io-uring.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kanchan Joshi <joshiiitr@gmail.com>
To: Christoph Hellwig <hch@lst.de>
Cc: "Kanchan Joshi" <joshi.k@samsung.com>,
	"Jens Axboe" <axboe@kernel.dk>, "Keith Busch" <kbusch@kernel.org>,
	"Pavel Begunkov" <asml.silence@gmail.com>,
	io-uring@vger.kernel.org, linux-nvme@lists.infradead.org,
	linux-block@vger.kernel.org, sbates@raithlin.com,
	logang@deltatee.com, "Pankaj Raghav" <pankydev8@gmail.com>,
	"Javier González" <javier@javigon.com>,
	"Luis Chamberlain" <mcgrof@kernel.org>,
	"Adam Manzanares" <a.manzanares@samsung.com>,
	"Anuj Gupta" <anuj20.g@samsung.com>
Subject: Re: [PATCH 17/17] nvme: enable non-inline passthru commands
Date: Fri, 11 Mar 2022 00:13:24 +0530	[thread overview]
Message-ID: <CA+1E3rL3Q2noHW-cD20SZyo9EqbzjF54F6TgZoUMMuZGkhkqnw@mail.gmail.com> (raw)
In-Reply-To: <20220310141945.GA890@lst.de>

On Thu, Mar 10, 2022 at 7:49 PM Christoph Hellwig <hch@lst.de> wrote:
>
> On Thu, Mar 10, 2022 at 05:20:13PM +0530, Kanchan Joshi wrote:
> > In sync ioctl, we always update this result field by doing put_user on
> > completion.
> > For async ioctl, since command is inside the the sqe, its lifetime is
> > only upto submission. SQE may get reused post submission, leaving no
> > way to update the "result" field on completion. Had this field been a
> > pointer, we could have saved this on submission and updated on
> > completion. But that would require redesigning this structure and
> > adding newer ioctl in nvme.
>
> Why would it required adding an ioctl to nvme?  The whole io_uring
> async_cmd infrastructure is completely independent from ioctls.

io_uring is sure not peeking into ioctl and its command-structure but
offering the facility to use its sqe to store that ioctl-command
inline.
Problem is, the inline facility does not go very well with this
particular nvme-passthru ioctl (NVME_IOCTL_IO64_CMD).
And that's because this ioctl requires additional "__u64 result;" to
be updated within "struct nvme_passthru_cmd64".
To update that during completion, we need, at the least, the result
field to be a pointer "__u64 result_ptr" inside the struct
nvme_passthru_cmd64.
Do you see that is possible without adding a new passthru ioctl in nvme?

> > Coming back, even though sync-ioctl alway updates this result to
> > user-space, only a few nvme io commands (e.g. zone-append, copy,
> > zone-mgmt-send) can return this additional result (spec-wise).
> > Therefore in nvme, when we are dealing with inline-sqe commands from
> > io_uring, we never attempt to update the result. And since we don't
> > update the result, we limit support to only read/write passthru
> > commands. And fail any other command during submission itself (Patch
> > 2).
>
> Yikes.  That is outright horrible.  passthrough needs to be command
> agnostic and future proof to any newly added nvme command.

This patch (along with patch 16) does exactly that. Makes it
command-agnostic and future-proof. All nvme-commands will work with
it.
Just that application needs to pass the pointer of ioctl-command and
not place it inline inside the sqe.

Overall, I think at io_uring infra level both submission makes sense:
big-sqe based inline submission (more efficient for <= 80 bytes) and
normal-sqe based non-inline/indirect submissions.
At nvme-level, we have to pick (depending on ioctl in hand). Currently
we are playing with both and constructing a sort of fast-path (for all
commands) and another faster-path (only for read/write commands).
Should we (at nvme-level) rather opt out and use only indirect
(because it works for all commands) or must we build a way to enable
inline-one for all commands?

> > > Overly long line.
> >
> > Under 100, but sure, can fold it under 80.
>
> You can only use 100 sparingly if it makes the code more readable.  Which
> I know is fuzzy, and in practice never does.  Certainly not in nvme and
> block code.

Clears up, thanks.

-- 
Kanchan

  reply	other threads:[~2022-03-10 18:43 UTC|newest]

Thread overview: 122+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20220308152651epcas5p1ebd2dc7fa01db43dd587c228a3695696@epcas5p1.samsung.com>
2022-03-08 15:20 ` [PATCH 00/17] io_uring passthru over nvme Kanchan Joshi
     [not found]   ` <CGME20220308152653epcas5p10c31f58cf6bff125cc0baa176b4d4fac@epcas5p1.samsung.com>
2022-03-08 15:20     ` [PATCH 01/17] io_uring: add support for 128-byte SQEs Kanchan Joshi
     [not found]   ` <CGME20220308152655epcas5p4ae47d715e1c15069e97152dcd283fd40@epcas5p4.samsung.com>
2022-03-08 15:20     ` [PATCH 02/17] fs: add file_operations->async_cmd() Kanchan Joshi
     [not found]   ` <CGME20220308152658epcas5p3929bd1fcf75edc505fec71901158d1b5@epcas5p3.samsung.com>
2022-03-08 15:20     ` [PATCH 03/17] io_uring: add infra and support for IORING_OP_URING_CMD Kanchan Joshi
2022-03-11  1:51       ` Luis Chamberlain
2022-03-11  2:43         ` Jens Axboe
2022-03-11 17:11           ` Luis Chamberlain
2022-03-11 18:47             ` Paul Moore
2022-03-11 20:57               ` Luis Chamberlain
2022-03-11 21:03                 ` Paul Moore
2022-03-14 16:25             ` Casey Schaufler
2022-03-14 16:32               ` Luis Chamberlain
2022-03-14 18:05                 ` Casey Schaufler
2022-03-14 19:40                   ` Luis Chamberlain
     [not found]   ` <CGME20220308152700epcas5p4130d20119a3a250a2515217d6552f668@epcas5p4.samsung.com>
2022-03-08 15:20     ` [PATCH 04/17] nvme: modify nvme_alloc_request to take an additional parameter Kanchan Joshi
2022-03-11  6:38       ` Christoph Hellwig
     [not found]   ` <CGME20220308152702epcas5p1eb1880e024ac8b9531c85a82f31a4e78@epcas5p1.samsung.com>
2022-03-08 15:20     ` [PATCH 05/17] nvme: wire-up support for async-passthru on char-device Kanchan Joshi
2022-03-10  0:02       ` Clay Mayers
2022-03-10  8:32         ` Kanchan Joshi
2022-03-11  7:01       ` Christoph Hellwig
2022-03-14 16:23         ` Kanchan Joshi
2022-03-15  8:54           ` Christoph Hellwig
2022-03-16  7:27             ` Kanchan Joshi
2022-03-24  6:22               ` Christoph Hellwig
2022-03-24 17:45                 ` Kanchan Joshi
2022-03-11 17:56       ` Luis Chamberlain
2022-03-11 18:53         ` Paul Moore
2022-03-11 21:02           ` Luis Chamberlain
2022-03-13 21:53       ` Sagi Grimberg
2022-03-14 17:54         ` Kanchan Joshi
2022-03-15  9:02           ` Sagi Grimberg
2022-03-16  9:21             ` Kanchan Joshi
2022-03-16 10:56               ` Sagi Grimberg
2022-03-16 11:51                 ` Kanchan Joshi
2022-03-16 13:52                   ` Sagi Grimberg
2022-03-16 14:35                     ` Jens Axboe
2022-03-16 14:50                       ` Sagi Grimberg
2022-03-24  6:20                         ` Christoph Hellwig
2022-03-24 10:42                           ` Sagi Grimberg
2022-03-22 15:18       ` Clay Mayers
2022-03-22 16:57         ` Kanchan Joshi
     [not found]   ` <CGME20220308152704epcas5p16610e1f50672b25fa1df5f7c5c261bb5@epcas5p1.samsung.com>
2022-03-08 15:20     ` [PATCH 06/17] io_uring: prep for fixed-buffer enabled uring-cmd Kanchan Joshi
     [not found]   ` <CGME20220308152707epcas5p430127761a7fd4bf90c2501eabe9ee96e@epcas5p4.samsung.com>
2022-03-08 15:20     ` [PATCH 07/17] io_uring: add support for uring_cmd with fixed-buffer Kanchan Joshi
     [not found]   ` <CGME20220308152709epcas5p1f9d274a0214dc462c22c278a72d8697c@epcas5p1.samsung.com>
2022-03-08 15:20     ` [PATCH 08/17] nvme: enable passthrough " Kanchan Joshi
2022-03-10  8:32       ` Christoph Hellwig
2022-03-11  6:43       ` Christoph Hellwig
2022-03-14 13:06         ` Kanchan Joshi
2022-03-15  8:55           ` Christoph Hellwig
2022-03-14 12:18       ` Ming Lei
2022-03-14 13:09         ` Kanchan Joshi
     [not found]   ` <CGME20220308152711epcas5p31de5d63f5de91fae94e61e5c857c0f13@epcas5p3.samsung.com>
2022-03-08 15:20     ` [PATCH 09/17] io_uring: plug for async bypass Kanchan Joshi
2022-03-10  8:33       ` Christoph Hellwig
2022-03-14 14:33         ` Ming Lei
2022-03-15  8:56           ` Christoph Hellwig
2022-03-11 17:15       ` Luis Chamberlain
     [not found]   ` <CGME20220308152714epcas5p4c5a0d16512fd7054c9a713ee28ede492@epcas5p4.samsung.com>
2022-03-08 15:20     ` [PATCH 10/17] block: wire-up support for plugging Kanchan Joshi
2022-03-10  8:34       ` Christoph Hellwig
2022-03-10 12:40         ` Kanchan Joshi
2022-03-14 14:40           ` Ming Lei
2022-03-21  7:02             ` Kanchan Joshi
2022-03-23  1:27               ` Ming Lei
2022-03-23  1:41                 ` Jens Axboe
2022-03-23  1:58                   ` Jens Axboe
2022-03-23  2:10                     ` Ming Lei
2022-03-23  2:17                       ` Jens Axboe
     [not found]   ` <CGME20220308152716epcas5p3d38d2372c184259f1a10c969f7e4396f@epcas5p3.samsung.com>
2022-03-08 15:20     ` [PATCH 11/17] block: factor out helper for bio allocation from cache Kanchan Joshi
2022-03-10  8:35       ` Christoph Hellwig
2022-03-10 12:25         ` Kanchan Joshi
2022-03-24  6:30           ` Christoph Hellwig
2022-03-24 17:45             ` Kanchan Joshi
2022-03-25  5:38               ` Christoph Hellwig
     [not found]   ` <CGME20220308152718epcas5p3afd2c8a628f4e9733572cbb39270989d@epcas5p3.samsung.com>
2022-03-08 15:21     ` [PATCH 12/17] nvme: enable bio-cache for fixed-buffer passthru Kanchan Joshi
2022-03-11  6:48       ` Christoph Hellwig
2022-03-14 18:18         ` Kanchan Joshi
2022-03-15  8:57           ` Christoph Hellwig
     [not found]   ` <CGME20220308152720epcas5p19653942458e160714444942ddb8b8579@epcas5p1.samsung.com>
2022-03-08 15:21     ` [PATCH 13/17] nvme: allow user passthrough commands to poll Kanchan Joshi
2022-03-08 17:08       ` Keith Busch
2022-03-09  7:03         ` Kanchan Joshi
2022-03-11  6:49           ` Christoph Hellwig
     [not found]   ` <CGME20220308152723epcas5p34460b4af720e515317f88dbb78295f06@epcas5p3.samsung.com>
2022-03-08 15:21     ` [PATCH 14/17] io_uring: add polling support for uring-cmd Kanchan Joshi
2022-03-11  6:50       ` Christoph Hellwig
2022-03-14 10:16         ` Kanchan Joshi
2022-03-15  8:57           ` Christoph Hellwig
2022-03-16  5:09             ` Kanchan Joshi
2022-03-24  6:30               ` Christoph Hellwig
     [not found]   ` <CGME20220308152725epcas5p36d1ce3269a47c1c22cc0d66bdc2b9eb3@epcas5p3.samsung.com>
2022-03-08 15:21     ` [PATCH 15/17] nvme: wire-up polling for uring-passthru Kanchan Joshi
     [not found]   ` <CGME20220308152727epcas5p20e605718dd99e97c94f9232d40d04d95@epcas5p2.samsung.com>
2022-03-08 15:21     ` [PATCH 16/17] io_uring: add support for non-inline uring-cmd Kanchan Joshi
     [not found]   ` <CGME20220308152729epcas5p17e82d59c68076eb46b5ef658619d65e3@epcas5p1.samsung.com>
2022-03-08 15:21     ` [PATCH 17/17] nvme: enable non-inline passthru commands Kanchan Joshi
2022-03-10  8:36       ` Christoph Hellwig
2022-03-10 11:50         ` Kanchan Joshi
2022-03-10 14:19           ` Christoph Hellwig
2022-03-10 18:43             ` Kanchan Joshi [this message]
2022-03-11  6:27               ` Christoph Hellwig
2022-03-22 17:10                 ` Kanchan Joshi
2022-03-24  6:32                   ` Christoph Hellwig
2022-03-25 13:39                     ` Kanchan Joshi
2022-03-28  4:44                       ` Kanchan Joshi
2022-03-30 12:59                         ` Christoph Hellwig
2022-03-30 13:02                       ` Christoph Hellwig
2022-03-30 13:14                         ` Kanchan Joshi
2022-04-01  1:25                           ` Jens Axboe
2022-04-01  2:33                             ` Kanchan Joshi
2022-04-01  2:44                               ` Jens Axboe
2022-04-01  3:05                                 ` Jens Axboe
2022-04-01  6:32                                 ` Kanchan Joshi
2022-04-19 17:31                                 ` Kanchan Joshi
2022-04-19 18:19                                   ` Jens Axboe
     [not found]                                     ` <CGME20220420152003epcas5p3991e6941773690bcb425fd9d817105c3@epcas5p3.samsung.com>
2022-04-20 15:14                                       ` Kanchan Joshi
2022-04-20 15:28                                         ` Kanchan Joshi
2022-04-01  1:23                         ` Jens Axboe
2022-04-01  1:22                   ` Jens Axboe
2022-04-01  6:29                     ` Kanchan Joshi
2022-03-24 21:09       ` Clay Mayers
2022-03-24 23:36         ` Jens Axboe
2022-03-10  8:29   ` [PATCH 00/17] io_uring passthru over nvme Christoph Hellwig
2022-03-10 10:05     ` Kanchan Joshi
2022-03-11 16:43       ` Luis Chamberlain
2022-03-11 23:35         ` Adam Manzanares
2022-03-12  2:27           ` Adam Manzanares
2022-03-13  5:07             ` Kanchan Joshi
2022-03-14 20:30               ` Adam Manzanares
2022-03-13  5:10         ` 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=CA+1E3rL3Q2noHW-cD20SZyo9EqbzjF54F6TgZoUMMuZGkhkqnw@mail.gmail.com \
    --to=joshiiitr@gmail.com \
    --cc=a.manzanares@samsung.com \
    --cc=anuj20.g@samsung.com \
    --cc=asml.silence@gmail.com \
    --cc=axboe@kernel.dk \
    --cc=hch@lst.de \
    --cc=io-uring@vger.kernel.org \
    --cc=javier@javigon.com \
    --cc=joshi.k@samsung.com \
    --cc=kbusch@kernel.org \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=logang@deltatee.com \
    --cc=mcgrof@kernel.org \
    --cc=pankydev8@gmail.com \
    --cc=sbates@raithlin.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).