linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Dev Jain <dev.jain@arm.com>
To: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: siddhartha@kenip.in, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org, mgorman@suse.de,
	Vlastimil Babka <vbabka@suse.cz>
Subject: Re: [PATCH] mm: limit THP alignment – performance gain observed in AI inference workloads
Date: Tue, 1 Jul 2025 12:00:21 +0530	[thread overview]
Message-ID: <80b849d4-faf3-47a9-8b8c-e8053299cfb2@arm.com> (raw)
In-Reply-To: <f746d3aa-17e7-4b42-9e08-97cdb2cad89b@lucifer.local>


On 01/07/25 11:23 am, Lorenzo Stoakes wrote:
> On Tue, Jul 01, 2025 at 11:15:25AM +0530, Dev Jain wrote:
>> Sorry I am not following, don't know in detail about the VMA merge stuff.
>> Are you saying the after the patch, the VMAs will eventually get merged?
>> Is it possible in the kernel to get a merge in the "future"; as I understand
>> it only happens at mmap() time?
>>
>> Suppose before the patch, you have two consecutive VMAs between (PMD, 2*PMD) size.
>> If they are able to get merged after the patch, why won't they be merged before the patch,
>> since the VMA characteristics are the same?
>>
>>
> Rik's patch aligned each to 2 MiB boundary. So you'd get gaps:
>
>
>    0            2MB                      4MB           6MB                      8MB          10MB
>    |-------------.------|                 |-------------.------|                 |-------------.------|
>    |             .      |		 |             .      |                 |             .      |
>    |             .      |		 |             .      |                 |             .      |
>    |-------------.------|		 |-------------.------|                 |-------------.------|
>      huge mapped  4k m'd

The effort to draw this is appreciated!

I understood the alignment, what I am asking is this:

In __get_unmapped_area(), we will return a THP-aligned addr from
thp_get_unmapped_area_vmflags(). Now for the diagram you have
drawn, suppose that before the patch, we first mmap() the
8MB-start chunk. Then we mmap the 4MB start chunk.
We go to __mmap_region(), and we see that the 8MB-start chunk
has mergeable characteristics, so we merge. So the gap goes away?

>
> If you don't force alignment then subsequent mappings will be adjacent to one
> another and those non-huge page parts can be merged.
>
> Vlasta's fix up means we only try to get the THP up-front if the length is
> already aligned at which point you won't end up with these gaps.


  reply	other threads:[~2025-07-01  6:31 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-27 10:39 siddhartha
2025-06-27 10:45 ` siddhartha
2025-06-27 15:30 ` Lorenzo Stoakes
2025-06-28  3:49   ` Dev Jain
2025-06-30  0:43     ` siddhartha
2025-06-30  5:25       ` Dev Jain
2025-06-30  5:28         ` Dev Jain
2025-06-30 10:54         ` Lorenzo Stoakes
2025-06-30 11:48           ` siddhartha
2025-07-01  5:23           ` Dev Jain
2025-07-01  5:28             ` Lorenzo Stoakes
2025-07-01  5:45               ` Dev Jain
2025-07-01  5:53                 ` Lorenzo Stoakes
2025-07-01  6:30                   ` Dev Jain [this message]
2025-07-01  6:50                     ` Lorenzo Stoakes
2025-07-01  6:58                       ` Dev Jain
2025-07-01 12:15                         ` siddhartha
2025-07-01 12:39                           ` Lorenzo Stoakes
2025-07-01 13:23                             ` siddhartha
2025-07-01 13:28                               ` Lorenzo Stoakes
2025-07-01 14:20                                 ` siddhartha
2025-07-01 16:20                             ` Dev Jain
2025-07-01 18:49                               ` Zi Yan
2025-07-07  8:56                                 ` Vlastimil Babka
2025-07-28  5:41                                   ` siddhartha
2025-07-28 11:00                                     ` Vlastimil Babka
2025-07-01 15:40                           ` Yang Shi
2025-08-11 22:14 siddhartha

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=80b849d4-faf3-47a9-8b8c-e8053299cfb2@arm.com \
    --to=dev.jain@arm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lorenzo.stoakes@oracle.com \
    --cc=mgorman@suse.de \
    --cc=siddhartha@kenip.in \
    --cc=vbabka@suse.cz \
    /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