All of lore.kernel.org
 help / color / mirror / Atom feed
From: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
To: Oleg Babin <obabin@virtuozzo.com>
Cc: netdev@vger.kernel.org, linux-sctp@vger.kernel.org,
	"David S. Miller" <davem@davemloft.net>,
	Vlad Yasevich <vyasevich@gmail.com>,
	Neil Horman <nhorman@tuxdriver.com>,
	Xin Long <lucien.xin@gmail.com>,
	Andrey Ryabinin <aryabinin@virtuozzo.com>
Subject: Re: [PATCH net-next 0/2] net/sctp: Avoid allocating high order memory with kmalloc()
Date: Mon, 23 Apr 2018 21:33:15 +0000	[thread overview]
Message-ID: <20180423213314.GG3711@localhost.localdomain> (raw)
In-Reply-To: <1524508866-317485-1-git-send-email-obabin@virtuozzo.com>

Hi,

On Mon, Apr 23, 2018 at 09:41:04PM +0300, Oleg Babin wrote:
> Each SCTP association can have up to 65535 input and output streams.
> For each stream type an array of sctp_stream_in or sctp_stream_out
> structures is allocated using kmalloc_array() function. This function
> allocates physically contiguous memory regions, so this can lead
> to allocation of memory regions of very high order, i.e.:
>
>   sizeof(struct sctp_stream_out) = 24,
>   ((65535 * 24) / 4096) = 383 memory pages (4096 byte per page),
>   which means 9th memory order.
>
> This can lead to a memory allocation failures on the systems
> under a memory stress.

Did you do performance tests while actually using these 65k streams
and with 256 (so it gets 2 pages)?

This will introduce another deref on each access to an element, but
I'm not expecting any impact due to it.

  Marcelo

WARNING: multiple messages have this Message-ID (diff)
From: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
To: Oleg Babin <obabin@virtuozzo.com>
Cc: netdev@vger.kernel.org, linux-sctp@vger.kernel.org,
	"David S. Miller" <davem@davemloft.net>,
	Vlad Yasevich <vyasevich@gmail.com>,
	Neil Horman <nhorman@tuxdriver.com>,
	Xin Long <lucien.xin@gmail.com>,
	Andrey Ryabinin <aryabinin@virtuozzo.com>
Subject: Re: [PATCH net-next 0/2] net/sctp: Avoid allocating high order memory with kmalloc()
Date: Mon, 23 Apr 2018 18:33:15 -0300	[thread overview]
Message-ID: <20180423213314.GG3711@localhost.localdomain> (raw)
In-Reply-To: <1524508866-317485-1-git-send-email-obabin@virtuozzo.com>

Hi,

On Mon, Apr 23, 2018 at 09:41:04PM +0300, Oleg Babin wrote:
> Each SCTP association can have up to 65535 input and output streams.
> For each stream type an array of sctp_stream_in or sctp_stream_out
> structures is allocated using kmalloc_array() function. This function
> allocates physically contiguous memory regions, so this can lead
> to allocation of memory regions of very high order, i.e.:
>
>   sizeof(struct sctp_stream_out) == 24,
>   ((65535 * 24) / 4096) == 383 memory pages (4096 byte per page),
>   which means 9th memory order.
>
> This can lead to a memory allocation failures on the systems
> under a memory stress.

Did you do performance tests while actually using these 65k streams
and with 256 (so it gets 2 pages)?

This will introduce another deref on each access to an element, but
I'm not expecting any impact due to it.

  Marcelo

  parent reply	other threads:[~2018-04-23 21:33 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-23 18:41 [PATCH net-next 0/2] net/sctp: Avoid allocating high order memory with kmalloc() Oleg Babin
2018-04-23 18:41 ` Oleg Babin
2018-04-23 18:41 ` [PATCH net-next 1/2] net/sctp: Make wrappers for accessing in/out streams Oleg Babin
2018-04-23 18:41   ` Oleg Babin
2018-04-23 21:33   ` Marcelo Ricardo Leitner
2018-04-23 21:33     ` Marcelo Ricardo Leitner
2018-04-26 22:19     ` Oleg Babin
2018-04-26 22:19       ` Oleg Babin
2018-04-23 18:41 ` [PATCH net-next 2/2] net/sctp: Replace in/out stream arrays with flex_array Oleg Babin
2018-04-23 18:41   ` Oleg Babin
2018-04-23 21:33 ` Marcelo Ricardo Leitner [this message]
2018-04-23 21:33   ` [PATCH net-next 0/2] net/sctp: Avoid allocating high order memory with kmalloc() Marcelo Ricardo Leitner
2018-04-26 22:14   ` Oleg Babin
2018-04-26 22:14     ` Oleg Babin
2018-04-26 22:28     ` Marcelo Ricardo Leitner
2018-04-26 22:28       ` Marcelo Ricardo Leitner
2018-04-26 22:45       ` Oleg Babin
2018-04-26 22:45         ` Oleg Babin
2018-07-24 15:35       ` Konstantin Khorenko
2018-07-24 15:35         ` Konstantin Khorenko
2018-07-24 17:36         ` Marcelo Ricardo Leitner
2018-07-24 17:36           ` Marcelo Ricardo Leitner
2018-08-03 16:21           ` [PATCH v2 " Konstantin Khorenko
2018-08-03 16:21             ` Konstantin Khorenko
2018-08-03 16:21             ` [PATCH v2 1/2] net/sctp: Make wrappers for accessing in/out streams Konstantin Khorenko
2018-08-03 16:21               ` Konstantin Khorenko
2018-08-03 16:41               ` David Laight
2018-08-03 16:41                 ` David Laight
2018-08-03 19:50               ` David Miller
2018-08-03 19:50                 ` David Miller
2018-08-09  8:39                 ` Konstantin Khorenko
2018-08-09  8:39                   ` Konstantin Khorenko
2018-08-03 20:40               ` Marcelo Ricardo Leitner
2018-08-03 20:40                 ` Marcelo Ricardo Leitner
2018-08-09  8:40                 ` Konstantin Khorenko
2018-08-09  8:40                   ` Konstantin Khorenko
2018-08-03 16:21             ` [PATCH v2 2/2] net/sctp: Replace in/out stream arrays with flex_array Konstantin Khorenko
2018-08-03 16:21               ` Konstantin Khorenko
2018-08-03 16:43             ` [PATCH v2 0/2] net/sctp: Avoid allocating high order memory with kmalloc() David Laight
2018-08-03 16:43               ` David Laight
2018-08-03 20:30               ` Marcelo Ricardo Leitner
2018-08-03 20:30                 ` Marcelo Ricardo Leitner
2018-08-03 20:56                 ` Michael Tuexen
2018-08-03 20:56                   ` Michael Tuexen
2018-08-06  9:34                   ` David Laight
2018-08-06  9:34                     ` David Laight
2018-08-08 14:48                     ` Marcelo Ricardo Leitner
2018-08-08 14:48                       ` Marcelo Ricardo Leitner
2018-08-03 23:36             ` Marcelo Ricardo Leitner
2018-08-03 23:36               ` Marcelo Ricardo Leitner
2018-08-09  8:43               ` Konstantin Khorenko
2018-08-09  8:43                 ` Konstantin Khorenko
2018-08-10 17:03                 ` Konstantin Khorenko
2018-08-10 17:03                   ` Konstantin Khorenko
2018-08-10 17:11                   ` [PATCH v3 " Konstantin Khorenko
2018-08-10 17:11                     ` Konstantin Khorenko
2018-08-10 17:11                     ` [PATCH v3 1/2] net/sctp: Make wrappers for accessing in/out streams Konstantin Khorenko
2018-08-10 17:11                       ` Konstantin Khorenko
2018-08-10 17:11                     ` [PATCH v3 2/2] net/sctp: Replace in/out stream arrays with flex_array Konstantin Khorenko
2018-08-10 17:11                       ` Konstantin Khorenko
2018-08-11 19:36                     ` [PATCH v3 0/2] net/sctp: Avoid allocating high order memory with kmalloc() David Miller
2018-08-11 19:36                       ` David Miller
2018-08-10 17:41                   ` [PATCH v2 " Marcelo Ricardo Leitner
2018-08-10 17:41                     ` Marcelo Ricardo Leitner

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=20180423213314.GG3711@localhost.localdomain \
    --to=marcelo.leitner@gmail.com \
    --cc=aryabinin@virtuozzo.com \
    --cc=davem@davemloft.net \
    --cc=linux-sctp@vger.kernel.org \
    --cc=lucien.xin@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=nhorman@tuxdriver.com \
    --cc=obabin@virtuozzo.com \
    --cc=vyasevich@gmail.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 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.