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]
Date:   Mon, 29 May 2017 13:41:38 +0200
From:   Michal Hocko <mhocko@...nel.org>
To:     Andrew Morton <akpm@...ux-foundation.org>
Cc:     linux-mm@...ck.org, Mel Gorman <mgorman@...e.de>,
        Vlastimil Babka <vbabka@...e.cz>,
        Andrea Arcangeli <aarcange@...hat.com>,
        Jerome Glisse <jglisse@...hat.com>,
        Reza Arbab <arbab@...ux.vnet.ibm.com>,
        Yasuaki Ishimatsu <yasu.isimatu@...il.com>,
        qiuxishi@...wei.com, Kani Toshimitsu <toshi.kani@....com>,
        slaoub@...il.com, Joonsoo Kim <js1304@...il.com>,
        Andi Kleen <ak@...ux.intel.com>,
        David Rientjes <rientjes@...gle.com>,
        Daniel Kiper <daniel.kiper@...cle.com>,
        Igor Mammedov <imammedo@...hat.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: [PATCH 0/2] remove CONFIG_MOVABLE_NODE

Hi,
this has been previously posted as an RFC [1]. As it hasn't met any
disagreement, I have integrated the feedback and resending it without
RFC and hope for a soon merging (ideally along with other memory hotplug
pile).

I am continuing to cleanup the memory hotplug code and
CONFIG_MOVABLE_NODE seems dubious at best. The following two patches
simply removes the flag and make it de-facto always enabled.

The current semantic of the config option is twofold 1) it automatically
binds hotplugable nodes to have memory in zone_movable by default when
movable_node is enabled 2) forbids memory hotplug to online all the memory
as movable when !CONFIG_MOVABLE_NODE.

The later restriction is quite dubious because there is no clear cut of
how much normal memory do we need for a reasonable system operation. A
single memory block which is sufficient to allow further movable
onlines is far from sufficient (e.g a node with >2GB and memblocks
128MB will fill up this zone with struct pages leaving nothing for
other allocations). Removing the config option will not only reduce the
configuration space it also removes quite some code.

The semantic of the movable_node command line parameter is preserved.

The first patch removes the restriction mentioned above and the second
one simply removes all the CONFIG_MOVABLE_NODE related stuff. The last
patch moves movable_node flag handling to memory_hotplug proper where
it belongs.

[1] http://lkml.kernel.org/r/20170524122411.25212-1-mhocko@kernel.org

Shortlog
Michal Hocko (3):
      mm, memory_hotplug: drop artificial restriction on online/offline
      mm, memory_hotplug: drop CONFIG_MOVABLE_NODE
      mm, memory_hotplug: move movable_node to the hotplug proper

Diffstat:
 Documentation/admin-guide/kernel-parameters.txt |  7 ++-
 drivers/base/node.c                             |  4 --
 include/linux/memblock.h                        | 25 ----------
 include/linux/memory_hotplug.h                  | 10 ++++
 include/linux/nodemask.h                        |  4 --
 mm/Kconfig                                      | 26 ----------
 mm/memblock.c                                   |  3 --
 mm/memory_hotplug.c                             | 64 ++-----------------------
 mm/page_alloc.c                                 |  2 -
 9 files changed, 19 insertions(+), 126 deletions(-)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ