From: Mike Rapoport <rppt@linux.vnet.ibm.com>
To: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Arnd Bergmann <arnd@arndb.de>,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Andrea Arcangeli <aarcange@redhat.com>,
linux-mm <linux-mm@kvack.org>,
lkml <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] mm: introduce MADV_CLR_HUGEPAGE
Date: Mon, 22 May 2017 11:12:25 +0300 [thread overview]
Message-ID: <20170522081223.GD27382@rapoport-lnx> (raw)
In-Reply-To: <8b21bb9a-4efc-288b-933d-be7e6a5e4a0a@linux.vnet.ibm.com>
On Mon, May 22, 2017 at 12:56:45PM +0530, Anshuman Khandual wrote:
> On 05/22/2017 11:42 AM, Mike Rapoport wrote:
> > Currently applications can explicitly enable or disable THP for a memory
> > region using MADV_HUGEPAGE or MADV_NOHUGEPAGE. However, once either of
> > these advises is used, the region will always have
> > VM_HUGEPAGE/VM_NOHUGEPAGE flag set in vma->vm_flags.
> > The MADV_CLR_HUGEPAGE resets both these flags and allows managing THP in
> > the region according to system-wide settings.
>
> Invoking madvise() for the first time with either MADV_HUGEPAGE or
> MADV_NOHUGEPAGE on the buffer will unsubscribe it from the system
> wide behavior for good. I am not saying we should not have a way
> to put it back into system wide mode but are there no other functions
> through madvise() or any other interface which may have the same
> situation.
There are madvise() interfaces that set or clear some of the vma->vm_flags,
e.g MADV_*FORK or MADV_*DUMP. The difference with MADV_*HUGEPAGE is that
it is using two flags and with current madvise() interface either of them
has to be set, but there is no interface to clear them both.
--
Sincerely yours,
Mike.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
WARNING: multiple messages have this Message-ID (diff)
From: Mike Rapoport <rppt@linux.vnet.ibm.com>
To: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Arnd Bergmann <arnd@arndb.de>,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Andrea Arcangeli <aarcange@redhat.com>,
linux-mm <linux-mm@kvack.org>,
lkml <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] mm: introduce MADV_CLR_HUGEPAGE
Date: Mon, 22 May 2017 11:12:25 +0300 [thread overview]
Message-ID: <20170522081223.GD27382@rapoport-lnx> (raw)
In-Reply-To: <8b21bb9a-4efc-288b-933d-be7e6a5e4a0a@linux.vnet.ibm.com>
On Mon, May 22, 2017 at 12:56:45PM +0530, Anshuman Khandual wrote:
> On 05/22/2017 11:42 AM, Mike Rapoport wrote:
> > Currently applications can explicitly enable or disable THP for a memory
> > region using MADV_HUGEPAGE or MADV_NOHUGEPAGE. However, once either of
> > these advises is used, the region will always have
> > VM_HUGEPAGE/VM_NOHUGEPAGE flag set in vma->vm_flags.
> > The MADV_CLR_HUGEPAGE resets both these flags and allows managing THP in
> > the region according to system-wide settings.
>
> Invoking madvise() for the first time with either MADV_HUGEPAGE or
> MADV_NOHUGEPAGE on the buffer will unsubscribe it from the system
> wide behavior for good. I am not saying we should not have a way
> to put it back into system wide mode but are there no other functions
> through madvise() or any other interface which may have the same
> situation.
There are madvise() interfaces that set or clear some of the vma->vm_flags,
e.g MADV_*FORK or MADV_*DUMP. The difference with MADV_*HUGEPAGE is that
it is using two flags and with current madvise() interface either of them
has to be set, but there is no interface to clear them both.
--
Sincerely yours,
Mike.
next prev parent reply other threads:[~2017-05-22 8:12 UTC|newest]
Thread overview: 117+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-05-22 6:12 [PATCH] mm: introduce MADV_CLR_HUGEPAGE Mike Rapoport
2017-05-22 6:12 ` Mike Rapoport
2017-05-22 7:26 ` Anshuman Khandual
2017-05-22 7:26 ` Anshuman Khandual
2017-05-22 8:12 ` Mike Rapoport [this message]
2017-05-22 8:12 ` Mike Rapoport
2017-05-22 11:42 ` Kirill A. Shutemov
2017-05-22 11:42 ` Kirill A. Shutemov
2017-05-22 13:36 ` Mike Rapoport
2017-05-22 13:36 ` Mike Rapoport
2017-05-22 13:44 ` Kirill A. Shutemov
2017-05-22 13:44 ` Kirill A. Shutemov
2017-05-22 13:55 ` Michal Hocko
2017-05-22 13:55 ` Michal Hocko
2017-05-22 14:29 ` Mike Rapoport
2017-05-22 14:29 ` Mike Rapoport
2017-05-22 15:52 ` Vlastimil Babka
2017-05-22 15:52 ` Vlastimil Babka
2017-05-22 17:51 ` Mike Rapoport
2017-05-22 17:51 ` Mike Rapoport
2017-05-24 7:50 ` Mike Rapoport
2017-05-24 7:50 ` Mike Rapoport
2017-05-24 7:58 ` Vlastimil Babka
2017-05-24 7:58 ` Vlastimil Babka
2017-05-24 7:58 ` Vlastimil Babka
2017-05-24 10:39 ` Mike Rapoport
2017-05-24 10:39 ` Mike Rapoport
2017-05-24 11:18 ` Michal Hocko
2017-05-24 11:18 ` Michal Hocko
2017-05-24 14:25 ` Pavel Emelyanov
2017-05-24 14:25 ` Pavel Emelyanov
2017-05-24 14:27 ` Mike Rapoport
2017-05-24 14:27 ` Mike Rapoport
2017-05-24 15:22 ` Andrea Arcangeli
2017-05-24 15:22 ` Andrea Arcangeli
2017-05-30 7:44 ` Michal Hocko
2017-05-30 7:44 ` Michal Hocko
2017-05-30 7:44 ` Michal Hocko
[not found] ` <20170530074408.GA7969-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-30 10:19 ` Mike Rapoport
2017-05-30 10:19 ` Mike Rapoport
2017-05-30 10:19 ` Mike Rapoport
2017-05-30 10:39 ` Michal Hocko
2017-05-30 10:39 ` Michal Hocko
[not found] ` <20170530103930.GB7969-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-30 14:04 ` Andrea Arcangeli
2017-05-30 14:04 ` Andrea Arcangeli
2017-05-30 14:04 ` Andrea Arcangeli
2017-05-30 14:39 ` Michal Hocko
2017-05-30 14:39 ` Michal Hocko
[not found] ` <20170530143941.GK7969-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-30 14:56 ` Michal Hocko
2017-05-30 14:56 ` Michal Hocko
2017-05-30 14:56 ` Michal Hocko
[not found] ` <20170530145632.GL7969-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-30 16:06 ` Andrea Arcangeli
2017-05-30 16:06 ` Andrea Arcangeli
2017-05-30 16:06 ` Andrea Arcangeli
2017-05-31 6:30 ` Vlastimil Babka
2017-05-31 6:30 ` Vlastimil Babka
2017-05-31 8:24 ` Michal Hocko
2017-05-31 8:24 ` Michal Hocko
2017-05-31 9:27 ` Mike Rapoport
2017-05-31 9:27 ` Mike Rapoport
2017-05-31 10:24 ` Michal Hocko
2017-05-31 10:24 ` Michal Hocko
[not found] ` <20170531082414.GB27783-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-31 10:22 ` Michal Hocko
2017-05-31 10:22 ` Michal Hocko
2017-05-31 10:22 ` Michal Hocko
2017-06-01 11:00 ` Mike Rapoport
2017-06-01 11:00 ` Mike Rapoport
2017-06-01 12:27 ` Michal Hocko
2017-06-01 12:27 ` Michal Hocko
2017-05-30 15:43 ` Andrea Arcangeli
2017-05-30 15:43 ` Andrea Arcangeli
2017-05-31 12:08 ` Michal Hocko
2017-05-31 12:08 ` Michal Hocko
[not found] ` <20170531120822.GL27783-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-05-31 12:39 ` Mike Rapoprt
2017-05-31 12:39 ` Mike Rapoprt
2017-05-31 12:39 ` Mike Rapoprt
2017-05-31 14:18 ` Andrea Arcangeli
2017-05-31 14:18 ` Andrea Arcangeli
[not found] ` <20170531141809.GB302-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2017-05-31 14:32 ` Michal Hocko
2017-05-31 14:32 ` Michal Hocko
2017-05-31 14:32 ` Michal Hocko
2017-05-31 15:46 ` Andrea Arcangeli
2017-05-31 15:46 ` Andrea Arcangeli
2017-06-01 6:58 ` Mike Rapoport
2017-06-01 6:58 ` Mike Rapoport
2017-06-01 6:58 ` Mike Rapoport
[not found] ` <8FA5E4C2-D289-4AF5-AA09-6C199E58F9A5-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2017-05-31 14:19 ` Michal Hocko
2017-05-31 14:19 ` Michal Hocko
2017-05-31 14:19 ` Michal Hocko
2017-06-01 6:53 ` Mike Rapoport
2017-06-01 6:53 ` Mike Rapoport
2017-06-01 8:09 ` Michal Hocko
2017-06-01 8:09 ` Michal Hocko
[not found] ` <20170601080909.GD32677-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-06-01 8:35 ` Mike Rapoport
2017-06-01 8:35 ` Mike Rapoport
2017-06-01 8:35 ` Mike Rapoport
2017-06-01 13:45 ` Andrea Arcangeli
2017-06-01 13:45 ` Andrea Arcangeli
2017-06-02 9:11 ` Mike Rapoport
2017-06-02 9:11 ` Mike Rapoport
2017-05-31 9:08 ` Mike Rapoport
2017-05-31 9:08 ` Mike Rapoport
2017-05-31 9:08 ` Mike Rapoport
2017-05-31 12:05 ` Michal Hocko
2017-05-31 12:05 ` Michal Hocko
2017-05-31 12:25 ` Mike Rapoprt
2017-05-31 12:25 ` Mike Rapoprt
2017-05-24 11:31 ` Vlastimil Babka
2017-05-24 11:31 ` Vlastimil Babka
2017-05-24 14:28 ` Pavel Emelyanov
2017-05-24 14:28 ` Pavel Emelyanov
2017-05-24 14:54 ` Vlastimil Babka
2017-05-24 14:54 ` Vlastimil Babka
2017-05-24 15:13 ` Mike Rapoport
2017-05-24 15:13 ` Mike Rapoport
2017-05-22 15:33 ` kbuild test robot
2017-05-22 15:33 ` kbuild test robot
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=20170522081223.GD27382@rapoport-lnx \
--to=rppt@linux.vnet.ibm.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=khandual@linux.vnet.ibm.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.