All of lore.kernel.org
 help / color / mirror / Atom feed
From: Will Deacon <will.deacon@arm.com>
To: huanglingyan <huanglingyan2@huawei.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>,
	liuyun01@kylinos.cn, linux-arm-kernel@lists.infradead.org,
	ard.biesheuvel@linaro.org
Subject: Re: [PATCH] arm64: lib: accelerate do_csum() with NEON instruction
Date: Mon, 3 Dec 2018 19:32:41 +0000	[thread overview]
Message-ID: <20181203193241.GH29028@arm.com> (raw)
In-Reply-To: <1542792065-98282-1-git-send-email-huanglingyan2@huawei.com>

[+ Ard and Jackie]

On Wed, Nov 21, 2018 at 05:21:05PM +0800, huanglingyan wrote:
> From: Lingyan Huang <huanglingyan2@huawei.com>
> 
> Function do_csum() in lib/checksum.c is used to compute checksum,
> which is turned out to be slowly and costs a lot of resources.
> Let's use neon instructions to accelerate the checksum computation
> for arm64.
> 
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will.deacon@arm.com>
> Signed-off-by: Lingyan Huang <huanglingyan2@huawei.com>
> ---
>  arch/arm64/include/asm/checksum.h |   8 ++
>  arch/arm64/lib/Makefile           |   3 +
>  arch/arm64/lib/checksum.c         |  30 +++++++
>  arch/arm64/lib/do_csum.S          | 182 ++++++++++++++++++++++++++++++++++++++
>  lib/checksum.c                    |   6 +-
>  5 files changed, 226 insertions(+), 3 deletions(-)
>  create mode 100644 arch/arm64/lib/checksum.c
>  create mode 100644 arch/arm64/lib/do_csum.S

I think we can avoid dropping into assembly for this if we build on top of
arm_neon.h for the core of the loop:

> +do_loop_64:
> +
> +	ldp	q5, q4, [x0], #32
> +	ldp	q7, q6, [x0], #32
> +
> +    uadalp v0.4s, v4.8h
> +    uadalp v1.4s, v5.8h
> +    uadalp v2.4s, v6.8h
> +    uadalp v3.4s, v7.8h

So please look at Jackie's patch for XOR checksumming as inspiration:

http://lists.infradead.org/pipermail/linux-arm-kernel/2018-November/615625.html

Thanks,

Will

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

      parent reply	other threads:[~2018-12-03 19:33 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-21  9:21 [PATCH] arm64: lib: accelerate do_csum() with NEON instruction huanglingyan
2018-11-21 14:41 ` Robin Murphy
2018-11-26 11:28   ` huanglingyan (A)
2018-11-28  1:47   ` huanglingyan (A)
2018-12-03 19:32 ` Will Deacon [this message]

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=20181203193241.GH29028@arm.com \
    --to=will.deacon@arm.com \
    --cc=ard.biesheuvel@linaro.org \
    --cc=catalin.marinas@arm.com \
    --cc=huanglingyan2@huawei.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=liuyun01@kylinos.cn \
    /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.