From: Jean-Louis Dupond <jean-louis@dupond.be>
To: Hanna Czenczek <hreitz@redhat.com>, qemu-devel@nongnu.org
Cc: Eric Blake <eblake@redhat.com>,
Markus Armbruster <armbru@redhat.com>,
Kevin Wolf <kwolf@redhat.com>,
qemu-block@nongnu.org
Subject: Re: [PATCH 1/3] block: add for_commit option to measure
Date: Tue, 13 May 2025 16:10:33 +0200 [thread overview]
Message-ID: <bd78687d-2be5-4ad1-9493-2513af223f6d@dupond.be> (raw)
In-Reply-To: <a1aa1173-5362-44f6-8b64-6d44f1b5eb82@redhat.com>
On 5/12/25 13:16, Hanna Czenczek wrote:
> On 16.04.25 09:16, Jean-Louis Dupond wrote:
>> To specify we use measure call for commit size calculations, we add a
>> new 'for_commit' option to the measure call.
>> This will be used in following commit to do a different measurement.
>
> Why not allow specifying the node name (or filename) of the commit
> target instead of this just being a boolean?
Honestly didn't even thought about that option :)
>
> (That was my main problem with the original series, that it wasn’t
> possible to specify the commit target (in an obvious way at least).
> I’m not a fan of deferring to JSON parameters for this, although I’m
> aware that to specify the node name of the commit target you would
> have to use JSON again, unless qemu-img can do some translation on
> behalf of the user: Looking at `commit -b`, that takes a filename; it
> could make sense to have the internal QAPI parameter use a node name,
> and qemu-img translating the filename parameter into a node name.
> Actually, `measure` could just accept the same `-b` option as
> `commit`, which it would then translate into the QAPI
> `for-commit=<node-name>` option.)
And you will need to use JSON parameters if you want to specify
something special, like for example to note that the commit target has
discard-no-unref enabled.
I think not many people will call the command with the for_commit
option, but it will be (always) called from code where the JSON
parameters give much more flexibility to add some additional options
(like discard-no-unref).
We could of course support both (JSON + for-commit=<node-name>).
But then we will also need to validate somehow that the for-commit
node-name is equal to the target specified in the JSON somehow.
Or have something like for-commit=true/false + json or
for-commit-node=<node-name> if you don't use a JSON?
Do you think it's worth the effort ? :)
>
> Hanna
>
>
Thanks for the review!
Jean-Louis
>
>> Signed-off-by: Jean-Louis Dupond <jean-louis@dupond.be>
>> ---
>> block/qcow2.c | 16 +++++++++++++
>> include/block/block_int-common.h | 4 ++++
>> qapi/block-core.json | 28 ++++++++++++++++++++++
>> qemu-img.c | 40 ++++++++++++++++++++++++++++----
>> 4 files changed, 83 insertions(+), 5 deletions(-)
>
next prev parent reply other threads:[~2025-05-13 14:11 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-16 7:16 [PATCH 0/3] Add a for_commit option to qemu-img measure Jean-Louis Dupond
2025-04-16 7:16 ` [PATCH 1/3] block: add for_commit option to measure Jean-Louis Dupond
2025-05-12 11:16 ` Hanna Czenczek
2025-05-13 14:10 ` Jean-Louis Dupond [this message]
2025-08-29 9:44 ` Jean-Louis Dupond
2025-04-16 7:16 ` [PATCH 2/3] qcow2: make measure for_commit aware Jean-Louis Dupond
2025-04-16 7:16 ` [PATCH 3/3] iotests/290: add test case for for_commit measure Jean-Louis Dupond
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=bd78687d-2be5-4ad1-9493-2513af223f6d@dupond.be \
--to=jean-louis@dupond.be \
--cc=armbru@redhat.com \
--cc=eblake@redhat.com \
--cc=hreitz@redhat.com \
--cc=kwolf@redhat.com \
--cc=qemu-block@nongnu.org \
--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).