qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: "Li, Liang Z" <liang.z.li@intel.com>,
	"quintela@redhat.com" <quintela@redhat.com>
Cc: "amit.shah@redhat.com" <amit.shah@redhat.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"mst@redhat.com" <mst@redhat.com>
Subject: Re: [Qemu-devel] [v2 0/2] add avx2 instruction optimization
Date: Thu, 12 Nov 2015 09:43:45 +0100	[thread overview]
Message-ID: <56445141.2070907@redhat.com> (raw)
In-Reply-To: <F2CBF3009FA73547804AE4C663CAB28E019A44B6@shsmsx102.ccr.corp.intel.com>



On 12/11/2015 03:49, Li, Liang Z wrote:
> I am very surprised about the live migration performance  result when
> I use your ' memeqzero4_paolo' instead of these SSE2 Intrinsics to
> check the zero pages.

What code were you using?  Remember I suggested using only unsigned long
checks, like

	unsigned long *p = ...
	if (p[0] || p[1] || p[2] || p[3]
	    || memcmp(p+4, p, size - 4 * sizeof(unsigned long)) != 0)
		return BUFFER_NOT_ZERO;
	else
		return BUFFER_ZERO;

> The total live migration time increased about
> 8%!   Not decreased.  Although in the unit test your '
> memeqzero4_paolo'  has better performance, any idea?

You only tested the case of zero pages.  But real pages usually are not
zero, even if they have a few zero bytes at the beginning.  It's very
important to optimize the initial check before the memcmp call.

Paolo

  reply	other threads:[~2015-11-12  8:43 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-10  2:51 [Qemu-devel] [v2 0/2] add avx2 instruction optimization Liang Li
2015-11-10  2:51 ` [Qemu-devel] [v2 1/2] cutils: " Liang Li
2015-11-12 10:08   ` Paolo Bonzini
2015-11-12 10:12     ` Li, Liang Z
2015-11-12 11:30     ` Juan Quintela
2015-11-13  2:49     ` Li, Liang Z
2015-11-13  9:30       ` Paolo Bonzini
2015-11-12 14:43   ` Richard Henderson
2015-11-10  2:51 ` [Qemu-devel] [v2 2/2] configure: add options to config avx2 Liang Li
2015-11-10  3:43 ` [Qemu-devel] [v2 0/2] add avx2 instruction optimization Eric Blake
2015-11-10  5:48   ` Li, Liang Z
2015-11-10  9:13     ` Juan Quintela
2015-11-10  9:26       ` Li, Liang Z
2015-11-10  9:35         ` Paolo Bonzini
2015-11-10  9:41           ` Li, Liang Z
2015-11-10  9:50             ` Paolo Bonzini
2015-11-10  9:56               ` Li, Liang Z
2015-11-10 10:00                 ` Paolo Bonzini
2015-11-10 10:04                   ` Li, Liang Z
2015-11-12  2:49           ` Li, Liang Z
2015-11-12  8:43             ` Paolo Bonzini [this message]
2015-11-12  8:53               ` Li, Liang Z
2015-11-12  9:04                 ` Paolo Bonzini
2015-11-12  9:40                   ` Li, Liang Z
2015-11-12  9:45                     ` Paolo Bonzini
2015-11-12  9:53                       ` Li, Liang Z
2015-11-12 11:34                         ` Juan Quintela
2015-11-12 11:42                           ` Li, Liang Z
2015-11-12 19:56                             ` Dr. David Alan Gilbert
2015-11-12 20:20                               ` Eric Blake
2016-04-07 11:09                                 ` Dr. David Alan Gilbert
2016-04-07 12:54                                   ` Michael S. Tsirkin
2016-04-07 13:42                                     ` Dr. David Alan Gilbert
2016-04-07 13:54                                     ` Paolo Bonzini
2015-11-10  9:30       ` Paolo Bonzini

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=56445141.2070907@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=amit.shah@redhat.com \
    --cc=liang.z.li@intel.com \
    --cc=mst@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@redhat.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).