public inbox for [email protected]
 help / color / mirror / Atom feed
From: Pavel Begunkov <[email protected]>
To: Bernd Schubert <[email protected]>, Miklos Szeredi <[email protected]>
Cc: Jens Axboe <[email protected]>,
	[email protected], [email protected],
	Joanne Koong <[email protected]>,
	Josef Bacik <[email protected]>,
	Amir Goldstein <[email protected]>,
	Ming Lei <[email protected]>, David Wei <[email protected]>,
	[email protected]
Subject: Re: [PATCH RFC v7 09/16] fuse: {uring} Add uring sqe commit and fetch support
Date: Tue, 3 Dec 2024 13:47:45 +0000	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>

On 11/27/24 13:40, Bernd Schubert wrote:
> This adds support for fuse request completion through ring SQEs
> (FUSE_URING_REQ_COMMIT_AND_FETCH handling). After committing
> the ring entry it becomes available for new fuse requests.
> Handling of requests through the ring (SQE/CQE handling)
> is complete now.
> 
> Fuse request data are copied through the mmaped ring buffer,
> there is no support for any zero copy yet.
> 
> Signed-off-by: Bernd Schubert <[email protected]>
> ---
...
> diff --git a/fs/fuse/dev_uring.c b/fs/fuse/dev_uring.c
> index af9c5f116ba1dcf6c01d0359d1a06491c92c32f9..7bb07f5ba436fcb89537f0821f08a7167da52902 100644
> --- a/fs/fuse/dev_uring.c
> +++ b/fs/fuse/dev_uring.c
> @@ -24,6 +24,19 @@ bool fuse_uring_enabled(void)
...
> +/*
> + * Write data to the ring buffer and send the request to userspace,
> + * userspace will read it
> + * This is comparable with classical read(/dev/fuse)
> + */
> +static int fuse_uring_send_next_to_ring(struct fuse_ring_ent *ring_ent)
> +{
> +	int err = 0;
> +	struct fuse_ring_queue *queue = ring_ent->queue;
> +
> +	err = fuse_uring_prepare_send(ring_ent);
> +	if (err)
> +		goto err;
> +
> +	spin_lock(&queue->lock);
> +	ring_ent->state = FRRS_USERSPACE;
> +	list_move(&ring_ent->list, &queue->ent_in_userspace);
> +	spin_unlock(&queue->lock);
> +
> +	io_uring_cmd_complete_in_task(ring_ent->cmd,
> +				      fuse_uring_async_send_to_ring);

Just io_uring_cmd_done should be enough.

io_uring_cmd_done(cmd, 0, 0, issue_flags);


> +	return 0;
> +
> +err:
> +	return err;
> +}
> +
>   /*
>    * Make a ring entry available for fuse_req assignment
>    */
> @@ -148,6 +385,189 @@ static void fuse_uring_ent_avail(struct fuse_ring_ent *ring_ent,
>   	ring_ent->state = FRRS_WAIT;
>   }
>   
...
> +
> +/* FUSE_URING_REQ_COMMIT_AND_FETCH handler */
> +static int fuse_uring_commit_fetch(struct io_uring_cmd *cmd, int issue_flags,
> +				   struct fuse_conn *fc)
> +{
> +	const struct fuse_uring_cmd_req *cmd_req = io_uring_sqe_cmd(cmd->sqe);

Same comment about checking SQE128, fuse_uring_cmd() sounds like a good
place for it.


> +	struct fuse_ring_ent *ring_ent;
> +	int err;
> +	struct fuse_ring *ring = fc->ring;
> +	struct fuse_ring_queue *queue;
> +	uint64_t commit_id = cmd_req->commit_id;
> +	struct fuse_pqueue fpq;
> +	struct fuse_req *req;
> +
> +	err = -ENOTCONN;
> +	if (!ring)
> +		return err;
> +
> +	queue = ring->queues[cmd_req->qid];

READ_ONCE()

> +	if (!queue)
> +		return err;
> +	fpq = queue->fpq;
> +

-- 
Pavel Begunkov


  reply	other threads:[~2024-12-03 13:46 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-11-27 13:40 [PATCH RFC v7 00/16] fuse: fuse-over-io-uring Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 01/16] fuse: rename to fuse_dev_end_requests and make non-static Bernd Schubert
2024-11-28  0:19   ` Joanne Koong
2024-11-27 13:40 ` [PATCH RFC v7 02/16] fuse: Move fuse_get_dev to header file Bernd Schubert
2024-11-28  0:20   ` Joanne Koong
2024-11-27 13:40 ` [PATCH RFC v7 03/16] fuse: Move request bits Bernd Schubert
2024-11-28  0:21   ` Joanne Koong
2024-11-27 13:40 ` [PATCH RFC v7 04/16] fuse: Add fuse-io-uring design documentation Bernd Schubert
2024-12-03 12:30   ` Pavel Begunkov
2024-11-27 13:40 ` [PATCH RFC v7 05/16] fuse: make args->in_args[0] to be always the header Bernd Schubert
2024-11-28  0:27   ` Joanne Koong
2024-11-27 13:40 ` [PATCH RFC v7 06/16] fuse: {uring} Handle SQEs - register commands Bernd Schubert
2024-11-28  2:23   ` Joanne Koong
2024-11-28 18:20     ` Bernd Schubert
2024-12-03 13:24   ` Pavel Begunkov
2024-12-03 13:49     ` Bernd Schubert
2024-12-03 14:16       ` Pavel Begunkov
2024-12-03 13:38   ` Pavel Begunkov
2024-11-27 13:40 ` [PATCH RFC v7 07/16] fuse: Make fuse_copy non static Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 08/16] fuse: Add fuse-io-uring handling into fuse_copy Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 09/16] fuse: {uring} Add uring sqe commit and fetch support Bernd Schubert
2024-12-03 13:47   ` Pavel Begunkov [this message]
2024-11-27 13:40 ` [PATCH RFC v7 10/16] fuse: {uring} Handle teardown of ring entries Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 11/16] fuse: {uring} Allow to queue fg requests through io-uring Bernd Schubert
2024-12-03 14:09   ` Pavel Begunkov
2024-12-03 22:46     ` Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 12/16] fuse: {uring} Allow to queue bg " Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 13/16] io_uring/cmd: let cmds to know about dying task Bernd Schubert
2024-12-03 12:15   ` Pavel Begunkov
2024-12-03 12:15     ` Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 14/16] fuse: {uring} Handle IO_URING_F_TASK_DEAD Bernd Schubert
2024-12-03 12:20   ` Pavel Begunkov
2024-11-27 13:40 ` [PATCH RFC v7 15/16] fuse: {io-uring} Prevent mount point hang on fuse-server termination Bernd Schubert
2024-11-27 13:40 ` [PATCH RFC v7 16/16] fuse: enable fuse-over-io-uring Bernd Schubert
2024-11-27 13:45 ` [PATCH RFC v7 00/16] fuse: fuse-over-io-uring Bernd Schubert
2024-12-03 14:24 ` Pavel Begunkov
2024-12-03 14:32   ` Bernd Schubert

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 \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    /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