On Sat, 2009-02-21 at 21:14 -0600, Matt Domsch wrote:
> Fedora Rawhide-in-Mock Build Results for i386
> based on rawhide as of Friday Feb 20, 2009.
> Most packages were actually built as of the F11 Alpha cut on or around
> Feb 1. Any newer packages in rawhide as of Feb 20 were then pulled in
> and rebuilt, and any previously-failed packages were also rebuilt, so
> I believe this list to be current and correct. I did have a couple
> servers fail during the builds, so it is possible that at most 8 of
> these are not true failures, but a result of builders crashing. Don't
> pin your hopes on that though. :-)
> Full logs at http://linux.dell.com/files/fedora/FixBuildRequires/
> ruby-libvirt-0.1.0-2.fc11 (build/make) lutter,stahnma,clalance,virtmaint
Loaded suite /usr/lib/ruby/gems/1.8/gems/rake-0.8.3/lib/rake/rake_test_loader
libvir: Test error : Domain not found
./tests/tc_connect.rb:165: [BUG] Segmentation fault
ruby 1.8.6 (2008-08-11) [i386-linux]
Command failed with status (): [/usr/bin/ruby -Ilib:ext/libvirt "/usr/lib/...]
(See full trace by running task with --trace)
2009-03-03 Feature freeze (18 days)
2009-03-10 Beta Freeze (25 days)
2009-04-14 Final freeze (60 days)
Fedora Weekly News
The latest instalment of Fedora Weekly News is out, again with a
detailed virt section:
Thanks to Dale Bewley.
Glauber has been working hard towards merging the qemu and kvm
This week most of the effort involved producing packages for
bochs-bios, openbios and vgabios. See this mail for more details:
The really nasty part - which Eduardo Hobkost ealier worked through
with the etherboot package - is that although the BIOS binaries need
to be built on a specific platform, they need to be available for
install on all architectures. For example, Bochs is built on x86, but
if you run qemu-system-x86_64 on ppc you need it there too.
As an interim measure, all these packages will be first built on the
required architecture, then the packager will download the new build,
pack the images in a tarball and then re-build the package with this
tarball. The second build will package the pre-built binaries on all
arches, rather than building from source.
In future, koji/rpm will support packages with ExclusiveArch that
produce noarch packages. So, e.g. Bochs will be built only on x86 and
the result will be a noarch package. See also:
The pacakge reviews for the three new packages are here:
PCI Device Assignment
The offline discussion around device resetting, unbinding and
filtering finally made it on-list here:
The upshot is that some further kernel work is required before device
assignment will be fully usable in Fedora.
David Woodhouse's work continued on resolving the DMAR issues seen on
some machines with intel_iommu=on:
David's first patch fixes a problem where one platform (Mobile 4
Series Chipset). The problem is that this chipset claims to not need
explicit write buffer flushing, when it does in fact need it. David's
patch just ignores the RWBF hint on that platform. This should fix the
issues seen with Lenovo X200 and the Dell Latitude E6400.
Another problem was found with the iwlwifi driver where the device is
writing to buffers that the driver marks as read-only. David has added
a fix for this issue to rawhide:
Shared Network Interface
David Lutterkort updated libvir-list with his progress:
After talking with Dan Williams, who is working in NetworkManager,
it is clear that the network interface functionality for libvirt
is also useful for NM and would help them with handling
system-wide interface configuration. To accomodate that, I will
implement network interface configuration in a separate library
(inspiringly called 'netcf', until somebody thinks of something
catchier) and add public API to libvirt that will use netcf to do
the actual work.
The F11 Feature page has also been updated:
Dan Walsh posted his thoughts on the sVirt patch set to libvir-list:
It appears the patch is in good shape already and could be merged very
soon. Dan Berrange replied with:
I've reviewed all the submissions James has posted for comments so
far, and am pretty happy with the way it now integrates with
If yourself & James are happy with what they're doing from a SELinux
/ security model point of view, then there's no reason they
shouldn't be posted for final merge now.
kvmclock and !constant_tsc
The fix to disable kvmclock on hosts with no constant_tsc has no been
pushed to Fedora 10:
Upstream, Avi has queued up the workaround for 2.6.29 here:
and the real fix is queued for 2.6.30 here:
Worryingly, Levente Farkas noted that he has constant_tsc and is
still seeing hangs which go away with clocksource=acpi_pm. Further
investigation of this is required.
The issue with pvmmu and CONFIG_SLAB_DEBUG was resolved this week:
pvmmu causes net driver oops during guest install (new_slab)
Marcelo came up with a guest kernel fix for pvmmu (only batch user pte
updates), built in koji and James Laska ran some test installs using
it. Unfortunately, James couldn't reproduce with latest rawhide and,
obviously, also didn't see the issue with Marcelo's patch.
Marcelo posted his patch upstream and some discussion with Jeremy
Fitzhardinge yielded a patch to add a missing flush_lazy_mmu_mode()
call to change_page_attr():
This patch has since been merged into Linus's tree for 2.6.29.
Dan Berrange pushed an libvirt-0.6.0-2 with the following fixes:
- Fix libvirtd --timeout usage
- Fix RPC call problems and QEMU startup handling (rhbz #484414)
- Fix unowned directories (rhbz #483442)
Further issues with this release include:
AVC denied errors when starting KVM guest
SELinux issue causing libvirtd launched dnsmasq to fail
Dan Berrange pointed out the selinux-policy changes
needed. Miroslav Grepl and Dan Walsh made the changes in
selinux-policy-3.5.13-45.fc10 which has been pushed to F10
libvirtd segfault in libdbus after starting kvm guest
Strangely, it proved to be very difficult to get a good trace of
In the end, the issue turned out to be nasty D-Bus/PolicyKit
related segfault in libvirtd which hasn't been resolved yet.
Starting kvm guest sometimes fails - qemu taking too long?
Guido Gunther committed a patch to fix this in upstream CVS.
libnuma: Warning: /sys not mounted or no numa system.
Bogus warning being printed at library load time because libvirt
now uses libnuma and libnuma was incorrectly complaining about
missing sysfs files on no-numa systems.
Neil Horman fixed in numactl-2.0.2-3.
Cole merged the VM 'Overview' tab into the 'Hardware' tab and renamed
it to 'Details'. He included some screenshots in his patch posting:
Cole also built a new package in rawhide which uses PolicyKit instead
Also, this bug gives some insight into plans for virt-manager:
root consolehelper breaks remote ssh virt-manager/libvirtd connections
Cole explains that this will be fixed when virt-manager no longer
runs as root.
David Cartwright points out this useful wiki page:
Marcelo has begun using patchwork to manage patches posted to the KVM
Too many patches isn't a bad complaint for a project to have, right?
Avi appears to have tagged kvm-84rc1 in git, so it looks like a new
release is on the way. Don't tell anyone I told you! :-)
Rich Jones took it upon himself to complete the merge reviews for
libvirt and gnome-applet-vm which had sat idle for 18 months:
That's the spirit!
Cole Robinson also took ownership of gnome-applet-vm from the original
author, Karel Zak, in the process.
Since package reviews are sometimes useful to refer back to, we've
collected together links to all the package reviews for virt related
Michael Young posted a test kernel RPM containing Dom0 support:
Some discussion on how complete the current dom0 patchset is. Gerd's
summary of the current situation is:
If it goes well .30 boots as dom0 kernel. I don't expect much of
the backend infrastructure makes it into .30 though (i.e. you need
patches to actually run guests). At this point the changes needed
might be non-invasive enough that it might be reasonable to carry
them as patches in the fedora kernel.
Michael is not to be deterred, though, and is preparing to start
pushing these packages through Koji.
Completely unrelated, Brad Smith brings attention to a common
complaint that "virsh start -c" is needed to allow users to easily get
to BIOS/pygrub screen at boot:
Xen and the Art of the NX Bit
Phill on the fedora-virt points out that the No Execute (NX) bit is
called the Execute Disable (XD) bit in Intel documentation:
By enabling NX, he was able to resolve his F10 guest boot failure:
Ian Campbell posted a patch upstream to fix the issue, but it looks
like it may not have been merged yet.
DOOM-O-METER: 192 open bugs last week, 193 this week. Hrmph!
qemu-kvm not working with evdev giving wrong key mappings
Some confusion about whether this is QEMU or evdev's fault,
which Dan Berrange cleared up with:
Yes, this *IS* a QEMU/KVM bug. It has hardcoded the assumption
that there is a 1-to-1 mapping of scancodes to keycodes. This just
happens to be true of the 'kbd' X driver, this is not true in
general, thus QEMU breaks with evdev. QEMU must be fixed.
block-rw-range-check.patch breaks qcow
Two issues were found with our single out of tree patch for
KVM. Eduardo, the hero that he is, promptly fixed both.
Guido Gunther supplied a patch and after consulting with the GNOME
HIG posted a second, better patch.
Qemu monitor "system_powerdown" command does not work
F11 alpha boots ugly (after install) on xen paravirt
Jeremy Fitzhardinge has fixed upstream.
Fix is in 2.6.29-rc4.git4
RFE: Support usb tablet in graphical setup via qemu
Xorg has been fixed to properly handle USB tablets.
We would like to be able to use USB tables by default for
Fedora guests, but some work needs to be done to check that
they are configured in absolute mode by default, rather than
On Tue, Feb 03, 2009 at 10:18:53PM -0200, Itamar Reis Peixoto wrote:
> I need to configure something to boot dom0 ?
Well you need to have dom0 kernel.. (there's no dom0 capable kernel in
fedora/rawhide yet, so you need to compile/build/get one yourself).
Nothing to configure really.. just configure grub.conf in the normal way to
boot Xen and dom0 kernel.
> On Tue, Feb 3, 2009 at 6:24 PM, Pasi Kärkkäinen <pasik(a)iki.fi> wrote:
> > Hello!
> > For those interested in testing pv_ops dom0 kernels it seems latest Xen
> > packages in rawhide contain bzImage dom0 kernel support in Xen hypervisor.
> > rawhide report: 20090203 changes
> > xen-3.3.1-3.fc11
> > ----------------
> > * Tue Feb 3 17:00:00 2009 Gerd Hoffmann <kraxel(a)redhat.com> - 3.3.1-3
> > - backport bzImage support for dom0 builder.
> > -- Pasi
> > _______________________________________________
> > Fedora-virt mailing list
> > Fedora-virt(a)redhat.com
> > https://www.redhat.com/mailman/listinfo/fedora-virt
> Itamar Reis Peixoto
> e-mail/msn: itamar(a)ispbrasil.com.br
> sip: itamar(a)ispbrasil.com.br
> skype: itamarjp
> icq: 81053601
> +55 11 4063 5033
> +55 34 3221 8599
2009-03-03 Feature freeze (25 days)
2009-03-10 Beta Freeze (32 days)
2009-04-14 Final freeze (67 days)
F11 Alpha, Take 3
The weird anaconda blocker which caused hang while moving to the
timezone screen was never figured out, but it turns out that later
rawhide didn't have this bug. Truly bizarre:
Because of this, rel-eng froze Alpha for the third time. After that
things got even more messy and even more confusing:
Another blocker cropped up, this time VNC installs failing:
In the end, after a mere two day slip, the installer gods were
appeased and Fedora 11 Alpha (blink) shipped to an eagerly waiting
On the virt side, the "disable kvmclock on machines with unsync TSC"
workaround ended up making it in at the last minute:
Unfortunately, what looks like a pvmmu bug didn't get fixed in time:
This causes some guest installs on an F11 Alpha host to oops during
heavy network activity.
Another edition of FWN, and another excellent virt section:
Thanks to Dale Bewley.
F11 Release Notes
Dale Bewley has begun work on the F11 release notes:
The "Unified Kernel Image" section provided some minior confusion on
the fedora-virt list:
'The kernel package in Fedora 11 supports booting as a guest domU,
but will not function as a dom0 until such support is provided
upstream. The most recent Fedora release with dom0 support is
To which one user responded:
As I write this, I am running F10 on a hardware virt enabled machine
... and have installed F10 and other distros as guests. Hence, the
statement above is unclear to me for F10 is running as both dom0 and
Am I not reading something right?
The confusion here is that folks use the terms "dom0" and domU to
refer to Xen, whereas the user in question is actually running
KVM. The equivalent terms for KVM tend to just be "KVM host" and "KVM
While on vacation, Daniel Veillard released libvirt 0.6.0:
* New features:
- thread safety of the API and event handling (Daniel Berrange)
- allow QEmu domains to survive daemon restart (Guido Gunther)
- extended logging capabilities (Daniel Veillard)
- support copy-on-write storage volumes (Daniel Berrange)
- support of storage cache control options for QEmu/KVM (Daniel Berrange)
Understandably, a number of post-release issues cropped up:
SELinux issue causing libvirtd launched dnsmasq to fail
This causes libvirt virtual networks to fail to start. The fix will
involve a small update to libvirt and selinux-policy changes.
libvir: Remote error : no call waiting for reply with serial 2
Dan Berrange has a fix for this upstream
Also, there are reports that if QEMU is too slow starting up, it can
A lively update was had as to whether the libvirt-0.6.0 update in F9
and F10 updates-testing should be withdrawn until some of the issues
are resolved. The current plan is to just fix the issues ASAP with
further updates to libvirt and selinux-policy.
All F11 virt features are now available under:
Glauber posted an initial qemu package, and solicited feedback. This
package did two things:
1) Updated to a snapshot of QEMU svn and
2) Split the package into multiple sub-package, with a sub-package
for each target architecture
There was some concern that updating to QEMU svn is not a good idea
until it is known that a QEMU release is due soon. By coincidence,
Anthony Liguori proposed doing a QEMU release at the end of February:
All involved were mightily relieved.
The discussion then moved on to a near-bikeshed argument about how to
split the package. In the end, it was agreed to go with:
- qemu-common, containing core bits
- qemu-system-* packages containing emulators for groups of
architectures, e.g. qemu-system-x86
- qemu-user, containing the Linux kernel ABI emulators for all
- qemu, a meta package which requires all sub-packages for
Further discussion was had on how to handle the building of BIOS
images. [Details redacted to avoid scaring the children]
The whole thread can be found here:
Also related, Eduardo added a kvm-tools sub-package containing
kvm_stat and kvmtrace:
The CONFIG_DMAR issue lumbered on this week. Firstly, Kyle McMartin's
patch to make intel_iommu=off the default is now in Andrew Morton's
queue for 2.6.29.
Progress on fixing the underlying issue is slow, but steady:
The focus is now on the fact that delaying for 2us after flushing the
I/O page tables from the CPU cache appears to resolve the
problem. Adar and Bhavesh from VMware are busily debugging the issue
on hardware which shows the problem. David Woodhouse and Chris Wright
have been working on patches to aid debugging. David has also included
Joerg Roedel's CONFIG_DMA_API_DEBUG patches in Fedora rawhide.
In other news, a minor circular locking issue was found and fixed:
Also, Weidong Han re-posted several device assignment fixes for QEMU
which had fallen through the cracks:
Some offline discussion was had around how to handle resetting PCI
devices before and after assignming them to guests. The details are
long and involved, but the core of the problem is that devices must be
reset before being assigned if they have been previously used in the
For example, with an e1000e NIC this manifests itself with "TX Unit
Hang" errors in the guest. Unfortunately, the preferred method for
resetting a PCI device (Function Level Reset, or FLR) is not
implemented by many devices. Alternatives include Secondary Bus Reset
and PCI PM D-state transitions. Both will need to be implemented in
addition to FLR support.
Some welcome interest was renewed in Xen Dom0 support on the
fedora-virt list this week with Gerd Hoffmann's addition of bzImage
loading support to the hypervisor:
Several fedora-virt users are now building and testing Dom0 kernels
using Jeremy Fitzhardinge's latest patch set.
Some changes to kernel architectures are on their way:
The main user-visible changes would be:
* 32-bit x86 would be built for i586 by default.
* The x86_64 kernel would be installed on compatible hardware,
even when installing a 32-bit OS
* The PAE kernel would be installed on other 32-bit hardware,
where it is supported
The proposal was accepted by FESCo and the kernel changes are already
gcc-4.4 is coming in F11:
Jakub Jelinek did a test mass-rebuild of rawhide this week:
It seemed to go pretty well. No virt packages had any problems.
DOOM-O-METER: 191 open bugs last week, 192 this week. Awww!
new_slab() oops from alloc_skb() with virtio_net, e1000 and rtl8139
Marcelo jumped on this and began investigating. At first it
looked like that it was a SLUB issue. However, the crafty
Marcelo came up with a crafty LD_PRELOAD lib for disabling
pvmmu for testing.
Using Marcelo's hack, James Laska confirmed that the culprit is
Both nVidia users who reported KVM MMU issues tryied to
reproduce without nvidia loaded.
The first reporter did indeed reproduce without nVidia and
Marcelo logged into his machine. The problem turned out to
be dodgy RAM though. Take home point - if something funny
is going on, try running memtest.
BUG_ON() in __shrink_dcache_sb() in rawhide x86_64 guest
Another oops reported by James Laska. Not yet investigated.
block-rw-range-check.patch breaks qcow2
Fedora's single not-upstream kvm-userspace patch causes us to fail
qemu bios images source code not packaged
The BIOS images in the qemu package aren't currently built
from source. This is fixed in the kvm package, so should be
resolved when we merge the two.
QEmu uses 100% of two assigned cores for XP guest
Perhaps the known issue with guest CPU usage accounting?
virbr0 loses it's routing when managed by NetworkManager
Reporter can't reproduce after an upgrade to f10, but it
seems fairly certain that the problem was something
The second request in recent times for libvirt autostart domains to
Sound under KVM requires exclusive access to the sound device
The debate continues on how to integrate with a user's sound
Issue with remote libvirt connections via SSH under virt-manager
Jon McCann noticed a virt-manager dialog:
Not Enough Free Space
The requested volume capacity will exceed the available
pool space when the volume is fully allocated. (4000M
requested capacity >2118M available)
I say "Yes!". Or maybe "No!". No, "Yes!" ... :-)
xen pv_ops DomU BUG() after alpha install
I've built a qemu image in here:
It contains the basic work to get qemu package updated. Basically,
we're talking about getting the qemu code from svn and creating multiple
packages, so one could grab just the architecture of interest.
Todo, is the creation of a meta-package to grab'em all at once.
This is also the first step in merging qemu and kvm packages, which
I'd like to do in two steps.
1) This one.
2) replace qemu-svn with kvm-userspace
The later is not as trivial as it seems. Nobody exercices this path
usually, and we're likely to find bugs that prevent the build to finish
(Well, I tested, and there _are_ bugs ;-) ). But I believe we can have
it sorted out.
If no one has any major opposition to the state of things, I intend to
commit it to CVS tomorrow (with some minor changes), and put it into
rawhide so we can start having a more serious testing out of it.
We're a bit behind on writing up descriptions for the virt features
coming in F11, but here's the first two pages:
Define a mapping of SASL authentication into the VNC protocol, and
implement it for QEMU and GTK-VNC, providing strongly authenticated,
securely encrypted remote access of virtual guest consoles.
Assign PCI devices from your KVM host machine to guest virtual
machines. A common example is assigning a network card to a guest.
Please jump in with comments, testing, wiki edits, etc.