All of lore.kernel.org
 help / color / mirror / Atom feed
From: Hanna Czenczek <hreitz@redhat.com>
To: Fabiano Rosas <farosas@suse.de>, qemu-devel@nongnu.org
Cc: qemu-block@nongnu.org, "Kevin Wolf" <kwolf@redhat.com>,
	"Markus Armbruster" <armbru@redhat.com>,
	"João Silva" <jsilva@suse.de>, "Lin Ma" <lma@suse.com>,
	"Claudio Fontana" <cfontana@suse.de>,
	"Dario Faggioli" <dfaggioli@suse.com>,
	"Eric Blake" <eblake@redhat.com>,
	"Dr. David Alan Gilbert" <dave@treblig.org>
Subject: Re: [PATCH v2 09/10] block: Convert qmp_query_block() to coroutine_fn
Date: Mon, 6 Nov 2023 16:02:52 +0100	[thread overview]
Message-ID: <b1f840cc-82c7-4ff5-8ab8-38382e425181@redhat.com> (raw)
In-Reply-To: <20230609201910.12100-10-farosas@suse.de>

On 09.06.23 22:19, Fabiano Rosas wrote:
> This is another caller of bdrv_get_allocated_file_size() that needs to
> be converted to a coroutine because that function will be made
> asynchronous when called (indirectly) from the QMP dispatcher.
>
> This QMP command is a candidate because it calls bdrv_do_query_node_info(),
> which in turn calls bdrv_get_allocated_file_size().
>
> We've determined bdrv_do_query_node_info() to be coroutine-safe (see
> previous commits), so we can just put this QMP command in a coroutine.
>
> Since qmp_query_block() now expects to run in a coroutine, its callers
> need to be converted as well. Convert hmp_info_block(), which calls
> only coroutine-safe code, including qmp_query_named_block_nodes()
> which has been converted to coroutine in the previous patches.
>
> Now that all callers of bdrv_[co_]block_device_info() are using the
> coroutine version, a few things happen:
>
>   - we can return to using bdrv_block_device_info() without a wrapper;
>
>   - bdrv_get_allocated_file_size() can stop being mixed;
>
>   - bdrv_co_get_allocated_file_size() needs to be put under the graph
>     lock because it is being called wthout the wrapper;
>
>   - bdrv_do_query_node_info() doesn't need to acquire the AioContext
>     because it doesn't call aio_poll anymore;
>
> Signed-off-by: Fabiano Rosas <farosas@suse.de>
> ---
>   block.c                        |  2 +-
>   block/monitor/block-hmp-cmds.c |  2 +-
>   block/qapi.c                   | 18 +++++++++---------
>   hmp-commands-info.hx           |  1 +
>   include/block/block-hmp-cmds.h |  2 +-
>   include/block/block-io.h       |  2 +-
>   include/block/qapi.h           | 12 ++++--------
>   qapi/block-core.json           |  2 +-
>   8 files changed, 19 insertions(+), 22 deletions(-)

After this series has been sent, we got some usages of 
GRAPH_RDLOCK_GUARD_MAINLOOP() that no longer fit with this patch – I’ve 
also mentioned one case on patch 7, not yet realizing that this was a 
new thing.  Those must now be fixed (e.g. in qmp_query_block(), or in 
bdrv_snapshot_list()), or they’ll crash.

Hanna



  parent reply	other threads:[~2023-11-06 15:04 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-09 20:19 [PATCH v2 00/10] block: Make raw_co_get_allocated_file_size asynchronous Fabiano Rosas
2023-06-09 20:19 ` [PATCH v2 01/10] block: Remove bdrv_query_block_node_info Fabiano Rosas
2023-06-09 20:19 ` [PATCH v2 02/10] block: Remove unnecessary variable in bdrv_block_device_info Fabiano Rosas
2023-07-03 14:04   ` Philippe Mathieu-Daudé
2023-06-09 20:19 ` [PATCH v2 03/10] block: Allow the wrapper script to see functions declared in qapi.h Fabiano Rosas
2023-11-06 14:51   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 04/10] block: Temporarily mark bdrv_co_get_allocated_file_size as mixed Fabiano Rosas
2023-11-06 14:51   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 05/10] block: Convert bdrv_query_block_graph_info to coroutine Fabiano Rosas
2023-11-06 14:52   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 06/10] block: Convert bdrv_block_device_info into co_wrapper Fabiano Rosas
2023-11-06 12:51   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 07/10] block: Convert qmp_query_named_block_nodes to coroutine Fabiano Rosas
2023-11-06 12:51   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 08/10] block: Don't query all block devices at hmp_nbd_server_start Fabiano Rosas
2023-11-06 13:07   ` Hanna Czenczek
2023-06-09 20:19 ` [PATCH v2 09/10] block: Convert qmp_query_block() to coroutine_fn Fabiano Rosas
2023-11-06 14:40   ` Hanna Czenczek
2023-11-06 15:02   ` Hanna Czenczek [this message]
2023-11-29 20:19     ` Fabiano Rosas
2023-06-09 20:19 ` [PATCH v2 10/10] block: Add a thread-pool version of fstat Fabiano Rosas
2023-11-06 14:52   ` Hanna Czenczek
2023-07-03 13:55 ` [PATCH v2 00/10] block: Make raw_co_get_allocated_file_size asynchronous Fabiano Rosas
2023-08-22 11:52 ` Claudio Fontana

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=b1f840cc-82c7-4ff5-8ab8-38382e425181@redhat.com \
    --to=hreitz@redhat.com \
    --cc=armbru@redhat.com \
    --cc=cfontana@suse.de \
    --cc=dave@treblig.org \
    --cc=dfaggioli@suse.com \
    --cc=eblake@redhat.com \
    --cc=farosas@suse.de \
    --cc=jsilva@suse.de \
    --cc=kwolf@redhat.com \
    --cc=lma@suse.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.