Sorry, you need to enable JavaScript to visit this website.

Linux Kernel Performance

Linux development evolves rapidly. The performance and scalability of the OS kernel has been a key part of its success. However, discussions have appeared on LKML (Linux Kernel Mailing List) regarding large performance regression between kernel versions. These discussions underscore the need for a systematic and disciplined way to characterize, improve, and test Linux kernel performance. Our goal is to work with the Linux community to further enhance the Linux kernel with consistent performance increases (avoiding degradations) across releases. The information available on this site gives community members better information about what 0-Day and LKP (Linux Kernel Performance) are doing to preserve performance integrity of the kernel.

0-Day CI Linux Kernel Performance Report (v5.9)

BY Rong Chen ON Oct 30, 2020
  1. Introduction

0-Day CI is an automated Linux kernel test service that provides comprehensive test coverage of the Linux kernel. It covers kernel build, static analysis, boot, functional, performance and power tests. This report shows the recent observations of kernel performance status on IA platform based on the test results from 0-Day CI service. It is structured in the following manner:

  • Section 2, merged regressions and improvements in v5.9 release candidates

  • Section 3, test parameter description

  • Section 4, captured regressions and improvements by shift-left testing during developers’ and maintainers’ tree during v5.9 release cycle

  • Section 5, performance comparison among different kernel releases

  • Section 6, test machine list

 

  1. Test parameters

Here are the descriptions for each parameter/field used in the tests. 

 

Classification

Name

Description

General

runtime

Run the test case within a certain time period (seconds or minutes)

 

nr_task

If it is an integer, which means the number of processes/threads (to run the workload) of this job. Default is 1.

If it is a percentage, e.g. 200% means the number of processes/threads is double of cpu number

 

nr_threads

Alias of nr_task

 

iterations

Number to repeat this job

 

test_size

Test disk size or memory size

 

set_nic_irq_affinity

Set NIC interrupt affinity

 

disable_latency_stats

Latency_stats may introduce too much noise if there are too many context switches, allow to disable it

 

transparent_hugepage

Set transparent hugepage policy (/sys/kernel/mm/transparent_hugepage)

 

boot_params:bp1_memmap

Boot parameters of memmap

 

disk:nr_pmem

number of pmem partitions used by test

 

swap:priority

Priority means the  priority  of  the  swap device. priority is a value between -1 and 32767, the default is -1 and higher priority with higher value. 

Test Machine

model

Name of Intel processor microarchitecture

 

brand

Brand name of cpu

 

cpu_number

Number of cpu

 

memory

Size of memory


 

  1. Linux Kernel v5.9 Release Test

The 5.9 release of the Linux kernel was on October 11, 2020. Linus has released the 5.9 kernel. "Ok, so I'll be honest - I had hoped for quite a bit fewer changes this last week, but at the same time there doesn't really seem to be anything particularly scary in here. It's just more commits and more lines changed than I would have wished for." Some of the significant features in this release are: x86 FSGSBASE supportcapacity awareness in the deadline scheduler, the close_range() system call, proactive compaction in the memory-management subsystem, the rationalization of kernel-thread priorities, and more. See the KernelNewbies 5.9 page for more details.

 

0-Day CI monitored the release closely to trace down the performance status on IA platform. 0-Day observed 7 regressions and 14 improvements during feature development phase for v5.9. We will share more detailed information together with correlated patches that led to the results. Note that the assessment is limited by the test coverage 0-Day has now. The list is summarized in the observation summary section.

  1. Observation Summary

0-Day CI observed 7 regressions and 14 improvements during the feature development phase for v5.9, which is in the time frame from v5.9-rc1 to v5.9 release. 

Test Indicator

Report

Test Scenario

Test Machine

Development Base

Status

aim7.jobs-per-min

[LKP] [f2fs] 79963d967b: 468.8% improvement

disk: 1BRD_48G

fs: f2fs

test: sync_disk_rw

load: 200

cpufreq_governor: performance

lkp-csl-2ap2

v5.7

merged at v5.9-rc1

fsmark.files_per_sec

[LKP] [nfsd4] 94415b06eb: -24.6% regression

iterations: 1x

nr_threads: 1t

disk: 1BRD_32G

fs: xfs

fs2: nfsv4

filesize: 4K

test_size: 4G

sync_method: fsyncBeforeClose

nr_files_per_directory: 1fpd

cpufreq_governor: performance

lkp-cpx-4s1

v5.8-rc5

merged at v5.9-rc1, fixed by commit ae3c57b5ca47 merged at v5.10-rc1

hackbench.throughput

[LKP] [mm] 2a9127fcf2: -69.2% regression

nr_threads: 1600%

iterations: 10

mode: process

ipc: socket

cpufreq_governor: performance

lkp-csl-2sp8

v5.8

merged at v5.9-rc1, Linus has explained: “Which is not surprising, I guess.”

stress-ng.dup.ops_per_sec

[LKP] [x86/entry] 4facb95b7a: -8.7% regression

nr_threads: 10%

disk: 1HDD

testtime: 30s

class: filesystem

cpufreq_governor: performance

fs: btrfs

lkp-csl-2sp5

v5.9-rc3

merged at v5.9-rc4, no response from author yet

stress-ng.sigq.ops_per_sec

[LKP] [rcu/tree] 53c72b590b: 18.8% improvement

nr_threads: 100%

disk: 1HDD

testtime: 30s

class: interrupt

cpufreq_governor: performance

lkp-csl-2sp7

v5.8-rc3

merged at v5.9-rc1

vm-scalability.throughput

[LKP] [sched/fair] 3edecfef02: 7.3% improvement

runtime: 300s

size: 8T

test: anon-w-seq-mt

cpufreq_governor: performance

lkp-skl-fpga01

v5.8-rc6

merged at v5.9-rc1

vm-scalability.throughput

[LKP] [mm] 2a9127fcf2: 130.0% improvement

runtime: 300s

size: 2T

test: shm-xread-seq-mt

cpufreq_governor: performance

lkp-cfl-e1

v5.8

merged at v5.9-rc1

vm-scalability.throughput

[LKP] [drm/mgag200] 913ec479bb: 26.2% improvement

runtime: 300s

size: 8T

test: anon-cow-seq-hugetlb

cpufreq_governor: performance

lkp-knm01

v5.7-rc1

merged at v5.9-rc1

vm-scalability.throughput

[LKP] [mm/workingset] 170b04b7ae: 11.2% improvement

runtime: 300s

size: 1T

test: lru-shm

cpufreq_governor: performance

lkp-skl-fpga01

v5.8

merged at v5.9-rc1

vm-scalability.throughput

[LKP] [mm] 09854ba94c: 31.4% improvement

runtime: 300s

size: 8T

test: anon-cow-seq

cpufreq_governor: performance

lkp-skl-fpga01

v5.8

merged at v5.9-rc4

vm-scalability.throughput

[LKP] [mm] 5ef64cc898: -20.6% regression

runtime: 300s

size: 2T

test: shm-xread-seq

cpufreq_governor: performance

lkp-cfl-e1

v5.9-rc5

merged at v5.9-rc6, Linus has  explained.

will-it-scale.per_process_ops

[LKP] [bpf] af7ec13833: -2.5% regression

nr_task: 16

mode: process

test: mmap1

cpufreq_governor: performance

lkp-csl-2ap2

v5.8-rc1

merged at v5.9-rc1, it’s a function align issue

will-it-scale.per_process_ops

[LKP] [perf] e17d43b93e: 5.8% improvement

nr_task: 50%

mode: process

test: signal1

cpufreq_governor: performance

lkp-csl-2ap3

v5.8-rc1

merged at v5.9-rc1

will-it-scale.per_process_ops

[LKP] [fsnotify] c738fbabb0: -9.5% regression

nr_task: 16

mode: process

test: open1

cpufreq_governor: performance

lkp-csl-2ap2

v5.8-rc4

merged at v5.9-rc1, the author is working on it.

will-it-scale.per_process_ops

[LKP] [fsnotify] 71d734103e: 4.2% improvement

nr_task: 50%

mode: process

test: eventfd1

cpufreq_governor: performance

lkp-csl-2ap3

v5.8-rc4

merged at v5.9-rc1

will-it-scale.per_process_ops

[LKP] [perf/x86/intel/lbr] 47125db27e: 229.0% improvement

nr_task: 50%

mode: process

test: context_switch1

cpufreq_governor: performance

lkp-csl-2ap3

v5.8-rc3

merged at v5.9-rc1

will-it-scale.per_process_ops

[LKP] [bpf] 9216477449: 196.3% improvementv

nr_task: 50%

mode: process

test: context_switch1

cpufreq_governor: performance

lkp-csl-2ap3

v5.8-rc4

merged at v5.9-rc1

will-it-scale.per_process_ops

[LKP] [core/entry] b6ec413461: 5.0% improvement

nr_task: 16

mode: process

test: lseek2

cpufreq_governor: performance

lkp-csl-2ap2

v5.9-rc5

merged at v5.9-rc6

will-it-scale.per_process_ops

[LKP] [kprobes] d002b8bc6d: 8.4% improvement

nr_task: 50%

mode: process

test: signal1

cpufreq_governor: performance

lkp-csl-2ap3

v5.8-rc1

merged at v5.9-rc1

will-it-scale.per_process_ops

[LKP] [mm] 70e806e4e6: 2.7% improvement

nr_task: 100%

mode: process

test: mmap2

cpufreq_governor: performance

lkp-skl-fpga01

v5.9-rc6

merged at v5.9-rc7 

will-it-scale.per_process_ops

[LKP] [mm] a4d63c3732: -4.2% regression

nr_task: 100%

mode: process

test: mmap2

cpufreq_governor: performance

lkp-skl-fpga01

v5.9-rc7

merged at v5.9-rc8

    1. fsmark.files_per_sec

The fsmark is a file system benchmark to test synchronous write workloads, for example, mail servers workload.

      1. scenario fsyncBeforeClose test on xfs

 

Commit 94415b06eb was reported to have -24.6% of fsmark.files_per_sec when comparing to v5.8-rc5. It was merged to mainline at v5.9-rc1.

 

Correlated commits

94415b06eb

nfsd4: a client's own opens needn't prevent delegations

branch

linus/master

report

[LKP] [nfsd4] 94415b06eb: -24.6% regression

test scenario

iterations: 1x

nr_threads: 1t

disk: 1BRD_32G

fs: xfs

fs2: nfsv4

filesize: 4K

test_size: 4G

sync_method: fsyncBeforeClose

nr_files_per_directory: 1fpd

cpufreq_governor: performance

test machine

lkp-cpx-4s1

status

merged at v5.9-rc1, fixed by commit ae3c57b5ca47 merged at v5.10-rc1

 

    1. vm-scalability.throughput

vm-scalability exercises functions and regions of the mm subsystem of the Linux kernel. We tested on multiple machines such as HSW EP server, during which we reported improvement on one test scenario.

 

  1. scenario: anon-cow-seq test

 

Commit 09854ba94c was reported to have 31.4% of vm-scalability.throughput when comparing to v5.8. It was merged to mainline at v5.9-rc4.

 

Correlated commits

09854ba94c

mm: do_wp_page() simplification

branch

linus/master

report

[LKP] [mm] 09854ba94c: 31.4% improvement

test scenario

runtime: 300s

size: 8T

test: anon-cow-seq

cpufreq_governor: performance

test machine

lkp-skl-fpga01

status

merged at v5.9-rc4

 

    1. will-it-scale.per_thread_ops

Will-it-scale takes a test case and runs it from 1 through to n parallel copies to see if the test case will scale. It builds both process and threads based tests in order to see any differences between the two.

    1. Scenario: process lseek2 test

 

  

Commit b6ec413461 was reported to have 5.0% of will-it-scale.per_process_ops when comparing to v5.9-rc5. It was merged to mainline at v5.9-rc6.

 

Correlated commits

b6ec413461

core/entry: Report syscall correctly for trace and audit

branch

linus/master

report

[LKP] [core/entry] b6ec413461: 5.0% improvement

test scenario

nr_task: 16

mode: process

test: lseek2

cpufreq_governor: performance

test machine

lkp-csl-2ap2

status

merged at v5.9-rc6


 

  1. Shift-Left Testing

Beyond testing trees in the upstream kernel, 0-Day CI also tests developers’ and maintainers’ trees, which can catch issues earlier and reduce wider impact. We call it “shift-left” testing. During the v5.9 release cycle, 0-Day CI had reported 19 major performance regressions and 10 major improvements by doing shift-left testing. We will share more detailed information together with possible code changes that led to this result for some of these, though the assessment is limited by the test coverage we have now. The whole list is summarized in the report summary section.

    1. Report Summary

0-Day CI had reported 19 performance regressions and 10 improvements by doing shift-left testing on developer and maintainer repos.

 

Test Indicator

Mail

Test Scenario

Test Machine

Status

aim7.jobs-per-min

LKP] [xfs] 456f9e38d2: -5.3% regression

disk: 1BRD_48G

fs: xfs

test: disk_rr

load: 3000

cpufreq_governor: performance

lkp-csl-2sp3

currently not merged, no response from author yet

aim7.jobs-per-min

[LKP] [xfs] 086c2d4ccc: -59.6% regression

disk: 4BRD_12G

md: RAID1

fs: xfs

test: disk_rw

load: 3000

cpufreq_governor: performance

lkp-csl-2sp2

currently not merged, no response from author yet

aim7.jobs-per-min

[LKP] [sched/fair] ec73240b16: 2.3% improvement

disk: 4BRD_12G

md: RAID1

fs: xfs

test: sync_disk_rw

load: 300

cpufreq_governor: performance

lkp-cpx-4s1

merged at v5.10-rc1

aim7.jobs-per-min

[LKP] [xfs] 41ee8a09a0: -37.4% regression

disk: 1BRD_48G

fs: xfs

test: disk_wrt

load: 3000

cpufreq_governor: performance

lkp-csl-2ap2

currently not merged, no response from author yet

fio.read_iops

[LKP] [mm/lru] 44c86dfabf: -11.8% regression

disk: 2pmem

fs: btrfs

runtime: 200s

nr_task: 50%

time_based: tb

rw: read

bs: 2M

ioengine: sync

test_size: 100G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged, the author is working on it.

fio.read_iops

[LKP] [mm/lru] 872102b4f7: -28.8% regression

disk: 2pmem

fs: xfs

runtime: 200s

nr_task: 50%

time_based: tb

rw: randread

bs: 2M

ioengine: mmap

test_size: 200G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged, the author is working on it.

fio.read_iops

[LKP] [mm/lru] 5197ae4fa3: -19.7% regression

disk: 2pmem

fs: xfs

runtime: 200s

nr_task: 50%

time_based: tb

rw: read

bs: 4k

ioengine: mmap

test_size: 200G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged, the author is working on it.

fio.read_iops

[LKP] [mm/lru] b37250464b: -32.2% regression

disk: 2pmem

fs: xfs

runtime: 200s

nr_task: 50%

time_based: tb

rw: randread

bs: 4k

ioengine: mmap

test_size: 200G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged, no response from author yet

fio.write_iops

[LKP] [btrfs] ba157afd0d: -69.4% regression

runtime: 300s

disk: 1SSD

fs: btrfs

nr_task: 100%

test_size: 128G

rw: randwrite

bs: 4k

ioengine: sync

cpufreq_governor: performance

lkp-cpx-4s1

currently not merged, no response from author yet

fio.write_iops

[LKP] [mm/lru] 0d6b06c622: -12.3% regression

disk: 2pmem

fs: xfs

runtime: 200s

nr_task: 50%

time_based: tb

rw: randrw

bs: 4k

ioengine: mmap

test_size: 200G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged, no response from author yet

fio.write_iops

[LKP] [ext4] 061113efe9: 364.6% improvement

disk: 2pmem

fs: ext4

mount_option: dax

runtime: 200s

nr_task: 50%

time_based: tb

rw: randwrite

bs: 4k

ioengine: sync

test_size: 200G

cpufreq_governor: performance

lkp-csl-2sp6

currently not merged

fsmark.files_per_sec

[LKP] [block] 8cfb68ac58: -17.9% regression

iterations: 1x

nr_threads: 32t

disk: 1SSD

fs: btrfs

filesize: 9B

test_size: 400M

sync_method: fsyncBeforeClose

nr_directories: 16d

nr_files_per_directory: 256fpd

cpufreq_governor: performance

lkp-csl-2sp7

currently not merged, no response from author yet

netperf.Throughput_Mbps

[LKP] [sched/fair] f6005b0d15: 9.4% improvement

ip: ipv4

runtime: 300s

nr_threads: 50%

cluster: cs-localhost

send_size: 5K

test: TCP_SENDFILE

cpufreq_governor: performance

lkp-cfl-e1

currently not merged

netperf.Throughput_Mbps

[LKP] [softirq] e29b4ce2af: -44.0% regression

ip: ipv4

runtime: 300s

nr_threads: 200%

cluster: cs-localhost

send_size: 10K

test: TCP_SENDFILE

cpufreq_governor: performance

lkp-csl-2ap4

currently not merged, no response from author yet

phoronix-test-suite.gluxmark.Fullscreen.1280x1024.Fragment.total_score

[LKP] [drm/i915] 55ea1cb178: 44.1% improvement

need_x: true

test: gluxmark-1.1.1

option_a: Fullscreen

option_b: Fragment

cpufreq_governor: performance

lkp-cfl-d1

currently not merged

phoronix-test-suite.supertuxkart.1024x768.Fullscreen.Ultimate.1.GranParadisoIsland.frames_per_second

[LKP] [drm/i915] 04ff178484: -30.4% regression

need_x: true

test: supertuxkart-1.5.2

option_a: Fullscreen

option_b: Ultimate

option_c: 1

option_d: Gran Paradiso Island [Approximately 275k triangles; advanced graphics]

cpufreq_governor: performance

lkp-kbl-nuc1

currently not merged, no response from author yet

phoronix-test-suite.supertuxkart.1280x1024.Fullscreen.Basic.1.GranParadisoIsland.frames_per_second

[LKP] [drm/i915] 55ea1cb178: -12.7% regression

need_x: true

test: supertuxkart-1.5.2

option_a: Fullscreen

option_b: Basic

option_c: 1

option_d: Gran Paradiso Island [Approximately 275k triangles; advanced graphics]

cpufreq_governor: performance

lkp-cfl-d1

currently not merged, no response from author yet

reaim.jobs_per_min

[LKP] [btrfs] c0aaf9b7a1: 21.8% improvement

runtime: 300s

nr_task: 100

disk: 1HDD

fs: btrfs

test: disk

cpufreq_governor: performance

lkp-csl-2sp7

currently not merged

reaim.jobs_per_min

[LKP] [locking/qspinlock] 5df52a74ab: 45.7% improvement

runtime: 300s

nr_task: 100%

test: fserver

cpufreq_governor: performance

lkp-csl-2sp5

currently not merged

unixbench.score

[LKP] [core/entry] 36ad6f6f66: 9.0% improvement

runtime: 300s

nr_task: 1

test: syscall

cpufreq_governor: performance

lkp-csl-2sp4

currently not merged

vm-scalability.throughput

[LKP] [mm/lru] f4ba6c0e1b: -27.1% regression

runtime: 300s

test: lru-file-readtwice

cpufreq_governor: performance

lkp-skl-fpga01

currently not merged, the author is working on it.

vm-scalability.throughput

[LKP] [mm/lru] f21e66f153: -9.1% regression

runtime: 300s

test: lru-file-readonce

cpufreq_governor: performance

lkp-skl-fpga01

currently not merged, no response from author yet

vm-scalability.throughput

[LKP] [mm/page_alloc] 7fef431be9: 87.8% improvement

runtime: 300s

size: 512G

test: anon-wx-rand-mt

cpufreq_governor: performance

lkp-csl-2ap4

merged at v5.10-rc1

will-it-scale.per_process_ops

[LKP] [softirq] 3e84c1dcb2: -13.5% regression

nr_task: 16

mode: process

test: writeseek2

cpufreq_governor: performance

lkp-csl-2ap2

currently not merged, no response from author yet

will-it-scale.per_process_ops

[LKP] [softirq] 56c21abbe6: -9.1% regression

nr_task: 50%

mode: process

test: unlink2

cpufreq_governor: performance

lkp-csl-2ap3

currently not merged, no response from author yet

will-it-scale.per_thread_ops

[LKP] [kernel/entry] 872a0a3f0b: -18.7% regression

nr_task: 100%

mode: thread

test: futex4

cpufreq_governor: performance

lkp-cpx-4s1

currently not merged, no response from author yet

will-it-scale.per_thread_ops

[LKP] [tracepoint] d25e37d89d: 1.4% improvement

nr_task: 100%

mode: thread

test: sched_yield

cpufreq_governor: performance

lkp-hsw-4ex1

merged at v5.10-rc1

will-it-scale.per_thread_ops

[LKP] [mm] 698ac7610f: 8.2% improvement

nr_task: 100%

mode: thread

test: mmap2

cpufreq_governor: performance

lkp-csl-2sp4

currently not merged

will-it-scale.per_thread_ops

[LKP] [sched] bdfcae1140: -37.0% regression

nr_task: 50%

mode: thread

test: context_switch1

cpufreq_governor: performance

lkp-skl-fpga01

currently not merged, the author is working on it.

    1. aim7.jobs-per-min

aim7 is a traditional UNIX system level benchmark suite which is used to test and measure the performance of a multiuser system. 

 

    1. scenario: disk_wrt test on xfs

 

Commit 41ee8a09a0 was reported to have -37.4% of aim7.jobs-per-min when comparing to v5.9-rc5.

 

Correlated commits

41ee8a09a0

xfs: deferred inode inactivation

branch

djwong-xfs/deferred-inactivation

report

[LKP] [xfs] 41ee8a09a0: -37.4% regression

test scenario

disk: 1BRD_48G

fs: xfs

test: disk_wrt

load: 3000

cpufreq_governor: performance

test machine

lkp-csl-2ap2

status

currently not merged, no response from author yet

 

  1. fio.write_iops

Fio is a tool that will spawn a number of threads or processes doing a particular type of I/O action as specified by the user.

      1. scenario: randwrite on btrfs

 

Commit ba157afd0d was reported to have -69.4% of fio.write_iops when comparing to v5.9-rc4.

 

Correlated commits

ba157afd0d

btrfs: switch extent buffer tree lock to rw_semaphore

branch

linux-next/master

report

[LKP] [btrfs] ba157afd0d: -69.4% regression

test scenario

runtime: 300s

disk: 1SSD

fs: btrfs

nr_task: 100%

test_size: 128G

rw: randwrite

bs: 4k

ioengine: sync

cpufreq_governor: performance

test machine

lkp-cpx-4s1

status

currently not merged, no response from author yet

 

    1. will-it-scale.per_process_ops

Will-it-scale takes a test case and runs it from 1 through to n parallel copies to see if the test case will scale. It builds both process and threads based tests in order to see any differences between the two.

    1. Scenario: thread futex4 test

 

  

Commit 872a0a3f0b was reported to have -18.7% of will-it-scale.per_thread_ops when comparing to v5.9-rc1.

 

Correlated commits

872a0a3f0b

kernel/entry: Add support for core-wide protection of kernel-mode

branch

linux-review/Julien-Desfossez/Core-scheduling-v7/20200829-035707

report

[LKP] [kernel/entry] 872a0a3f0b: -18.7% regression

test scenario

nr_task: 100%

mode: thread

test: futex4

cpufreq_governor: performance

test machine

lkp-cpx-4s1

status

currently not merged, no response from author yet

 

  1. Latest Release Performance Comparing

 

This session gives some information about the performance difference among different kernel releases, especially between v5.9 and v5.8. There are 50+ performance benchmarks running in 0-Day CI, and we selected 9 benchmarks which historically showed the most regressions/improvements reported by 0-Day CI. Some typical configuration/parameters are used to run the test. For some of the regressions from the comparison, 0-Day did not successfully bisect it thus no related report sent out during the release development period, but it is still worth checking. The root cause to cause the regressions won’t be covered in this session. 

 

In the following figures, the value on the Y-axis is the relative performance number. We used the v5.8 data as the base (performance number is 100).

    1. test suite: vm-scalability

vm-scalability exercises functions and regions of the mm subsystem of the Linux kernel. Below 2 tests show the typical test results. 

 

vm-scalability Test 1

vm-scalability Test 2

 

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                           

 

vm-scalability Test 1 

vm-scalability Test 2

test machine

model: Haswell-EX

brand: Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz

cpu_number: 144

memory: 512G

model: Skylake

cpu_number: 104

memory: 192G

runtime

300s

300s

size

1T

8T

vm-scalability test parameter

test case: msync-mt

test case: anon-cow-seq

performance summary

vm-scalability.throughput on kernel v5.9 has -46.97% regression when comparing to v5.8

vm-scalability.throughput on kernel v5.9 has 35.09% improvement when comparing to v5.8

 

    1. test suite: will-it-scale

Will-it-scale takes a test case and runs it from 1 through to n parallel copies to see if the test case will scale. It builds both process and threads based tests in order to see any differences between the two.

 

will-it-scale  Test 1

Will-it-scale Test 2

will-it-scale  Test 3

will-it-scale  Test 4

 

Here are the parameters and performance test summary for above tests:                                                                                                                                                                                                                          

 

will-it-scale Test 1 

will-it-scale Test 2 

will-it-scale Test 3 

will-it-scale Test 4 

test machine

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

nr_task

50%

100%

50%

16

will-it-scale test parameter

mode: process

test case: brk1

mode: process

test: page_fault3

mode: process

test: eventfd1

mode: process

test: mmap1

summary

will-it-scale.per_process_ops on kernel v5.9 has -11.13% regression when comparing to v5.8

will-it-scale.per_process_ops on kernel v5.9 has -8.39% regression when comparing to v5.8

will-it-scale.per_process_ops on kernel v5.9 has 105.61% improvement when comparing to v5.8

will-it-scale.per_process_ops on kernel v5.9 has 505.74% improvement when comparing to v5.8

 

    1. test suite: unixbench

UnixBench is a system benchmark to provide a basic indicator of the performance of a Unix-like system.

 

Unixbench Test 1

Unixbench Test 2

 

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                           

 

Unixbench Test 1 

Unixbench Test 2

test machine

model: Cascade Lake

brand: Intel(R) Xeon(R) CPU @ 2.30GHz

cpu_number: 96

memory: 128G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

runtime

300s

300s

nr_task

30%

30%

unixbench test parameter

test case: context1

test case: context1

performance summary

unixbench.score on kernel v5.9 has -5.57% regression when comparing to v5.8

unixbench.score on kernel v5.9 has 4.4% improvement when comparing to v5.8 

 

  1. test suite: reaim

reaim updates and improves the existing Open Source AIM 7 benchmark. aim7 is a traditional UNIX system level benchmark suite which is used to test and measure the performance of a multiuser system.

 

reaim Test 1                             

reaim Test 2

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                                      

 

reaim Test 1 

reaim Test 2

test machine

model: Cascade Lake

brand: Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz

cpu_number: 96

memory: 192G

model: Cascade Lake

brand: Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz

cpu_number: 96

memory: 256G

runtime

300s

300s

nr_task

100%

100

disk

1HDD

1HDD

fs

xfs

ext4

reaim test parameter

test case: disk

test case: disk

performance  summary

reaim.jobs_per_min on kernel v5.9 has 3.97% improvement when comparing to v5.8

reaim.jobs_per_min on kernel v5.9 has 7.32% improvement when comparing to v5.8

                                                                                                                                                                                                     

    1. test suite: pigz

pigz, which stands for Parallel Implementation of GZip, is a fully functional replacement for gzip that exploits multiple processors and multiple cores to the hilt when compressing data.

pigz Test 1 


 

 

Here are the test configuration and performance test summary for above tests:          

 

 

pigz Test 1

test machine

model: Knights Mill

brand: Intel(R) Xeon Phi(TM) CPU 7295 @ 1.50GHz

cpu_number: 288

memory: 80G

nr_threads

100%

pigz Test parameter

blocksize: 128K

performance  summary

pigz.throughput on kernel v5.9 has 6.01% improvement when comparing to v5.8

            

    1. test suite: netperf

Netperf is a benchmark that can be used to measure the performance of many different types of networking. It provides tests for both unidirectional throughput, and end-to-end latency.

 

netperf Test 1

netperf Test 2

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                                          

 

netperf Test 1 

netperf Test 2 

test machine

model: Cooper Lake

brand: Intel(R) Xeon(R) Gold 5318H CPU @ 2.50GHz

cpu_number: 144

memory: 128G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

disable_latency_stats

1

1

set_nic_irq_affinity

1

1

runtime

300s

300s

nr_threads

16

50%

ip

ipv4

ipv4

netperf test parameter

test case: TCP_CRR

test case: TCP_RR

performance  summary

netperf.Throughput_tps on kernel v5.9 has -31.17% regression when comparing to v5.8

netperf.Throughput_tps on kernel v5.9 has 11.49% improvement when comparing to v5.8

 

    1. test suite: hackbench

Hackbench is both a benchmark and a stress test for the Linux kernel scheduler. It's  main job  is  to  create a specified number of pairs of schedulable entities (either threads or traditional processes) which communicate via either sockets or pipes and time how long  it takes for each pair to send data back and forth.

hackbench Test 1

hackbench Test 2

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                           

 

hackbench Test 1 

hackbench Test 2 

test machine

model: Cascade Lake

brand: Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz

cpu_number: 96

memory: 192G

model: Cooper Lake

brand: Intel(R) Xeon(R) Gold 5318H CPU @ 2.50GHz

cpu_number: 144

memory: 128G

disable_latency_stats

1

1

nr_task

100%

100%

unixbench test parameter

iterations: 10

mode: process

ipc: socket

iterations: 4

mode: threads

ipc: pipe

performance summary

hackbench.throughput on kernel v5.9 has -82.54% regression when comparing to v5.8

hackbench.throughput on kernel v5.9 has 7.43% improvement when comparing to v5.8

            

    1. test suite: fio

Fio was originally written to save me the hassle of writing special test case programs when I wanted to test a specific workload, either for performance reasons or to find/reproduce a bug.

fio Test 1 

  

fio Test 2

Here are the test configuration and performance test summary for above tests:                                                                                                                                                                                                                          

 

fio Test 1

fio Test 2

test machine

model: Cascade Lake

brand: Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz

cpu_number: 96

memory: 256G

model: Cascade Lake

brand: Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz

cpu_number: 192

memory: 192G

runtime

200s

300s

file system

ext4

btrfs

disk

2pmem

1SSD

boot_params

bp1_memmap: 104G!8G

bp2_memmap: 104G!132G

No requirement

nr_task

50%

8

time_based

tb

No requirement

fio test parameter

fio-setup-basic:

  rw: randread

  bs: 2M

  ioengine: mmap

  test_size: 200G

fio-setup-basic:

  rw: write

  bs: 4k

  ioengine: sync

  test_size: 256g

performance  summary

fio.read_bw_MBps on kernel v5.9 has 9.97% improvement when comparing to v5.8

fio.write_bw_MBps on kernel v5.9 has 46.0% improvement when comparing to v5.8

 

    1. test suite: ebizzy

ebizzy is designed to generate a workload resembling common web application server workloads. It is highly threaded, has a large in-memory working set, and allocates and deallocates memory frequently.

ebizzy Test 1

 

 

Here are the test configuration and performance test summary for above test:                                                                                                                                                                                                                          

 

ebizzy Test 1

test machine

model: Cooper Lake

brand: 

cpu_number: 192

memory: 128G

transparent_hugepage

thp_enabled: always

thp_defrag: always

nr_threads

200%

iterations

100x

ebizzy test parameter

duration: 10s

performance  summary

ebizzy.throughput on kernel v5.9 is almost the same as that in v5.8


 

  1. Test Machines

    1. IVB Desktop

model

Ivy Bridge

brand

Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz

cpu number

8

memory

16G

 

model

Ivy Bridge

brand

Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz

cpu number

4

memory

8G

 

    1. SKL SP

model

Skylake

brand

Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz

cpu number

80

memory

64G

 

    1. BDW EP

model

Broadwell-EP

brand

Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz

cpu number

88

memory

128G

 

    1. HSW EP

model

Haswell-EP

brand

Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz

cpu number

72

memory

128G

 

    1. IVB EP

model

Ivy Bridge-EP

brand

Intel(R) Xeon(R) CPU E5-2690 v2 @ 3.00GHz

cpu number

40

memory

384G

 

model

Ivytown Ivy Bridge-EP

brand

Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz

cpu number

48

memory

64G

 

    1. HSX EX

model

Brickland Haswell-EX

brand

Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz

cpu number

144

memory

512G