netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Ira W. Snyder" <iws@ovro.caltech.edu>
To: Gregory Haskins <ghaskins@novell.com>
Cc: Arnd Bergmann <arnd@arndb.de>,
	paulmck@linux.vnet.ibm.com,
	alacrityvm-devel@lists.sourceforge.net,
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org
Subject: Re: [PATCH 1/7] shm-signal: shared-memory signals
Date: Thu, 6 Aug 2009 13:51:09 -0700	[thread overview]
Message-ID: <20090806205109.GA1330@ovro.caltech.edu> (raw)
In-Reply-To: <4A7ABA530200005A00051C18@sinclair.provo.novell.com>

On Thu, Aug 06, 2009 at 09:11:15AM -0600, Gregory Haskins wrote:
> Hi Arnd,
> 
> >>> On 8/6/2009 at  9:56 AM, in message <200908061556.55390.arnd@arndb.de>, Arnd
> Bergmann <arnd@arndb.de> wrote: 
> > On Monday 03 August 2009, Gregory Haskins wrote:
> >> shm-signal provides a generic shared-memory based bidirectional
> >> signaling mechanism.  It is used in conjunction with an existing
> >> signal transport (such as posix-signals, interrupts, pipes, etc) to
> >> increase the efficiency of the transport since the state information
> >> is directly accessible to both sides of the link.  The shared-memory
> >> design provides very cheap access to features such as event-masking
> >> and spurious delivery mititgation, and is useful implementing higher
> >> level shared-memory constructs such as rings.
> > 
> > Looks like a very useful feature in general.
> 
> Thanks, I was hoping that would be the case.
> 
> > 
> >> +struct shm_signal_irq {
> >> +       __u8                  enabled;
> >> +       __u8                  pending;
> >> +       __u8                  dirty;
> >> +};
> > 
> > Won't this layout cause cache line ping pong? Other schemes I have
> > seen try to separate the bits so that each cache line is written to
> > by only one side.
> 
> It could possibly use some optimization in that regard.  I generally consider myself an expert at concurrent programming, but this lockless stuff is, um, hard ;)  I was going for correctness first.
> 
> Long story short, any suggestions on ways to split this up are welcome (particularly now, before the ABI is sealed ;)
> 
> > This gets much more interesting if the two sides
> > are on remote ends of an I/O link, e.g. using a nontransparent
> > PCI bridge, where you only want to send stores over the wire, but
> > never fetches or even read-modify-write cycles.
> 
> /me head explodes ;)
> 

I've actually implemented this idea for virtio. Read the virtio-over-PCI
patches I posted, and you'll see that the entire virtqueue
implementation NEVER uses reads across the PCI bus, only writes. The
slowpath configuration space uses reads, but the virtqueues themselves
are write-only.

Some trivial benchmarking against an earlier driver that did
writes+reads across the PCI bus showed that the write-only driver was
about 2x as fast. (Throughput increased from ~30MB/sec to ~65MB/sec).

I'm sure the write-only design was not the only change responsible for
the speedup, but it was definitely a contributing factor.

Ira

  reply	other threads:[~2009-08-06 20:51 UTC|newest]

Thread overview: 59+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-03 17:17 [PATCH 0/7] AlacrityVM guest drivers Gregory Haskins
2009-08-03 17:17 ` [PATCH 1/7] shm-signal: shared-memory signals Gregory Haskins
2009-08-06 13:56   ` Arnd Bergmann
2009-08-06 15:11     ` Gregory Haskins
2009-08-06 20:51       ` Ira W. Snyder [this message]
2009-08-03 17:17 ` [PATCH 2/7] ioq: Add basic definitions for a shared-memory, lockless queue Gregory Haskins
2009-08-03 17:17 ` [PATCH 3/7] vbus: add a "vbus-proxy" bus model for vbus_driver objects Gregory Haskins
2009-08-03 17:17 ` [PATCH 4/7] vbus-proxy: add a pci-to-vbus bridge Gregory Haskins
2009-08-06 14:42   ` Arnd Bergmann
2009-08-06 15:59     ` Gregory Haskins
2009-08-06 17:03       ` Arnd Bergmann
2009-08-06 21:04         ` Gregory Haskins
2009-08-06 22:57           ` Arnd Bergmann
2009-08-07  4:42             ` Gregory Haskins
2009-08-07 14:57               ` Arnd Bergmann
2009-08-07 15:44                 ` Gregory Haskins
2009-08-07 15:55               ` Ira W. Snyder
2009-08-07 18:25                 ` Gregory Haskins
2009-08-03 17:17 ` [PATCH 5/7] ioq: add driver-side vbus helpers Gregory Haskins
2009-08-03 17:18 ` [PATCH 6/7] net: Add vbus_enet driver Gregory Haskins
2009-08-03 18:30   ` Stephen Hemminger
2009-08-03 20:10     ` Gregory Haskins
2009-08-03 20:19       ` Stephen Hemminger
2009-08-03 20:24         ` Gregory Haskins
2009-08-03 20:29           ` Stephen Hemminger
2009-08-04  1:14   ` [PATCH v2] " Gregory Haskins
2009-08-04  2:38     ` David Miller
2009-08-04 13:57       ` [Alacrityvm-devel] " Gregory Haskins
2009-10-02 15:33     ` [PATCH v3] " Gregory Haskins
2009-08-03 17:18 ` [PATCH 7/7] venet: add scatter-gather/GSO support Gregory Haskins
2009-08-03 18:32   ` Stephen Hemminger
2009-08-03 19:30     ` Gregory Haskins
2009-08-03 18:33   ` Stephen Hemminger
2009-08-03 19:57     ` Gregory Haskins
2009-08-06  8:19 ` [PATCH 0/7] AlacrityVM guest drivers Reply-To: Michael S. Tsirkin
2009-08-06 10:17   ` Michael S. Tsirkin
2009-08-06 12:09     ` Gregory Haskins
2009-08-06 12:08   ` Gregory Haskins
2009-08-06 12:24     ` Michael S. Tsirkin
2009-08-06 13:00       ` Gregory Haskins
2009-08-06 12:54     ` Avi Kivity
2009-08-06 13:03       ` Gregory Haskins
2009-08-06 13:44         ` Avi Kivity
2009-08-06 13:45           ` Gregory Haskins
2009-08-06 13:57             ` Avi Kivity
2009-08-06 14:06               ` Gregory Haskins
2009-08-06 15:40                 ` Arnd Bergmann
2009-08-06 15:45                   ` Michael S. Tsirkin
2009-08-06 15:50                   ` Avi Kivity
2009-08-06 16:55                     ` Gregory Haskins
2009-08-09  7:48                       ` Avi Kivity
2009-08-06 16:29                   ` Gregory Haskins
2009-08-06 23:23                     ` Ira W. Snyder
2009-08-06 13:59             ` Michael S. Tsirkin
2009-08-06 14:07               ` Gregory Haskins
2009-08-07 14:19   ` Anthony Liguori
2009-08-07 15:05     ` [PATCH 0/7] AlacrityVM guest drivers Gregory Haskins
2009-08-07 15:46       ` Anthony Liguori
2009-08-07 18:04         ` Gregory Haskins

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=20090806205109.GA1330@ovro.caltech.edu \
    --to=iws@ovro.caltech.edu \
    --cc=alacrityvm-devel@lists.sourceforge.net \
    --cc=arnd@arndb.de \
    --cc=ghaskins@novell.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=paulmck@linux.vnet.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).