linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@infradead.org>
To: Pankaj Raghav <p.raghav@samsung.com>
Cc: Christoph Hellwig <hch@infradead.org>,
	hubcap@omnibond.com, senozhatsky@chromium.org,
	martin@omnibond.com, willy@infradead.org, minchan@kernel.org,
	viro@zeniv.linux.org.uk, brauner@kernel.org, axboe@kernel.dk,
	akpm@linux-foundation.org, linux-kernel@vger.kernel.org,
	linux-block@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-mm@kvack.org, gost.dev@samsung.com, mcgrof@kernel.org,
	devel@lists.orangefs.org
Subject: Re: [RFC PATCH 1/3] filemap: convert page_endio to folio_endio
Date: Thu, 16 Mar 2023 08:24:04 -0700	[thread overview]
Message-ID: <ZBM0lPycYm2X0Tfp@infradead.org> (raw)
In-Reply-To: <d6cde35e-359a-e837-d2e0-f2bd362f2c3e@samsung.com>

On Thu, Mar 16, 2023 at 11:04:54AM +0100, Pankaj Raghav wrote:
> It looks like this endio function is called when alloc_page is used (for
> partial IO) to trigger writeback from the user space `echo "idle" >
> /sys/block/zram0/writeback`.


Yes.

> I don't understand when you say the harm might not be horrible if we don't
> call folio_endio here. Do you think it is just safe to remove the call to
> folio_endio function?

I suspect so.  It doesn't seem like the involved pages are ever locked
or have the writeback set, so it should be fine.

> +               while ((folio = readahead_folio(rac))) {
> +                       folio_clear_uptodate(folio);
> +                       folio_set_error(folio);
> +                       folio_unlock(folio);
> +               }
> +               return;
> +       }
> +
> +       while ((folio = readahead_folio(rac))) {
> +               folio_mark_uptodate(folio);
> +               folio_unlock(folio);
>         }
>  }

Looks good.

> @@ -59,7 +59,8 @@ static void mpage_end_io(struct bio *bio)
> static struct bio *mpage_bio_submit(struct bio *bio)
>  {
> -       bio->bi_end_io = mpage_end_io;
> +       bio->bi_end_io = (op_is_write(bio_op(bio))) ? mpage_write_end_io :
> +                                                     mpage_read_end_io;
>         guard_bio_eod(bio);
>         submit_bio(bio);
>         return NULL;
> And mpage_{write,read}_end_io will iterate over the folio and call the
> respective functions.

Yes, although I'd do it with a good old if/else and with less braces.
It might make sense to split mpage_bio_submit as well, though.


  reply	other threads:[~2023-03-16 15:24 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20230315123234eucas1p179bf8c0583a71d91bef7e909d7ec6504@eucas1p1.samsung.com>
2023-03-15 12:32 ` [RFC PATCH 0/3] convert page_endio to folio_endio Pankaj Raghav
     [not found]   ` <CGME20230315123234eucas1p2503d83ad0180cecde02e924d7b143535@eucas1p2.samsung.com>
2023-03-15 12:32     ` [RFC PATCH 1/3] filemap: " Pankaj Raghav
2023-03-15 14:46       ` Luis Chamberlain
2023-03-15 14:51       ` Hannes Reinecke
2023-03-15 14:56       ` Christoph Hellwig
2023-03-16 10:04         ` Pankaj Raghav
2023-03-16 15:24           ` Christoph Hellwig [this message]
2023-03-17 15:31           ` Matthew Wilcox
2023-03-17 16:14             ` Pankaj Raghav
2023-03-21 13:51               ` Pankaj Raghav
     [not found]   ` <CGME20230315123235eucas1p1bd62cb2aab435727880769f2e57624fd@eucas1p1.samsung.com>
2023-03-15 12:32     ` [RFC PATCH 2/3] mpage: use bio_for_each_folio_all in mpage_end_io() Pankaj Raghav
2023-03-15 14:47       ` Luis Chamberlain
2023-03-15 14:52       ` Hannes Reinecke
2023-03-15 16:08         ` Matthew Wilcox
2023-03-16 10:07           ` Pankaj Raghav
     [not found]   ` <CGME20230315123236eucas1p1116e1b8537191310bd03dd267b9f8eb8@eucas1p1.samsung.com>
2023-03-15 12:32     ` [RFC PATCH 3/3] orangefs: use folio in orangefs_readahead() Pankaj Raghav
2023-03-15 14:48       ` Luis Chamberlain
2023-03-15 16:05       ` Matthew Wilcox

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=ZBM0lPycYm2X0Tfp@infradead.org \
    --to=hch@infradead.org \
    --cc=akpm@linux-foundation.org \
    --cc=axboe@kernel.dk \
    --cc=brauner@kernel.org \
    --cc=devel@lists.orangefs.org \
    --cc=gost.dev@samsung.com \
    --cc=hubcap@omnibond.com \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=martin@omnibond.com \
    --cc=mcgrof@kernel.org \
    --cc=minchan@kernel.org \
    --cc=p.raghav@samsung.com \
    --cc=senozhatsky@chromium.org \
    --cc=viro@zeniv.linux.org.uk \
    --cc=willy@infradead.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 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).