From: Anthony Liguori <anthony@codemonkey.ws>
To: Alexander Graf <agraf@suse.de>
Cc: Gleb Natapov <gleb@redhat.com>,
kvm@vger.kernel.org, qemu-devel@nongnu.org,
"Richard W.M. Jones" <rjones@redhat.com>,
Avi Kivity <avi@redhat.com>, Gerd Hoffmann <kraxel@redhat.com>
Subject: Re: [Qemu-devel] Anyone seeing huge slowdown launching qemu with Linux 2.6.35?
Date: Wed, 04 Aug 2010 12:26:06 -0500 [thread overview]
Message-ID: <4C59A2AE.2060602@codemonkey.ws> (raw)
In-Reply-To: <721C3046-D6A7-47B2-A18C-4E59F2B68797@suse.de>
On 08/04/2010 11:45 AM, Alexander Graf wrote:
> Frankly, I partially agreed to your point when we were talking about 300ms vs. 2 seconds. Now that we're talking 8 seconds that whole point is moot. We chose the wrong interface to transfer kernel+initrd data into the guest.
>
> Now the question is how to fix that. I would veto against anything normally guest-OS-visible. By occupying the floppy, you lose a floppy drive in the guest. By occupying a disk, you see an unwanted disk in the guest.
Introduce a new virtio device type (say, id 6). Teach SeaBIOS that 6 is
exactly like virtio-blk (id 2). Make it clear that id 6 is only to be
used by firmware and that normal guest drivers should not be written for
id 6.
Problem is now solved and everyone's happy. Now we can all go back to
making slides for next week :-)
Regards,
Anthony Liguori
> By taking virtio-serial you see an unwanted virtio-serial line in the guest. fw_cfg is great because it's a private interface nobody else accesses.
>
> I see two alternatives out of this mess:
>
> 1) Speed up string PIO so we're actually fast again.
> 2) Using a different interface (that could also be DMA fw_cfg - remember, we're on a private interface anyways)
>
> Admittedly 1 would also help in more cases than just booting with -kernel and -initrd, but if that won't get us to acceptable levels (and yes, 8 seconds for 100MB is unacceptable) I don't see any way around 2.
>
>
> Alex
>
>
next prev parent reply other threads:[~2010-08-04 17:26 UTC|newest]
Thread overview: 151+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-03 11:13 [Qemu-devel] Anyone seeing huge slowdown launching qemu with Linux 2.6.35? Richard W.M. Jones
2010-08-03 11:33 ` Gleb Natapov
2010-08-03 12:10 ` Richard W.M. Jones
2010-08-03 12:37 ` Gleb Natapov
2010-08-03 12:48 ` Richard W.M. Jones
2010-08-03 13:19 ` Avi Kivity
2010-08-03 14:05 ` Richard W.M. Jones
2010-08-03 14:38 ` Avi Kivity
2010-08-03 14:53 ` Richard W.M. Jones
2010-08-03 16:10 ` Avi Kivity
2010-08-03 16:28 ` Richard W.M. Jones
2010-08-03 16:44 ` Avi Kivity
2010-08-03 16:46 ` Anthony Liguori
2010-08-03 16:50 ` Avi Kivity
2010-08-03 16:53 ` Anthony Liguori
2010-08-03 17:01 ` Avi Kivity
2010-08-03 17:42 ` Anthony Liguori
2010-08-03 17:58 ` Avi Kivity
2010-08-03 18:11 ` Richard W.M. Jones
2010-08-03 18:26 ` Anthony Liguori
2010-08-03 18:43 ` Avi Kivity
2010-08-03 18:47 ` Avi Kivity
2010-08-03 18:55 ` Anthony Liguori
2010-08-03 19:00 ` Avi Kivity
2010-08-03 19:05 ` Gleb Natapov
2010-08-03 19:09 ` Avi Kivity
2010-08-03 19:15 ` Anthony Liguori
2010-08-03 19:24 ` Avi Kivity
2010-08-03 19:38 ` Anthony Liguori
2010-08-03 19:41 ` Avi Kivity
2010-08-03 19:47 ` Anthony Liguori
2010-08-04 5:47 ` Avi Kivity
2010-08-03 21:24 ` Gerd Hoffmann
2010-08-03 21:20 ` Gerd Hoffmann
2010-08-04 5:53 ` Avi Kivity
2010-08-04 7:56 ` Gerd Hoffmann
2010-08-04 8:17 ` Avi Kivity
2010-08-04 8:43 ` Gleb Natapov
2010-08-04 9:22 ` Gerd Hoffmann
2010-08-04 13:04 ` Anthony Liguori
2010-08-04 13:07 ` Gleb Natapov
2010-08-04 13:15 ` Anthony Liguori
2010-08-04 13:24 ` Richard W.M. Jones
2010-08-04 13:26 ` Gleb Natapov
2010-08-04 14:22 ` Anthony Liguori
2010-08-04 14:38 ` Gleb Natapov
2010-08-04 14:50 ` Anthony Liguori
2010-08-04 15:01 ` Gleb Natapov
2010-08-04 15:07 ` Anthony Liguori
2010-08-04 15:15 ` Gleb Natapov
2010-08-04 22:41 ` Kevin O'Connor
2010-08-04 16:26 ` Avi Kivity
2010-08-04 13:34 ` Gleb Natapov
2010-08-04 13:52 ` Anthony Liguori
2010-08-04 14:00 ` Gleb Natapov
2010-08-04 14:14 ` Anthony Liguori
2010-08-04 14:36 ` Gleb Natapov
2010-08-04 14:22 ` Paolo Bonzini
2010-08-04 14:39 ` Anthony Liguori
2010-08-04 16:33 ` Avi Kivity
2010-08-04 16:30 ` Avi Kivity
2010-08-04 16:36 ` Avi Kivity
2010-08-04 16:44 ` Anthony Liguori
2010-08-04 16:52 ` Avi Kivity
2010-08-04 17:37 ` Gleb Natapov
2010-08-05 7:28 ` Gerd Hoffmann
2010-08-05 7:34 ` Gleb Natapov
2010-08-05 7:56 ` Avi Kivity
2010-08-05 7:59 ` Gleb Natapov
2010-08-05 8:45 ` Avi Kivity
2010-08-05 8:48 ` Gleb Natapov
2010-08-05 13:43 ` Anthony Liguori
2010-08-04 16:45 ` Alexander Graf
2010-08-04 16:54 ` Avi Kivity
2010-08-04 17:01 ` Alexander Graf
2010-08-04 17:14 ` Avi Kivity
2010-08-04 17:27 ` Alexander Graf
2010-08-04 17:34 ` Avi Kivity
2010-08-04 20:06 ` David S. Ahern
2010-08-04 20:16 ` Richard W.M. Jones
2010-08-05 2:38 ` Avi Kivity
2010-08-04 17:26 ` Anthony Liguori [this message]
2010-08-04 17:31 ` Alexander Graf
2010-08-04 17:35 ` Avi Kivity
2010-08-04 17:36 ` Anthony Liguori
2010-08-04 17:36 ` Alexander Graf
2010-08-04 17:46 ` Richard W.M. Jones
2010-08-04 17:50 ` Avi Kivity
2010-08-04 18:13 ` Alexander Graf
2010-08-04 18:16 ` Anthony Liguori
2010-08-04 18:18 ` Alexander Graf
2010-08-04 18:19 ` Avi Kivity
2010-08-04 18:18 ` Avi Kivity
2010-08-04 16:42 ` Anthony Liguori
2010-08-04 13:22 ` Richard W.M. Jones
2010-08-04 13:29 ` Gleb Natapov
2010-08-04 16:25 ` Avi Kivity
2010-08-03 22:06 ` Richard W.M. Jones
2010-08-04 5:54 ` Avi Kivity
2010-08-04 9:24 ` Richard W.M. Jones
2010-08-04 9:27 ` Gleb Natapov
2010-08-04 9:52 ` Avi Kivity
2010-08-04 11:33 ` Richard W.M. Jones
2010-08-04 11:36 ` Avi Kivity
2010-08-04 12:07 ` Gleb Natapov
2010-08-04 12:59 ` Anthony Liguori
2010-08-03 19:26 ` Gleb Natapov
2010-08-03 19:13 ` Richard W.M. Jones
2010-08-03 19:17 ` Gleb Natapov
2010-08-03 19:19 ` Anthony Liguori
2010-08-03 19:22 ` Avi Kivity
2010-08-03 20:00 ` Richard W.M. Jones
2010-08-03 20:49 ` Anthony Liguori
2010-08-03 21:13 ` Paolo Bonzini
2010-08-03 21:34 ` Anthony Liguori
2010-08-04 7:57 ` Paolo Bonzini
2010-08-04 8:19 ` Avi Kivity
2010-08-04 12:53 ` Anthony Liguori
2010-08-04 16:44 ` Avi Kivity
2010-08-04 16:46 ` Anthony Liguori
2010-08-04 16:48 ` Alexander Graf
2010-08-04 16:49 ` Anthony Liguori
2010-08-04 16:51 ` Alexander Graf
2010-08-04 17:01 ` Paolo Bonzini
2010-08-04 17:19 ` Avi Kivity
2010-08-04 17:25 ` Alexander Graf
2010-08-04 17:27 ` Anthony Liguori
2010-08-04 17:37 ` Avi Kivity
2010-08-04 17:53 ` Anthony Liguori
2010-08-04 18:05 ` Alexander Graf
2010-08-04 5:56 ` Avi Kivity
2010-08-04 1:17 ` Jamie Lokier
2010-08-04 8:21 ` Avi Kivity
2010-08-04 14:51 ` David S. Ahern
2010-08-04 14:57 ` Anthony Liguori
2010-08-04 15:25 ` Gleb Natapov
2010-08-04 15:31 ` Alexander Graf
2010-08-04 15:48 ` Gleb Natapov
2010-08-04 15:59 ` Alexander Graf
2010-08-04 16:08 ` Gleb Natapov
2010-08-04 16:48 ` Avi Kivity
2010-08-04 23:17 ` Kevin O'Connor
2010-08-05 5:26 ` Gleb Natapov
2010-08-03 16:56 ` Anthony Liguori
2010-08-03 16:48 ` Avi Kivity
2010-08-03 17:00 ` Richard W.M. Jones
2010-08-03 17:05 ` Avi Kivity
2010-08-03 16:56 ` Richard W.M. Jones
2010-08-03 17:08 ` Avi Kivity
2010-08-03 16:39 ` Anthony Liguori
2010-08-03 16:43 ` Richard W.M. Jones
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=4C59A2AE.2060602@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=agraf@suse.de \
--cc=avi@redhat.com \
--cc=gleb@redhat.com \
--cc=kraxel@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=qemu-devel@nongnu.org \
--cc=rjones@redhat.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).