From: "Jörn Engel" <joern@lazybastard.org>
To: Arnd Bergmann <arnd@arndb.de>
Cc: Ulrich Drepper <drepper@redhat.com>,
Anton Altaparmakov <aia21@cam.ac.uk>,
Christoph Hellwig <hch@infradead.org>,
Dave Kleikamp <shaggy@linux.vnet.ibm.com>,
Andrew Morton <akpm@linux-foundation.org>,
"Amit K. Arora" <aarora@linux.vnet.ibm.com>,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-ext4@vger.kernel.org, suparna@in.ibm.com, cmm@us.ibm.com,
alex@clusterfs.com, suzuki@in.ibm.com
Subject: Re: [RFC] Heads up on sys_fallocate()
Date: Mon, 5 Mar 2007 12:41:09 +0100 [thread overview]
Message-ID: <20070305114109.GA454@lazybastard.org> (raw)
In-Reply-To: <200703050136.37181.arnd@arndb.de>
On Mon, 5 March 2007 01:36:36 +0100, Arnd Bergmann wrote:
>
> Using the current glibc implementation on a compressed file system ideally
> should be a very expensive no-op because you won't actually allocate much
> space for a file when writing zeroes to it. You also don't benefit of a
> contiguous allocation in logfs, since flash has uniform seek times over
> all the medium.
>
> I'd suggest you implement posix_fallocate as an real nop and just return
> success without doing anything. You could also return ENOSPC in case
> the blocks requested by posix_fallocate don't fit on the medium without
> compression, but that is more or less just guesswork (like statfs is).
Quoting POSIX_FALLOCATE(3):
The function posix_fallocate() ensures that disk space is allocated for
the file referred to by the descriptor fd for the bytes in the range
starting at offset and continuing for len bytes. After a successful
call to posix_fallocate(), subsequent writes to bytes in the specified
range are guaranteed not to fail because of lack of disk space.
If the size of the file is less than offset+len, then the file is
increased to this size; otherwise the file size is left unchanged.
Afaics, the (main) purpose of this function is not to decrease
fragmentation but to ensure mmap() won't cause any problems because the
medium fills up. That problem exists for LogFS as well, once rw mmap()
is supported.
Simply returning success without doing anything would be a bug. -ENOSPC
is a better choice, but still a lame implementation. And falling back
on libc to write zeroes in a loop is an exercise in futility.
Does the allocation have to be persistent beyond lifetime of the file
descriptor? It would be fairly simple to support the write guarantee
while the file is open (or rather the inode remains cached) and drop it
afterwards.
Jörn
--
"[One] doesn't need to know [...] how to cause a headache in order
to take an aspirin."
-- Scott Culp, Manager of the Microsoft Security Response Center, 2001
next prev parent reply other threads:[~2007-03-05 11:41 UTC|newest]
Thread overview: 303+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20070117094658.GA17390@amitarora.in.ibm.com>
2007-02-07 7:48 ` Testing ext4 persistent preallocation patches for 64 bit features Amit K. Arora
2007-02-07 8:25 ` Mingming Cao
2007-02-07 10:36 ` Suparna Bhattacharya
2007-02-07 21:11 ` Andreas Dilger
2007-02-08 8:52 ` Amit K. Arora
2007-02-08 10:51 ` Amit K. Arora
[not found] ` <20070225022326.137b4875.akpm@linux-foundation.org>
2007-03-01 18:34 ` [RFC] Heads up on sys_fallocate() Amit K. Arora
2007-03-01 19:15 ` Eric Sandeen
2007-03-02 10:45 ` Andreas Dilger
2007-03-02 13:17 ` Dave Kleikamp
2007-03-01 20:23 ` Jeff Garzik
2007-03-01 20:31 ` Jeremy Allison
2007-03-01 21:14 ` Jeremy Fitzhardinge
2007-03-01 22:58 ` Alan
2007-03-01 22:05 ` Jeremy Fitzhardinge
2007-03-01 23:11 ` Alan
2007-03-01 22:15 ` Jeremy Fitzhardinge
2007-03-01 22:25 ` Andrew Morton
2007-03-01 22:40 ` Nathan Scott
2007-03-01 22:39 ` Eric Sandeen
2007-03-01 22:52 ` Andrew Morton
2007-03-02 18:28 ` Mingming Cao
2007-03-05 12:27 ` Jan Kara
2007-03-05 20:02 ` Mingming Cao
2007-03-06 7:28 ` Christoph Hellwig
2007-03-06 14:36 ` Ulrich Drepper
2007-03-06 14:47 ` Christoph Hellwig
2007-03-06 14:50 ` Jan Kara
2007-03-06 18:23 ` Eric Sandeen
2007-03-07 8:51 ` Jan Kara
2007-03-07 11:30 ` Jörn Engel
2007-03-06 16:46 ` Eric Sandeen
2007-03-13 23:46 ` David Chinner
2007-03-05 21:41 ` Eric Sandeen
2007-03-01 22:41 ` Anton Blanchard
2007-03-01 22:44 ` Dave Kleikamp
2007-03-01 22:59 ` Andrew Morton
2007-03-01 23:09 ` Dave Kleikamp
2007-03-02 13:41 ` Jan Engelhardt
2007-03-02 18:09 ` Mingming Cao
2007-03-02 7:09 ` Ulrich Drepper
2007-03-01 23:38 ` Christoph Hellwig
2007-03-03 22:45 ` Arnd Bergmann
2007-03-04 20:11 ` Anton Altaparmakov
2007-03-04 20:53 ` Arnd Bergmann
2007-03-04 22:38 ` Ulrich Drepper
2007-03-04 23:22 ` Anton Altaparmakov
2007-03-05 14:37 ` Theodore Tso
2007-03-05 15:07 ` Anton Altaparmakov
2007-03-05 15:15 ` Ulrich Drepper
2007-03-05 15:35 ` Christoph Hellwig
2007-03-05 16:01 ` Theodore Tso
2007-03-05 16:07 ` Ulrich Drepper
2007-03-05 0:16 ` Jörn Engel
2007-03-05 0:32 ` Anton Altaparmakov
2007-03-05 0:35 ` Anton Altaparmakov
2007-03-05 0:44 ` Arnd Bergmann
2007-03-05 11:49 ` Jörn Engel
2007-03-05 15:09 ` Ulrich Drepper
2007-03-05 0:36 ` Arnd Bergmann
2007-03-05 11:41 ` Jörn Engel [this message]
2007-03-05 15:08 ` Ulrich Drepper
2007-03-05 15:33 ` Jörn Engel
2007-03-05 15:48 ` Ulrich Drepper
2007-03-05 22:00 ` Eric Sandeen
2007-03-05 4:23 ` Christoph Hellwig
2007-03-05 13:18 ` Christoph Hellwig
2007-03-01 23:29 ` Eric Sandeen
2007-03-01 23:51 ` Christoph Hellwig
2007-03-01 23:36 ` Christoph Hellwig
2007-03-02 6:03 ` Badari Pulavarty
2007-03-02 6:16 ` Andrew Morton
2007-03-02 13:23 ` Dave Kleikamp
2007-03-02 15:29 ` Ulrich Drepper
2007-03-02 15:16 ` Eric Sandeen
2007-03-02 16:13 ` Badari Pulavarty
2007-03-02 17:01 ` Andrew Morton
2007-03-02 17:19 ` Eric Sandeen
2007-03-16 14:31 ` [RFC][PATCH] sys_fallocate() system call Amit K. Arora
2007-03-16 15:21 ` Heiko Carstens
2007-03-19 9:24 ` Amit K. Arora
2007-03-19 11:23 ` Heiko Carstens
2007-03-16 16:17 ` Heiko Carstens
2007-03-17 9:59 ` Paul Mackerras
2007-03-17 11:07 ` Matthew Wilcox
2007-03-17 14:30 ` Heiko Carstens
2007-03-17 14:38 ` Stephen Rothwell
2007-03-17 14:42 ` Stephen Rothwell
2007-03-17 11:10 ` Matthew Wilcox
2007-03-21 12:04 ` Amit K. Arora
2007-03-21 21:35 ` Chris Wedgwood
2007-03-29 11:51 ` Interface for the new fallocate() " Amit K. Arora
2007-03-29 16:35 ` Chris Wedgwood
2007-03-29 17:01 ` Jan Engelhardt
2007-03-29 17:18 ` linux-os (Dick Johnson)
2007-03-29 18:05 ` Jan Engelhardt
2007-03-29 18:37 ` Linus Torvalds
2007-03-30 7:00 ` Heiko Carstens
2007-03-29 17:10 ` Andrew Morton
2007-03-30 7:14 ` Jakub Jelinek
2007-03-30 8:39 ` Heiko Carstens
2007-03-30 9:15 ` Paul Mackerras
2007-04-05 11:26 ` Amit K. Arora
2007-04-05 11:44 ` Amit K. Arora
2007-04-05 15:50 ` Randy Dunlap
2007-04-06 9:58 ` Andreas Dilger
2007-04-17 12:55 ` Amit K. Arora
2007-04-18 13:06 ` Andreas Dilger
2007-04-20 13:51 ` Amit K. Arora
2007-04-20 14:59 ` Jakub Jelinek
2007-04-24 12:16 ` Amit K. Arora
2007-04-26 17:50 ` [PATCH 0/5] fallocate " Amit K. Arora
2007-04-26 18:03 ` [PATCH 1/5] fallocate() implementation in i86, x86_64 and powerpc Amit K. Arora
2007-05-04 4:29 ` Andrew Morton
2007-05-04 4:41 ` Paul Mackerras
2007-05-09 10:15 ` Suparna Bhattacharya
2007-05-09 10:50 ` Paul Mackerras
2007-05-09 11:10 ` Suparna Bhattacharya
2007-05-09 11:37 ` Paul Mackerras
2007-05-09 12:00 ` Martin Schwidefsky
2007-05-09 12:05 ` Amit K. Arora
2007-05-04 4:55 ` Andrew Morton
2007-05-04 6:07 ` David Chinner
2007-05-04 6:28 ` Andrew Morton
2007-05-04 6:56 ` Jakub Jelinek
2007-05-07 13:08 ` Ulrich Drepper
2007-05-04 7:27 ` David Chinner
2007-05-07 11:10 ` Amit K. Arora
2007-05-07 11:03 ` Amit K. Arora
2007-05-09 16:01 ` Amit K. Arora
2007-05-09 16:54 ` Andreas Dilger
2007-05-09 17:07 ` Mingming Cao
2007-05-10 0:59 ` David Chinner
2007-05-10 11:56 ` Amit K. Arora
2007-05-10 22:39 ` David Chinner
2007-05-11 11:03 ` Suparna Bhattacharya
2007-05-12 8:01 ` David Chinner
2007-06-12 6:16 ` Amit K. Arora
2007-06-12 8:11 ` David Chinner
2007-06-13 23:52 ` David Chinner
2007-06-14 9:14 ` Andreas Dilger
2007-06-14 12:04 ` David Chinner
2007-06-14 19:33 ` Andreas Dilger
2007-06-25 13:28 ` [PATCH 0/6][TAKE5] fallocate system call Amit K. Arora
2007-06-25 13:40 ` [PATCH 1/7][TAKE5] fallocate() implementation on i386, x86_64 and powerpc Amit K. Arora
2007-06-26 19:38 ` Heiko Carstens
2007-06-25 13:42 ` [PATCH 2/7][TAKE5] fallocate() on s390(x) Amit K. Arora
2007-06-26 15:15 ` Heiko Carstens
2007-06-25 13:43 ` [PATCH 3/7][TAKE5] fallocate() on ia64 Amit K. Arora
2007-06-25 13:45 ` [PATCH 4/7][TAKE5] support new modes in fallocate Amit K. Arora
2007-06-25 15:03 ` Amit K. Arora
2007-06-25 21:46 ` Andreas Dilger
2007-06-26 10:32 ` Amit K. Arora
2007-06-26 15:34 ` Andreas Dilger
2007-06-26 19:09 ` Amit K. Arora
2007-06-26 23:18 ` David Chinner
2007-06-28 18:19 ` Amit K. Arora
2007-06-28 23:39 ` Nathan Scott
2007-06-29 1:03 ` David Chinner
2007-06-30 10:21 ` Christoph Hellwig
2007-06-30 16:52 ` Andreas Dilger
2007-07-03 10:08 ` Amit K. Arora
2007-07-03 10:31 ` Christoph Hellwig
2007-07-03 11:46 ` Amit K. Arora
2007-07-04 5:37 ` Timothy Shimmin
2007-07-11 9:04 ` Christoph Hellwig
2007-07-11 9:03 ` Christoph Hellwig
2007-07-12 7:28 ` Suparna Bhattacharya
2007-07-12 8:26 ` Amit K. Arora
2007-07-12 14:40 ` Andreas Dilger
2007-07-12 13:13 ` David Chinner
2007-07-12 14:15 ` Amit K. Arora
2007-07-01 22:55 ` David Chinner
2007-07-02 11:47 ` Amit K. Arora
2007-07-11 9:05 ` Christoph Hellwig
2007-06-26 23:14 ` David Chinner
2007-06-27 3:49 ` Andreas Dilger
2007-06-27 13:36 ` David Chinner
2007-06-27 23:28 ` Nathan Scott
2007-06-28 0:39 ` David Chinner
2007-06-28 0:53 ` Nathan Scott
2007-06-30 10:26 ` Christoph Hellwig
2007-06-25 21:52 ` Andreas Dilger
2007-06-26 10:45 ` Amit K. Arora
2007-06-26 15:42 ` Andreas Dilger
2007-06-26 19:12 ` Amit K. Arora
2007-06-26 23:32 ` David Chinner
2007-06-26 23:26 ` David Chinner
2007-06-25 13:48 ` [PATCH 5/7][TAKE5] ext4: fallocate support in ext4 Amit K. Arora
2007-06-25 13:49 ` [PATCH 6/7][TAKE5] ext4: write support for preallocated blocks Amit K. Arora
2007-06-25 13:50 ` [PATCH 7/7][TAKE5] ext4: support new modes Amit K. Arora
2007-06-25 21:56 ` Andreas Dilger
2007-06-26 12:07 ` Amit K. Arora
2007-06-26 16:14 ` Andreas Dilger
2007-06-26 19:29 ` Amit K. Arora
2007-06-27 0:04 ` David Chinner
2007-06-28 18:07 ` Amit K. Arora
2007-06-26 23:15 ` [PATCH 0/6][TAKE5] fallocate system call David Chinner
2007-06-28 9:55 ` Andrew Morton
2007-06-28 17:36 ` Mingming Cao
2007-06-28 17:57 ` Amit K. Arora
2007-06-28 18:33 ` Andrew Morton
2007-06-28 18:45 ` Dave Kleikamp
2007-06-28 18:57 ` Jeff Garzik
2007-06-29 7:20 ` Christoph Hellwig
2007-06-29 13:56 ` Theodore Tso
2007-06-29 14:29 ` Jeff Garzik
2007-06-29 17:42 ` Theodore Tso
2007-06-29 15:50 ` Mingming Caoc
2007-06-29 20:57 ` Andrew Morton
2007-07-01 7:35 ` Ext4 patches for 2.6.22-rc6 Mingming Cao
2007-06-28 20:34 ` [PATCH 0/6][TAKE5] fallocate system call Andreas Dilger
2007-06-30 10:14 ` [PATCH 1/5] fallocate() implementation in i86, x86_64 and powerpc Christoph Hellwig
2007-04-26 18:07 ` [PATCH 2/5] fallocate() on s390 Amit K. Arora
2007-04-26 18:11 ` [PATCH 3/5] ext4: Extent overlap bugfix Amit K. Arora
2007-05-04 4:30 ` Andrew Morton
2007-05-07 11:46 ` Amit K. Arora
2007-04-26 18:13 ` [PATCH 4/5] ext4: fallocate support in ext4 Amit K. Arora
2007-05-04 4:31 ` Andrew Morton
2007-05-07 11:37 ` Andreas Dilger
2007-05-07 20:58 ` Andrew Morton
2007-05-07 22:21 ` Andreas Dilger
2007-05-07 22:38 ` Andrew Morton
2007-05-07 23:14 ` Theodore Tso
2007-05-07 23:31 ` Andrew Morton
2007-05-08 0:30 ` Mingming Cao
2007-05-07 23:02 ` Jeff Garzik
2007-05-07 23:36 ` Theodore Tso
2007-05-08 1:07 ` Andreas Dilger
2007-05-08 1:25 ` Jeff Garzik
2007-05-08 0:00 ` Mingming Cao
2007-05-08 0:15 ` Andrew Morton
2007-05-08 0:41 ` Mingming Cao
2007-05-08 1:43 ` Theodore Tso
2007-05-08 16:52 ` Andreas Dilger
2007-05-08 17:46 ` Mingming Cao
2007-05-14 13:34 ` Jan Kara
2007-05-07 12:07 ` Amit K. Arora
2007-05-07 15:24 ` Dave Kleikamp
2007-05-08 10:52 ` Amit K. Arora
2007-05-08 14:47 ` Dave Kleikamp
2007-04-26 18:16 ` [PATCH 5/5] ext4: write support for preallocated blocks/extents Amit K. Arora
2007-05-04 4:32 ` Andrew Morton
2007-05-07 12:11 ` Amit K. Arora
2007-05-07 12:40 ` Pekka Enberg
2007-05-07 13:04 ` Amit K. Arora
2007-04-27 12:10 ` [PATCH 0/5] fallocate system call Heiko Carstens
2007-04-27 14:43 ` Jörn Engel
2007-04-27 17:46 ` Heiko Carstens
2007-04-27 20:42 ` Chris Wedgwood
2007-04-30 0:47 ` David Chinner
2007-04-30 3:09 ` [PATCH] ia64 fallocate syscall David Chinner
2007-04-30 3:11 ` [PATCH] XFS ->fallocate() support David Chinner
2007-04-30 3:14 ` [PATCH] Add preallocation beyond EOF to fallocate David Chinner
2007-04-30 5:25 ` [PATCH 0/5] fallocate system call Chris Wedgwood
2007-04-30 5:56 ` David Chinner
2007-04-30 6:01 ` Chris Wedgwood
2007-05-02 12:53 ` Amit K. Arora
2007-05-03 10:34 ` Andreas Dilger
2007-05-14 13:29 ` [PATCH 0/5][TAKE2] " Amit K. Arora
[not found] ` <20070514142820.GA31468@amitarora.in.ibm.com>
2007-05-14 14:45 ` [PATCH 1/5][TAKE2] fallocate() implementation on i86, x86_64 and powerpc Amit K. Arora
2007-05-14 23:44 ` Stephen Rothwell
2007-05-15 13:23 ` Amit K. Arora
2007-05-14 14:48 ` [PATCH 2/5][TAKE2] fallocate() on s390 Amit K. Arora
2007-05-14 15:33 ` [PATCH 2/5][TAKE2] fallocate() on s390 - glibc wrapper Amit K. Arora
2007-05-14 14:50 ` [PATCH 3/5][TAKE2] ext4: Extent overlap bugfix Amit K. Arora
2007-05-14 14:52 ` [PATCH 4/5][TAKE2] ext4: fallocate support in ext4 Amit K. Arora
2007-05-14 14:54 ` [PATCH 5/5][TAKE2] ext4: write support for preallocated blocks Amit K. Arora
2007-05-15 6:31 ` [PATCH 0/5][TAKE2] fallocate system call Andreas Dilger
2007-05-15 12:40 ` Amit K. Arora
2007-05-15 19:37 ` [PATCH 0/5][TAKE3] " Amit K. Arora
[not found] ` <20070515195421.GA2948@amitarora.in.ibm.com>
2007-05-15 20:03 ` [PATCH 1/5][TAKE3] fallocate() implementation on i86, x86_64 and powerpc Amit K. Arora
2007-05-16 0:42 ` Mingming Cao
2007-05-16 12:31 ` Amit K. Arora
2007-05-16 3:16 ` David Chinner
2007-05-16 12:21 ` Dave Kleikamp
2007-05-16 12:37 ` Amit K. Arora
2007-05-16 23:40 ` David Chinner
2007-05-17 12:10 ` Dave Kleikamp
2007-05-17 12:28 ` Amit K. Arora
2007-05-15 20:10 ` [PATCH 2/5][TAKE3] fallocate() on s390 Amit K. Arora
2007-05-15 20:13 ` [PATCH 3/5][TAKE3] ext4: Extent overlap bugfix Amit K. Arora
2007-05-15 20:16 ` [PATCH 4/5][TAKE3] ext4: fallocate support in ext4 Amit K. Arora
2007-05-15 20:18 ` [PATCH 5/5][TAKE3] ext4: write support for preallocated blocks Amit K. Arora
2007-05-15 23:52 ` [PATCH 0/5][TAKE3] fallocate system call Mingming Cao
2007-05-17 14:11 ` [PATCH 0/6][TAKE4] " Amit K. Arora
[not found] ` <20070517141458.GA26641@amitarora.in.ibm.com>
2007-05-17 14:23 ` [PATCH 1/6][TAKE4] fallocate() implementation on i86, x86_64 and powerpc Amit K. Arora
2007-05-17 14:25 ` [PATCH 2/6][TAKE4] fallocate() on s390 Amit K. Arora
2007-05-17 14:25 ` [PATCH 3/6][TAKE4] fallocate() on ia64 Amit K. Arora
2007-05-17 14:26 ` [PATCH 4/6][TAKE4] ext4: Extent overlap bugfix Amit K. Arora
2007-05-17 14:29 ` [PATCH 5/6][TAKE4] ext4: fallocate support in ext4 Amit K. Arora
2007-05-17 14:30 ` [PATCH 6/6][TAKE4] ext4: write support for preallocated blocks Amit K. Arora
2007-05-19 6:44 ` [PATCH 0/6][TAKE4] fallocate system call Andrew Morton
2007-05-21 5:24 ` Mingming Cao
2007-03-30 7:19 ` Interface for the new fallocate() " Heiko Carstens
2007-03-30 9:15 ` Paul Mackerras
[not found] ` <17932.54606.323431.491736@cargo.ozlabs.ibm.com>
2007-03-30 10:44 ` Jörn Engel
2007-03-30 12:55 ` Heiko Carstens
2007-04-09 13:01 ` Paul Mackerras
2007-04-09 16:34 ` Jörn Engel
2007-03-17 5:33 ` [RFC][PATCH] sys_fallocate() " Stephen Rothwell
2007-03-19 9:30 ` Amit K. Arora
2007-03-17 14:53 ` Russell King
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=20070305114109.GA454@lazybastard.org \
--to=joern@lazybastard.org \
--cc=aarora@linux.vnet.ibm.com \
--cc=aia21@cam.ac.uk \
--cc=akpm@linux-foundation.org \
--cc=alex@clusterfs.com \
--cc=arnd@arndb.de \
--cc=cmm@us.ibm.com \
--cc=drepper@redhat.com \
--cc=hch@infradead.org \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=shaggy@linux.vnet.ibm.com \
--cc=suparna@in.ibm.com \
--cc=suzuki@in.ibm.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).