Date:   Tue,  4 Jan 2022 13:22:19 -0700
From:   Yu Zhao <yuzhao@...gle.com>
To:     Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     Andi Kleen <ak@...ux.intel.com>,
        Catalin Marinas <catalin.marinas@....com>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        Hillf Danton <hdanton@...a.com>, Jens Axboe <axboe@...nel.dk>,
        Jesse Barnes <jsbarnes@...gle.com>,
        Johannes Weiner <hannes@...xchg.org>,
        Jonathan Corbet <corbet@....net>,
        Matthew Wilcox <willy@...radead.org>,
        Mel Gorman <mgorman@...e.de>,
        Michael Larabel <Michael@...haellarabel.com>,
        Michal Hocko <mhocko@...nel.org>,
        Rik van Riel <riel@...riel.com>,
        Vlastimil Babka <vbabka@...e.cz>,
        Will Deacon <will@...nel.org>,
        Ying Huang <ying.huang@...el.com>,
        linux-arm-kernel@...ts.infradead.org, linux-doc@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        page-reclaim@...gle.com, x86@...nel.org,
        Yu Zhao <yuzhao@...gle.com>
Subject: [PATCH v6 0/9] Multigenerational LRU Framework

TLDR
====
The current page reclaim is too expensive in terms of CPU usage and it
often makes poor choices about what to evict. This patchset offers an
alternative solution that is performant, versatile and
straightforward.

Design objectives
=================
The design objectives are:
1. Better representation of access recency
2. Try to profit from spatial locality
3. Clear fast path making obvious choices
4. Simple self-correcting heuristics

The representation of access recency is at the core of all LRU
approximations. The multigenerational LRU (MGLRU) divides pages into
multiple lists (generations), each having bounded access recency (a
time interval). Generations establish a common frame of reference and
help make better choices, e.g., between different memcgs on a computer
or different computers in a data center (for cluster job scheduling).

Exploiting spatial locality improves the efficiency when gathering the
accessed bit. A rmap walk targets a single page and doesn't try to
profit from discovering an accessed PTE. A page table walk can sweep
all hotspots in an address space, but its search space can be too
large to make a profit. The key is to optimize both methods and use
them in combination. (PMU is another option for further exploration.)

Fast path reduces code complexity and runtime overhead. Unmapped pages
don't require TLB flushes; clean pages don't require writeback. These
facts are only helpful when other conditions, e.g., access recency,
are similar. With generations as a common frame of reference,
additional factors stand out. But obvious choices might not be good
choices; thus self-correction is required (the next objective).

The benefits of simple self-correcting heuristics are self-evident.
Again with generations as a common frame of reference, this becomes
attainable. Specifically, pages in the same generation are categorized
based on additional factors, and a closed-loop control statistically
compares the refault percentages across all categories and throttles
the eviction of those that have higher percentages.

Patchset overview
=================
1. mm: x86, arm64: add arch_has_hw_pte_young()
2. mm: x86: add CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG
Materializing hardware optimizations when trying to clear the accessed
bit in many PTEs. If hardware automatically sets the accessed bit in
PTEs, there is no need to worry about bursty page faults (emulating
the accessed bit). If it also sets the accessed bit in non-leaf PMD
entries, there is no need to search the PTE table pointed to by a PMD
entry that doesn't have the accessed bit set.

3. mm/vmscan.c: refactor shrink_node()
A minor refactor.

4. mm: multigenerational lru: groundwork
Adding the basic data structure and the functions to initialize it and
insert/remove pages.

5. mm: multigenerational lru: mm_struct list
An infra keeps track of mm_struct's for page table walkers and
provides them with optimizations, i.e., switch_mm() tracking and Bloom
filters.

6. mm: multigenerational lru: aging
7. mm: multigenerational lru: eviction
"The page reclaim" is a producer/consumer model. "The aging" produces
cold pages, whereas "the eviction " consumes them. Cold pages flow
through generations. The aging uses the mm_struct list infra to sweep
dense hotspots in page tables. During a page table walk, the aging
clears the accessed bit and tags accessed pages with the youngest
generation number. The eviction sorts those pages when it encounters
them. For pages in the oldest generation, eviction walks the rmap to
check the accessed bit one more time before evicting them. During an
rmap walk, the eviction feeds dense hotspots back to the aging. Dense
hotspots flow through the Bloom filters. For pages not mapped in page
tables, the eviction uses the PID controller to statistically
determine whether they have higher refaults. If so, the eviction
throttles their eviction by moving them to the next generation (the
second oldest).

8. mm: multigenerational lru: user interface
The knobs to turn on/off MGLRU and provide the userspace with
thrashing prevention, working set estimation (the aging) and proactive
reclaim (the eviction).

9. mm: multigenerational lru: Kconfig
The Kconfig options.

Benchmark results
=================
Independent lab results
-----------------------
Based on the popularity of searches [01] and the memory usage in
Google's public cloud, the most popular open-source memory-hungry
applications, in alphabetical order, are:
      Apache Cassandra      Memcached
      Apache Hadoop         MongoDB
      Apache Spark          PostgreSQL
      MariaDB (MySQL)       Redis

An independent lab evaluated MGLRU with the most widely used benchmark
suites for the above applications. They posted 960 data points along
with kernel metrics and perf profiles collected over more than 500
hours of total benchmark time. Their final reports show that, with 95%
confidence intervals (CIs), the above applications all performed
significantly better for at least part of their benchmark matrices.

On 5.14:
1. Apache Spark [02] took 95% CIs [9.28, 11.19]% and [12.20, 14.93]%
   less wall time to sort three billion random integers, respectively,
   under the medium- and the high-concurrency conditions, when
   overcommitting memory. There were no statistically significant
   changes in wall time for the rest of the benchmark matrix.
2. MariaDB [03] achieved 95% CIs [5.24, 10.71]% and [20.22, 25.97]%
   more transactions per minute (TPM), respectively, under the medium-
   and the high-concurrency conditions, when overcommitting memory.
   There were no statistically significant changes in TPM for the rest
   of the benchmark matrix.
3. Memcached [04] achieved 95% CIs [23.54, 32.25]%, [20.76, 41.61]%
   and [21.59, 30.02]% more operations per second (OPS), respectively,
   for sequential access, random access and Gaussian (distribution)
   access, when THP=always; 95% CIs [13.85, 15.97]% and
   [23.94, 29.92]% more OPS, respectively, for random access and
   Gaussian access, when THP=never. There were no statistically
   significant changes in OPS for the rest of the benchmark matrix.
4. MongoDB [05] achieved 95% CIs [2.23, 3.44]%, [6.97, 9.73]% and
   [2.16, 3.55]% more operations per second (OPS), respectively, for
   exponential (distribution) access, random access and Zipfian
   (distribution) access, when underutilizing memory; 95% CIs
   [8.83, 10.03]%, [21.12, 23.14]% and [5.53, 6.46]% more OPS,
   respectively, for exponential access, random access and Zipfian
   access, when overcommitting memory.

On 5.15:
5. Apache Cassandra [06] achieved 95% CIs [1.06, 4.10]%, [1.94, 5.43]%
   and [4.11, 7.50]% more operations per second (OPS), respectively,
   for exponential (distribution) access, random access and Zipfian
   (distribution) access, when swap was off; 95% CIs [0.50, 2.60]%,
   [6.51, 8.77]% and [3.29, 6.75]% more OPS, respectively, for
   exponential access, random access and Zipfian access, when swap was
   on.
6. Apache Hadoop [07] took 95% CIs [5.31, 9.69]% and [2.02, 7.86]%
   less average wall time to finish twelve parallel TeraSort jobs,
   respectively, under the medium- and the high-concurrency
   conditions, when swap was on. There were no statistically
   significant changes in average wall time for the rest of the
   benchmark matrix.
7. PostgreSQL [08] achieved 95% CI [1.75, 6.42]% more transactions per
   minute (TPM) under the high-concurrency condition, when swap was
   off; 95% CIs [12.82, 18.69]% and [22.70, 46.86]% more TPM,
   respectively, under the medium- and the high-concurrency
   conditions, when swap was on. There were no statistically
   significant changes in TPM for the rest of the benchmark matrix.
8. Redis [09] achieved 95% CIs [0.58, 5.94]%, [6.55, 14.58]% and
   [11.47, 19.36]% more total operations per second (OPS),
   respectively, for sequential access, random access and Gaussian
   (distribution) access, when THP=always; 95% CIs [1.27, 3.54]%,
   [10.11, 14.81]% and [8.75, 13.64]% more total OPS, respectively,
   for sequential access, random access and Gaussian access, when
   THP=never.

Our lab results
---------------
To supplement the above results, we ran the following benchmark suites
on 5.16-rc7 and found no regressions [10]. (These synthetic benchmarks
are popular among MM developers, but we prefer large-scale A/B
experiments to validate improvements.)
      fs_fio_bench_hdd_mq      pft
      fs_lmbench               pgsql-hammerdb
      fs_parallelio            redis
      fs_postmark              stream
      hackbench                sysbenchthread
      kernbench                tpcc_spark
      memcached                unixbench
      multichase               vm-scalability
      mutilate                 will-it-scale
      nginx

[01] https://trends.google.com
[02] https://lore.kernel.org/linux-mm/20211102002002.92051-1-bot@edi.works/
[03] https://lore.kernel.org/linux-mm/20211009054315.47073-1-bot@edi.works/
[04] https://lore.kernel.org/linux-mm/20211021194103.65648-1-bot@edi.works/
[05] https://lore.kernel.org/linux-mm/20211109021346.50266-1-bot@edi.works/
[06] https://lore.kernel.org/linux-mm/20211202062806.80365-1-bot@edi.works/
[07] https://lore.kernel.org/linux-mm/20211209072416.33606-1-bot@edi.works/
[08] https://lore.kernel.org/linux-mm/20211218071041.24077-1-bot@edi.works/
[09] https://lore.kernel.org/linux-mm/20211122053248.57311-1-bot@edi.works/
[10] https://lore.kernel.org/linux-mm/20220104202247.2903702-1-yuzhao@google.com/

Read-world applications
=======================
Third-party testimonials
------------------------
Konstantin wrote [11]:
   I have Archlinux with 8G RAM + zswap + swap. While developing, I
   have lots of apps opened such as multiple LSP-servers for different
   langs, chats, two browsers, etc... Usually, my system gets quickly
   to a point of SWAP-storms, where I have to kill LSP-servers,
   restart browsers to free memory, etc, otherwise the system lags
   heavily and is barely usable.
   
   1.5 day ago I migrated from 5.11.15 kernel to 5.12 + the LRU
   patchset, and I started up by opening lots of apps to create memory
   pressure, and worked for a day like this. Till now I had *not a
   single SWAP-storm*, and mind you I got 3.4G in SWAP. I was never
   getting to the point of 3G in SWAP before without a single
   SWAP-storm.

The Arch Linux Zen kernel [12] has been using MGLRU since 5.12. Many
of its users reported their positive experiences to me, e.g., Shivodit
wrote:
   I've tried the latest Zen kernel (5.14.13-zen1-1-zen in the
   archlinux testing repos), everything's been smooth so far. I also
   decided to copy a large volume of files to check performance under
   I/O load, and everything went smoothly - no stuttering was present,
   everything was responsive.

Large-scale deployments
-----------------------
We've rolled out MGLRU to tens of millions of Chrome OS users and
about a million Android users. Google's fleetwide profiling [13] shows
an overall 40% decrease in kswapd CPU usage, in addition to
improvements in other UX metrics, e.g., an 85% decrease in the number
of low-memory kills at the 75th percentile and an 18% decrease in
rendering latency at the 50th percentile.

[11] https://lore.kernel.org/linux-mm/140226722f2032c86301fbd326d91baefe3d7d23.camel@yandex.ru/
[12] https://github.com/zen-kernel/zen-kernel/
[13] https://research.google/pubs/pub44271/

Summery
=======
The facts are:
1. The independent lab results and the real-world applications
   indicate substantial improvements; there are no known regressions.
2. Thrashing prevention, working set estimation and proactive reclaim
   work out of the box; there are no equivalent solutions.
3. There is a lot of new code; nobody has demonstrated smaller changes
   with similar effects.

Our options, accordingly, are:
1. Given the amount of evidence, the reported improvements will likely
   materialize for a wide range of workloads.
2. Gauging the interest from the past discussions [14][15][16], the
   new features will likely be put to use for both personal computers
   and data centers.
3. Based on Google's track record, the new code will likely be well
   maintained in the long term. It'd be more difficult if not
   impossible to achieve similar effects on top of the existing
   design.

[14] https://lore.kernel.org/lkml/20201005081313.732745-1-andrea.righi@canonical.com/
[15] https://lore.kernel.org/lkml/20210716081449.22187-1-sj38.park@gmail.com/
[16] https://lore.kernel.org/lkml/20211130201652.2218636d@mail.inbox.lv/

Yu Zhao (9):
  mm: x86, arm64: add arch_has_hw_pte_young()
  mm: x86: add CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG
  mm/vmscan.c: refactor shrink_node()
  mm: multigenerational lru: groundwork
  mm: multigenerational lru: mm_struct list
  mm: multigenerational lru: aging
  mm: multigenerational lru: eviction
  mm: multigenerational lru: user interface
  mm: multigenerational lru: Kconfig

 Documentation/vm/index.rst          |    1 +
 Documentation/vm/multigen_lru.rst   |   80 +
 arch/Kconfig                        |    9 +
 arch/arm64/include/asm/cpufeature.h |    5 +
 arch/arm64/include/asm/pgtable.h    |   13 +-
 arch/arm64/kernel/cpufeature.c      |   19 +
 arch/arm64/tools/cpucaps            |    1 +
 arch/x86/Kconfig                    |    1 +
 arch/x86/include/asm/pgtable.h      |    9 +-
 arch/x86/mm/pgtable.c               |    5 +-
 fs/exec.c                           |    2 +
 fs/fuse/dev.c                       |    3 +-
 include/linux/cgroup.h              |   15 +-
 include/linux/memcontrol.h          |   11 +
 include/linux/mm.h                  |   42 +
 include/linux/mm_inline.h           |  204 ++
 include/linux/mm_types.h            |   78 +
 include/linux/mmzone.h              |  175 ++
 include/linux/nodemask.h            |    1 +
 include/linux/oom.h                 |   16 +
 include/linux/page-flags-layout.h   |   19 +-
 include/linux/page-flags.h          |    4 +-
 include/linux/pgtable.h             |   17 +-
 include/linux/sched.h               |    4 +
 include/linux/swap.h                |    4 +
 kernel/bounds.c                     |    3 +
 kernel/cgroup/cgroup-internal.h     |    1 -
 kernel/exit.c                       |    1 +
 kernel/fork.c                       |    9 +
 kernel/sched/core.c                 |    1 +
 mm/Kconfig                          |   48 +
 mm/huge_memory.c                    |    3 +-
 mm/memcontrol.c                     |   26 +
 mm/memory.c                         |   21 +-
 mm/mm_init.c                        |    6 +-
 mm/oom_kill.c                       |    4 +-
 mm/page_alloc.c                     |    1 +
 mm/rmap.c                           |    7 +
 mm/swap.c                           |   51 +-
 mm/vmscan.c                         | 2691 ++++++++++++++++++++++++++-
 mm/workingset.c                     |  119 +-
 41 files changed, 3591 insertions(+), 139 deletions(-)
 create mode 100644 Documentation/vm/multigen_lru.rst

-- 
2.34.1.448.ga2b2bfdf31-goog

Powered by blists - more mailing lists