All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: Avi Kivity <avi@scylladb.com>, io-uring@vger.kernel.org
Subject: Re: memory access op ideas
Date: Sat, 23 Apr 2022 12:11:52 -0600	[thread overview]
Message-ID: <3035b054-dfe3-29b8-7d00-a810d15ab696@kernel.dk> (raw)
In-Reply-To: <b5b7a49a-fc43-2af3-f3c8-988ea47ae986@kernel.dk>

On 4/23/22 12:02 PM, Jens Axboe wrote:
> On 4/23/22 11:32 AM, Jens Axboe wrote:
>>> I guess copy_to_user saves us from having to consider endianness.
>>
>> I was considering that too, definitely something that should be
>> investigated. Making it a 1/2/4/8 switch and using put_user() is
>> probably a better idea. Easy enough to benchmark.
> 
> FWIW, this is the current version. Some quick benchmarking doesn't show
> any difference between copy_to_user and put_user, but that may depend on
> the arch as well (using aarch64). But we might as well use put user and
> combine it with the length check, so we explicitly only support 1/2/4/8
> sizes.

In terms of performance, on this laptop, I can do about 36-37M NOP
requests per second. If I use IORING_OP_MEMCPY with immediate mode, it's
around 15M ops/sec. This is regardless of the size, get about the same
whether it's 1 or 8 byte memory writes.

-- 
Jens Axboe


  reply	other threads:[~2022-04-23 18:12 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-13 10:33 memory access op ideas Avi Kivity
2022-04-22 12:52 ` Hao Xu
2022-04-22 13:24   ` Hao Xu
2022-04-22 13:38   ` Jens Axboe
2022-04-23  7:19     ` Hao Xu
2022-04-23 16:14   ` Avi Kivity
2022-04-22 14:50 ` Jens Axboe
2022-04-22 15:03   ` Jens Axboe
2022-04-23 16:30     ` Avi Kivity
2022-04-23 17:32       ` Jens Axboe
2022-04-23 18:02         ` Jens Axboe
2022-04-23 18:11           ` Jens Axboe [this message]
2022-04-22 20:03   ` Walker, Benjamin
2022-04-23 10:19     ` Pavel Begunkov
2022-04-23 13:20     ` Jens Axboe
2022-04-23 16:23   ` Avi Kivity
2022-04-23 17:30     ` Jens Axboe
2022-04-24 13:04       ` Avi Kivity
2022-04-24 13:30         ` Jens Axboe
2022-04-24 14:56           ` Avi Kivity
2022-04-25  0:45             ` Jens Axboe
2022-04-25 18:05               ` Walker, Benjamin

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=3035b054-dfe3-29b8-7d00-a810d15ab696@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=avi@scylladb.com \
    --cc=io-uring@vger.kernel.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.