From: Changman Lee <[email protected]>
To: Jens Axboe <[email protected]>
Cc: [email protected]
Subject: Re: io uring support for character device
Date: Thu, 26 May 2022 10:49:01 +0900 [thread overview]
Message-ID: <CAN863Pv6zVTvgZA30VcuSvPZojUCvNCB3We--Na8ZPNQjA_9Jg@mail.gmail.com> (raw)
In-Reply-To: <[email protected]>
I have both ->read() and ->read_iter(). I'll negate a ->read() and
a_ops->direct_IO() because those were for testing.
I'll go over my io_uring test code in user space.
Thanks for your advice.
2022년 5월 25일 (수) 오후 12:00, Jens Axboe <[email protected]>님이 작성:
>
> On 5/24/22 6:56 PM, Changman Lee wrote:
> > Hi List,
> >
> > I'm CM Lee. I'm developing a custom character device managing pcie dma.
> > I've tried to use io uring for the char device which supports readv
> > and writev with synchronous and blocking manner and seek.
> > When I use a io uring with IORING_SETUP_IOPOLL and IORING_SETUP_SQPOLL
> > for reducing syscall overhead, a readv of the char device driver seems
> > to be not called. So I added a_ops->direct_IO when the device is
> > opened with O_DIRECT. But the result was the same.
> > This is my question.
> > Q1: Does io uring support a character device ?
> > Q2: Is it better to reimplement a device driver as block device type ?
>
> io_uring doesn't care what file type it is, I suspect your problem lies
> elsewhere. Do you have a ->read() defined as well? If you do, the vfs
> will pick that over ->read_iter().
>
> If regular read/write works with O_DIRECT and reading from the device in
> general, then io_uring will too.
>
> --
> Jens Axboe
>
prev parent reply other threads:[~2022-05-26 1:49 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-25 0:56 io uring support for character device Changman Lee
2022-05-25 3:00 ` Jens Axboe
2022-05-26 1:49 ` Changman Lee [this message]
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=CAN863Pv6zVTvgZA30VcuSvPZojUCvNCB3We--Na8ZPNQjA_9Jg@mail.gmail.com \
[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