lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAD14+f349Fsx9syBzEtgbOXEts0FRrHWU8vUgbpRO-SJ8DMmJw@mail.gmail.com>
Date: Fri, 6 Sep 2024 23:52:49 +0900
From: Juhyung Park <qkrwngud825@...il.com>
To: Wu Bo <bo.wu@...o.com>
Cc: Wu Bo <wubo.oduw@...il.com>, Jaegeuk Kim <jaegeuk@...nel.org>, Chao Yu <chao@...nel.org>, 
	linux-kernel@...r.kernel.org, linux-f2fs-devel@...ts.sourceforge.net
Subject: Re: [f2fs-dev] [PATCH 00/13] f2fs: introduce inline tail

Hi Wu,

On Fri, Sep 6, 2024 at 7:42 PM Wu Bo <bo.wu@...o.com> wrote:
>
> On Fri, Sep 06, 2024 at 06:02:42PM +0900, Juhyung Park wrote:
> > Hi Wu,
> >
> > Interesting patch-set.
> >
> > A quick test here on my daily-driver phone's data (785558 inodes) with
> > both compression and encryption disabled, copied via rsync to 2 fresh
> > f2fs partitions with and without inline tail:
> > Before: 170064928KiB
> > After: 169249780KiB
> >
> > So about 0.48% saved.
> >
> Hi Juhyung,
>
> Thanks for your test. If it's not too much trouble, please help check the f2fs
> statistics:
> cat /sys/kernel/debug/f2fs/status | grep Utilization

# w/o inline tail
Utilization: 70% (42153224 valid blocks, 0 discard blocks)
# w/ inline tail
Utilization: 69% (41949437 valid blocks, 0 discard blocks)

>
> However, it’s more likely that the benefits of inline tail are indeed limited on
> mobile devices. I previously evaluated the gains on my own phone, and they were
> less than 1% too.
>
> From the data on your phone, the average size is 170064928KiB/785558≈200KiB.
> Storage space on phones is primarily consumed by multimedia files, so the
> proportion of files smaller than 64KB is quite small.
>
> Therefore, compared to phone storage, scenarios like storing Linux source code,
> which involve many small files, are more likely to yield noticeable benefits.

Mmm.

I do have my own Linux setups using f2fs as well with extended node bitmap:
# Workstation, 2.5T/3.5T, 55602404 inodes
Utilization: 70% (657421720 valid blocks, 902 discard blocks)
# Laptop, 736G/1.9T, 12229380 inodes
Utilization: 39% (190980868 valid blocks, 2887 discard blocks)

I do have a lot of small files here, it'll be interesting to test this
out but I can't afford to run an unstable kernel for those atm. (Not
to mention finding a new SSD for migration.)

>
> However, don't be too disappointed with it, as it can still double the benefits
> based on the existing gains from inline data.
> Inline data:
> https://lore.kernel.org/linux-f2fs-devel/1384096401-25169-1-git-send-email-huajun.li.lee@gmail.com/T/#u
>
> Thanks,
> Wu Bo
>
> > Let me know if this is an unexpected result.
> >
> > Thanks,
> >
> > On Tue, Sep 3, 2024 at 5:42 PM Wu Bo via Linux-f2fs-devel
> > <linux-f2fs-devel@...ts.sourceforge.net> wrote:
> > >
> > > The inode in F2FS occupies an entire 4k block. For many small files, this means
> > > they consume much more space than their actual size. Therefore, there is
> > > significant potential to better utilize the inode block space.
> > >
> > > Currently, F2FS has two features to make use of the inode block space: inline
> > > data and inline xattr.
> > >
> > > Inline data stores file which size is smaller then 3.5k in inode block. However,
> > > for slightly larger small files, there still have much waste.
> > > For example, a 5k file requires 3 blocks, totaling 12k of space, which is
> > > more than twice the size of the file itself!
> > >
> > > Additionally, the end of a file often does not occupy an entire block. If we can
> > > store the end of the file data within the inode block, we can save an entire
> > > block for the file. This is particularly important for small files.
> > >
> > > In fact, the current inline data is a special case of inline tail, and
> > > inline tail is an extension of inline data.
> > >
> > > To make it simple, inline tail only on small files(<64k). And for larger files,
> > > inline tails don't provide any significant benefits.
> > >
> > > The layout of an inline tail inode block is following:
> > >
> > > | inode block     | 4096 |     inline tail enable    |
> > > | --------------- | ---- | --------------------------|
> > > | inode info      | 360  |                           |
> > > | --------------- | ---- | --------------------------|
> > > |                 |      | extra info         | 0~36 |
> > > |                 |      | **compact_addr[16] | 64   |
> > > | addr table[923] | 3692 | reserved           | 4    |
> > > |                 |      | **tail data        |      |
> > > |                 |      | inline_xattr       | 200  |
> > > | --------------- | ---- | --------------------------|
> > > | nid table[5]    | 20   |
> > > | node footer     | 24   |
> > >
> > > F2fs-tools to support inline tail:
> > > https://lore.kernel.org/linux-f2fs-devel/20240903075931.3339584-1-bo.wu@vivo.com
> > >
> > > I tested inline tail by copying the source code of Linux 6.9.7. The storage
> > > space was reduced by approximately 8%. Additionally, due to the reduced IO, the
> > > copy time also reduced by around 10%.
> > >
> > > This patch series has been tested with xfstests by running 'kvm-xfstests -c f2fs
> > > -g quick' both with and without the patch; no regressions were observed.
> > > The test result is:
> > > f2fs/default: 583 tests, 6 failures, 213 skipped, 650 seconds
> > >   Failures: generic/050 generic/064 generic/250 generic/252 generic/563
> > >       generic/735
> > >       Totals: 607 tests, 213 skipped, 30 failures, 0 errors, 579s
> > >
> > > Wu Bo (13):
> > >   f2fs: add inline tail mount option
> > >   f2fs: add inline tail disk layout definition
> > >   f2fs: implement inline tail write & truncate
> > >   f2fs: implement inline tail read & fiemap
> > >   f2fs: set inline tail flag when create inode
> > >   f2fs: fix address info has been truncated
> > >   f2fs: support seek for inline tail
> > >   f2fs: convert inline tail when inode expand
> > >   f2fs: fix data loss during inline tail writing
> > >   f2fs: avoid inlining quota files
> > >   f2fs: fix inline tail data lost
> > >   f2fs: convert inline tails to avoid potential issues
> > >   f2fs: implement inline tail forward recovery
> > >
> > >  fs/f2fs/data.c     |  93 +++++++++++++++++++++++++-
> > >  fs/f2fs/f2fs.h     |  46 ++++++++++++-
> > >  fs/f2fs/file.c     |  85 +++++++++++++++++++++++-
> > >  fs/f2fs/inline.c   | 159 +++++++++++++++++++++++++++++++++++++++------
> > >  fs/f2fs/inode.c    |   6 ++
> > >  fs/f2fs/namei.c    |   3 +
> > >  fs/f2fs/node.c     |   6 +-
> > >  fs/f2fs/recovery.c |   9 ++-
> > >  fs/f2fs/super.c    |  25 +++++++
> > >  fs/f2fs/verity.c   |   4 ++
> > >  10 files changed, 409 insertions(+), 27 deletions(-)
> > >
> > > --
> > > 2.35.3
> > >
> > >
> > >
> > > _______________________________________________
> > > Linux-f2fs-devel mailing list
> > > Linux-f2fs-devel@...ts.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
> >
> >
> > _______________________________________________
> > Linux-f2fs-devel mailing list
> > Linux-f2fs-devel@...ts.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

Powered by blists - more mailing lists