All of lore.kernel.org
 help / color / mirror / Atom feed
From: Song Liu <song@kernel.org>
To: Guoqing Jiang <jgq516@gmail.com>
Cc: linux-raid <linux-raid@vger.kernel.org>
Subject: Re: [UPDATE PATCH V2 3/7] md: the latest try for improve io stats accounting
Date: Sun, 23 May 2021 22:49:58 -0700	[thread overview]
Message-ID: <CAPhsuW7wpWR5cUspUjgLoC78AFd3X=-7BpUDQCeqZ3UZTNcj9Q@mail.gmail.com> (raw)
In-Reply-To: <20210521080036.723295-1-jiangguoqing@kylinos.cn>

On Fri, May 21, 2021 at 1:01 AM Guoqing Jiang <jgq516@gmail.com> wrote:
>
> From: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
>
> Use generic io accounting functions to manage io stats. There was an
> attempt to do this earlier in commit 18c0b223cf990172 ("md: use generic
> io stats accounting functions to simplify io stat accounting"), but it
> did not include a call to generic_end_io_acct() and caused issues with
> tracking in-flight IOs, so it was later removed in commit 74672d069b29
> ("md: fix md io stats accounting broken").
>
> This patch attempts to fix this by using both generic_start_io_acct()
> and generic_end_io_acct(). To make it possible, in md_make_request() a
> bio is cloned with additional data - struct md_io, which includes the io
> start_time. A new bioset is introduced for this purpose. We call
> generic_start_io_acct() and pass the clone instead of the original to
> md_handle_request(). When it completes, we call generic_end_io_acct()
> and complete the original bio.
>
> This adds correct statistics about in-flight IOs and IO processing time,
> interpreted e.g. in iostat as await, svctm, aqu-sz and %util.
>
> It also fixes a situation where too many IOs where reported if a bio was
> re-submitted to the mddev, because io accounting is now performed only
> on newly arriving bios.
>
> Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
> [Guoqing: rebase and make generic accounting applies to personalities
>           which don't have clone infrastructure]
> Signed-off-by: Guoqing Jiang <jiangguoqing@kylinos.cn>
> ---
> Delete not necessary bioset_exit in md_integrity_register, thanks for
> Artur's review.
>
>  drivers/md/md.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++++-
>  drivers/md/md.h |  1 +
>  2 files changed, 55 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/md/md.c b/drivers/md/md.c
> index 7ba00e4c862d..d8823db843db 100644
> --- a/drivers/md/md.c
> +++ b/drivers/md/md.c
> @@ -441,6 +441,25 @@ void md_handle_request(struct mddev *mddev, struct bio *bio)
>  }
>  EXPORT_SYMBOL(md_handle_request);
>
> +struct md_io {
> +       struct mddev *mddev;
> +       struct bio *orig_bio;
> +       unsigned long start_time;
> +       struct bio bio_clone;
> +};
> +
> +static void md_end_io(struct bio *bio)
> +{
> +       struct md_io *md_io = bio->bi_private;
> +       struct bio *orig_bio = md_io->orig_bio;
> +
> +       orig_bio->bi_status = bio->bi_status;
> +
> +       bio_end_io_acct(orig_bio, md_io->start_time);
> +       bio_put(bio);
> +       bio_endio(orig_bio);
> +}
> +
>  static blk_qc_t md_submit_bio(struct bio *bio)
>  {
>         const int rw = bio_data_dir(bio);
> @@ -465,6 +484,30 @@ static blk_qc_t md_submit_bio(struct bio *bio)
>                 return BLK_QC_T_NONE;
>         }
>
> +       /*
> +        * clone bio under conditions:
> +        * 1. QUEUE_FLAG_IO_STAT flag is set.
> +        * 2. bio just enters md and it is not split from personality.
> +        */

We had iostat on by default. So, let set QUEUE_FLAG_IO_STAT in
md_run().

Thanks,
Song

  reply	other threads:[~2021-05-24  5:50 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-21  0:55 [PATCH V2 0/7] md: io stats accounting Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 1/7] md: revert " Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 2/7] md: add accounting_bio for raid0 and raid5 Guoqing Jiang
2021-05-24  5:48   ` Song Liu
2021-05-24  8:45   ` Christoph Hellwig
2021-05-21  0:55 ` [PATCH V2 3/7] md: the latest try for improve io stats accounting Guoqing Jiang
2021-05-21  7:32   ` Artur Paszkiewicz
2021-05-21  7:43     ` Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 4/7] md/raid1: rename print_msg with r1bio_existed Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 5/7] md/raid1: enable io accounting Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 6/7] md/raid10: " Guoqing Jiang
2021-05-21  0:55 ` [PATCH V2 7/7] md: mark some personalities as deprecated Guoqing Jiang
2021-05-21  8:00 ` [UPDATE PATCH V2 3/7] md: the latest try for improve io stats accounting Guoqing Jiang
2021-05-24  5:49   ` Song Liu [this message]
2021-05-24  6:04 ` [PATCH V2 0/7] md: " Song Liu
2021-05-27 14:09   ` Paweł Wiejacha

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='CAPhsuW7wpWR5cUspUjgLoC78AFd3X=-7BpUDQCeqZ3UZTNcj9Q@mail.gmail.com' \
    --to=song@kernel.org \
    --cc=jgq516@gmail.com \
    --cc=linux-raid@vger.kernel.org \
    /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.