All of lore.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: Kanchan Joshi <joshiiitr@gmail.com>
Cc: Jens Axboe <axboe@kernel.dk>, Kanchan Joshi <joshi.k@samsung.com>,
	Christoph Hellwig <hch@lst.de>, Keith Busch <kbusch@kernel.org>,
	Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>,
	io-uring@vger.kernel.org, linux-nvme@lists.infradead.org,
	anuj20.g@samsung.com, Javier Gonzalez <javier.gonz@samsung.com>,
	Nitesh Shetty <nj.shetty@samsung.com>,
	Selvakumar S <selvakuma.s1@samsung.com>
Subject: Re: [RFC PATCH v3 1/3] io_uring: add helper for uring_cmd completion in submitter-task
Date: Thu, 18 Mar 2021 06:48:07 +0100	[thread overview]
Message-ID: <20210318054807.GA28576@lst.de> (raw)
In-Reply-To: <CA+1E3r+Mt7KKeFeYf7WY3CoKwnkXT-jE2EgJSTE6zaAfJX0dzQ@mail.gmail.com>

On Thu, Mar 18, 2021 at 10:55:55AM +0530, Kanchan Joshi wrote:
> I started with that, but the problem was implementing the driver callback .
> The callbacks receive only one argument which is "struct callback_head
> *", and the driver needs to extract "io_uring_cmd *" out of it.
> This part -
> +static void uring_cmd_work(struct callback_head *cb)
> +{
> +     struct io_kiocb *req = container_of(cb, struct io_kiocb, task_work);
> +     struct io_uring_cmd *cmd = &req->uring_cmd;
> 
> If the callback has to move to the driver (nvme), the driver needs
> visibility to "struct io_kiocb" which is uring-local.
> Do you see a better way to handle this?

Can't you just add a helper in io_uring.c that does something like this:

struct io_uring_cmd *callback_to_io_uring_cmd(struct callback_head *cb)
{
	return &container_of(cb, struct io_kiocb, task_work)->uring_cmd;
}
EXPORT_SYMBOL_GPL(callback_to_io_uring_cmd);

WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@lst.de>
To: Kanchan Joshi <joshiiitr@gmail.com>
Cc: Jens Axboe <axboe@kernel.dk>, Kanchan Joshi <joshi.k@samsung.com>,
	Christoph Hellwig <hch@lst.de>, Keith Busch <kbusch@kernel.org>,
	Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>,
	io-uring@vger.kernel.org, linux-nvme@lists.infradead.org,
	anuj20.g@samsung.com, Javier Gonzalez <javier.gonz@samsung.com>,
	Nitesh Shetty <nj.shetty@samsung.com>,
	Selvakumar S <selvakuma.s1@samsung.com>
Subject: Re: [RFC PATCH v3 1/3] io_uring: add helper for uring_cmd completion in submitter-task
Date: Thu, 18 Mar 2021 06:48:07 +0100	[thread overview]
Message-ID: <20210318054807.GA28576@lst.de> (raw)
In-Reply-To: <CA+1E3r+Mt7KKeFeYf7WY3CoKwnkXT-jE2EgJSTE6zaAfJX0dzQ@mail.gmail.com>

On Thu, Mar 18, 2021 at 10:55:55AM +0530, Kanchan Joshi wrote:
> I started with that, but the problem was implementing the driver callback .
> The callbacks receive only one argument which is "struct callback_head
> *", and the driver needs to extract "io_uring_cmd *" out of it.
> This part -
> +static void uring_cmd_work(struct callback_head *cb)
> +{
> +     struct io_kiocb *req = container_of(cb, struct io_kiocb, task_work);
> +     struct io_uring_cmd *cmd = &req->uring_cmd;
> 
> If the callback has to move to the driver (nvme), the driver needs
> visibility to "struct io_kiocb" which is uring-local.
> Do you see a better way to handle this?

Can't you just add a helper in io_uring.c that does something like this:

struct io_uring_cmd *callback_to_io_uring_cmd(struct callback_head *cb)
{
	return &container_of(cb, struct io_kiocb, task_work)->uring_cmd;
}
EXPORT_SYMBOL_GPL(callback_to_io_uring_cmd);

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

  reply	other threads:[~2021-03-18  5:49 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20210316140229epcas5p23d68a4c9694bbf7759b5901115a4309b@epcas5p2.samsung.com>
2021-03-16 14:01 ` [RFC PATCH v3 0/3] Async nvme passthrough over io_uring Kanchan Joshi
2021-03-16 14:01   ` Kanchan Joshi
     [not found]   ` <CGME20210316140233epcas5p372405e7cb302c61dba5e1094fa796513@epcas5p3.samsung.com>
2021-03-16 14:01     ` [RFC PATCH v3 1/3] io_uring: add helper for uring_cmd completion in submitter-task Kanchan Joshi
2021-03-16 14:01       ` Kanchan Joshi
2021-03-16 15:43       ` Stefan Metzmacher
2021-03-16 15:43         ` Stefan Metzmacher
2021-03-18  1:57       ` Jens Axboe
2021-03-18  1:57         ` Jens Axboe
2021-03-18  5:25         ` Kanchan Joshi
2021-03-18  5:25           ` Kanchan Joshi
2021-03-18  5:48           ` Christoph Hellwig [this message]
2021-03-18  5:48             ` Christoph Hellwig
2021-03-18  6:14             ` Kanchan Joshi
2021-03-18  6:14               ` Kanchan Joshi
     [not found]   ` <CGME20210316140236epcas5p4de087ee51a862402146fbbc621d4d4c6@epcas5p4.samsung.com>
2021-03-16 14:01     ` [RFC PATCH v3 2/3] nvme: keep nvme_command instead of pointer to it Kanchan Joshi
2021-03-16 14:01       ` Kanchan Joshi
2021-03-16 17:16       ` Keith Busch
2021-03-16 17:16         ` Keith Busch
2021-03-17  9:38         ` Kanchan Joshi
2021-03-17  9:38           ` Kanchan Joshi
2021-03-17 14:17           ` Keith Busch
2021-03-17 14:17             ` Keith Busch
     [not found]   ` <CGME20210316140240epcas5p3e71bfe2afecd728c5af60056f21cc9b7@epcas5p3.samsung.com>
2021-03-16 14:01     ` [RFC PATCH v3 3/3] nvme: wire up support for async passthrough Kanchan Joshi
2021-03-16 14:01       ` Kanchan Joshi
2021-03-17  8:52       ` Christoph Hellwig
2021-03-17  8:52         ` Christoph Hellwig
2021-03-17 16:49         ` Jens Axboe
2021-03-17 16:49           ` Jens Axboe
2021-03-17 16:59           ` Christoph Hellwig
2021-03-17 16:59             ` Christoph Hellwig
2021-03-17 17:21             ` Jens Axboe
2021-03-17 17:21               ` Jens Axboe
2021-03-17 18:59               ` Jens Axboe
2021-03-17 18:59                 ` Jens Axboe
2021-03-18  5:54         ` Kanchan Joshi
2021-03-18  5:54           ` Kanchan Joshi
2021-03-17 16:45       ` Keith Busch
2021-03-17 16:45         ` Keith Busch
2021-03-17 17:02         ` Christoph Hellwig
2021-03-17 17:02           ` Christoph Hellwig
2021-03-16 15:51   ` [RFC PATCH v3 0/3] Async nvme passthrough over io_uring Jens Axboe
2021-03-16 15:51     ` Jens Axboe
2021-03-17  9:31     ` Kanchan Joshi
2021-03-17  9:31       ` Kanchan Joshi
2021-03-18  1:58       ` Jens Axboe
2021-03-18  1:58         ` Jens Axboe
2021-03-18  7:47         ` Kanchan Joshi
2021-03-18  7:47           ` Kanchan Joshi

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=20210318054807.GA28576@lst.de \
    --to=hch@lst.de \
    --cc=anuj20.g@samsung.com \
    --cc=axboe@kernel.dk \
    --cc=chaitanya.kulkarni@wdc.com \
    --cc=io-uring@vger.kernel.org \
    --cc=javier.gonz@samsung.com \
    --cc=joshi.k@samsung.com \
    --cc=joshiiitr@gmail.com \
    --cc=kbusch@kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=nj.shetty@samsung.com \
    --cc=selvakuma.s1@samsung.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 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.