From: Mike Kravetz <mike.kravetz@oracle.com>
To: Dave Hansen <dave.hansen@intel.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
linux-api@vger.kernel.org
Cc: Marek Szyprowski <m.szyprowski@samsung.com>,
Michal Nazarewicz <mina86@mina86.com>,
"Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>,
Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Guy Shattah <sguy@mellanox.com>, Christoph Lameter <cl@linux.com>
Subject: Re: [RFC] mmap(MAP_CONTIG)
Date: Tue, 24 Oct 2017 15:49:12 -0700 [thread overview]
Message-ID: <29f88dc2-81f7-4042-50c3-f1c3bc957af8@oracle.com> (raw)
In-Reply-To: <b2dee13d-a19a-2b53-7317-7227749375d9@intel.com>
On 10/23/2017 03:10 PM, Dave Hansen wrote:
> On 10/03/2017 04:56 PM, Mike Kravetz wrote:
>> mmap(MAP_CONTIG) would have the following semantics:
>> - The entire mapping (length size) would be backed by physically contiguous
>> pages.
>> - If 'length' physically contiguous pages can not be allocated, then mmap
>> will fail.
>> - MAP_CONTIG only works with MAP_ANONYMOUS mappings.
>> - MAP_CONTIG will lock the associated pages in memory. As such, the same
>> privileges and limits that apply to mlock will also apply to MAP_CONTIG.
>> - A MAP_CONTIG mapping can not be expanded.
>
> Do you also need to lock out the NUMA migration APIs somehow? What
> about KSM (or does it already ignore VM_LOCKED)?
Yes, and no.
The primary use case driving this request is RDMA. As such, the pages
can not move while being used for this purpose.
When this thread was started the thought was that generic mmap would
handle the contiguous allocations. The resulting allocated pages would
be handed to the driver for additional setup based on it's specific needs.
Since then, the thought is that the driver should handle contiguous
allocations as well. I am looking at making the existing contiguous memory
allocator more usable for driver writers.
--
Mike Kravetz
>
>> - At fork time, private MAP_CONTIG mappings will be converted to regular
>> (non-MAP_CONTIG) mapping in the child. As such a COW fault in the child
>> will not require a contiguous allocation.
> Maybe we should just define it as acting as if it had MADV_DONTFORK set
> on it, and also that it doesn't allow MADV_DONTFORK to be called on it.
>
> --
> 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>
--
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>
prev parent reply other threads:[~2017-10-24 22:49 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-10-03 23:56 [RFC] mmap(MAP_CONTIG) Mike Kravetz
2017-10-04 11:54 ` Michal Nazarewicz
2017-10-04 17:08 ` Mike Kravetz
2017-10-04 21:29 ` Laura Abbott
2017-10-04 13:49 ` Anshuman Khandual
2017-10-04 16:05 ` Christopher Lameter
2017-10-04 17:38 ` Mike Kravetz
2017-10-04 17:35 ` Mike Kravetz
2017-10-05 7:06 ` Vlastimil Babka
2017-10-05 8:58 ` Guy Shattah
2017-10-05 12:36 ` Guy Shattah
2017-10-05 14:30 ` Christopher Lameter
2017-10-12 1:46 ` [RFC PATCH 0/3] Add mmap(MAP_CONTIG) support Mike Kravetz
2017-10-12 1:46 ` [RFC PATCH 1/3] mm/map_contig: Add VM_CONTIG flag to vma struct Mike Kravetz
2017-10-12 1:46 ` [RFC PATCH 2/3] mm/map_contig: Use pre-allocated pages for VM_CONTIG mappings Mike Kravetz
2017-10-12 11:04 ` Anshuman Khandual
2017-10-12 1:46 ` [RFC PATCH 3/3] mm/map_contig: Add mmap(MAP_CONTIG) support Mike Kravetz
2017-10-12 11:22 ` Anshuman Khandual
2017-10-13 15:14 ` Christopher Lameter
2017-10-12 14:37 ` Michal Hocko
2017-10-12 17:19 ` Mike Kravetz
2017-10-13 8:40 ` Michal Hocko
2017-10-13 15:20 ` Christopher Lameter
2017-10-13 15:28 ` Michal Hocko
2017-10-13 15:42 ` Christopher Lameter
2017-10-13 15:47 ` Michal Hocko
[not found] ` <20171013154747.2jv7rtfqyyagiodn-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-10-13 15:56 ` Christopher Lameter
2017-10-13 16:17 ` Michal Hocko
2017-10-15 7:50 ` Guy Shattah
2017-10-16 8:24 ` Michal Hocko
[not found] ` <20171016082456.no6ux63uy2rmj4fe-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-10-16 9:11 ` Guy Shattah
2017-10-16 12:32 ` Michal Hocko
[not found] ` <20171016123248.csntl6luxgafst6q-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-10-16 16:00 ` Christopher Lameter
2017-10-16 17:42 ` Michal Hocko
[not found] ` <20171016174229.pz3o4uhzz3qbrp6n-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-10-16 17:56 ` Christopher Lameter
2017-10-16 18:17 ` Michal Hocko
2017-10-23 15:25 ` David Nellans
2017-10-17 10:50 ` Guy Shattah
[not found] ` <AM6PR0502MB378375AF8B569DBCCFE20D7DBD4C0-md96bDB8+JV1k1TWM4Wt8cDSnupUy6xnnBOFsp37pqbUKgpGm//BTAC/G2K4zDHf@public.gmane.org>
2017-10-17 10:59 ` Michal Hocko
2017-10-17 13:22 ` Michal Nazarewicz
2017-10-17 14:20 ` Guy Shattah
2017-10-17 17:44 ` Vlastimil Babka
2017-10-17 18:23 ` Mike Kravetz
2017-10-17 19:56 ` Vlastimil Babka
[not found] ` <752b49eb-55c6-5a34-ab41-6e91dd93ea70-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2017-10-16 10:33 ` Michal Nazarewicz
[not found] ` <xa1t60bfxtzw.fsf-deATy8a+UHjQT0dZR+AlfA@public.gmane.org>
2017-10-16 11:09 ` Guy Shattah
2017-10-16 17:43 ` Mike Kravetz
[not found] ` <aff6b405-6a06-f84d-c9b1-c6fb166dff81-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
2017-10-16 18:07 ` Michal Hocko
2017-10-16 20:32 ` Mike Kravetz
2017-10-16 20:58 ` Michal Hocko
2017-10-16 21:03 ` Laura Abbott
2017-10-16 21:18 ` Mike Kravetz
[not found] ` <e8cf6227-003d-8a82-8b4d-07176b43810c-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
2017-10-17 6:59 ` Vlastimil Babka
2017-10-15 6:58 ` Pavel Machek
2017-10-16 8:18 ` Michal Hocko
2017-10-16 9:54 ` Pavel Machek
2017-10-16 12:18 ` Michal Hocko
[not found] ` <20171016121808.m4sq3g5nxeyxoymc-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org>
2017-10-16 16:02 ` Christopher Lameter
2017-10-16 17:33 ` Michal Hocko
2017-10-16 17:53 ` Christopher Lameter
2017-10-15 8:07 ` Guy Shattah
2017-10-12 10:36 ` [RFC PATCH 0/3] " Anshuman Khandual
2017-10-12 14:25 ` Anshuman Khandual
2017-10-23 22:10 ` [RFC] mmap(MAP_CONTIG) Dave Hansen
2017-10-24 22:49 ` Mike Kravetz [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=29f88dc2-81f7-4042-50c3-f1c3bc957af8@oracle.com \
--to=mike.kravetz@oracle.com \
--cc=aneesh.kumar@linux.vnet.ibm.com \
--cc=cl@linux.com \
--cc=dave.hansen@intel.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=m.szyprowski@samsung.com \
--cc=mina86@mina86.com \
--cc=sguy@mellanox.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;
as well as URLs for NNTP newsgroup(s).