qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Eric Blake <eblake@redhat.com>
To: Peter Lieven <pl@kamp.de>
Cc: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCHv2 7/9] migration: do not sent zero pages in bulk stage
Date: Tue, 19 Mar 2013 11:36:48 -0600	[thread overview]
Message-ID: <5148A230.60908@redhat.com> (raw)
In-Reply-To: <1363362619-3190-8-git-send-email-pl@kamp.de>

[-- Attachment #1: Type: text/plain, Size: 1853 bytes --]

On 03/15/2013 09:50 AM, Peter Lieven wrote:
> during bulk stage of ram migration if a page is a
> zero page do not send it at all.
> the memory at the destination reads as zero anyway.
> 
> even if there is an madvise with QEMU_MADV_DONTNEED
> at the target upon receival of a zero page I have observed

s/receival/receipt/

> that the target starts swapping if the memory is overcommitted.
> it seems that the pages are dropped asynchronously.
> 
> Signed-off-by: Peter Lieven <pl@kamp.de>
> ---
>  arch_init.c |    8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)

> +++ b/arch_init.c
> @@ -432,9 +432,11 @@ static int ram_save_block(QEMUFile *f, bool last_stage)
>              bytes_sent = -1;
>              if (buffer_is_zero(p, TARGET_PAGE_SIZE)) {
>                  acct_info.dup_pages++;
> -                bytes_sent = save_block_hdr(f, block, offset, cont,
> -                                            RAM_SAVE_FLAG_COMPRESS);
> -                qemu_put_byte(f, *p);
> +                if (!ram_bulk_stage) {
> +                    bytes_sent = save_block_hdr(f, block, offset, cont,
> +                                                RAM_SAVE_FLAG_COMPRESS);
> +                    qemu_put_byte(f, *p);

Hmm, in patch 5/9, why not 'qemu_put_byte(f, 0)' instead of spending
time dereferencing *p, since you already know the byte is 0?

> +                }
>                  bytes_sent += 1;

Your accounting is now off.  This increments bytes_sent even when
nothing was sent in the bulk phase.  When touching this line, consider
using ++ instead of += 1.

>              } else if (migrate_use_xbzrle()) {
>                  current_addr = block->offset + offset;
> 

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 621 bytes --]

  reply	other threads:[~2013-03-19 17:37 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-15 15:50 [Qemu-devel] [PATCHv2 0/9] buffer_is_zero / migration optimizations Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 1/9] move vector definitions to qemu-common.h Peter Lieven
2013-03-19 15:35   ` Eric Blake
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 2/9] cutils: add a function to find non-zero content in a buffer Peter Lieven
2013-03-19 15:54   ` Eric Blake
2013-03-19 16:18     ` Peter Lieven
2013-03-19 16:43       ` Eric Blake
2013-03-19 19:42         ` Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 3/9] buffer_is_zero: use vector optimizations if possible Peter Lieven
2013-03-19 16:08   ` Eric Blake
2013-03-19 16:14     ` Peter Lieven
2013-03-19 19:44     ` Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 4/9] bitops: use vector algorithm to optimize find_next_bit() Peter Lieven
2013-03-19 16:49   ` Eric Blake
2013-03-19 19:40     ` Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 5/9] migration: search for zero instead of dup pages Peter Lieven
2013-03-19 16:55   ` Eric Blake
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 6/9] migration: add an indicator for bulk state of ram migration Peter Lieven
2013-03-19 17:32   ` Eric Blake
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 7/9] migration: do not sent zero pages in bulk stage Peter Lieven
2013-03-19 17:36   ` Eric Blake [this message]
2013-03-19 19:35     ` Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 8/9] migration: do not search dirty " Peter Lieven
2013-03-19 17:40   ` Eric Blake
2013-03-19 19:29     ` Peter Lieven
2013-03-15 15:50 ` [Qemu-devel] [PATCHv2 9/9] migration: use XBZRLE only after " Peter Lieven
2013-03-19 17:43   ` Eric Blake

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=5148A230.60908@redhat.com \
    --to=eblake@redhat.com \
    --cc=pl@kamp.de \
    --cc=qemu-devel@nongnu.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 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).