From: Max Reitz <mreitz@redhat.com>
To: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
qemu-devel@nongnu.org, qemu-block@nongnu.org
Cc: kwolf@redhat.com, den@openvz.org, armbru@redhat.com, stefanha@redhat.com
Subject: Re: [Qemu-devel] [PATCH v2 1/4] block: introduce aio task pool
Date: Tue, 13 Aug 2019 22:47:29 +0200 [thread overview]
Message-ID: <565ee3a9-d606-cc64-3748-21f9c54cffb7@redhat.com> (raw)
In-Reply-To: <20190730141826.709849-2-vsementsov@virtuozzo.com>
[-- Attachment #1.1: Type: text/plain, Size: 2711 bytes --]
On 30.07.19 16:18, Vladimir Sementsov-Ogievskiy wrote:
> Common interface for aio task loops. To be used for improving
> performance of synchronous io loops in qcow2, block-stream,
> copy-on-read, and may be other places.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> ---
Looks good to me overall.
> block/aio_task.h | 52 +++++++++++++++++++
I’ve move this to include/block/.
> block/aio_task.c | 119 ++++++++++++++++++++++++++++++++++++++++++++
> block/Makefile.objs | 2 +
> 3 files changed, 173 insertions(+)
> create mode 100644 block/aio_task.h
> create mode 100644 block/aio_task.c
>
> diff --git a/block/aio_task.h b/block/aio_task.h
> new file mode 100644
> index 0000000000..933af1d8e7
> --- /dev/null
> +++ b/block/aio_task.h
[...]
> +typedef struct AioTaskPool AioTaskPool;
> +typedef struct AioTask AioTask;
> +typedef int (*AioTaskFunc)(AioTask *task);
+coroutine_fn
> +struct AioTask {
> + AioTaskPool *pool;
> + AioTaskFunc func;
> + int ret;
> +};
> +
> +/*
> + * aio_task_pool_new
> + *
> + * The caller is responsible to g_free AioTaskPool pointer after use.
s/to g_free/for g_freeing/ or something similar.
Or you’d just add aio_task_pool_free().
> + */
> +AioTaskPool *aio_task_pool_new(int max_busy_tasks);
> +int aio_task_pool_status(AioTaskPool *pool);
A comment wouldn’t hurt. It wasn’t immediately clear to me that status
refers to the error code of a failing task (or 0), although it wasn’t
too much of a surprise either.
> +bool aio_task_pool_empty(AioTaskPool *pool);
> +void aio_task_pool_start_task(AioTaskPool *pool, AioTask *task);
Maybe make a note that task->pool will be set automatically?
> +void aio_task_pool_wait_slot(AioTaskPool *pool);
> +void aio_task_pool_wait_one(AioTaskPool *pool);
> +void aio_task_pool_wait_all(AioTaskPool *pool);
Shouldn’t all of these but aio_task_pool_empty() and
aio_task_pool_status() be coroutine_fns?
> +#endif /* BLOCK_AIO_TASK_H */
> diff --git a/block/aio_task.c b/block/aio_task.c
> new file mode 100644
> index 0000000000..807be8deb5
> --- /dev/null
> +++ b/block/aio_task.c
[...]
> +static void aio_task_co(void *opaque)
+coroutine_fn
[...]
> +void aio_task_pool_wait_one(AioTaskPool *pool)
> +{
> + assert(pool->busy_tasks > 0);
> + assert(qemu_coroutine_self() == pool->main_co);
> +
> + pool->wait_done = true;
Hmmm, but the wait actually isn’t done yet. :-)
Maybe s/wait_done/waiting/?
Max
> + qemu_coroutine_yield();
> +
> + assert(!pool->wait_done);
> + assert(pool->busy_tasks < pool->max_busy_tasks);
> +}
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2019-08-13 20:48 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-30 14:18 [Qemu-devel] [PATCH v2 0/4] qcow2: async handling of fragmented io Vladimir Sementsov-Ogievskiy
2019-07-30 14:18 ` [Qemu-devel] [PATCH v2 1/4] block: introduce aio task pool Vladimir Sementsov-Ogievskiy
2019-08-13 20:47 ` Max Reitz [this message]
2019-08-14 8:18 ` Vladimir Sementsov-Ogievskiy
2019-07-30 14:18 ` [Qemu-devel] [PATCH v2 2/4] block/qcow2: refactor qcow2_co_preadv_part Vladimir Sementsov-Ogievskiy
2019-08-13 21:31 ` Max Reitz
2019-08-14 9:11 ` Vladimir Sementsov-Ogievskiy
2019-08-14 15:03 ` Max Reitz
2019-08-14 15:15 ` Eric Blake
2019-08-14 15:58 ` Max Reitz
2019-07-30 14:18 ` [Qemu-devel] [PATCH v2 3/4] block/qcow2: refactor qcow2_co_pwritev_part Vladimir Sementsov-Ogievskiy
2019-08-14 15:55 ` Max Reitz
2019-08-14 16:23 ` Max Reitz
2019-07-30 14:18 ` [Qemu-devel] [PATCH v2 4/4] block/qcow2: introduce parallel subrequest handling in read and write Vladimir Sementsov-Ogievskiy
2019-08-14 16:24 ` 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=565ee3a9-d606-cc64-3748-21f9c54cffb7@redhat.com \
--to=mreitz@redhat.com \
--cc=armbru@redhat.com \
--cc=den@openvz.org \
--cc=kwolf@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).