From: Matthew Wilcox <willy@infradead.org>
To: "Darrick J. Wong" <djwong@kernel.org>
Cc: "Ritesh Harjani (IBM)" <ritesh.list@gmail.com>,
linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
Christoph Hellwig <hch@infradead.org>,
Aravinda Herle <araherle@in.ibm.com>
Subject: Re: [RFC 2/2] iomap: Support subpage size dirty tracking to improve write performance
Date: Fri, 28 Oct 2022 19:15:41 +0100 [thread overview]
Message-ID: <Y1wcTUxlo5zinsg3@casper.infradead.org> (raw)
In-Reply-To: <Y1wK3x7IketHl+DQ@magnolia>
On Fri, Oct 28, 2022 at 10:01:19AM -0700, Darrick J. Wong wrote:
> On Fri, Oct 28, 2022 at 10:00:33AM +0530, Ritesh Harjani (IBM) wrote:
> > Performance testing of below fio workload reveals ~16x performance
> > improvement on nvme with XFS (4k blocksize) on Power (64K pagesize)
> > FIO reported write bw scores improved from around ~28 MBps to ~452 MBps.
> >
> > <test_randwrite.fio>
> > [global]
> > ioengine=psync
> > rw=randwrite
> > overwrite=1
> > pre_read=1
> > direct=0
> > bs=4k
> > size=1G
> > dir=./
> > numjobs=8
> > fdatasync=1
> > runtime=60
> > iodepth=64
> > group_reporting=1
>
> Admittedly I'm not thrilled at the reintroduction of page and iop dirty
> state that are updated in separate places, but OTOH the write
> amplification here is demonstrably horrifying as you point out so it's
> clearly necessary.
Well, *something* is necessary. I worked on a different approach that
would have similar effects for this exact workload, which was to submit
the I/O for O_SYNC while we still know which part of the page we
dirtied.
Previous discussion:
https://lore.kernel.org/all/YQlgjh2R8OzJkFoB@casper.infradead.org/
Actual patches:
https://lore.kernel.org/all/20220503064008.3682332-1-willy@infradead.org/
next prev parent reply other threads:[~2022-10-28 18:15 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-28 4:30 [RFC 0/2] iomap: Add support for subpage dirty state tracking to improve write performance Ritesh Harjani (IBM)
2022-10-28 4:30 ` [RFC 1/2] iomap: Change uptodate variable name to state Ritesh Harjani (IBM)
2022-10-28 16:31 ` Darrick J. Wong
2022-10-29 3:09 ` Ritesh Harjani (IBM)
2022-10-28 4:30 ` [RFC 2/2] iomap: Support subpage size dirty tracking to improve write performance Ritesh Harjani (IBM)
2022-10-28 12:42 ` Matthew Wilcox
2022-10-29 3:05 ` Ritesh Harjani (IBM)
2022-10-28 17:01 ` Darrick J. Wong
2022-10-28 18:15 ` Matthew Wilcox [this message]
2022-10-29 3:25 ` Ritesh Harjani (IBM)
2022-10-28 21:04 ` Dave Chinner
2022-10-30 3:27 ` Ritesh Harjani (IBM)
2022-10-30 22:31 ` Dave Chinner
2022-10-31 3:43 ` Matthew Wilcox
2022-10-31 7:08 ` Dave Chinner
2022-10-31 10:27 ` Matthew Wilcox
2022-11-02 8:57 ` Christoph Hellwig
2022-11-03 0:38 ` Dave Chinner
2022-11-02 9:03 ` Christoph Hellwig
2022-11-02 17:35 ` Darrick J. Wong
2022-11-04 7:27 ` Christoph Hellwig
2022-11-04 14:15 ` Ritesh Harjani (IBM)
2022-11-03 14:51 ` David Howells
2022-11-04 7:30 ` Christoph Hellwig
2022-11-07 13:03 ` David Howells
2022-11-04 11:28 ` Ritesh Harjani (IBM)
2022-11-03 14:12 ` David Howells
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=Y1wcTUxlo5zinsg3@casper.infradead.org \
--to=willy@infradead.org \
--cc=araherle@in.ibm.com \
--cc=djwong@kernel.org \
--cc=hch@infradead.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=ritesh.list@gmail.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).