mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Matthias Andree <matthias.andree@gmx.de>
To: lkml <linux-kernel@vger.kernel.org>
Cc: marcelo <marcelo@conectiva.com.br>, Andrew Morton <akpm@digeo.com>
Subject: Re: Linux 2.4.21-pre1
Date: Wed, 11 Dec 2002 09:55:53 +0100	[thread overview]
Message-ID: <20021211085553.GB8740@merlin.emma.line.org> (raw)
In-Reply-To: <Pine.LNX.4.50L.0212101834240.23096-100000@freak.distro.conectiva>

On Tue, 10 Dec 2002, Marcelo Tosatti wrote:

> So here goes the first pre of 2.4.21 including the new IDE code merged
> from Alan's tree.
> 
> Test it carefully, since the new IDE code is not yet fully tested.
> 
> Do not use it with critical data.

Oh, speaking of "carefully", could Andrew Morton's mount -o dirsync
feature patch for 2.4.21 please be merged early in v2.4.21-pre? User
space (util-linux, e2fsprogs' lsattr/chattr) has been supporting it for
a while, just the kernel lacked it.

This -o dirsync feature is necessary to give applications that are
either unaware of Linux' default -o async or that do not want to cater
for deliberate incompatibilities a big performance boost: these
applications (Postfix on ext2fs but not on 2.4 ext3fs or reiserfs, qmail
on any Linux fs, probably a lot others I haven't even looked at)
traditionally *require* chattr +S or mount -o sync to work reliably,
after the merge, they could instead go with mount -o dirsync or chattr
+D, which improves write speed quite a bit.

Not applying the dirsync patch means that a stupid BSD ffs in default
configuration will beat Linux (that needs -o sync) hands-down without
any tuning the admin could do. One might also consider making -o dirsync
the default while one's there, to give users a default-reliable setting,
but this is going to rouse the performance fetishists who don't care
about data persistence and who don't care to collect their new files
from lost+found on ext2fs but rather cheat benchmarks.

I can elaborate a bit more, but this has been discussed multiple times
and after all, the code is there.

> Summary of changes from v2.4.20 to v2.4.21-pre1

...

  parent reply	other threads:[~2002-12-11  8:48 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-12-10 20:37 Marcelo Tosatti
2002-12-11  1:14 ` Adrian Bunk
2002-12-11  8:55 ` Matthias Andree [this message]
2002-12-11  9:20   ` Marc-Christian Petersen
2002-12-11  9:08 ` Ralf Hildebrandt
2002-12-11 16:07   ` Alan Cox
2002-12-11 15:34     ` Ralf Hildebrandt
2002-12-11 15:56       ` Ralf Hildebrandt
2002-12-11 17:29         ` Alan Cox
2002-12-15  9:56           ` Ralf Hildebrandt
2002-12-16  6:34             ` Alan Cox
2002-12-17  8:02   ` Ralf Hildebrandt
2002-12-11  9:37 ` Linux 2.4.21-pre1 compile failure: drivers/net/pcmcia/fmvj18x_cs.c Eyal Lebedinsky
2002-12-11 12:51   ` Adrian Bunk
2002-12-11 10:50 ` Linux 2.4.21-pre1 - unresolved symbols: scx200_gpio.o Eyal Lebedinsky
2002-12-12  1:35 ` Linux 2.4.21-pre1 IDE Erik Andersen
2002-12-12 13:10   ` Alan Cox
2002-12-13 14:23     ` Tomas Szepe
2002-12-13 17:42       ` Alan Cox
2002-12-11  3:58 Linux 2.4.21-pre1 Alex Davis
2002-12-12 15:12 ` Adrian Bunk
2002-12-11 17:51 Petr Vandrovec

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=20021211085553.GB8740@merlin.emma.line.org \
    --to=matthias.andree@gmx.de \
    --cc=akpm@digeo.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=marcelo@conectiva.com.br \
    /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

all inboxes | Powered by JetHome®