linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Garzik <jeff@garzik.org>
To: Mark Lord <lkml@rtr.ca>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
	Ray Lee <ray-lk@madrabbit.org>, Hua Zhong <hzhong@gmail.com>,
	Theodore Tso <tytso@mit.edu>, Jens Axboe <jens.axboe@oracle.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 0/8][RFC] IO latency/throughput fixes
Date: Tue, 07 Apr 2009 15:45:31 -0400	[thread overview]
Message-ID: <49DBAD5B.7010107@garzik.org> (raw)
In-Reply-To: <49DBA87F.6070501@rtr.ca>

Mark Lord wrote:
> Linus Torvalds wrote:
>>
>> On Tue, 7 Apr 2009, Mark Lord wrote:
>>> What happens with ext3 "writeback", and ext4 "whatever",
>>> when one does the quickie reboot method:
>>>
>>>   ALT-SYSRQ-S ALT-SYSRQ-U ALT-SYSRQ-S ALT-SYSRQ-B
>>>
>>> ???
>>
>> Since 's' syncs (I think 'u' does too, as part of making things 
>> read-only), the data blocks will be on disk after the boot regardless 
>> of any other ordering.
> ..
> 
> I was thinking more about delayed allocation in ext4, though.
> If it hasn't allocated the blocks, then sync() has nothing to write out.
> Or do they have hooks into the block layer to force alloc/commit when
> somebody does a sync() ??

sync(2) doesn't just sync dirty buffers...  it sync's inodes, which 
pokes the filesystem to do something intelligent, perhaps triggering (a) 
write-out of data, (b) write-out of zeroed blocks, or (c) annotation in 
filesystem metadata that certain blocks are allocated, but not initialized.

	Jeff




  reply	other threads:[~2009-04-07 19:46 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-04-06 12:48 [PATCH 0/8][RFC] IO latency/throughput fixes Jens Axboe
2009-04-06 12:48 ` [PATCH 1/8] block: change the request allocation/congestion logic to be sync/async based Jens Axboe
2009-04-06 12:48 ` [PATCH 2/8] Add WRITE_SYNC_PLUG and SWRITE_SYNC_PLUG Jens Axboe
2009-04-06 12:48 ` [PATCH 3/8] block: fsync_buffers_list() should use SWRITE_SYNC_PLUG Jens Axboe
2009-04-06 12:48 ` [PATCH 4/8] jbd: use WRITE_SYNC_PLUG instead of WRITE_SYNC Jens Axboe
2009-04-06 12:48 ` [PATCH 5/8] jbd2: " Jens Axboe
2009-04-06 12:48 ` [PATCH 6/8] block: enabling plugging on SSD devices that don't do queuing Jens Axboe
2009-04-06 12:48 ` [PATCH 7/8] block: Add flag for telling the IO schedulers NOT to anticipate more IO Jens Axboe
2009-04-06 12:48 ` [PATCH 8/8] block: switch sync_dirty_buffer() over to WRITE_SYNC Jens Axboe
2009-04-06 13:04 ` [PATCH 0/8][RFC] IO latency/throughput fixes Jens Axboe
2009-04-06 13:13   ` Jens Axboe
2009-04-06 15:37   ` Linus Torvalds
2009-04-06 16:57     ` Jens Axboe
2009-04-07  3:28     ` Chris Mason
2009-04-06 15:04 ` Linus Torvalds
2009-04-06 15:10   ` Jens Axboe
2009-04-06 15:45     ` Linus Torvalds
2009-04-06 17:01       ` Jens Axboe
2009-04-06 18:31       ` Theodore Tso
2009-04-06 19:57         ` Linus Torvalds
2009-04-06 20:10           ` Linus Torvalds
2009-04-06 21:26             ` Theodore Tso
2009-04-06 20:12           ` Hua Zhong
2009-04-06 20:20             ` Linus Torvalds
2009-04-06 21:19             ` Theodore Tso
2009-04-06 21:35               ` Hua Zhong
2009-04-06 22:04                 ` Ray Lee
2009-04-06 22:17                   ` Linus Torvalds
2009-04-06 23:10                     ` Linus Torvalds
2009-04-07  7:51                       ` Geert Uytterhoeven
2009-04-07 10:36                         ` Ingo Molnar
2009-04-07 14:10                           ` Diego Calleja
2009-04-08 12:04                             ` Ingo Molnar
2009-04-08 12:56                           ` Denys Vlasenko
2009-04-08 13:27                             ` Ingo Molnar
2009-04-07 13:35                       ` Mark Lord
2009-04-07 14:33                         ` Linus Torvalds
2009-04-07 19:24                           ` Mark Lord
2009-04-07 19:45                             ` Jeff Garzik [this message]
2009-04-07 20:53                           ` Mike Galbraith
2009-04-09  2:40                       ` Eric Sandeen
2009-04-09 14:01                         ` Ric Wheeler
2009-04-06 22:25                   ` Hua Zhong
2009-04-06 22:48                     ` Ray Lee
2009-04-06 22:52                       ` Hua Zhong
2009-04-06 23:19                       ` Alan Cox
2009-04-07  3:52               ` Chris Mason
2009-04-07  4:13                 ` Trenton D. Adams
2009-04-07  4:27                   ` Linus Torvalds
2009-04-07  4:48                     ` Trenton D. Adams
2009-04-07  5:02                       ` Linus Torvalds
2009-04-07  5:23                         ` Hua Zhong
2009-04-07  6:27                         ` Trenton D. Adams

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=49DBAD5B.7010107@garzik.org \
    --to=jeff@garzik.org \
    --cc=hzhong@gmail.com \
    --cc=jens.axboe@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lkml@rtr.ca \
    --cc=ray-lk@madrabbit.org \
    --cc=torvalds@linux-foundation.org \
    --cc=tytso@mit.edu \
    /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).