From: Roman Kagan <rvkagan@yandex-team.ru>
To: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Cc: kwolf@redhat.com, Fam Zheng <fam@euphon.net>,
qemu-block@nongnu.org, qemu-devel@nongnu.org, mreitz@redhat.com,
Stefan Hajnoczi <stefanha@redhat.com>,
den@openvz.org
Subject: Re: [PATCH v3 06/33] util/async: aio_co_schedule(): support reschedule in same ctx
Date: Fri, 23 Apr 2021 13:09:58 +0300 [thread overview]
Message-ID: <YIKc9pCDmN2FRDV4@rvkaganb> (raw)
In-Reply-To: <20210416080911.83197-7-vsementsov@virtuozzo.com>
On Fri, Apr 16, 2021 at 11:08:44AM +0300, Vladimir Sementsov-Ogievskiy wrote:
> With the following patch we want to call wake coroutine from thread.
> And it doesn't work with aio_co_wake:
> Assume we have no iothreads.
> Assume we have a coroutine A, which waits in the yield point for
> external aio_co_wake(), and no progress can be done until it happen.
> Main thread is in blocking aio_poll() (for example, in blk_read()).
>
> Now, in a separate thread we do aio_co_wake(). It calls aio_co_enter(),
> which goes through last "else" branch and do aio_context_acquire(ctx).
>
> Now we have a deadlock, as aio_poll() will not release the context lock
> until some progress is done, and progress can't be done until
> aio_co_wake() wake the coroutine A. And it can't because it wait for
> aio_context_acquire().
>
> Still, aio_co_schedule() works well in parallel with blocking
> aio_poll(). So we want use it. Let's add a possibility of rescheduling
> coroutine in same ctx where it was yield'ed.
>
> Fetch co->ctx in same way as it is done in aio_co_wake().
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> ---
> include/block/aio.h | 2 +-
> util/async.c | 8 ++++++++
> 2 files changed, 9 insertions(+), 1 deletion(-)
>
> diff --git a/include/block/aio.h b/include/block/aio.h
> index 5f342267d5..744b695525 100644
> --- a/include/block/aio.h
> +++ b/include/block/aio.h
> @@ -643,7 +643,7 @@ static inline bool aio_node_check(AioContext *ctx, bool is_external)
>
> /**
> * aio_co_schedule:
> - * @ctx: the aio context
> + * @ctx: the aio context, if NULL, the current ctx of @co will be used.
> * @co: the coroutine
> *
> * Start a coroutine on a remote AioContext.
> diff --git a/util/async.c b/util/async.c
> index 674dbefb7c..750be555c6 100644
> --- a/util/async.c
> +++ b/util/async.c
> @@ -545,6 +545,14 @@ fail:
>
> void aio_co_schedule(AioContext *ctx, Coroutine *co)
> {
> + if (!ctx) {
> + /*
> + * Read coroutine before co->ctx. Matches smp_wmb in
> + * qemu_coroutine_enter.
> + */
> + smp_read_barrier_depends();
> + ctx = qatomic_read(&co->ctx);
> + }
I'd rather not extend this interface, but add a new one on top. And
document how it's different from aio_co_wake().
Roman.
> trace_aio_co_schedule(ctx, co);
> const char *scheduled = qatomic_cmpxchg(&co->scheduled, NULL,
> __func__);
> --
> 2.29.2
>
next prev parent reply other threads:[~2021-04-23 10:12 UTC|newest]
Thread overview: 121+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-16 8:08 [PATCH v3 00/33] block/nbd: rework client connection Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 01/33] block/nbd: fix channel object leak Vladimir Sementsov-Ogievskiy
2021-05-24 21:31 ` Eric Blake
2021-05-25 4:47 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 02/33] block/nbd: fix how state is cleared on nbd_open() failure paths Vladimir Sementsov-Ogievskiy
2021-04-21 14:00 ` Roman Kagan
2021-04-21 22:27 ` Vladimir Sementsov-Ogievskiy
2021-04-22 8:49 ` Roman Kagan
2021-06-01 21:39 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 03/33] block/nbd: ensure ->connection_thread is always valid Vladimir Sementsov-Ogievskiy
2021-06-01 21:41 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 04/33] block/nbd: nbd_client_handshake(): fix leak of s->ioc Vladimir Sementsov-Ogievskiy
2021-04-22 8:59 ` Roman Kagan
2021-04-16 8:08 ` [PATCH v3 05/33] block/nbd: BDRVNBDState: drop unused connect_err and connect_status Vladimir Sementsov-Ogievskiy
2021-06-01 21:43 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 06/33] util/async: aio_co_schedule(): support reschedule in same ctx Vladimir Sementsov-Ogievskiy
2021-04-23 10:09 ` Roman Kagan [this message]
2021-04-26 8:52 ` Vladimir Sementsov-Ogievskiy
2021-05-12 6:56 ` Paolo Bonzini
2021-05-12 7:15 ` Vladimir Sementsov-Ogievskiy
2021-05-13 21:04 ` Paolo Bonzini
2021-05-14 17:27 ` Roman Kagan
2021-05-14 21:19 ` Paolo Bonzini
2021-06-08 18:45 ` Vladimir Sementsov-Ogievskiy
2021-06-09 9:35 ` Paolo Bonzini
2021-06-09 10:24 ` Vladimir Sementsov-Ogievskiy
2021-06-09 12:17 ` Paolo Bonzini
2021-04-16 8:08 ` [PATCH v3 07/33] block/nbd: simplify waking of nbd_co_establish_connection() Vladimir Sementsov-Ogievskiy
2021-04-27 21:44 ` Roman Kagan
2021-06-02 19:05 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 08/33] block/nbd: drop thr->state Vladimir Sementsov-Ogievskiy
2021-04-27 22:23 ` Roman Kagan
2021-04-28 8:01 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 09/33] block/nbd: bs-independent interface for nbd_co_establish_connection() Vladimir Sementsov-Ogievskiy
2021-06-02 19:14 ` Eric Blake
2021-06-08 10:12 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 10/33] block/nbd: make nbd_co_establish_connection_cancel() bs-independent Vladimir Sementsov-Ogievskiy
2021-06-02 21:18 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 11/33] block/nbd: rename NBDConnectThread to NBDClientConnection Vladimir Sementsov-Ogievskiy
2021-04-27 22:28 ` Roman Kagan
2021-06-02 21:21 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 12/33] block/nbd: introduce nbd_client_connection_new() Vladimir Sementsov-Ogievskiy
2021-06-02 21:22 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 13/33] block/nbd: introduce nbd_client_connection_release() Vladimir Sementsov-Ogievskiy
2021-04-27 22:35 ` Roman Kagan
2021-04-28 8:06 ` Vladimir Sementsov-Ogievskiy
2021-06-02 21:27 ` Eric Blake
2021-06-03 11:59 ` Vladimir Sementsov-Ogievskiy
2021-06-08 10:00 ` Vladimir Sementsov-Ogievskiy
2021-06-08 14:18 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 14/33] nbd: move connection code from block/nbd to nbd/client-connection Vladimir Sementsov-Ogievskiy
2021-04-27 22:45 ` Roman Kagan
2021-04-28 8:14 ` Vladimir Sementsov-Ogievskiy
2021-06-09 15:49 ` Vladimir Sementsov-Ogievskiy
2021-06-03 15:55 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 15/33] nbd/client-connection: use QEMU_LOCK_GUARD Vladimir Sementsov-Ogievskiy
2021-04-28 6:08 ` Roman Kagan
2021-04-28 8:17 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 16/33] nbd/client-connection: add possibility of negotiation Vladimir Sementsov-Ogievskiy
2021-05-11 10:45 ` Roman Kagan
2021-05-12 6:42 ` Vladimir Sementsov-Ogievskiy
2021-06-08 19:23 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 17/33] nbd/client-connection: implement connection retry Vladimir Sementsov-Ogievskiy
2021-05-11 20:54 ` Roman Kagan
2021-06-08 10:24 ` Vladimir Sementsov-Ogievskiy
2021-06-03 16:17 ` Eric Blake
2021-06-03 17:49 ` Vladimir Sementsov-Ogievskiy
2021-06-08 10:38 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:08 ` [PATCH v3 18/33] nbd/client-connection: shutdown connection on release Vladimir Sementsov-Ogievskiy
2021-05-11 21:08 ` Roman Kagan
2021-05-12 6:39 ` Vladimir Sementsov-Ogievskiy
2021-06-03 16:20 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 19/33] block/nbd: split nbd_handle_updated_info out of nbd_client_handshake() Vladimir Sementsov-Ogievskiy
2021-05-12 8:40 ` Roman Kagan
2021-06-03 16:29 ` Eric Blake
2021-06-09 17:23 ` Vladimir Sementsov-Ogievskiy
2021-06-09 18:28 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 20/33] block/nbd: use negotiation of NBDClientConnection Vladimir Sementsov-Ogievskiy
2021-06-03 18:05 ` Eric Blake
2021-04-16 8:08 ` [PATCH v3 21/33] qemu-socket: pass monitor link to socket_get_fd directly Vladimir Sementsov-Ogievskiy
2021-04-19 9:34 ` Daniel P. Berrangé
2021-04-19 10:09 ` Vladimir Sementsov-Ogievskiy
2021-05-12 9:40 ` Roman Kagan
2021-05-12 9:59 ` Daniel P. Berrangé
2021-05-13 11:02 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 22/33] block/nbd: pass monitor directly to connection thread Vladimir Sementsov-Ogievskiy
2021-06-03 18:16 ` Eric Blake
2021-06-03 18:31 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 23/33] block/nbd: nbd_teardown_connection() don't touch s->sioc Vladimir Sementsov-Ogievskiy
2021-06-03 19:04 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 24/33] block/nbd: drop BDRVNBDState::sioc Vladimir Sementsov-Ogievskiy
2021-06-03 19:12 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 25/33] nbd/client-connection: return only one io channel Vladimir Sementsov-Ogievskiy
2021-06-03 19:58 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 26/33] block-coroutine-wrapper: allow non bdrv_ prefix Vladimir Sementsov-Ogievskiy
2021-06-03 20:00 ` Eric Blake
2021-06-03 20:53 ` Eric Blake
2021-06-04 5:29 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 27/33] block/nbd: split nbd_co_do_establish_connection out of nbd_reconnect_attempt Vladimir Sementsov-Ogievskiy
2021-06-03 20:04 ` Eric Blake
2021-06-04 5:30 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 28/33] nbd/client-connection: do qio_channel_set_delay(false) Vladimir Sementsov-Ogievskiy
2021-06-03 20:48 ` Eric Blake
2021-06-04 5:32 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 29/33] nbd/client-connection: add option for non-blocking connection attempt Vladimir Sementsov-Ogievskiy
2021-06-03 20:51 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 30/33] block/nbd: reuse nbd_co_do_establish_connection() in nbd_open() Vladimir Sementsov-Ogievskiy
2021-06-03 20:57 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 31/33] block/nbd: add nbd_clinent_connected() helper Vladimir Sementsov-Ogievskiy
2021-05-12 7:06 ` Paolo Bonzini
2021-05-12 7:19 ` Vladimir Sementsov-Ogievskiy
2021-06-03 21:08 ` Eric Blake
2021-04-16 8:09 ` [PATCH v3 32/33] block/nbd: safer transition to receiving request Vladimir Sementsov-Ogievskiy
2021-06-03 21:11 ` Eric Blake
2021-06-04 5:36 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:09 ` [PATCH v3 33/33] block/nbd: drop connection_co Vladimir Sementsov-Ogievskiy
2021-04-16 8:14 ` Vladimir Sementsov-Ogievskiy
2021-04-16 8:21 ` Vladimir Sementsov-Ogievskiy
2021-06-03 21:27 ` Eric Blake
2021-06-04 5:39 ` Vladimir Sementsov-Ogievskiy
2021-05-12 6:54 ` [PATCH v3 00/33] block/nbd: rework client connection Paolo Bonzini
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=YIKc9pCDmN2FRDV4@rvkaganb \
--to=rvkagan@yandex-team.ru \
--cc=den@openvz.org \
--cc=fam@euphon.net \
--cc=kwolf@redhat.com \
--cc=mreitz@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@redhat.com \
--cc=vsementsov@virtuozzo.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).