterogeneous computing
where GPU, DSP or FPGA are use to perform various computations.
This document is divided as follow, in the first section i expose the problems
related to the use of a device specific allocator. The second section i expose
the hardware limitations that are inherent to many platforms. The third section
gives an overview of HMM designs. The fourth section explains how CPU page-
table mirroring works and what is HMM purpose in this context. Fifth section
deals with how device memory is represented inside the kernel. Finaly the last
section present the new migration helper that allow to leverage the device DMA
engine.
1) Problems of using device specific memory allocator:
2) System bus, device memory characteristics
3) Share address space and migration
4) Address space mirroring implementation and API
5) Represent and manage device memory from core kernel point of view
6) Migrate to and from device memory
7) Memory cgroup (memcg) and rss accounting
-------------------------------------------------------------------------------
1) Problems of using device specific memory allocator:
Device with large amount of on board memory (several giga bytes) like GPU have
historically manage their memory through dedicated driver specific API. This
creates a disconnect between memory allocated and managed by device driver and
regular application memory (private anonymous, share memory or regular file
back memory). From here on i will refer to this aspect as split address space.
I use share address space to refer to the opposite situation ie one in which
any memory region can be use by device transparently.
Split address space because device can only access memory allocated through the
device specific API. This imply that all memory object in a program are not
equal from device point of view which complicate large program that rely on a
wide set of libraries.
Concretly this means that code that wants to leverage device like GPU need to
copy object between genericly allocated memory (malloc, mmap private/share/)
and memory allocated through the device driver API (this still end up with an
mmap but of the device file).
For flat dataset (array, grid, image, ...) this isn't too hard to achieve but
complex data-set (list, tree, ...) are hard to get right. Duplicating a complex
data-set need to re-map all the pointer relations between each of its elements.
This is error prone and program gets harder to debug because of the duplicate
data-set.
Split address space also means that library can not transparently use data they
are getting from core program or other library and thus each library might have
to duplicate its input data-set using specific memory allocator. Large project
suffer from this and waste resources because of the various memory copy.
Duplicating each library API to accept as input or output memory allocted by
each device specific allocator is not a viable option. It would lead to a
combinatorial explosions in the library entry points.
Finaly with the advance of high level language constructs (in C++ but in other
language too) it is now possible for compiler to leverage GPU or other devices
without even the programmer knowledge. Some of compiler identified patterns are
only do-able with a share address. It is as well more reasonable to use a share
address space for all the other patterns.
-------------------------------------------------------------------------------
2) System bus, device memory characteristics
System bus cripple share address due to few limitations. Most system bus only
allow basic memory access from device to main memory, even cache coherency is
often optional. Access to device memory from CPU is even more limited, most
often than not it is not cache coherent.
If we only consider the PCIE bus than device can access main memory (often
through an IOMMU) and be cache coherent with the CPUs. However it only allows
a limited set of atomic operation from device on main memory. This is worse
in the other direction the CPUs can only access a limited range of the device
memory and can not perform atomic operations on it. Thus device memory can not
be consider like regular memory from kernel point of view.
Another crippling factor is the limited bandwidth (~32GBytes/s with PCIE 4.0
and 16 lanes). This is 33 times less that fastest GPU memory (1 TBytes/s).
The final limitation is latency, access to main memory from the device has an
order of magnitude higher latency than when the device access its own memory.
Some platform are developing new system bus or additions/modifications to PCIE
to address some of those limitations (OpenCAPI, CCIX). They mainly allow two
way cache coherency between CPU and device and allow all atomic operations the
architecture supports. Saddly not all platform are following this trends and
some major architecture are left without hardware solutions to those problems.
So for share address space to make sense not only we must allow device to
access any memory memory but we must also permit any memory to be migrated to
device memory while device is using it (blocking CPU access while it happens).
-------------------------------------------------------------------------------
3) Share address space and migration
HMM intends to provide two main features. First one is to share the address
space by duplication the CPU page table into the device page table so same
address point to same memory and this for any valid main memory address in
the process address space.
To achieve this, HMM offer a set of helpers to populate the device page table
while keeping track of CPU page table updates. Device page table updates are
not as easy as CPU page table updates. To update the device page table you must
allow a buffer (or use a pool of pre-allocated buffer) and write GPU specifics
commands in it to perform the update (unmap, cache invalidations and flush,
...). This can not be done through common code for all device. Hence why HMM
provides helpers to factor out everything that can be while leaving the gory
details to the device driver.
The second mechanism HMM provide is a new kind of ZONE_DEVICE memory that does
allow to allocate a struct page for each page of the device memory. Those page
are special because the CPU can not map them. They however allow to migrate
main memory to device memory using exhisting migration mechanism and everything
looks like if page was swap out to disk from CPU point of view. Using a struct
page gives the easiest and cleanest integration with existing mm mechanisms.
Again here HMM only provide helpers, first to hotplug new ZONE_DEVICE memory
for the device memory and second to perform migration. Policy decision of what
and when to migrate things is left to the device driver.
Note that any CPU access to a device page trigger a page fault and a migration
back to main memory ie when a page backing an given address A is migrated from
a main memory page to a device page then any CPU access to address A trigger a
page fault and initiate a migration back to main memory.
With this two features, HMM not only allow a device to mirror a process address
space and keeps both CPU and device page table synchronize, but also allow to
leverage device memory by migrating part of data-set that is actively use by a
device.
-------------------------------------------------------------------------------
4) Address space mirroring implementation and API
Address space mirroring main objective is to allow to duplicate range of CPU
page table into a device page table and HMM helps keeping both synchronize. A
device driver that want to mirror a process address space must start with the
registration of an hmm_mirror struct:
int hmm_mirror_register(struct hmm_mirror *mirror,
struct mm_struct *mm);
int hmm_mirror_register_locked(struct hmm_mirror *mirror,
struct mm_struct *mm);
The locked variant is to be use when the driver is already holding the mmap_sem
of the mm in write mode. The mirror struct has a set of callback that are use
to propagate CPU page table:
struct hmm_mirror_ops {
/* sync_cpu_device_pagetables() - synchronize page tables
*
* @mirror: pointer to struct hmm_mirror
* @update_type: type of update that occurred to the CPU page table
* @start: virtual start address of the range to update
* @end: virtual end address of the range to update
*
* This callback ultimately originates from mmu_notifiers when the CPU
* page table is updated. The device driver must update its page table
* in response to this callback. The update argument tells what action
* to perform.
*
* The device driver must not return from this callback until the device
* page tables are completely updated (TLBs flushed, etc); this is a
* synchronous call.
*/
void (*update)(struct hmm_mirror *mirror,
enum hmm_update action,
unsigned long start,
unsigned long end);
};
Device driver must perform update to the range following action (turn range
read only, or fully unmap, ...). Once driver callback returns the device must
be done with the update.
When device driver wants to populate a range of virtual address it can use
either:
int hmm_vma_get_pfns(struct vm_area_struct *vma,
struct hmm_range *range,
unsigned long start,
unsigned long end,
hmm_pfn_t *pfns);
int hmm_vma_fault(struct vm_area_struct *vma,
struct hmm_range *range,
unsigned long start,
unsigned long end,
hmm_pfn_t *pfns,
bool write,
bool block);
First one (hmm_vma_get_pfns()) will only fetch present CPU page table entry and
will not trigger a page fault on missing or non present entry. The second one
do trigger page fault on missing or read only entry if write parameter is true.
Page fault use the generic mm page fault code path just like a CPU page fault.
Both function copy CPU page table into their pfns array argument. Each entry in
that array correspond to an address in the virtual range. HMM provide a set of
flags to help driver identify special CPU page table entries.
Locking with the update() callback is the most important aspect the driver must
respect in order to keep things properly synchronize. The usage pattern is :
int driver_populate_range(...)
{
struct hmm_range range;
...
again:
ret = hmm_vma_get_pfns(vma, &range, start, end, pfns);
if (ret)
return ret;
take_lock(driver->update);
if (!hmm_vma_range_done(vma, &range)) {
release_lock(driver->update);
goto again;
}
// Use pfns array content to update device page table
release_lock(driver->update);
return 0;
}
The driver->update lock is the same lock that driver takes inside its update()
callback. That lock must be call before hmm_vma_range_done() to avoid any race
with a concurrent CPU page table update.
HMM implements all this on top of the mmu_notifier API because we wanted to a
simpler API and also to be able to perform optimization latter own like doing
concurrent device update in multi-devices scenario.
HMM also serve as an impedence missmatch between how CPU page table update are
done (by CPU write to the page table and TLB flushes) from how device update
their own page table. Device update is a multi-step process, first appropriate
commands are write to a buffer, then this buffer is schedule for execution on
the device. It is only once the device has executed commands in the buffer that
the update is done. Creating and scheduling update command buffer can happen
concurrently for multiple devices. Waiting for each device to report commands
as executed is serialize (there is no point in doing this concurrently).
-------------------------------------------------------------------------------
5) Represent and manage device memory from core kernel point of view
Several differents design were try to support device memory. First one use
device specific data structure to keep information about migrated memory and
HMM hooked itself in various place of mm code to handle any access to address
that were back by device memory. It turns out that this ended up replicating
most of the fields of struct page and also needed many kernel code path to be
updated to understand this new kind of memory.
Thing is most kernel code path never try to access the memory behind a page
but only care about struct page contents. Because of this HMM switchted to
directly using struct page for device memory which left most kernel code path
un-aware of the difference. We only need to make sure that no one ever try to
map those page from the CPU side.
HMM provide a set of helpers to register and hotplug device memory as a new
region needing struct page. This is offer through a very simple API:
struct hmm_devmem *hmm_devmem_add(const struct hmm_devmem_ops *ops,
struct device *device,
unsigned long size);
void hmm_devmem_remove(struct hmm_devmem *devmem);
The hmm_devmem_ops is where most of the important things are:
struct hmm_devmem_ops {
void (*free)(struct hmm_devmem *devmem, struct page *page);
int (*fault)(struct hmm_devmem *devmem,
struct vm_area_struct *vma,
unsigned long addr,
struct page *page,
unsigned flags,
pmd_t *pmdp);
};
The first callback (free()) happens when the last reference on a device page is
drop. This means the device page is now free and no longer use by anyone. The
second callback happens whenever CPU try to access a device page which it can
not do. This second callback must trigger a migration back to system memory.
-------------------------------------------------------------------------------
6) Migrate to and from device memory
Because CPU can not access device memory, migration must use device DMA engine
to perform copy from and to device memory. For this we need a new migration
helper:
int migrate_vma(const struct migrate_vma_ops *ops,
struct vm_area_struct *vma,
unsigned long mentries,
unsigned long start,
unsigned long end,
unsigned long *src,
unsigned long *dst,
void *private);
Unlike other migration function it works on a range of virtual address, there
is two reasons for that. First device DMA copy has a high setup overhead cost
and thus batching multiple pages is needed as otherwise the migration overhead
make the whole excersie pointless. The second reason is because driver trigger
such migration base on range of address the device is actively accessing.
The migrate_vma_ops struct define two callbacks. First one (alloc_and_copy())
control destination memory allocation and copy operation. Second one is there
to allow device driver to perform cleanup operation after migration.
struct migrate_vma_ops {
void (*alloc_and_copy)(struct vm_area_struct *vma,
const unsigned long *src,
unsigned long *dst,
unsigned long start,
unsigned long end,
void *private);
void (*finalize_and_map)(struct vm_area_struct *vma,
const unsigned long *src,
const unsigned long *dst,
unsigned long start,
unsigned long end,
void *private);
};
It is important to stress that this migration helpers allow for hole in the
virtual address range. Some pages in the range might not be migrated for all
the usual reasons (page is pin, page is lock, ...). This helper does not fail
but just skip over those pages.
The alloc_and_copy() might as well decide to not migrate all pages in the
range (for reasons under the callback control). For those the callback just
have to leave the corresponding dst entry empty.
Finaly the migration of the struct page might fails (for file back page) for
various reasons (failure to freeze reference, or update page cache, ...). If
that happens then the finalize_and_map() can catch any pages that was not
migrated. Note those page were still copied to new page and thus we wasted
bandwidth but this is considered as a rare event and a price that we are
willing to pay to keep all the code simpler.
-------------------------------------------------------------------------------
7) Memory cgroup (memcg) and rss accounting
For now device memory is accounted as any regular page in rss counters (either
anonymous if device page is use for anonymous, file if device page is use for
file back page or shmem if device page is use for share memory). This is a
deliberate choice to keep existing application that might start using device
memory without knowing about it to keep runing unimpacted.
Drawbacks is that OOM killer might kill an application using a lot of device
memory and not a lot of regular system memory and thus not freeing much system
memory. We want to gather more real world experience on how application and
system react under memory pressure in the presence of device memory before
deciding to account device memory differently.
Same decision was made for memory cgroup. Device memory page are accounted
against same memory cgroup a regular page would be accounted to. This does
simplify migration to and from device memory. This also means that migration
back from device memory to regular memory can not fail because it would
go above memory cgroup limit. We might revisit this choice latter on once we
get more experience in how device memory is use and its impact on memory
resource control.
Note that device memory can never be pin nor by device driver nor through GUP
and thus such memory is always free upon process exit. Or when last reference
is drop in case of share memory or file back memory.

7
MAINTAINERS
View File

@ -7775,6 +7775,13 @@ M: Sasha Levin <alexander.levin@verizon.com>
S: Maintained
F: tools/lib/lockdep/
HMM - Heterogeneous Memory Management
M: Jérôme Glisse <jglisse@redhat.com>
L: linux-mm@kvack.org
S: Maintained
F: mm/hmm*
F: include/linux/hmm*
LIBNVDIMM BLK: MMIO-APERTURE DRIVER
M: Ross Zwisler <ross.zwisler@linux.intel.com>
L: linux-nvdimm@lists.01.org