linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Anshuman Khandual <anshuman.khandual@arm.com>
To: linux-mm@kvack.org, akpm@linux-foundation.org
Cc: mhocko@kernel.org, kirill@shutemov.name,
	kirill.shutemov@linux.intel.com, vbabka@suse.cz,
	will.deacon@arm.com, catalin.marinas@arm.com,
	dave.hansen@intel.com
Subject: [RFC 0/4] mm: Introduce lazy exec permission setting on a page
Date: Wed, 13 Feb 2019 13:36:27 +0530	[thread overview]
Message-ID: <1550045191-27483-1-git-send-email-anshuman.khandual@arm.com> (raw)

Setting an exec permission on a page normally triggers I-cache invalidation
which might be expensive. I-cache invalidation is not mandatory on a given
page if there is no immediate exec access on it. Non-fault modification of
user page table from generic memory paths like migration can be improved if
setting of the exec permission on the page can be deferred till actual use.
There was a performance report [1] which highlighted the problem.

This introduces [pte|pmd]_mklazyexec() which clears the exec permission on
a page during migration. This exec permission deferral must be enabled back
with maybe_[pmd]_mkexec() during exec page fault (FAULT_FLAG_INSTRUCTION)
if the corresponding VMA contains exec flag (VM_EXEC).

This framework is encapsulated under CONFIG_ARCH_SUPPORTS_LAZY_EXEC so that
non-subscribing architectures don't take any performance hit. For now only
generic memory migration path will be using this framework but later it can
be extended to other generic memory paths as well.

This enables CONFIG_ARCH_SUPPORTS_LAZY_EXEC on arm64 and defines required
helper functions in this regard while changing ptep_set_access_flags() to
allow non-exec to exec transition.

[1] http://lists.infradead.org/pipermail/linux-arm-kernel/2018-December/620357.html

Anshuman Khandual (4):
  mm: Introduce lazy exec permission setting on a page
  arm64/mm: Identify user level instruction faults
  arm64/mm: Allow non-exec to exec transition in ptep_set_access_flags()
  arm64/mm: Enable ARCH_SUPPORTS_LAZY_EXEC

 arch/arm64/Kconfig               |  1 +
 arch/arm64/include/asm/pgtable.h | 17 +++++++++++++++++
 arch/arm64/mm/fault.c            | 22 ++++++++++++++--------
 include/asm-generic/pgtable.h    | 12 ++++++++++++
 include/linux/mm.h               | 26 ++++++++++++++++++++++++++
 mm/Kconfig                       |  9 +++++++++
 mm/huge_memory.c                 |  5 +++++
 mm/hugetlb.c                     |  2 ++
 mm/memory.c                      |  4 ++++
 mm/migrate.c                     |  2 ++
 10 files changed, 92 insertions(+), 8 deletions(-)

-- 
2.7.4


             reply	other threads:[~2019-02-13  8:06 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-02-13  8:06 Anshuman Khandual [this message]
2019-02-13  8:06 ` [RFC 1/4] " Anshuman Khandual
2019-02-13 13:17   ` Matthew Wilcox
2019-02-13 13:53     ` Anshuman Khandual
2019-02-14  9:06       ` Mike Rapoport
2019-02-15  8:11         ` Anshuman Khandual
2019-02-15  9:49           ` Catalin Marinas
2019-02-13  8:06 ` [RFC 2/4] arm64/mm: Identify user level instruction faults Anshuman Khandual
2019-02-13  8:06 ` [RFC 3/4] arm64/mm: Allow non-exec to exec transition in ptep_set_access_flags() Anshuman Khandual
2019-02-13  8:06 ` [RFC 4/4] arm64/mm: Enable ARCH_SUPPORTS_LAZY_EXEC Anshuman Khandual
2019-02-13 11:21 ` [RFC 0/4] mm: Introduce lazy exec permission setting on a page Catalin Marinas
2019-02-13 15:38   ` Michal Hocko
2019-02-14  6:04     ` Anshuman Khandual
2019-02-14  8:38       ` Michal Hocko
2019-02-14 10:19         ` Catalin Marinas
2019-02-14 12:28           ` Michal Hocko
2019-02-15  8:45             ` Anshuman Khandual
2019-02-15  9:27               ` Michal Hocko
2019-02-18  3:07                 ` Anshuman Khandual
2019-02-14 15:38       ` Dave Hansen
2019-02-18  3:19         ` Anshuman Khandual
2019-02-13 15:44 ` Dave Hansen
2019-02-14  4:12   ` Anshuman Khandual
2019-02-14 16:55     ` Dave Hansen
2019-02-18  8:31       ` Anshuman Khandual
2019-02-18  9:04         ` Catalin Marinas
2019-02-18  9:16           ` Anshuman Khandual
2019-02-18 18:20         ` Dave Hansen

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=1550045191-27483-1-git-send-email-anshuman.khandual@arm.com \
    --to=anshuman.khandual@arm.com \
    --cc=akpm@linux-foundation.org \
    --cc=catalin.marinas@arm.com \
    --cc=dave.hansen@intel.com \
    --cc=kirill.shutemov@linux.intel.com \
    --cc=kirill@shutemov.name \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@kernel.org \
    --cc=vbabka@suse.cz \
    --cc=will.deacon@arm.com \
    /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