linux-kernel - [v4 0/6] optimize memory hotplug

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20180215165920.8570-1-pasha.tatashin@oracle.com>
Date:   Thu, 15 Feb 2018 11:59:14 -0500
From:   Pavel Tatashin <pasha.tatashin@...cle.com>
To:     steven.sistare@...cle.com, daniel.m.jordan@...cle.com,
        akpm@...ux-foundation.org, mgorman@...hsingularity.net,
        mhocko@...e.com, linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        gregkh@...uxfoundation.org, vbabka@...e.cz,
        bharata@...ux.vnet.ibm.com, tglx@...utronix.de, mingo@...hat.com,
        hpa@...or.com, x86@...nel.org, dan.j.williams@...el.com,
        kirill.shutemov@...ux.intel.com, bhe@...hat.com
Subject: [v4 0/6] optimize memory hotplug

Changelog:
	v3 - v4
	Addressed comments from Ingo Molnar and from Michal Hocko
	Splitted 4th patch into three patches
	Instead of using section table to save node ids, saving node id in
	the first page of every section.

	v2 - v3
	Fixed two issues found during testing
	Addressed Kbuild warning reports

	v1 - v2
	Added struct page poisoning checking in order to verify that
	struct pages are never accessed until initialized during memory
	hotplug

This patchset:
- Improves hotplug performance by eliminating a number of
struct page traverses during memory hotplug.

- Fixes some issues with hotplugging, where boundaries
were not properly checked. And on x86 block size was not properly aligned
with end of memory

- Also, potentially improves boot performance by eliminating condition from
  __init_single_page().

- Adds robustness by verifying that that struct pages are correctly
  poisoned when flags are accessed.

The following experiments were performed on Xeon(R) CPU E7-8895 v3 @ 2.60GHz
with 1T RAM:

booting in qemu with 960G of memory, time to initialize struct pages:

no-kvm:
	TRY1		TRY2
BEFORE:	39.433668	39.39705
AFTER:	36.903781	36.989329

with-kvm:
BEFORE:	10.977447	11.103164
AFTER:	10.929072	10.751885

Hotplug 896G memory:
no-kvm:
	TRY1		TRY2
BEFORE: 848.740000	846.910000
AFTER:  783.070000	786.560000

with-kvm:
	TRY1		TRY2
BEFORE: 34.410000	33.57
AFTER:	29.810000	29.580000

Pavel Tatashin (6):
  mm/memory_hotplug: enforce block size aligned range check
  x86/mm/memory_hotplug: determine block size based on the end of boot
    memory
  mm: uninitialized struct page poisoning sanity checking
  mm/memory_hotplug: optimize probe routine
  mm/memory_hotplug: don't read nid from struct page during hotplug
  mm/memory_hotplug: optimize memory hotplug

 arch/x86/mm/init_64.c      | 33 +++++++++++++++++++++++++++++----
 drivers/base/memory.c      | 38 +++++++++++++++++++++-----------------
 drivers/base/node.c        | 22 +++++++++++++++-------
 include/linux/mm.h         |  4 +++-
 include/linux/node.h       |  4 ++--
 include/linux/page-flags.h | 22 +++++++++++++++++-----
 mm/memblock.c              |  2 +-
 mm/memory_hotplug.c        | 36 ++++++++++++++----------------------
 mm/page_alloc.c            | 28 ++++++++++------------------
 mm/sparse.c                |  9 ++++++++-
 10 files changed, 120 insertions(+), 78 deletions(-)

-- 
2.16.1