9fans - fans of the OS Plan 9 from Bell Labs
 help / color / mirror / Atom feed
From: erik quanstrom <quanstro@coraid.com>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] XML
Date: Fri, 25 May 2007 18:11:35 -0400	[thread overview]
Message-ID: <cc31f831e72570b452e31604265a25af@coraid.com> (raw)
In-Reply-To: <13426df10705251115g494cb819k881854e8a7952fe8@mail.gmail.com>

On Fri May 25 14:15:01 EDT 2007, rminnich@gmail.com wrote:
> It hit me that the dom0 could export its tcp stack to dom1 as a
> paravirtual device. you could bypass the silly virtual enet emulation
> that way. Your /net would go right to the tcp, not via some odd
> pseudo-device. That would save some delay and overhead, and, not
> incidentally, would make my mp3 player smoother in dom1 ...

muxing a /net heirarchy from plan9 via 9p is certainly the most flexable
way to go.  this is a very good idea.  (i assume that you really didn't
mean importing tcp only from dom0.)

however, there are some hardware models that could be very efficient
to emulate in /a/ hypervisor.  the myricom 10gbe moves all data
by dma.  the only mmio is for commands, send descriptors (16 bytes)
and receive descriptors (8 bytes).  assuming the guest doesn't have
real pci window addresses, dom0 would only need to rewrite the
descriptors and the card could still dma directly from the guest.

it's significantly less general than the 9p way of doing things, but
it could make zero-copy sends (and zero-copy tco, if you care about
such things) doable.

i don't know enough about xen's inner workings to know if this would
work.

- erik

p.s.  perhaps the way to put generality back in is to marshal 9p between
dom0 and guests with the same trick.  use an pseudo-mmio descriptor for
the 9p "header" which points to the memory to be transfered.


  parent reply	other threads:[~2007-05-25 22:11 UTC|newest]

Thread overview: 77+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-05-21 16:57 ron minnich
2007-05-21 17:08 ` W B Hacker
2007-05-21 17:11 ` Roman Shaposhnik
2007-05-21 17:27   ` W B Hacker
2007-05-21 17:37     ` Francisco J Ballesteros
2007-05-22 10:47       ` Charles Forsyth
2007-05-22 11:31         ` erik quanstrom
2007-05-25 15:03         ` David Leimbach
2007-05-25 15:08           ` erik quanstrom
2007-05-25 17:02           ` Charles Forsyth
2007-05-25 17:32             ` Francisco J Ballesteros
2007-05-25 18:15               ` ron minnich
2007-05-25 19:58                 ` Uriel
2007-05-25 22:11                 ` erik quanstrom [this message]
2007-05-25 19:35               ` Steve Simon
2007-05-25 20:36                 ` Francisco J Ballesteros
2007-05-25 18:30         ` ozan s. yigit
2007-05-25 20:23           ` Charles Forsyth
2007-05-21 17:33   ` ron minnich
2007-05-21 20:18     ` David Leimbach
2007-05-21 20:25       ` erik quanstrom
2007-05-22  4:13         ` lucio
2007-05-21 18:17   ` erik quanstrom
2007-05-21 18:29     ` Uriel
2007-05-21 18:21   ` Uriel
2007-05-22  1:25     ` LiteStar numnums
2007-05-21 18:26   ` Skip Tavakkolian
2007-05-21 18:39     ` Jack Johnson
2007-05-21 18:48       ` Uriel
2007-05-21 19:26       ` Skip Tavakkolian
2007-05-21 20:17   ` David Leimbach
2007-05-21 22:34   ` Steve Simon
2007-05-21 17:32 ` Bakul Shah
2007-05-21 18:23 ` matt
2007-05-21 18:34 ` Geoffrey Avila
2007-05-21 18:41   ` Uriel
2007-05-21 19:27     ` Geoffrey Avila
2007-05-22 22:24       ` Roman Shaposhnik
2007-05-22  4:08     ` lucio
2007-05-21 23:29   ` W B Hacker
2007-05-21 23:34     ` erik quanstrom
2007-05-22 14:02       ` Wes Kussmaul
2007-05-22 14:03         ` erik quanstrom
2007-05-22 14:55           ` Bruce Ellis
2007-05-22 15:05             ` Jack Johnson
2007-05-23  7:12             ` Lluís Batlle
2007-05-23  8:11               ` Lluís Batlle
2007-05-23  8:29                 ` lucio
2007-05-22 15:22           ` Wes Kussmaul
2007-05-22 14:53         ` Jack Johnson
2007-05-22  9:53   ` Charles Forsyth
2010-06-28 16:19 [9fans] xml hugo rivera
2010-06-28 16:55 ` Russ Cox
2010-06-28 17:06 ` Ethan Grammatikidis
2010-06-28 18:26   ` Rob Pike
2010-06-28 18:31     ` David Leimbach
2010-06-28 18:40       ` Rob Pike
2010-06-28 19:06         ` erik quanstrom
2010-06-28 19:14           ` Corey Thomasson
2010-06-30 14:47             ` hugo rivera
2010-06-30 18:25               ` David Leimbach
2010-06-30 18:43                 ` ron minnich
2010-06-30 19:54                   ` Pietro Gagliardi
2010-06-30 20:04                     ` Corey Thomasson
2010-06-30 20:21                       ` erik quanstrom
2010-06-30 20:53                         ` Wes Kussmaul
2010-07-01 15:06                           ` Ethan Grammatikidis
2010-07-01 17:12                             ` Russ Cox
2010-07-01 17:43                               ` Bakul Shah
2010-07-01 18:37                                 ` Jorden M
2010-07-01 18:49                                   ` Rob Pike
2010-07-01 19:33                                     ` Bakul Shah
2010-07-01 19:34                                     ` Jorden M
2010-07-02 10:45                                       ` roger peppe
2010-06-30 20:07                     ` LiteStar numnums
2010-06-30 18:59                 ` LiteStar numnums
2010-07-03 15:51 Dean Bittner

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=cc31f831e72570b452e31604265a25af@coraid.com \
    --to=quanstro@coraid.com \
    --cc=9fans@cse.psu.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).