From: Max Reitz <mreitz@redhat.com>
To: Eric Blake <eblake@redhat.com>, qemu-devel@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>, Peter Lieven <pl@kamp.de>,
Stefan Hajnoczi <stefanha@redhat.com>
Subject: Re: [Qemu-devel] [PATCH 05/21] qcow2: Refcount overflow and qcow2_alloc_bytes()
Date: Wed, 12 Nov 2014 09:52:22 +0100 [thread overview]
Message-ID: <54631FC6.40703@redhat.com> (raw)
In-Reply-To: <5462683E.9000206@redhat.com>
On 2014-11-11 at 20:49, Eric Blake wrote:
> On 11/11/2014 09:18 AM, Max Reitz wrote:
>
>>> No, I was envisioning that we have a brand new image with one cluster
>>> allocated (cluster 1 has refcount 1), then 5 times in a row we do
>>> 'savevm' to take an internal snapshot. If I understand your code
>>> correctly, the first two snapshots increase the refcount, so cluster 1
>>> has a refcount of 3. Then the next snapshot can't increase the refcount,
>>> so it instead copies the contents to cluster 2.
>> No, it just errors out.
>>
>> qcow2_alloc_bytes() is only used for allocating space for a compressed
>> cluster. When taking a snapshot, update_refcount() will be called to
>> increase the clusters' refcounts, and that function will simply throw an
>> error.
> That's okay for now (always better for an initial feature to be
> conservative, then expand it later if there is demand). But I wonder if
> we could be made smarter in the future and auto-COW any cluster that
> would otherwise exceed max refcount. Thus, for a refcount_order=0
> (width=1) image, a snapshot now doubles the size of the image (as every
> single cluster would COW into a new cluster) rather than erroring out.
> Food for thought; maybe worth injecting comments into this series
> (whether in code or in commit messages, as appropriate) pointing out
> that we thought about the future possibility even though we chose not to
> allow it for now.
Ah, right, thank you. Yes, that sounds like a good idea, I'll see to it
at some later point in time.
I think adding comments will be hard because the snapshot functions
aren't really modified. They just try to increase the refcount and that
may now fail earlier than it did for a refcount width of 16 bits, so
there's no real change in behavior there, it's just that it's now
reasonably possible to hit that case. I will add appropriate comments to
the test case (which tests this snapshotting issue), though.
Max
next prev parent reply other threads:[~2014-11-12 8:52 UTC|newest]
Thread overview: 75+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-11-10 13:45 [Qemu-devel] [PATCH 00/21] qcow2: Support refcount orders != 4 Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 01/21] qcow2: Add two new fields to BDRVQcowState Max Reitz
2014-11-10 19:00 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 02/21] qcow2: Add refcount_width to format-specific info Max Reitz
2014-11-10 19:06 ` Eric Blake
2014-11-11 8:11 ` Max Reitz
2014-11-11 15:49 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 03/21] qcow2: Use 64 bits for refcount values Max Reitz
2014-11-10 20:59 ` Eric Blake
2014-11-11 8:12 ` Max Reitz
2014-11-11 9:22 ` Kevin Wolf
2014-11-11 9:25 ` Max Reitz
2014-11-11 9:26 ` Max Reitz
2014-11-11 9:36 ` Kevin Wolf
2014-11-10 13:45 ` [Qemu-devel] [PATCH 04/21] qcow2: Respect error in qcow2_alloc_bytes() Max Reitz
2014-11-10 21:05 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 05/21] qcow2: Refcount overflow and qcow2_alloc_bytes() Max Reitz
2014-11-10 21:12 ` Eric Blake
2014-11-11 8:22 ` Max Reitz
2014-11-11 16:13 ` Eric Blake
2014-11-11 16:18 ` Max Reitz
2014-11-11 19:49 ` Eric Blake
2014-11-12 8:52 ` Max Reitz [this message]
2014-11-10 13:45 ` [Qemu-devel] [PATCH 06/21] qcow2: Helper function for refcount modification Max Reitz
2014-11-10 22:30 ` Eric Blake
2014-11-11 8:35 ` Max Reitz
2014-11-11 9:43 ` Max Reitz
2014-11-11 10:56 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 07/21] qcow2: Helper for refcount array size calculation Max Reitz
2014-11-10 22:49 ` Eric Blake
2014-11-11 8:37 ` Max Reitz
2014-11-11 10:08 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 08/21] qcow2: More helpers for refcount modification Max Reitz
2014-11-11 0:29 ` Eric Blake
2014-11-11 8:42 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 09/21] qcow2: Open images with refcount order != 4 Max Reitz
2014-11-10 17:03 ` Eric Blake
2014-11-10 17:06 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 10/21] qcow2: refcount_order parameter for qcow2_create2 Max Reitz
2014-11-11 5:40 ` Eric Blake
2014-11-11 8:48 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 11/21] iotests: Prepare for refcount_width option Max Reitz
2014-11-11 17:57 ` Eric Blake
2014-11-12 8:41 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 12/21] qcow2: Allow creation with refcount order != 4 Max Reitz
2014-11-11 18:05 ` Eric Blake
2014-11-12 8:47 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 13/21] block: Add opaque value to the amend CB Max Reitz
2014-11-11 18:08 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 14/21] qcow2: Use error_report() in qcow2_amend_options() Max Reitz
2014-11-11 18:11 ` Eric Blake
2014-11-12 8:47 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 15/21] qcow2: Use abort() instead of assert(false) Max Reitz
2014-11-11 18:12 ` Eric Blake
2014-11-12 8:48 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 16/21] qcow2: Split upgrade/downgrade paths for amend Max Reitz
2014-11-11 18:14 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 17/21] qcow2: Use intermediate helper CB " Max Reitz
2014-11-11 21:05 ` Eric Blake
2014-11-12 9:10 ` Max Reitz
2014-11-10 13:45 ` [Qemu-devel] [PATCH 18/21] qcow2: Add function for refcount order amendment Max Reitz
2014-11-12 4:15 ` Eric Blake
2014-11-12 9:55 ` Max Reitz
2014-11-12 13:50 ` Eric Blake
2014-11-12 14:19 ` Eric Blake
2014-11-12 14:21 ` Max Reitz
2014-11-12 17:45 ` Max Reitz
2014-11-12 20:21 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 19/21] qcow2: Invoke refcount order amendment function Max Reitz
2014-11-12 4:36 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 20/21] qcow2: Point to amend function in check Max Reitz
2014-11-12 4:38 ` Eric Blake
2014-11-10 13:45 ` [Qemu-devel] [PATCH 21/21] iotests: Add test for different refcount widths Max Reitz
2014-11-11 19:53 ` Eric Blake
2014-11-12 8:58 ` Max Reitz
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=54631FC6.40703@redhat.com \
--to=mreitz@redhat.com \
--cc=eblake@redhat.com \
--cc=kwolf@redhat.com \
--cc=pl@kamp.de \
--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).