From: Paul Brook <paul@codesourcery.com>
To: Christoph Hellwig <hch@lst.de>
Cc: qemu-devel@nongnu.org, Anthony Liguori <anthony@codemonkey.ws>,
kvm@vger.kernel.org, Rusty Russell <rusty@rustcorp.com.au>,
Christian Borntraeger <borntraeger@de.ibm.com>,
Hannes Reinecke <hare@suse.de>
Subject: Re: [Qemu-devel] Re: [PATCH] virtio-blk: add SGI_IO passthru support
Date: Wed, 29 Apr 2009 12:37:20 +0100 [thread overview]
Message-ID: <200904291237.21558.paul@codesourcery.com> (raw)
In-Reply-To: <20090429112130.GA11241@lst.de>
On Wednesday 29 April 2009, Christoph Hellwig wrote:
> On Wed, Apr 29, 2009 at 12:11:19PM +0100, Paul Brook wrote:
> > Is this actually measurably faster, or just infinitesimally faster in
> > theory?
>
> On normal disks it's rather theoretical. On high-end SSDs and arrays the
> impact is noticeable, mostly due to the additional latency.
How exactly does it introduce additional latency? A scsi command block is
hardly large or complicated. Are you suggesting that a 16/32byte scsi command
takes significantly longer to process than a 16byte virtio command
descriptor? I'd expect any extra processing to be a small fraction of the
host syscall latency, let alone the latency of the physical host adapter. It
probably even fits on the same CPU cache line.
Paul
WARNING: multiple messages have this Message-ID (diff)
From: Paul Brook <paul@codesourcery.com>
To: Christoph Hellwig <hch@lst.de>
Cc: kvm@vger.kernel.org, Rusty Russell <rusty@rustcorp.com.au>,
qemu-devel@nongnu.org,
Christian Borntraeger <borntraeger@de.ibm.com>,
Hannes Reinecke <hare@suse.de>
Subject: Re: [Qemu-devel] Re: [PATCH] virtio-blk: add SGI_IO passthru support
Date: Wed, 29 Apr 2009 12:37:20 +0100 [thread overview]
Message-ID: <200904291237.21558.paul@codesourcery.com> (raw)
In-Reply-To: <20090429112130.GA11241@lst.de>
On Wednesday 29 April 2009, Christoph Hellwig wrote:
> On Wed, Apr 29, 2009 at 12:11:19PM +0100, Paul Brook wrote:
> > Is this actually measurably faster, or just infinitesimally faster in
> > theory?
>
> On normal disks it's rather theoretical. On high-end SSDs and arrays the
> impact is noticeable, mostly due to the additional latency.
How exactly does it introduce additional latency? A scsi command block is
hardly large or complicated. Are you suggesting that a 16/32byte scsi command
takes significantly longer to process than a 16byte virtio command
descriptor? I'd expect any extra processing to be a small fraction of the
host syscall latency, let alone the latency of the physical host adapter. It
probably even fits on the same CPU cache line.
Paul
next prev parent reply other threads:[~2009-04-29 11:37 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-04-27 8:26 [PATCH] virtio-blk: add SGI_IO passthru support Christoph Hellwig
2009-04-27 8:29 ` Christoph Hellwig
2009-04-27 8:29 ` [Qemu-devel] " Christoph Hellwig
2009-04-27 9:15 ` Avi Kivity
2009-04-27 9:15 ` [Qemu-devel] " Avi Kivity
2009-04-28 9:47 ` Christoph Hellwig
2009-04-28 9:47 ` Christoph Hellwig
2009-04-27 14:36 ` Anthony Liguori
2009-04-27 14:36 ` [Qemu-devel] " Anthony Liguori
2009-04-28 9:51 ` Christoph Hellwig
2009-04-28 9:51 ` Christoph Hellwig
2009-04-28 16:37 ` Anthony Liguori
2009-04-28 16:52 ` Christian Borntraeger
2009-04-28 16:52 ` Christian Borntraeger
2009-04-29 10:48 ` Christoph Hellwig
2009-04-29 10:48 ` Christoph Hellwig
2009-04-29 11:11 ` Paul Brook
2009-04-29 11:11 ` Paul Brook
2009-04-29 11:19 ` Christian Borntraeger
2009-04-29 11:19 ` Christian Borntraeger
2009-04-29 11:29 ` Paul Brook
2009-04-29 11:29 ` Paul Brook
2009-04-29 11:21 ` Christoph Hellwig
2009-04-29 11:21 ` Christoph Hellwig
2009-04-29 11:37 ` Paul Brook [this message]
2009-04-29 11:37 ` Paul Brook
2009-04-30 20:13 ` Christoph Hellwig
2009-04-30 20:13 ` Christoph Hellwig
2009-04-30 20:55 ` Nicholas A. Bellinger
2009-04-30 20:55 ` Nicholas A. Bellinger
2009-04-30 21:49 ` Paul Brook
2009-04-30 21:49 ` Paul Brook
2009-04-30 21:56 ` Javier Guerra
2009-04-30 21:56 ` Javier Guerra
2009-05-01 7:24 ` Christoph Hellwig
2009-05-01 7:24 ` Christoph Hellwig
2009-05-01 11:08 ` Jamie Lokier
2009-05-01 11:08 ` Jamie Lokier
2009-05-01 14:28 ` Christoph Hellwig
2009-05-01 14:28 ` Christoph Hellwig
2009-05-05 5:21 ` Rusty Russell
2009-05-05 5:21 ` Rusty Russell
2009-04-28 19:09 ` Christian Borntraeger
2009-04-28 19:09 ` Christian Borntraeger
2009-04-29 10:50 ` Christoph Hellwig
2009-04-29 10:50 ` Christoph Hellwig
2009-04-29 11:07 ` Christian Borntraeger
2009-04-29 11:07 ` Christian Borntraeger
2009-04-28 9:57 ` [PATCH 2/2] virtio-blk: add SG_IO " Christoph Hellwig
2009-04-28 9:57 ` [Qemu-devel] " Christoph Hellwig
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=200904291237.21558.paul@codesourcery.com \
--to=paul@codesourcery.com \
--cc=anthony@codemonkey.ws \
--cc=borntraeger@de.ibm.com \
--cc=hare@suse.de \
--cc=hch@lst.de \
--cc=kvm@vger.kernel.org \
--cc=qemu-devel@nongnu.org \
--cc=rusty@rustcorp.com.au \
/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.