linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Zi Yan <zi.yan@cs.rutgers.edu>
To: Balbir Singh <bsingharora@gmail.com>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	akpm@linux-foundation.org, minchan@kernel.org, vbabka@suse.cz,
	mgorman@techsingularity.net, kirill.shutemov@linux.intel.com,
	n-horiguchi@ah.jp.nec.com, khandual@linux.vnet.ibm.com
Subject: Re: [PATCH 0/5] Parallel hugepage migration optimization
Date: Mon, 28 Nov 2016 10:22:45 -0500	[thread overview]
Message-ID: <B5954713-0912-40E6-898E-BF4A05B7E5CB@cs.rutgers.edu> (raw)
In-Reply-To: <9cf7f4c6-6dde-9dbb-cf93-7874437a442d@gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1808 bytes --]

On 24 Nov 2016, at 18:59, Balbir Singh wrote:

> On 23/11/16 03:25, Zi Yan wrote:
>> From: Zi Yan <zi.yan@cs.rutgers.edu>
>>
>> Hi all,
>>
>> This patchset boosts the hugepage migration throughput and helps THP migration
>> which is added by Naoya's patches: https://lwn.net/Articles/705879/.
>>
>> Motivation
>> ===============================
>>
>> In x86, 4KB page migrations are underutilizing the memory bandwidth compared
>> to 2MB THP migrations. I did some page migration benchmarking on a two-socket
>> Intel Xeon E5-2640v3 box, which has 23.4GB/s bandwidth, and discover
>> there are big throughput gap, ~3x, between 4KB and 2MB page migrations.
>>
>> Here are the throughput numbers for different page sizes and page numbers:
>>         | 512 4KB pages | 1 2MB THP  |  1 4KB page
>> x86_64  |  0.98GB/s     |  2.97GB/s  |   0.06GB/s
>>
>> As Linux currently use single-threaded page migration, the throughput is still
>> much lower than the hardware bandwidth, 2.97GB/s vs 23.4GB/s. So I parallelize
>> the copy_page() part of THP migration with workqueue and achieve 2.8x throughput.
>>
>> Here are the throughput numbers of 2MB page migration:
>>            |  single-threaded   | 8-thread
>> x86_64 2MB |    2.97GB/s        | 8.58GB/s
>>
>
> Whats the impact on CPU utilization? Is there a huge impact?
>
> Balbir Singh.

It depends on the throughput we can achieve.

For single-threaded copy, the current routine, it takes one CPU 2MB/(2.97GB/s) = 657.6 us
to copy one 2MB page.

For 8-thread copy, it take 8 CPUs 2MB/(8.58GB/s) = 227.6 us to copy one 2MB page.

If we have 8 idle CPUs, I think it worths using them.

I am going to add code to check idle_cpu() in the system before doing the copy.
If no idle CPUs are present, I can fall back to single-threaded copy.


--
Best Regards
Yan Zi

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 496 bytes --]

      reply	other threads:[~2016-11-28 15:22 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-22 16:25 Zi Yan
2016-11-22 16:25 ` [PATCH 1/5] mm: migrate: Add mode parameter to support additional page copy routines Zi Yan
2016-11-22 19:56   ` kbuild test robot
2016-11-24  6:38     ` Anshuman Khandual
2016-11-24  6:50   ` Anshuman Khandual
2016-11-24 23:56   ` Balbir Singh
2016-11-28 15:13     ` Zi Yan
2016-11-29  5:10       ` Anshuman Khandual
2016-11-22 16:25 ` [PATCH 2/5] mm: migrate: Change migrate_mode to support combination migration modes Zi Yan
2016-11-24  8:15   ` Anshuman Khandual
2016-11-28 14:31     ` Zi Yan
2016-11-25  0:06   ` Balbir Singh
2016-11-22 16:25 ` [PATCH 3/5] migrate: Add copy_page_mt to use multi-threaded page migration Zi Yan
2016-11-24  9:26   ` Anshuman Khandual
2016-11-28 15:03     ` Zi Yan
2016-11-29  6:03       ` Anshuman Khandual
2016-11-22 16:25 ` [PATCH 4/5] mm: migrate: Add copy_page_mt into migrate_pages Zi Yan
2016-11-24  9:57   ` Anshuman Khandual
2016-11-22 16:25 ` [PATCH 5/5] mm: migrate: Add vm.accel_page_copy in sysfs to control whether to use multi-threaded to accelerate page copy Zi Yan
2016-11-24 10:09   ` Anshuman Khandual
2016-11-28 15:11     ` Zi Yan
2016-11-24 23:59 ` [PATCH 0/5] Parallel hugepage migration optimization Balbir Singh
2016-11-28 15:22   ` Zi Yan [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=B5954713-0912-40E6-898E-BF4A05B7E5CB@cs.rutgers.edu \
    --to=zi.yan@cs.rutgers.edu \
    --cc=akpm@linux-foundation.org \
    --cc=bsingharora@gmail.com \
    --cc=khandual@linux.vnet.ibm.com \
    --cc=kirill.shutemov@linux.intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mgorman@techsingularity.net \
    --cc=minchan@kernel.org \
    --cc=n-horiguchi@ah.jp.nec.com \
    --cc=vbabka@suse.cz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox