qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Orit Wasserman <owasserm@redhat.com>
To: Juan Quintela <quintela@redhat.com>
Cc: chegu_vinod@hp.com, qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH 6/7] Exit loop if we have been there too long
Date: Thu, 14 Jun 2012 14:36:55 +0300	[thread overview]
Message-ID: <4FD9CCD7.3030407@redhat.com> (raw)
In-Reply-To: <189d97ce396dd1b4e8b497fbd1b276086ac05cbc.1337710679.git.quintela@redhat.com>

On 05/22/2012 09:32 PM, Juan Quintela wrote:
> cheking each 64 pages is a random magic number as good as any other.
s/cheking/checking
> We don't want to test too many times, but on the other hand,
> qemu_get_clock_ns() is not so expensive either.  We want to be sure
> that we spent less than 50ms (half of buffered_file timer), if we
> spent more than 100ms, all the accounting got wrong.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
>  arch_init.c |   15 +++++++++++++++
>  1 files changed, 15 insertions(+), 0 deletions(-)
> 
> diff --git a/arch_init.c b/arch_init.c
> index 76a3d4e..2aa77ff 100644
> --- a/arch_init.c
> +++ b/arch_init.c
> @@ -296,6 +296,7 @@ int ram_save_live(QEMUFile *f, int stage, void *opaque)
>      uint64_t bytes_transferred_last;
>      double bwidth = 0;
>      int ret;
> +    int i;
> 
>      if (stage < 0) {
>          memory_global_dirty_log_stop();
> @@ -335,6 +336,7 @@ int ram_save_live(QEMUFile *f, int stage, void *opaque)
>      bytes_transferred_last = bytes_transferred;
>      bwidth = qemu_get_clock_ns(rt_clock);
> 
> +    i = 0;
>      while ((ret = qemu_file_rate_limit(f)) == 0) {
>          int bytes_sent;
> 
> @@ -343,6 +345,19 @@ int ram_save_live(QEMUFile *f, int stage, void *opaque)
>          if (bytes_sent == 0) { /* no more blocks */
>              break;
>          }
> +        /* we want to check in the 1st loop, just in case it was the 1st time
> +           and we had to sync the dirty bitmap.
> +           qemu_get_clock_ns() is a bit expensive, so we only check each some
> +           iterations
> +        */
> +        if ((i & 63) == 0) {
> +            uint64_t t1 = (qemu_get_clock_ns(rt_clock) - bwidth) / 1000000;
> +            if (t1 > 50) { /* 50ms, half buffered_file limit */
can't we use a constant ?
> +                printf("big delay %ld milliseconds, %d iterations\n", t1, i);
printf ? 

Orit
> +                break;
> +            }
> +        }
> +        i++;
>      }
> 
>      if (ret < 0) {

  reply	other threads:[~2012-06-14 11:36 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-22 18:32 [Qemu-devel] [RFC 0/7] Fix migration with lots of memory Juan Quintela
2012-05-22 18:32 ` [Qemu-devel] [PATCH 1/7] Add spent time for migration Juan Quintela
2012-06-14 10:52   ` Orit Wasserman
2012-05-22 18:32 ` [Qemu-devel] [PATCH 2/7] Add tracepoints for savevm section start/end Juan Quintela
2012-06-14 11:00   ` Orit Wasserman
2012-05-22 18:32 ` [Qemu-devel] [PATCH 3/7] No need to iterate if we already are over the limit Juan Quintela
2012-06-14 11:03   ` Orit Wasserman
2012-05-22 18:32 ` [Qemu-devel] [PATCH 4/7] Only TCG needs TLB handling Juan Quintela
2012-06-14 11:15   ` Orit Wasserman
2012-05-22 18:32 ` [Qemu-devel] [PATCH 5/7] Only calculate expected_time for stage 2 Juan Quintela
2012-06-14 11:31   ` Orit Wasserman
2012-05-22 18:32 ` [Qemu-devel] [PATCH 6/7] Exit loop if we have been there too long Juan Quintela
2012-06-14 11:36   ` Orit Wasserman [this message]
2012-06-21 19:34     ` Juan Quintela
2012-06-22  2:42       ` 陳韋任 (Wei-Ren Chen)
2012-06-22 12:44         ` Juan Quintela
2012-05-22 18:32 ` [Qemu-devel] [PATCH 7/7] Maintaing number of dirty pages Juan Quintela
2012-06-14 11:42   ` Orit Wasserman
2012-06-11  3:56 ` [Qemu-devel] [RFC 0/7] Fix migration with lots of memory Chegu Vinod

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=4FD9CCD7.3030407@redhat.com \
    --to=owasserm@redhat.com \
    --cc=chegu_vinod@hp.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).