qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: Eric Blake <eblake@redhat.com>
Cc: kwolf@redhat.com, jcody@redhat.com, qemu-devel@nongnu.org,
	stefanha@redhat.com
Subject: Re: [Qemu-devel] [PATCH 10/20] mirror: add buf-size argument to drive-mirror
Date: Sat, 15 Dec 2012 04:09:06 -0500 (EST)	[thread overview]
Message-ID: <867170921.24992648.1355562546865.JavaMail.root@redhat.com> (raw)
In-Reply-To: <50CBA6B3.2040801@redhat.com>

> On 12/12/2012 06:46 AM, Paolo Bonzini wrote:
> > This makes sense when the next commit starts using the extra buffer
> > space
> > to perform many I/O operations asynchronously.
> > 
> > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> > ---
> >  block/mirror.c         |  6 +++---
> >  block_int.h            |  5 +++--
> >  blockdev.c             |  9 ++++++++-
> >  hmp.c                  |  2 +-
> >  qapi-schema.json       |  5 ++++-
> >  qmp-commands.hx        |  4 +++-
> >  tests/qemu-iotests/041 | 31 +++++++++++++++++++++++++++++++
> >  7 files changed, 53 insertions(+), 9 deletions(-)
> > 
> 
> > @@ -447,7 +447,7 @@ void mirror_start(BlockDriverState *bs,
> > BlockDriverState *target,
> >      s->target = target;
> >      s->mode = mode;
> >      s->granularity = granularity;
> > -    s->buf_size = granularity;
> > +    s->buf_size = MAX(buf_size, granularity);
> 
> So you silently clamp the buffer size if the user gives something
> larger than granularity.

I silently make it large if it gives something *smaller*.  The job
always has to copy one granularity-sized chunk.

> > +++ b/qapi-schema.json
> > @@ -1641,6 +1641,9 @@
> >  #               are smaller than that, else the cluster size.
> >   Must be a
> >  #               power of 2 between 512 and 64M.
> >  #
> > +# @buf-size: #optional maximum amount of data in flight from
> > source to
> > +#            target.
> 
> Mention that it was added in 1.4.  Also, is it worth mentioning
> reasonable bounds (such as granularity), and whether it must be a
> power of two?

It is arbitrary, though making it too big makes little sense.

> >  - "granularity": granularity of the dirty bitmap (json-int,
> >  optional)
> > +- "buf_size": maximum amount of data in flight from source to
> > target
> > +  (json-int, default 10M)
> 
> Oh, so unlike granularity, this one does NOT have to be a power of 2.
> But why is the default 10M if you clamp it to granularity, which
> defaults to 64k?

I don't clamp it, I make it _at least_ as big as the granularity.

Paolo

  reply	other threads:[~2012-12-15  9:09 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-12 13:46 [Qemu-devel] [PATCH 00/20] Block device mirroring enhancements, 12-12-12 edition Paolo Bonzini
2012-12-12 13:46 ` [Qemu-devel] [PATCH 01/20] host-utils: add ffsl Paolo Bonzini
2012-12-12 23:41   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 02/20] add hierarchical bitmap data type and test cases Paolo Bonzini
2012-12-14  0:04   ` Eric Blake
2013-01-11 18:27   ` Stefan Hajnoczi
2012-12-12 13:46 ` [Qemu-devel] [PATCH 03/20] block: implement dirty bitmap using HBitmap Paolo Bonzini
2012-12-14  0:27   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 04/20] block: make round_to_clusters public Paolo Bonzini
2012-12-14 20:13   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 05/20] mirror: perform COW if the cluster size is bigger than the granularity Paolo Bonzini
2012-12-14 20:21   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 06/20] block: return count of dirty sectors, not chunks Paolo Bonzini
2012-12-14 20:49   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 07/20] block: allow customizing the granularity of the dirty bitmap Paolo Bonzini
2012-12-14 21:27   ` Eric Blake
2012-12-15  9:11     ` Paolo Bonzini
2012-12-12 13:46 ` [Qemu-devel] [PATCH 08/20] mirror: allow customizing the granularity Paolo Bonzini
2012-12-14 22:01   ` Eric Blake
2013-01-14 11:28   ` Stefan Hajnoczi
2012-12-12 13:46 ` [Qemu-devel] [PATCH 09/20] mirror: switch mirror_iteration to AIO Paolo Bonzini
2012-12-14 22:11   ` Eric Blake
2012-12-15  9:09     ` Paolo Bonzini
2012-12-15 13:05       ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 10/20] mirror: add buf-size argument to drive-mirror Paolo Bonzini
2012-12-14 22:22   ` Eric Blake
2012-12-15  9:09     ` Paolo Bonzini [this message]
2013-01-14 11:41   ` Stefan Hajnoczi
2012-12-12 13:46 ` [Qemu-devel] [PATCH 11/20] mirror: support more than one in-flight AIO operation Paolo Bonzini
2012-12-14 22:32   ` Eric Blake
2013-01-14 12:56   ` Stefan Hajnoczi
2013-01-14 13:28     ` Paolo Bonzini
2012-12-12 13:46 ` [Qemu-devel] [PATCH 12/20] mirror: support arbitrarily-sized iterations Paolo Bonzini
2012-12-14 22:39   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 13/20] oslib: add a wrapper for mmap/munmap Paolo Bonzini
2012-12-14 22:54   ` Eric Blake
2012-12-15  9:06     ` Paolo Bonzini
2012-12-12 13:46 ` [Qemu-devel] [PATCH 14/20] hbitmap: add hbitmap_alloc_with_data and hbitmap_required_size Paolo Bonzini
2012-12-17 17:14   ` Eric Blake
2012-12-17 17:18     ` Paolo Bonzini
2012-12-12 13:46 ` [Qemu-devel] [PATCH 15/20] hbitmap: add hbitmap_copy Paolo Bonzini
2012-12-17 18:25   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 16/20] block: split bdrv_enable_dirty_tracking and bdrv_disable_dirty_tracking Paolo Bonzini
2012-12-20 18:26   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 17/20] block: support a persistent dirty bitmap Paolo Bonzini
2012-12-20 23:03   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 18/20] mirror: add support for " Paolo Bonzini
2012-12-20 23:49   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 19/20] block: choose the default dirty bitmap granularity in bdrv_enable_dirty_tracking Paolo Bonzini
2012-12-20 23:53   ` Eric Blake
2012-12-12 13:46 ` [Qemu-devel] [PATCH 20/20] monitor: add commands to start/stop dirty bitmap Paolo Bonzini
2012-12-21 18:30   ` Eric Blake
2013-01-14 13:02 ` [Qemu-devel] [PATCH 00/20] Block device mirroring enhancements, 12-12-12 edition Stefan Hajnoczi

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=867170921.24992648.1355562546865.JavaMail.root@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=eblake@redhat.com \
    --cc=jcody@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@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).