From: "Jörn Engel" <joern@logfs.org>
To: Arnd Bergmann <arnd@arndb.de>
Cc: linux-fsdevel@vger.kernel.org, linux-mtd@lists.infradead.org,
linux-kernel@vger.kernel.org
Subject: Re: [patch 4/15] fs/logfs/compr.c
Date: Fri, 11 Apr 2008 12:41:28 +0200 [thread overview]
Message-ID: <20080411104128.GG25462@logfs.org> (raw)
In-Reply-To: <200804101613.48465.arnd@arndb.de>
On Thu, 10 April 2008 16:13:47 +0200, Arnd Bergmann wrote:
>
> How much time is spent in the compression code? Would it be helpful
> to have a percpu stream and do preempt_disable around it instead
> of the mutex?
>
> That would improve parallelism and cache-locality of the compression,
> at the cost of adding a source of latency to preemptible kernels.
Several years back (before logfs) I was planning to do just that. It
turned out that ipsec doesn't work in such a fashion, because each
connection is a stream. And the two block-oriented users, jffs2 and
cramfs just weren't worth it.
Might be a good idea by now. I guess it should become self-contained
code, so that jffs2, cramfs and ubifs can use it as well.
Jörn
--
Write programs that do one thing and do it well. Write programs to work
together. Write programs to handle text streams, because that is a
universal interface.
-- Doug MacIlroy
WARNING: multiple messages have this Message-ID (diff)
From: "Jörn Engel" <joern@logfs.org>
To: Arnd Bergmann <arnd@arndb.de>
Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
linux-mtd@lists.infradead.org
Subject: Re: [patch 4/15] fs/logfs/compr.c
Date: Fri, 11 Apr 2008 12:41:28 +0200 [thread overview]
Message-ID: <20080411104128.GG25462@logfs.org> (raw)
In-Reply-To: <200804101613.48465.arnd@arndb.de>
On Thu, 10 April 2008 16:13:47 +0200, Arnd Bergmann wrote:
>
> How much time is spent in the compression code? Would it be helpful
> to have a percpu stream and do preempt_disable around it instead
> of the mutex?
>
> That would improve parallelism and cache-locality of the compression,
> at the cost of adding a source of latency to preemptible kernels.
Several years back (before logfs) I was planning to do just that. It
turned out that ipsec doesn't work in such a fashion, because each
connection is a stream. And the two block-oriented users, jffs2 and
cramfs just weren't worth it.
Might be a good idea by now. I guess it should become self-contained
code, so that jffs2, cramfs and ubifs can use it as well.
Jörn
--
Write programs that do one thing and do it well. Write programs to work
together. Write programs to handle text streams, because that is a
universal interface.
-- Doug MacIlroy
next prev parent reply other threads:[~2008-04-11 10:41 UTC|newest]
Thread overview: 85+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-01 18:13 [patch 0/15] LogFS take five joern
2008-04-01 18:13 ` joern
2008-04-01 18:13 ` [patch 14/15] fs/logfs/dev_bdev.c joern
2008-04-01 18:13 ` [patch 9/15] fs/logfs/journal.c joern
2008-04-01 18:13 ` [patch 11/15] fs/logfs/readwrite.c joern
2008-04-01 18:13 ` [patch 5/15] fs/logfs/dir.c joern
2008-04-04 6:22 ` Kyungmin Park
2008-04-04 6:22 ` Kyungmin Park
2008-04-01 18:13 ` [patch 6/15] fs/logfs/file.c joern
2008-04-01 18:13 ` [patch 3/15] fs/logfs/logfs.h joern
2008-04-08 0:35 ` Arnd Bergmann
2008-04-08 0:35 ` Arnd Bergmann
2008-04-08 0:35 ` Arnd Bergmann
2008-04-08 9:41 ` Jörn Engel
2008-04-08 9:41 ` Jörn Engel
2008-04-08 9:41 ` Jörn Engel
2008-04-01 18:13 ` [patch 8/15] fs/logfs/inode.c joern
2008-04-04 6:57 ` Kyungmin Park
2008-04-04 6:57 ` Kyungmin Park
2008-04-07 11:12 ` Jörn Engel
2008-04-07 11:12 ` Jörn Engel
2008-04-07 11:12 ` Jörn Engel
2008-04-01 18:13 ` [patch 15/15] fs/logfs/dev_mtd.c joern
2008-04-01 18:13 ` [patch 2/15] fs/logfs/logfs_abi.h joern
2008-04-08 0:24 ` Arnd Bergmann
2008-04-08 0:24 ` Arnd Bergmann
2008-04-08 0:24 ` Arnd Bergmann
2008-04-08 9:39 ` Jörn Engel
2008-04-08 9:39 ` Jörn Engel
2008-04-08 9:39 ` Jörn Engel
2008-04-08 21:52 ` Andres Salomon
2008-04-08 21:52 ` Andres Salomon
2008-04-09 12:08 ` Jörn Engel
2008-04-09 12:08 ` Jörn Engel
2008-04-09 12:08 ` Jörn Engel
2008-04-01 18:13 ` [patch 12/15] fs/logfs/segment.c joern
2008-04-01 18:13 ` [patch 4/15] fs/logfs/compr.c joern
2008-04-10 14:13 ` Arnd Bergmann
2008-04-10 14:13 ` Arnd Bergmann
2008-04-10 14:13 ` Arnd Bergmann
2008-04-11 10:41 ` Jörn Engel [this message]
2008-04-11 10:41 ` Jörn Engel
2008-04-01 18:13 ` [patch 13/15] fs/logfs/super.c joern
2008-04-01 18:13 ` [patch 1/15] Makefiles and Kconfig joern
2008-04-07 8:28 ` Christian Borntraeger
2008-04-07 8:28 ` Christian Borntraeger
2008-04-07 8:40 ` Jörn Engel
2008-04-07 8:40 ` Jörn Engel
2008-04-07 8:40 ` Jörn Engel
2008-04-08 0:30 ` Arnd Bergmann
2008-04-08 0:30 ` Arnd Bergmann
2008-04-08 8:33 ` Jörn Engel
2008-04-08 8:33 ` Jörn Engel
2008-04-08 8:33 ` Jörn Engel
2008-04-08 13:41 ` Arnd Bergmann
2008-04-08 13:41 ` Arnd Bergmann
2008-04-08 13:41 ` Arnd Bergmann
2008-04-08 13:52 ` Jörn Engel
2008-04-08 13:52 ` Jörn Engel
2008-04-08 13:52 ` Jörn Engel
2008-04-01 18:13 ` [patch 7/15] fs/logfs/gc.c joern
2008-04-01 18:13 ` [patch 10/15] fs/logfs/memtree.c joern
2008-04-10 14:07 ` Arnd Bergmann
2008-04-10 14:07 ` Arnd Bergmann
2008-04-10 14:07 ` Arnd Bergmann
2008-04-11 10:37 ` Jörn Engel
2008-04-11 10:37 ` Jörn Engel
2008-04-11 10:37 ` Jörn Engel
2008-04-03 17:13 ` [patch 0/15] LogFS take five^Wsix Jörn Engel
2008-04-03 17:13 ` Jörn Engel
2008-04-03 17:13 ` Jörn Engel
2008-04-04 11:46 ` [patch 0/15] LogFS take five Jens Axboe
2008-04-04 11:46 ` Jens Axboe
2008-04-07 8:22 ` Jörn Engel
2008-04-07 8:22 ` Jörn Engel
2008-04-07 8:22 ` Jörn Engel
2008-04-07 8:28 ` Jens Axboe
2008-04-07 8:28 ` Jens Axboe
2008-04-07 8:28 ` Jens Axboe
2008-04-07 9:10 ` Jörn Engel
2008-04-07 9:10 ` Jörn Engel
2008-04-07 9:10 ` Jörn Engel
2008-04-07 9:17 ` Jens Axboe
2008-04-07 9:17 ` Jens Axboe
2008-04-07 9:17 ` Jens Axboe
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=20080411104128.GG25462@logfs.org \
--to=joern@logfs.org \
--cc=arnd@arndb.de \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mtd@lists.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 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.