Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

qemu success with -current

5 views
Skip to first unread message

Antti Kantee

unread,
Feb 22, 2011, 1:48:14 PM2/22/11
to
Hi,

I just noticed that a -current iso from the NetBSD daily builds installs
and boots cleanly with anita/qemu.
(-current used to be broken at least for me, although nb5 worked)

If someone has some spare computing power, it would be cool to get
sparc represented here:
http://releng.netbsd.org/test-results.html

--
älä karot toivorikkauttas, kyl rätei ja lumpui piisaa

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de

matthew green

unread,
Feb 22, 2011, 3:40:55 PM2/22/11
to

> I just noticed that a -current iso from the NetBSD daily builds installs
> and boots cleanly with anita/qemu.
> (-current used to be broken at least for me, although nb5 worked)
>
> If someone has some spare computing power, it would be cool to get
> sparc represented here:
> http://releng.netbsd.org/test-results.html


i tried to setup qemu with a MP sparc system, but our kernel
fails on it pretty quickly. it would be great if someone fixed
this, too :-)


.mrg.

Artyom Tarasenko

unread,
Mar 2, 2011, 4:20:02 AM3/2/11
to
On Tue, Feb 22, 2011 at 9:40 PM, matthew green <m...@eterna.com.au> wrote:
>
>> I just noticed that a -current iso from the NetBSD daily builds installs
>> and boots cleanly with anita/qemu.
>> (-current used to be broken at least for me, although nb5 worked)
>>
>> If someone has some spare computing power, it would be cool to get
>> sparc represented here:
>> http://releng.netbsd.org/test-results.html
>
>
> i tried to setup qemu with a MP sparc system, but our kernel
> fails on it pretty quickly.  it would be great if someone fixed
> this, too :-)

You mean you started qemu with -cpu 2 ? Then it's probably a qemu bug,
linux has problems with it too.


--
Regards,
Artyom Tarasenko

solaris/sparc under qemu blog: http://tyom.blogspot.com/

matthew green

unread,
Mar 2, 2011, 6:56:37 AM3/2/11
to

> >> I just noticed that a -current iso from the NetBSD daily builds installs
> >> and boots cleanly with anita/qemu.
> >> (-current used to be broken at least for me, although nb5 worked)
> >>
> >> If someone has some spare computing power, it would be cool to get
> >> sparc represented here:
> >> http://releng.netbsd.org/test-results.html
> >
> >
> > i tried to setup qemu with a MP sparc system, but our kernel
> > fails on it pretty quickly.  it would be great if someone fixed
> > this, too :-)
>
> You mean you started qemu with -cpu 2 ? Then it's probably a qemu bug,
> linux has problems with it too.

oh yeah - i didn't mean to imply anything else.


.mrg.

Antti Kantee

unread,
Mar 5, 2011, 9:12:14 AM3/5/11
to
On Tue Feb 22 2011 at 20:48:14 +0200, Antti Kantee wrote:
> Hi,
>
> I just noticed that a -current iso from the NetBSD daily builds installs
> and boots cleanly with anita/qemu.
> (-current used to be broken at least for me, although nb5 worked)
>
> If someone has some spare computing power, it would be cool to get
> sparc represented here:
> http://releng.netbsd.org/test-results.html

I did a one-off qemu test run on my desktop. There are a few interesting
failing tests, such as a ipf bus error and SA_RESETHAND. Also, it
seems that sparc has the same problem as powerpc with mutexes, i.e.
trying lock a competed mutex hangs occasionally (see PR powerpc/44387).
That causes a few extra failures in addition to the pthread tests failing,
or at least that's my theory when I xref the results with macppc tests.

I uploaded the test results here:
http://ftp.netbsd.org/pub/NetBSD/misc/pooka/atf-misc/sparc-201103041300Z.html

(A few moments ago I fixed newfs_msdos, so the msdosfs tests no longer
fail).

p.s. qemu-system-sparc takes a full CPU on the host regardless of
guest load. is that a bug or a feature?

Manuel Bouyer

unread,
Mar 5, 2011, 3:41:47 PM3/5/11
to
On Sat, Mar 05, 2011 at 04:12:14PM +0200, Antti Kantee wrote:
> [...]

> p.s. qemu-system-sparc takes a full CPU on the host regardless of
> guest load. is that a bug or a feature?

It depends on the system you're emulating.
Some sparc CPUs don't have a idle instruction so the kernel just
spins waiting to have something to do.

The apc(4) driver should allow some sparc to go to sleep waiting for
the next interrupt; it's not enabled by default (it's reported that some
system hang trying to use it).

apc(4) works fine with the pkgsrc qemu emulating a SPARCstation 5, and
allows the qemu process to go idle when the guest is idle.

--
Manuel Bouyer <bou...@antioche.eu.org>
NetBSD: 26 ans d'experience feront toujours la difference
--

0 new messages