public inbox for linux-arm-kernel@lists.infradead.org 
 help / color / mirror / Atom feed
From: fgenfb@yahoo•com (Harm Hanemaaijer)
To: linux-arm-kernel@lists•infradead.org
Subject: Call for testing/opinions: Optimized memset/memcpy
Date: Sat, 13 Jul 2013 21:51:18 +0000 (UTC)	[thread overview]
Message-ID: <loom.20130713T232124-511@post.gmane.org> (raw)
In-Reply-To: 20130713172445.GL32054@1wt.eu

Willy Tarreau <w <at> 1wt.eu> writes:

> OK I've run bench.script on the following platforms :

Thanks, that's incredibly helpful!

Note that Thumb2 mode usually doesn't do much in synthetic benchmarks,
because the benchmark code will fit into the L1 instruction cache; the
benefit of Thumb2 happens in real-world usage when the active code
footprint becomes larger.

To summarize, memset seems to be in good shape and also the "fast path"
for common word-aligned memcpy of size <= 256 seems to be working well.

However, the copy_page and memcpy results for larger sizes seem to suggest
that the prefetch strategy isn't working well on these platforms. Note also
that on the quad core the existing copy_page is also highly sub-optimal.

Fixing the preload strategy for these platforms may simply be a case of
changing the configurable constant PREFETCH_DISTANCE from 3 to 2 (from an
offset of 192 bytes to 128 bytes), which more closely mimics the original
kernel memcpy. I have added PREFETCH_DISTANCE as a configurable parameter
in the Makefile in the latest version of test-arm-kernel-memcpy. It will
be interesting to see the results of testing with a PREFETCH_DISTANCE
of 2 especially on the quad-core platform or a similar one.

  reply	other threads:[~2013-07-13 21:51 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-13 15:51 Call for testing/opinions: Optimized memset/memcpy Harm Hanemaaijer
2013-07-13 16:48 ` Dr. David Alan Gilbert
2013-07-13 21:13   ` Harm Hanemaaijer
2013-07-15 13:15     ` Catalin Marinas
2013-07-14 11:19   ` Harm Hanemaaijer
2013-07-14 11:32     ` Dr. David Alan Gilbert
2013-07-14 11:37     ` Ard Biesheuvel
2013-07-14 13:13       ` Russell King - ARM Linux
2013-07-14 13:33       ` Harm Hanemaaijer
2013-07-14 14:09         ` Ard Biesheuvel
2013-07-14 14:32           ` Russell King - ARM Linux
2013-07-13 17:24 ` Willy Tarreau
2013-07-13 21:51   ` Harm Hanemaaijer [this message]
2013-07-14  6:13     ` Willy Tarreau
2013-07-14 11:00       ` Harm Hanemaaijer
2013-07-14 13:09         ` Russell King - ARM Linux
2013-07-14 13:59           ` Harm Hanemaaijer
2013-07-14 15:21         ` Siarhei Siamashka

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=loom.20130713T232124-511@post.gmane.org \
    --to=fgenfb@yahoo$(echo .)com \
    --cc=linux-arm-kernel@lists$(echo .)infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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