All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sergey Senozhatsky <senozhatsky@chromium.org>
To: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Nitin Gupta <ngupta@vflare.org>,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCHv4 2/9] zram: Add recompression algorithm sysfs knob
Date: Tue, 8 Nov 2022 09:40:15 +0900	[thread overview]
Message-ID: <Y2mlb6F7G8KskRch@google.com> (raw)
In-Reply-To: <Y2lXt9E6YqB7fRbQ@google.com>

On (22/11/07 11:08), Minchan Kim wrote:
[..]
> I am thinking like this:
> 
> * Without recomp_algo setup, user can do whatever they want on the fly
> 
> 
>     echo "type=idle threshold=3000 algo=zstd" > recompress
> 
> Later they could do
> 
>     echo "type=idle threshold=3000 algo=deflate" > recompress

By "without recomp_algo setup" you mean that user doesn't configure
anything before `echo XG > zramX/disksize`? Currently algorithm and
recomp algorithm need to be selected at the same time - before zram
device is initialised, because we use the same code and same approaches:
we need to have zcomp back-ends in per-CPU data in zram read/write/recompress.

Creating per-CPU zcomps on the fly is probably going to be a little bit
intrusive to zram.



What I currently have is as follows.

A copy paste from my test script:

- init device
        echo "algo=lz4 priority=1" > /sys/block/zram0/recomp_algorithm
        echo "algo=zstd priority=2" > /sys/block/zram0/recomp_algorithm
        echo "algo=deflate priority=3" > /sys/block/zram0/recomp_algorithm
        echo 5G > /sys/block/zram0/disksize


Various recompression use cases:

- recompress huge pages using all secondary algos in order of priority
        echo "type=huge" > /sys/block/zram0/recompress

- recompress huge pages using zstd only
        echo "type=huge algo=zstd" > /sys/block/zram0/recompress

- recompress all pages using lz4
        echo "algo=lz4" > /sys/block/zram0/recompress

- recompress idle pages, use all algos in priority order, with threshold
        echo "type=idle threshold=3000" > /sys/block/zram0/recompress

- recompress idle pages, using zstd only, with threshold
        echo "algo=zstd type=idle threshold=2000" > /sys/block/zram0/recompress

  reply	other threads:[~2022-11-08  0:40 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-18  4:55 [PATCHv4 0/9] zram: Support multiple compression streams Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 1/9] zram: Preparation for multi-zcomp support Sergey Senozhatsky
2022-11-02 20:13   ` Minchan Kim
2022-11-03  2:40     ` Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 2/9] zram: Add recompression algorithm sysfs knob Sergey Senozhatsky
2022-11-02 20:15   ` Minchan Kim
2022-11-03  3:05     ` Sergey Senozhatsky
2022-11-03  3:54       ` Sergey Senozhatsky
2022-11-03 17:10         ` Minchan Kim
2022-11-03  4:09       ` Sergey Senozhatsky
2022-11-03  5:36         ` Sergey Senozhatsky
2022-11-03 17:11         ` Minchan Kim
2022-11-03 16:34       ` Minchan Kim
2022-11-04  3:18         ` Sergey Senozhatsky
2022-11-04  4:53           ` Sergey Senozhatsky
2022-11-04 17:43             ` Minchan Kim
2022-11-04 23:41               ` Sergey Senozhatsky
2022-11-05  0:00                 ` Sergey Senozhatsky
2022-11-07 19:08                   ` Minchan Kim
2022-11-08  0:40                     ` Sergey Senozhatsky [this message]
2022-11-05  0:01                 ` Minchan Kim
2022-11-05  1:30                   ` Sergey Senozhatsky
2022-11-04 16:34           ` Minchan Kim
2022-11-04 23:25             ` Sergey Senozhatsky
2022-11-04 23:40               ` Minchan Kim
2022-11-04 23:44                 ` Sergey Senozhatsky
2022-11-05  0:02                   ` Minchan Kim
2022-10-18  4:55 ` [PATCHv4 3/9] zram: Factor out WB and non-WB zram read functions Sergey Senozhatsky
2022-11-02 20:20   ` Minchan Kim
2022-11-03  2:43     ` Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 4/9] zram: Introduce recompress sysfs knob Sergey Senozhatsky
2022-11-02 21:06   ` Minchan Kim
2022-11-03  3:25     ` Sergey Senozhatsky
2022-11-03  6:03       ` Sergey Senozhatsky
2022-11-03 17:00       ` Minchan Kim
2022-11-04  3:48         ` Sergey Senozhatsky
2022-11-04  7:12           ` Sergey Senozhatsky
2022-11-04 17:53             ` Minchan Kim
2022-11-04 17:27           ` Minchan Kim
2022-11-04 23:22             ` Sergey Senozhatsky
2022-11-04  7:53         ` Sergey Senozhatsky
2022-11-04  8:08           ` Sergey Senozhatsky
2022-11-04 17:47           ` Minchan Kim
2022-10-18  4:55 ` [PATCHv4 5/9] documentation: Add recompression documentation Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 6/9] zram: Add recompression algorithm choice to Kconfig Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 7/9] zram: Add recompress flag to read_block_state() Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 8/9] zram: Clarify writeback_store() comment Sergey Senozhatsky
2022-10-18  4:55 ` [PATCHv4 9/9] zram: Use IS_ERR_VALUE() to check for zs_malloc() errors Sergey Senozhatsky
2022-11-02 20:07 ` [PATCHv4 0/9] zram: Support multiple compression streams Minchan Kim
2022-11-03  3:36   ` Sergey Senozhatsky

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=Y2mlb6F7G8KskRch@google.com \
    --to=senozhatsky@chromium.org \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=minchan@kernel.org \
    --cc=ngupta@vflare.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.