From: Jens Axboe <axboe@kernel.dk>
To: "Elliott, Robert (Persistent Memory)" <elliott@hpe.com>,
"Gavriliuk, Anton (HPS Ukraine)" <anton.gavriliuk@hpe.com>,
Rebecca Cran <rebecca@bluestop.org>,
Sitsofe Wheeler <sitsofe@gmail.com>
Cc: "fio@vger.kernel.org" <fio@vger.kernel.org>,
"Kani, Toshimitsu" <toshi.kani@hpe.com>
Subject: Re: fio 3.2
Date: Wed, 29 Nov 2017 21:21:30 -0700 [thread overview]
Message-ID: <aa9efb3f-9fd5-5322-7614-aa8c773dfa29@kernel.dk> (raw)
In-Reply-To: <AT5PR8401MB03875F6DB61311D1D8B40419AB3B0@AT5PR8401MB0387.NAMPRD84.PROD.OUTLOOK.COM>
On 11/28/2017 09:13 PM, Elliott, Robert (Persistent Memory) wrote:
> small_content_scramble has hardly been touched since 2011, so it probably
> hasn't had much performance analysis.
That's fair, would be a good thing to look at, especially since it's on
by default.
> One of the few changes made was to add an integer divide by 1000, which
> is always slow (painfully slow on some CPUs):
>
> offset = ((io_u->start_time.tv_nsec/1000) ^ boffset) & 511;
>
> perf top doesn't show that taking time - I think the compiler realized
> it could pull that calculation out of the loop and just do it once. Different
> compilers and compiler options might not realize that.
Yeah, int divides suck. This didn't used to be like this, it got added
with the nsec conversion. I have fixed this up now.
--
Jens Axboe
next prev parent reply other threads:[~2017-11-30 4:21 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <TU4PR8401MB099112614914A9232A07490DEC270@TU4PR8401MB0991.NAMPRD84.PROD.OUTLOOK.COM>
[not found] ` <903d418b-bac4-104b-28e5-3c529efab7f5@kernel.dk>
2017-11-26 5:30 ` fio 3.2 Gavriliuk, Anton (HPS Ukraine)
2017-11-26 15:28 ` Sitsofe Wheeler
2017-11-27 12:39 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-27 19:38 ` Sitsofe Wheeler
2017-11-27 22:48 ` Rebecca Cran
2017-11-28 1:12 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-28 4:51 ` Elliott, Robert (Persistent Memory)
2017-11-29 3:35 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-29 4:13 ` Elliott, Robert (Persistent Memory)
2017-11-29 4:44 ` Rebecca Cran
2017-11-30 4:04 ` Jens Axboe
2017-11-29 6:55 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-29 7:40 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-29 8:22 ` Elliott, Robert (Persistent Memory)
2017-11-29 13:12 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-29 17:17 ` Sitsofe Wheeler
2017-11-29 20:24 ` Elliott, Robert (Persistent Memory)
2017-11-30 4:21 ` Jens Axboe [this message]
2017-11-30 14:17 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-30 21:23 ` Jens Axboe
2017-12-01 5:21 ` Gavriliuk, Anton (HPS Ukraine)
2017-12-01 7:15 ` Robert Elliott (Persistent Memory)
2017-12-03 9:35 ` Sitsofe Wheeler
2017-12-03 17:10 ` Jens Axboe
2017-12-05 20:29 ` Elliott, Robert (Persistent Memory)
2017-12-03 3:24 ` Gavriliuk, Anton (HPS Ukraine)
2017-12-04 17:12 ` Gavriliuk, Anton (HPS Ukraine)
2017-12-04 17:30 ` Robert Elliott (Persistent Memory)
2017-12-04 18:25 ` Gavriliuk, Anton (HPS Ukraine)
2017-12-04 18:50 ` Jens Axboe
2017-12-04 19:17 ` Jeff Furlong
2017-12-04 19:29 ` Gavriliuk, Anton (HPS Ukraine)
2017-12-04 18:41 ` Jens Axboe
2017-12-04 17:30 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-30 16:13 ` Jens Axboe
2017-11-30 16:18 ` Jens Axboe
[not found] ` <F5206176-8A92-46B4-9092-F901C132E22C@bluestop.org>
2017-11-28 8:38 ` Sitsofe Wheeler
2017-11-28 14:25 ` Gavriliuk, Anton (HPS Ukraine)
2017-11-29 0:04 ` Sitsofe Wheeler
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=aa9efb3f-9fd5-5322-7614-aa8c773dfa29@kernel.dk \
--to=axboe@kernel.dk \
--cc=anton.gavriliuk@hpe.com \
--cc=elliott@hpe.com \
--cc=fio@vger.kernel.org \
--cc=rebecca@bluestop.org \
--cc=sitsofe@gmail.com \
--cc=toshi.kani@hpe.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.