public inbox for [email protected]
 help / color / mirror / Atom feed
From: Christoph Hellwig <[email protected]>
To: Jens Axboe <[email protected]>
Cc: [email protected], [email protected], [email protected],
	[email protected], [email protected]
Subject: Re: [PATCH 6/6] iomap: support IOCB_DIO_DEFER
Date: Thu, 20 Jul 2023 06:59:19 +0200	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>

> +	if (dio->flags & IOMAP_DIO_DEFER_COMP) {
> +		/* only polled IO cares about private cleared */
> +		iocb->private = dio;

FYI, I find this comment a bit weird as it comments on what we're
not doing in a path where it is irreleant.  I'd rather only clear
the private data in the path where polling is applicable and have
a comment there why it is cleared.  That probably belongs into the
first patch restructuring the function.

> @@ -277,12 +308,15 @@ static loff_t iomap_dio_bio_iter(const struct iomap_iter *iter,
>  		 * data IO that doesn't require any metadata updates (including
>  		 * after IO completion such as unwritten extent conversion) and
>  		 * the underlying device supports FUA. This allows us to avoid
> -		 * cache flushes on IO completion.
> +		 * cache flushes on IO completion. If we can't use FUA and
> +		 * need to sync, disable in-task completions.

... because iomap_dio_complete will have to call generic_write_sync to
do a blocking ->fsync call.

> @@ -308,6 +342,8 @@ static loff_t iomap_dio_bio_iter(const struct iomap_iter *iter,
>  		pad = pos & (fs_block_size - 1);
>  		if (pad)
>  			iomap_dio_zero(iter, dio, pos - pad, pad);
> +
> +		dio->flags &= ~IOMAP_DIO_DEFER_COMP;

Why does zeroing disable the deferred completions?  I can't really think
of why, which is probably a strong indicator why this needs a comment.


  reply	other threads:[~2023-07-20  4:59 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-19 19:54 [PATCHSET v3 0/6] Improve async iomap DIO performance Jens Axboe
2023-07-19 19:54 ` [PATCH 1/6] iomap: cleanup up iomap_dio_bio_end_io() Jens Axboe
2023-07-20  4:50   ` Christoph Hellwig
2023-07-20 16:13     ` Jens Axboe
2023-07-19 19:54 ` [PATCH 2/6] iomap: add IOMAP_DIO_INLINE_COMP Jens Axboe
2023-07-20  4:51   ` Christoph Hellwig
2023-07-20 16:19     ` Jens Axboe
2023-07-19 19:54 ` [PATCH 3/6] iomap: treat a write through cache the same as FUA Jens Axboe
2023-07-20  4:54   ` Christoph Hellwig
2023-07-20 16:23     ` Jens Axboe
2023-07-19 19:54 ` [PATCH 4/6] fs: add IOCB flags related to passing back dio completions Jens Axboe
2023-07-20  5:01   ` Christoph Hellwig
2023-07-20 16:25     ` Jens Axboe
2023-07-19 19:54 ` [PATCH 5/6] io_uring/rw: add write support for IOCB_DIO_DEFER Jens Axboe
2023-07-19 19:54 ` [PATCH 6/6] iomap: support IOCB_DIO_DEFER Jens Axboe
2023-07-20  4:59   ` Christoph Hellwig [this message]
2023-07-20 16:27     ` Jens Axboe

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] \
    /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