From: Matthew Wilcox <willy@linux.intel.com>
To: Valdis.Kletnieks@vt.edu
Cc: Matthew Wilcox <willy@linux.intel.com>,
Matthew Wilcox <matthew.r.wilcox@intel.com>,
linux-fsdevel@vger.kernel.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v11 00/21] Add support for NV-DIMMs to ext4
Date: Tue, 30 Sep 2014 12:08:41 -0400 [thread overview]
Message-ID: <20140930160841.GB5098@wil.cx> (raw)
In-Reply-To: <123795.1412088827@turing-police.cc.vt.edu>
On Tue, Sep 30, 2014 at 10:53:47AM -0400, Valdis.Kletnieks@vt.edu wrote:
> On Tue, 30 Sep 2014 10:48:54 -0400, Matthew Wilcox said:
>
> > No, it doesn't try to do that. Wouldn't you be better served with an
> > LD_PRELOAD that forces O_DIRECT on?
>
> Not when you don't want it on every file, and users are creating and
> deleting files once in a while. A chattr-like command is easier and
> more scalable than rebuilding the LD_PRELOAD every time the list of
> files gets changed....
The more I think about this, the more I think this is a bad idea.
When you have a file open with O_DIRECT, your I/O has to be done in
512-byte multiples, and it has to be aligned to 512-byte boundaries
in memory. If an unsuspecting application has O_DIRECT forced on it,
it isn't going to know to do that, and so all its I/Os will fail.
It'll also be horribly inefficient if a program has the file mmaped.
What problem are you really trying to solve? Some big files hogging
the page cache?
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
WARNING: multiple messages have this Message-ID (diff)
From: Matthew Wilcox <willy@linux.intel.com>
To: Valdis.Kletnieks@vt.edu
Cc: Matthew Wilcox <willy@linux.intel.com>,
Matthew Wilcox <matthew.r.wilcox@intel.com>,
linux-fsdevel@vger.kernel.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v11 00/21] Add support for NV-DIMMs to ext4
Date: Tue, 30 Sep 2014 12:08:41 -0400 [thread overview]
Message-ID: <20140930160841.GB5098@wil.cx> (raw)
In-Reply-To: <123795.1412088827@turing-police.cc.vt.edu>
On Tue, Sep 30, 2014 at 10:53:47AM -0400, Valdis.Kletnieks@vt.edu wrote:
> On Tue, 30 Sep 2014 10:48:54 -0400, Matthew Wilcox said:
>
> > No, it doesn't try to do that. Wouldn't you be better served with an
> > LD_PRELOAD that forces O_DIRECT on?
>
> Not when you don't want it on every file, and users are creating and
> deleting files once in a while. A chattr-like command is easier and
> more scalable than rebuilding the LD_PRELOAD every time the list of
> files gets changed....
The more I think about this, the more I think this is a bad idea.
When you have a file open with O_DIRECT, your I/O has to be done in
512-byte multiples, and it has to be aligned to 512-byte boundaries
in memory. If an unsuspecting application has O_DIRECT forced on it,
it isn't going to know to do that, and so all its I/Os will fail.
It'll also be horribly inefficient if a program has the file mmaped.
What problem are you really trying to solve? Some big files hogging
the page cache?
next prev parent reply other threads:[~2014-09-30 16:08 UTC|newest]
Thread overview: 169+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-25 20:33 [PATCH v11 00/21] Add support for NV-DIMMs to ext4 Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 01/21] axonram: Fix bug in direct_access Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 7:52 ` Mathieu Desnoyers
2014-10-16 7:52 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 02/21] block: Change direct_access calling convention Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 8:45 ` Mathieu Desnoyers
2014-10-16 8:45 ` Mathieu Desnoyers
2014-10-16 19:39 ` Matthew Wilcox
2014-10-16 19:39 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 03/21] mm: Fix XIP fault vs truncate race Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 8:56 ` Mathieu Desnoyers
2014-10-16 8:56 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 04/21] mm: Allow page fault handlers to perform the COW Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 9:12 ` Mathieu Desnoyers
2014-10-16 9:12 ` Mathieu Desnoyers
2014-10-16 19:48 ` Matthew Wilcox
2014-10-16 19:48 ` Matthew Wilcox
2014-10-17 15:35 ` Mathieu Desnoyers
2014-10-17 15:35 ` Mathieu Desnoyers
2014-10-18 17:22 ` Matthew Wilcox
2014-10-18 17:22 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 05/21] vfs,ext2: Introduce IS_DAX(inode) Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 9:35 ` Mathieu Desnoyers
2014-10-16 9:35 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 06/21] vfs: Add copy_to_iter(), copy_from_iter() and iov_iter_zero() Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 13:33 ` Mathieu Desnoyers
2014-10-16 13:33 ` Mathieu Desnoyers
2014-10-16 13:59 ` Matthew Wilcox
2014-10-16 13:59 ` Matthew Wilcox
2014-10-16 14:12 ` Mathieu Desnoyers
2014-10-16 14:12 ` Mathieu Desnoyers
2014-10-16 22:21 ` Matthew Wilcox
2014-10-16 22:21 ` Matthew Wilcox
2014-10-17 15:39 ` Mathieu Desnoyers
2014-10-17 15:39 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 07/21] dax,ext2: Replace XIP read and write with DAX I/O Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 9:50 ` Mathieu Desnoyers
2014-10-16 9:50 ` Mathieu Desnoyers
2014-10-16 19:51 ` Matthew Wilcox
2014-10-16 19:51 ` Matthew Wilcox
2014-10-16 22:33 ` Matthew Wilcox
2014-10-16 22:33 ` Matthew Wilcox
2014-10-17 15:52 ` Mathieu Desnoyers
2014-10-17 15:52 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 08/21] dax,ext2: Replace ext2_clear_xip_target with dax_clear_blocks Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 10:05 ` Mathieu Desnoyers
2014-10-16 10:05 ` Mathieu Desnoyers
2014-10-16 21:22 ` Matthew Wilcox
2014-10-16 21:22 ` Matthew Wilcox
2014-10-17 15:45 ` Mathieu Desnoyers
2014-10-17 15:45 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 09/21] dax,ext2: Replace the XIP page fault handler with the DAX page fault handler Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 10:20 ` Mathieu Desnoyers
2014-10-16 10:20 ` Mathieu Desnoyers
2014-10-16 21:29 ` Matthew Wilcox
2014-10-16 21:29 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 10/21] dax,ext2: Replace xip_truncate_page with dax_truncate_page Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 10:28 ` Mathieu Desnoyers
2014-10-16 10:28 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 11/21] dax: Replace XIP documentation with DAX documentation Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:08 ` Mathieu Desnoyers
2014-10-16 12:08 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 12/21] vfs: Remove get_xip_mem Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:14 ` Mathieu Desnoyers
2014-10-16 12:14 ` Mathieu Desnoyers
2014-10-16 21:44 ` Matthew Wilcox
2014-10-16 21:44 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 13/21] ext2: Remove ext2_xip_verify_sb() Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:18 ` Mathieu Desnoyers
2014-10-16 12:18 ` Mathieu Desnoyers
2014-10-16 21:45 ` Matthew Wilcox
2014-10-16 21:45 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 14/21] ext2: Remove ext2_use_xip Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:20 ` Mathieu Desnoyers
2014-10-16 12:20 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 15/21] ext2: Remove xip.c and xip.h Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:21 ` Mathieu Desnoyers
2014-10-16 12:21 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 16/21] vfs,ext2: Remove CONFIG_EXT2_FS_XIP and rename CONFIG_FS_XIP to CONFIG_FS_DAX Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:26 ` Mathieu Desnoyers
2014-10-16 12:26 ` Mathieu Desnoyers
2014-10-16 21:52 ` Matthew Wilcox
2014-10-16 21:52 ` Matthew Wilcox
2014-09-25 20:33 ` [PATCH v11 17/21] ext2: Remove ext2_aops_xip Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:29 ` Mathieu Desnoyers
2014-10-16 12:29 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 18/21] ext2: Get rid of most mentions of XIP in ext2 Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:32 ` Mathieu Desnoyers
2014-10-16 12:32 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 19/21] dax: Add dax_zero_page_range Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:38 ` Mathieu Desnoyers
2014-10-16 12:38 ` Mathieu Desnoyers
2014-10-16 22:01 ` Matthew Wilcox
2014-10-16 22:01 ` Matthew Wilcox
2014-10-17 15:49 ` Mathieu Desnoyers
2014-10-17 15:49 ` Mathieu Desnoyers
2014-10-18 17:41 ` Matthew Wilcox
2014-10-18 17:41 ` Matthew Wilcox
2014-10-18 21:16 ` Mathieu Desnoyers
2014-10-18 21:16 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 20/21] ext4: Add DAX functionality Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 12:56 ` Mathieu Desnoyers
2014-10-16 12:56 ` Mathieu Desnoyers
2014-10-16 22:16 ` Matthew Wilcox
2014-10-16 22:16 ` Matthew Wilcox
2014-10-17 15:42 ` Mathieu Desnoyers
2014-10-17 15:42 ` Mathieu Desnoyers
2014-09-25 20:33 ` [PATCH v11 21/21] brd: Rename XIP to DAX Matthew Wilcox
2014-09-25 20:33 ` Matthew Wilcox
2014-10-16 13:00 ` Mathieu Desnoyers
2014-10-16 13:00 ` Mathieu Desnoyers
2015-03-24 18:50 ` Matt Mullins
2015-03-24 18:50 ` Matt Mullins
2015-03-25 3:25 ` Dave Chinner
2015-03-25 3:25 ` Dave Chinner
2015-03-26 17:09 ` Should implementations of ->direct_access be allowed to sleep? Matthew Wilcox
2015-03-26 17:09 ` Matthew Wilcox
2015-03-26 19:32 ` Dave Chinner
2015-03-26 19:32 ` Dave Chinner
2015-03-29 8:02 ` Boaz Harrosh
2015-03-29 8:02 ` Boaz Harrosh
2015-03-29 9:13 ` Boaz Harrosh
2015-03-29 9:13 ` Boaz Harrosh
2014-09-25 20:47 ` [PATCH v11 00/21] Add support for NV-DIMMs to ext4 Matthew Wilcox
2014-09-25 20:47 ` Matthew Wilcox
2014-09-30 9:45 ` Valdis.Kletnieks
2014-09-30 14:48 ` Matthew Wilcox
2014-09-30 14:48 ` Matthew Wilcox
2014-09-30 14:53 ` Valdis.Kletnieks
2014-09-30 16:08 ` Matthew Wilcox [this message]
2014-09-30 16:08 ` Matthew Wilcox
2014-09-30 17:10 ` Zuckerman, Boris
2014-09-30 17:10 ` Zuckerman, Boris
2014-09-30 19:24 ` Matthew Wilcox
2014-09-30 19:24 ` Matthew Wilcox
2014-09-30 19:31 ` Zuckerman, Boris
2014-09-30 19:31 ` Zuckerman, Boris
2014-09-30 20:37 ` Valdis.Kletnieks
2014-09-30 21:25 ` Andreas Dilger
2014-09-30 21:52 ` Valdis.Kletnieks
2014-10-01 15:45 ` Jeff Moyer
2014-10-01 15:45 ` Jeff Moyer
2014-10-01 17:10 ` Valdis.Kletnieks
2014-10-01 17:17 ` Valdis.Kletnieks
2014-10-16 7:39 ` Mathieu Desnoyers
2014-10-16 7:39 ` Mathieu Desnoyers
2014-10-16 14:11 ` Matthew Wilcox
2014-10-16 14:11 ` Matthew Wilcox
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=20140930160841.GB5098@wil.cx \
--to=willy@linux.intel.com \
--cc=Valdis.Kletnieks@vt.edu \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=matthew.r.wilcox@intel.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 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.