From: Ronald <look@reply.to>
To: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: The QEMU Accelerator Module
Date: Sat, 12 Feb 2005 06:37:34 +0100 [thread overview]
Message-ID: <pan.2005.02.12.05.37.34.249027@reply.to> (raw)
In-Reply-To: 420BEEB7.1010906@bellard.org
Le Fri, 11 Feb 2005 00:31:03 +0100, Fabrice Bellard a écrit :
> Hi,
>
> I just commited the first alpha release of the QEMU Accelerator Module
> (aka KQEMU) in the CVS. It gives better performance for the "x86 on x86"
> case by running most of the application code as is. It works only for a
> Linux x86 host running a 2.4 or 2.6 kernel. Linux 2.4 and 2.6 kernels and
> Windows 2000 have been runnning as guest OSes, but other OSes may work as
> well. As with every alpha kernel driver testing, it is better to backup
> your data before trying it.
>
> KQEMU is _not_ open source as the rest of QEMU. It is a proprietary kernel
> module (read the LICENSE file) and will stay so until a gentle company
> decides to subsidy the QEMU project.
>
> KQEMU usage is optional: you can disable it at compilation or run time, so
> no one is forced to use it.
>
> Fabrice.
Strange that nobody has posted some comparisons between qemu with and
without this module, maybe it's uninteresting but I do that:
I have tried to play a big divx file with mplayer under qemu, host is
running 2.6.10, guest 2.6.9, mplayer is build without optimisations (no
mmx or sse) and launched with -ao alsa -framedrop.
Without using the module the file is far away to be played correctly: no
picture in the 3 minutes before I stop, sound is not even played correctly
: a huge number of "resetting stream".
When using the kqemu module that's a lot better, sound is decoded and
played perfectly, and the video stream is played too, just some a/v sync
problems that's why I have used -framedrop but I bet this is more a
display thing than a decoding one. I have tried some different -vo x11,dga
and sdl, the first two are ok, sdl one makes mplayer crash with a "bad
usage of CPU/FPU/RAM" message (with or without kqemu), but this may be
related to the flags¹ I have used when building qemu.
The following is the output from mplayer for the streams properties:
AVI file format detected.
VIDEO: [DIVX] 640x368 16bpp 25.000 fps 682.8 kbps (83.4 kbyte/s)
==========================================================================
Opening audio decoder: [mp3lib] MPEG layer-2, layer-3
MP3lib: init layer2&3 finished, tables done
AUDIO: 48000 Hz, 2 ch, 16 bit (0x10), ratio: 40000->192000 (320.0 kbit)
Selected audio codec: [mp3] afm:mp3lib (mp3lib MPEG layer-2, layer-3)
==========================================================================
[...]
==========================================================================
Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family Selected
video codec: [ffodivx] vfm:ffmpeg (FFmpeg MPEG-4)
==========================================================================
Checking audio filter chain for 48000Hz/2ch/16bit -> 48000Hz/2ch/16bit...
AF_pre: af format: 2 bps, 2 ch, 48000 hz, little endian signed int AF_pre:
48000Hz 2ch Signed 16-bit (Little-Endian)
I can imagine that some more games could have a chance to run on a windows
guest now :)
¹This is what I have passed to make for this one (kqemu should be build
again after a clean distclean):
OP_CFLAGS='-Wall -O2 -g -fno-strict-aliasing -fomit-frame-pointer
-mpreferred-stack-boundary=2 -falign-functions=0 -fno-gcse
-fno-reorder-blocks -fno-optimize-sibling-calls -march=pentium
-mcpu=athlon-xp -mmmx' CFLAGS='-Wall -O2 -g -fno-strict-aliasing
-fomit-frame-pointer -march=athlon-xp -mcpu=athlon-xp -mfpmath=sse -msse
-m3dnow' HELPER_CFLAGS='-Wall -O2 -g -fno-strict-aliasing
-fomit-frame-pointer -march=pentium -mcpu=athlon-xp -mmmx'
next prev parent reply other threads:[~2005-02-12 5:52 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-02-10 23:31 [Qemu-devel] The QEMU Accelerator Module Fabrice Bellard
2005-02-10 23:57 ` Grzegorz Kulewski
2005-02-11 1:00 ` Lennert Buytenhek
2005-02-11 1:27 ` Tim
2005-02-12 14:10 ` Fabrice Bellard
2005-02-12 15:28 ` Fabrice Bellard
2005-02-11 0:38 ` Darryl Dixon
2005-02-11 1:01 ` Hetz Ben Hamo
2005-02-11 1:20 ` Leigh Dyer
2005-02-11 2:11 ` James Mastros
2005-02-11 5:15 ` Tom Marble
2005-02-11 9:23 ` Karel Gardas
2005-02-11 10:52 ` Lionel Ulmer
2005-02-11 12:05 ` Karel Gardas
2005-02-11 13:44 ` Lionel Ulmer
2005-02-11 13:47 ` Grzegorz Kulewski
2005-02-11 9:54 ` [Qemu-devel] KQemu logic and marketing goals Jean-Michel POURE
2005-02-11 11:37 ` Lionel Ulmer
2005-02-11 12:27 ` Julian Seward
2005-02-11 12:48 ` Lennert Buytenhek
2005-02-11 13:17 ` Johannes Schindelin
2005-02-11 14:16 ` Lennert Buytenhek
2005-02-11 14:05 ` Jean-Michel POURE
2005-02-11 12:02 ` [Qemu-devel] The QEMU Accelerator Module Johannes Schindelin
2005-02-11 12:38 ` Jens Arm
2005-02-11 16:09 ` Derek Fawcus
2005-02-11 17:17 ` Jim C. Brown
2005-02-11 16:16 ` [Qemu-devel] " Anand Kumria
2005-02-11 21:33 ` Antti-Juhani Kaijanaho
2005-02-12 2:01 ` Johannes Schindelin
2005-02-12 3:23 ` Darryl Dixon
2005-02-12 4:06 ` Hetz Ben Hamo
2005-02-12 6:14 ` [Qemu-devel] " Anand Kumria
2005-02-12 16:19 ` Natalia Portillo
2005-02-12 17:04 ` Jean-Christian de Rivaz
2005-02-12 17:29 ` René Korthaus
2005-02-12 19:48 ` Natalia Portillo
2005-02-12 20:51 ` Jean-Christian de Rivaz
2005-02-13 0:07 ` René Korthaus
2005-02-11 16:22 ` [Qemu-devel] " Alexander E. Patrakov
2005-02-12 5:37 ` Ronald [this message]
2005-02-12 17:12 ` [Qemu-devel] " Felipe Sanchez
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=pan.2005.02.12.05.37.34.249027@reply.to \
--to=look@reply.to \
--cc=daimon55@free.fr \
--cc=qemu-devel@nongnu.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;
as well as URLs for NNTP newsgroup(s).