mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
To: Christoph Hellwig <hch@infradead.org>,
	lkml <linux-kernel@vger.kernel.org>,
	David Miller <davem@davemloft.net>,
	Ulrich Drepper <drepper@redhat.com>,
	netdev <netdev@vger.kernel.org>
Subject: Re: [3/4] kevent: AIO, aio_sendfile() implementation.
Date: Wed, 26 Jul 2006 14:19:21 +0400	[thread overview]
Message-ID: <20060726101919.GB2715@2ka.mipt.ru> (raw)
In-Reply-To: <20060726100431.GA7518@infradead.org>

On Wed, Jul 26, 2006 at 11:04:31AM +0100, Christoph Hellwig (hch@infradead.org) wrote:
> On Wed, Jul 26, 2006 at 01:18:15PM +0400, Evgeniy Polyakov wrote:
> > 
> > This patch includes asynchronous propagation of file's data into VFS
> > cache and aio_sendfile() implementation.
> > Network aio_sendfile() works lazily - it asynchronously populates pages
> > into the VFS cache (which can be used for various tricks with adaptive
> > readahead) and then uses usual ->sendfile() callback.
> 
> And please don't base this on sendfile.  Please make the splice infrastructure
> aynschronous without duplicating all the code but rather make the existing
> code aynch and the existing synchronous call wait on them to finish, similar
> to how we handle async/sync direct I/O.  And to be honest, I don't think
> adding all this code is acceptable if it can't replace the existing aio
> code while keeping the interface.  So while you interface looks pretty
> sane the implementation needs a lot of work still :)

Kevent was created quite before splice and friends, so I used what there
were :)

I stopped to work on AIO, since neither existing, nor mine
implementation were able to outperform sync speeds (one of the major problems
in my implementation is get_user_pages() overhead, which can be
completely eliminated with physical memory allocation being done in
advance in userspace, like Ulrich described).
My personal opinion on existing AIO is that it is not the right design.
Benjamin LaHaise agree with me (if I understood him right), but he
failed to move AIO outside repeated-call model (2.4 had state machine
based one, and out-of-the tree 2.6 patches have that design too).
In theory existing AIO (with all posix userspace API) can be replaced
with kevent (it will even take less space), but I would present it as a
TODO item, since kevent itself has nothing to do with AIO.

Kevent is a generic event processing mechanism, AIO, network AIO and all
others are just kernel users of it's functionality.

-- 
	Evgeniy Polyakov

  parent reply	other threads:[~2006-07-26 10:20 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-07-25 19:23 async network I/O, event channels, etc Ulrich Drepper
2006-07-25 20:25 ` Stephen Hemminger
2006-07-25 21:24   ` Ulrich Drepper
2006-07-25 22:01 ` David Miller
2006-07-25 22:55   ` Nicholas Miell
2006-07-26  6:28   ` Evgeniy Polyakov
2006-07-26  9:18     ` [0/4] kevent: generic event processing subsystem Evgeniy Polyakov
2006-07-26  9:18       ` [1/4] kevent: core files Evgeniy Polyakov
2006-07-26  9:18         ` [2/4] kevent: network AIO, socket notifications Evgeniy Polyakov
2006-07-26  9:18           ` [3/4] kevent: AIO, aio_sendfile() implementation Evgeniy Polyakov
2006-07-26  9:18             ` [4/4] kevent: poll/select() notifications. Timer notifications Evgeniy Polyakov
2006-07-26 10:00             ` [3/4] kevent: AIO, aio_sendfile() implementation Christoph Hellwig
2006-07-26 10:08               ` Evgeniy Polyakov
2006-07-26 10:13                 ` Christoph Hellwig
2006-07-26 10:25                   ` Evgeniy Polyakov
2006-07-26 10:04             ` Christoph Hellwig
2006-07-26 10:12               ` David Miller
2006-07-26 10:15                 ` Christoph Hellwig
2006-07-26 20:21                   ` Phillip Susi
2006-07-26 14:14                 ` Avi Kivity
2006-07-26 10:19               ` Evgeniy Polyakov [this message]
2006-07-26 10:30                 ` Christoph Hellwig
2006-07-26 14:28                   ` Ulrich Drepper
2006-07-26 16:22                     ` Badari Pulavarty
2006-07-27  6:49                       ` Sébastien Dugué
2006-07-27 15:28                         ` Badari Pulavarty
2006-07-27 18:14                           ` Zach Brown
2006-07-27 18:29                             ` Badari Pulavarty
2006-07-27 18:44                               ` Ulrich Drepper
2006-07-27 21:02                                 ` Badari Pulavarty
2006-07-28  7:31                                   ` Sébastien Dugué
2006-07-28 12:58                                   ` Sébastien Dugué
2006-08-11 19:45                                     ` Ulrich Drepper
2006-08-12 18:29                                       ` Kernel patches enabling better POSIX AIO (Was Re: [3/4] kevent: AIO, aio_sendfile) Suparna Bhattacharya
2006-08-12 19:10                                         ` Ulrich Drepper
2006-08-12 19:28                                           ` Jakub Jelinek
2006-09-04 14:37                                             ` Sébastien Dugué
2006-08-14  7:02                                           ` Suparna Bhattacharya
2006-08-14 16:38                                             ` Ulrich Drepper
2006-08-15  2:06                                               ` Nicholas Miell
2006-09-04 14:36                                           ` Sébastien Dugué
2006-09-04 14:28                                         ` Sébastien Dugué
2006-07-28  7:29                                 ` [3/4] kevent: AIO, aio_sendfile() implementation Sébastien Dugué
2006-07-31 10:11                                 ` Suparna Bhattacharya
2006-07-28  7:26                           ` Sébastien Dugué
2006-07-26 10:31         ` [1/4] kevent: core files Andrew Morton
2006-07-26 10:37           ` Evgeniy Polyakov
2006-07-26 10:44         ` Evgeniy Polyakov
2006-07-27  6:10     ` async network I/O, event channels, etc David Miller
2006-07-27  7:49       ` Evgeniy Polyakov
2006-07-27  8:02         ` David Miller
2006-07-27  8:09           ` Jens Axboe
2006-07-27  8:11             ` Jens Axboe
2006-07-27  8:20               ` David Miller
2006-07-27  8:29                 ` Jens Axboe
2006-07-27  8:37                   ` David Miller
2006-07-27  8:39                     ` Jens Axboe
2006-07-27  8:58           ` Evgeniy Polyakov
2006-07-27  9:31             ` David Miller
2006-07-27  9:37               ` Evgeniy Polyakov

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=20060726101919.GB2715@2ka.mipt.ru \
    --to=johnpol@2ka.mipt.ru \
    --cc=davem@davemloft.net \
    --cc=drepper@redhat.com \
    --cc=hch@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.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 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®