On 03/20/2015 10:39 PM, Pavel Hrdina wrote:
From: Luyao Huang
Commit e105dc9 fix setting vcpus for offline domain, but forget check
if vm is active when pass VIR_DOMAIN_AFFECT_LIVE flags.
# virsh setvcpus test3 4 --live
error: Failed to create controller cpu for group: No such file or dir
On 03/18/2015 08:36 AM, Daniel P. Berrange wrote:
> Currently the QEMU processes inherit their core dump rlimit
> from libvirtd, which is really suboptimal. This change allows
> their limit to be directly controller from qemu.conf instead.
> ---
> src/libvirt_private.syms | 2 ++
> sr
On 03/18/2015 08:36 AM, Daniel P. Berrange wrote:
> When parsing integer values, we only used 'long' data type
> in the virConfValue struct. This is insufficiently large
> to deal with things like guest memory sizes on 32-bit platforms
> which are using PAE for addressing > 4 GB of RAM.
> ---
>
On Mon, Mar 23, 2015 at 05:32:30PM +0100, Marc-André Lureau wrote:
> The QEMU machine vmport option allows to set the VMWare IO port
> emulation. This emulation is useful for absolute pointer input when the
> guest has vmware input drivers, and is enabled by default for kvm.
>
> However it is unne
On Sat, Mar 21, 2015 at 11:45:03AM +0100, Guido Günther wrote:
> When using QEMU's 9pfs the target "dir" element is not necessarily an
> absolute path. If it is not validation currently fails with the
> misleaading
In fact the target "dir" is not actually a directory at all. With
KVM is is an opaq
On 03/17/2015 07:41 AM, Ján Tomko wrote:
> Create a sorted array of virtio-serial controllers.
> Each of the elements contains the controller index
> and a bitmap of available ports.
>
> Buses are not tracked, because they aren't supported by QEMU.
> ---
> src/conf/domain_addr.c | 348
>
On 03/17/2015 07:41 AM, Ján Tomko wrote:
> Instead of always using controller 0 and incrementing port number,
> respect the maximum port numbers of controllers and use all of them.
>
> Ports for virtio consoles are quietly reserved, but not formatted
> (neither in XML nor on QEMU command line).
These two functions are identical, so no sense in having the
duplication. I resisted the temptation to replace calls to
virNetDevMacVLanDelete() with calls to virNetlinkDelLink() just in
case some mythical future platform has macvtap devices that aren't
managed with netlink (or in case we some day
https://bugzilla.redhat.com/show_bug.cgi?id=1125755
reported that a stray bridge device was left on the system when a
libvirt network failed to start due to an illegal iptables rule caused
by bad config. Apparently the reason this was happening was that
NetworkManager was noticing immediately when
Just as it is possible to delete a bridge device with the netlink
RTM_DELLINK message, one can be created with the RTM_NEWLINK
message. Because of differences in the format of the message, it's not
as straightforward as with virNetlinkDelLink() to create a single
utility function that can be used t
libvirt has always used the netlink RTM_DELLINK message to delete
macvtap/macvlan devices, but it can actually be used to delete other
types of network devices, such as bonds and bridges. This patch makes
virNetDevMacVLanDelete() available as a generic function so it can
intelligibly be called to d
This patch series comes out of a comment in:
https://bugzilla.redhat.com/show_bug.cgi?id=1125755
that pointed out that sioctl(SIOCBRDELBR) won't delete a bridge
interface if it is IFF_UP, but the netlink RTM_DELLINK message doesn't
care - it will delete it anyway. In these patches we switch to u
On 3/23/15, Stefan Hajnoczi wrote:
> I have CCed the libvirt mailing list, since KVM is a component here but
> your question seems to be mainly about libvirt, virt-manager,
> virt-install, etc.
Apologies for posting to the wrong list, I assumed it would be KVM
related as the guest could run but c
>From 73172eeed1fcffcfae088a3059fbca0689b7437f Mon Sep 17 00:00:00 2001
From: Amy Fong
Date: Mon, 23 Mar 2015 13:44:03 -0400
Subject: [PATCH] libvirt: 'src/cpu/cpu_map.xml': No such file or directory'
In some circumstances where the build tree differs from the source,
libvirt's compile will try t
The virNodeDeviceDettach API only works on PCI devices.
Originally added by commit 10d3272e, but the API never
supported USB devices.
Reported by: Martin Polednik
---
docs/formatdomain.html.in | 19 +--
tools/virsh.pod | 17 -
2 files changed, 17 insert
The QEMU machine vmport option allows to set the VMWare IO port
emulation. This emulation is useful for absolute pointer input when the
guest has vmware input drivers, and is enabled by default for kvm.
However it is unnecessary for Spice-enabled VM, since the agent already
handles absolute pointe
Bug 1: The the next element in the pcidevs is skipped after we
virPCIDeviceListDel the previous element.
Bug 2: virHostdevNetConfigRestore is called for store the hostdevs
which may be used by other domain.
Signed-off-by: Huanle Han
---
src/util/virhostdev.c | 25 +
1 fi
On Thu, Mar 19, 2015 at 18:41:31 -0400, John Ferlan wrote:
>
>
> On 03/17/2015 10:19 AM, Peter Krempa wrote:
> > This version includes review feedback changes from John and also fixes the
> > element documentation and code that calculates it to support
> > possible
> > non-NUMA configs with me
On Mon, Mar 23, 2015 at 09:41:23 -0400, John Ferlan wrote:
>
>
> On 03/23/2015 09:23 AM, Peter Krempa wrote:
> > In the last section if the function determines that the config is
> > invalid when QEMU doesn't support the memory device the JSON config
> > object would be returned even if it doesn'
On Fri, Mar 20, 2015 at 07:40:21 -0400, John Ferlan wrote:
>
> >>> I have tested your series with our qemu memory hot remove patch series,
> >>> here would be a possible error.
> >>>
> >>> When hotplug a memory device, its size has been aligned. So the
> >>> compare for
> >>> size here would fail
On 03/23/2015 09:23 AM, Peter Krempa wrote:
> In the last section if the function determines that the config is
> invalid when QEMU doesn't support the memory device the JSON config
> object would be returned even if it doesn't make sense.
>
> Assign the object to be returned only on success.
>
Since all APIs are also RPC calls, we automatically get all APIs covered
with thread jobs.
Signed-off-by: Jiri Denemark
---
daemon/remote.c| 1 +
src/locking/lock_daemon_dispatch.c | 1 +
src/qemu/qemu_domain.c | 1 +
src/rpc/gendispatch.pl | 6 +-
Every thread created as a worker thread within a pool gets a name
according to virThreadPoolJobFunc name.
Signed-off-by: Jiri Denemark
---
src/libvirt_private.syms | 2 +-
src/util/virthreadpool.c | 44 +++-
src/util/virthreadpool.h | 14 +-
3
Signed-off-by: Jiri Denemark
---
po/POTFILES.in | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/po/POTFILES.in b/po/POTFILES.in
index fd8bf33..ab6ae3b 100644
--- a/po/POTFILES.in
+++ b/po/POTFILES.in
@@ -178,6 +178,8 @@ src/util/virconf.c
src/util/vircrypto.c
src/util/vi
This is very helpful when we want to log and report why we could not
acquire a state change lock. Reporting what job keeps it locked helps
with understanding the issue. Moreover, after calling
virDomainGetControlInfo, it's possible to tell whether libvirt is just
stuck somewhere within the API (or
Each thread can use a thread local variable to keep the name of a job
which is currently running in the job.
The virThreadJobSetWorker API is supposed to be called once by any
thread which is used as a worker, i.e., it is waiting in a pool, woken
up to do a job, and returned back to the pool.
The
https://bugzilla.redhat.com/show_bug.cgi?id=853839
Jiri Denemark (8):
POTFILES.in: Sort
Add support for tracking thread jobs
Force usage of virThreadCreate
virThread: Set thread job
virThreadPool: Set thread worker name
Set thread job for every RPC call
qemu: Track the API which star
Automatically assign a job to every thread created by virThreadCreate.
The name of the virThreadFunc function passed to virThreadCreate is used
as the job or worker name in case no name is explicitly passed.
Signed-off-by: Jiri Denemark
---
src/libvirt_private.syms | 2 +-
src/util/virthread.c
Whenever we fail to acquire a job, we can report how long ago it was
locked by another API.
https://bugzilla.redhat.com/show_bug.cgi?id=853839
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_domain.c | 20 ++--
src/qemu/qemu_domain.h | 2 ++
2 files changed, 20 insertions(+), 2
We want all threads to be set as workers or to have a job assigned to
them, which can easily be achieved in virThreadCreate wrapper to
pthread_create. Let's make sure we always use the wrapper.
Signed-off-by: Jiri Denemark
---
cfg.mk | 9 +
src/nwfilter/nwfi
In the last section if the function determines that the config is
invalid when QEMU doesn't support the memory device the JSON config
object would be returned even if it doesn't make sense.
Assign the object to be returned only on success.
---
src/qemu/qemu_command.c | 7 ---
1 file changed,
On Thu, Mar 19, 2015 at 18:40:36 -0400, John Ferlan wrote:
>
>
> On 03/17/2015 10:20 AM, Peter Krempa wrote:
> > Add code to hot-add memory devices to running qemu instances.
> > ---
> >
> > Notes:
> > Version 3:
> > - added comment to clarify that @mem is always consumed by
> > qemuDom
On 03/23/2015 06:12 AM, Pavel Hrdina wrote:
> On Mon, Mar 23, 2015 at 12:29:42PM +0100, Ján Tomko wrote:
>
> Just a question: where is the typo mentioned in $subject?
>
>> by rewriting it completely from:
>> error: unsupported configuration: virtio only support device address
>> type 'PCI'
'supp
On Mon, Mar 23, 2015 at 12:29:42PM +0100, Ján Tomko wrote:
Just a question: where is the typo mentioned in $subject?
> by rewriting it completely from:
> error: unsupported configuration: virtio only support device address
> type 'PCI'
>
> to:
>
> error: unsupported configuration: virtio disk c
On 03/23/2015 06:10 PM, Peter Krempa wrote:
On Thu, Mar 19, 2015 at 18:13:04 +0800, Luyao Huang wrote:
5bba61f introduce a issue : when start a vm with settings and
restart libvirtd, numa settings will disappear. Because when parse the
vm states file, there is no node in "/domain/cpu/numa" thi
(just ccing the other tools maintainers, in particular Stefano who knows
what this stuff is supposed to do...)
On Fri, 2015-03-20 at 17:10 -0600, Jim Fehlig wrote:
> Recent testing on large memory systems revealed a bug in the Xen xl
> tool's freemem() function. When autoballooning is enabled, fr
by rewriting it completely from:
error: unsupported configuration: virtio only support device address
type 'PCI'
to:
error: unsupported configuration: virtio disk cannot have an address of type
drive
Since we now support CCW addresses as well.
---
src/qemu/qemu_command.c | 5 +++--
1 file chang
On 20.03.2015 16:01, Boris Fiuczynski wrote:
> When no model is specified in the domain definition for
> a scsi controller and the architectur is s390 than virtio-scsi
> is set as default model.
>
> Signed-off-by: Boris Fiuczynski
> Reviewed-by: Daniel Hansel
> Reviewed-by: Stefan Zimmermann
>
On 23.03.2015 10:31, zhang bo wrote:
> The problem we encountered is:
> 1) use qemu-agent to shutdown a domain.
> 2) libvirt will block in "qemuAgentShutdown()", if the domain itself got
> stucked when it powers-off.
>
> It's the *guest domain*'s fault that it stucks when shutdown. However, we
>
On Fri, 2015-03-20 at 22:13 +0100, Marek Marczykowski-Górecki wrote:
> I'll definitely do. But above raises a question - how can I set extra
> arguments for qemu? In case of qemu in dom0, it's not a problem because
> I can create some wrapper script. But in case of qemu in stubdom, the
> only way i
On Sat, Mar 21, 2015 at 01:50:46AM +0800, Emmanuel Noobadmin wrote:
> Running
> 3.18.9-200.fc21.x86_64
> qemu 2:2.1.3-3.fc21
> libvirt 1.2.9.2-1.fc21
> System is a Thinkpad X250 with Intel i7-5600u Broadwell GT2
>
> I'm trying to replace the Win7 installation on my laptop with Fedora
> 21 and virt
On Mon, Mar 23, 2015 at 09:58:40AM +0100, Ján Tomko wrote:
On Fri, Mar 20, 2015 at 02:40:05PM +0100, Martin Kletzander wrote:
I've left out "ok" (which should be OK, but it looks like nobody cares
and there were *so many* false positives in the code), and there's
"(in)dependant" where it sugges
On Thu, Mar 19, 2015 at 18:13:04 +0800, Luyao Huang wrote:
> 5bba61f introduce a issue : when start a vm with settings and
> restart libvirtd, numa settings will disappear. Because when parse the
> vm states file, there is no node in "/domain/cpu/numa" this place.
>
> Change to use ./cpu/numa ins
On Mon, Mar 23, 2015 at 10:11:14AM +0100, Natanael Copa wrote:
> Always add udev linker flags when WITH_UDEV is enabled to avoid
> underlinking.
>
> See commit 43dbcb15 (interface: always build all available backends)
>
> Signed-off-by: Natanael Copa
> ---
> src/Makefile.am | 3 +--
> 1 file ch
On 13.03.2015 01:34, Michael Chapman wrote:
> Commit cf54c60699833b3791a5d0eb3eb5a1948c267f6b introduced the ability
> to create missing storage volumes during migration. For network disks,
> however, we may not necessarily be able to detect whether they already
> exist -- there is no straight-forw
The problem we encountered is:
1) use qemu-agent to shutdown a domain.
2) libvirt will block in "qemuAgentShutdown()", if the domain itself got
stucked when it powers-off.
It's the *guest domain*'s fault that it stucks when shutdown. However, we could
not handle the domain in libvirt anymore, ex
Always add udev linker flags when WITH_UDEV is enabled to avoid
underlinking.
See commit 43dbcb15 (interface: always build all available backends)
Signed-off-by: Natanael Copa
---
src/Makefile.am | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/src/Makefile.am b/src/Makefil
On 20.03.2015 20:59, John Ferlan wrote:
>
>
> On 03/16/2015 12:42 PM, Michal Privoznik wrote:
>> This function does not make any sense now, that network driver is
>> (almost) dropped. I mean, previously, when threads were
>> serialized, this function was there to check, if no other network
>> wit
On Fri, Mar 20, 2015 at 02:40:05PM +0100, Martin Kletzander wrote:
>
> I've left out "ok" (which should be OK, but it looks like nobody cares
> and there were *so many* false positives in the code), and there's
> "(in)dependant" where it suggests s/dant/dent/ and I couldn't find
> anywhere that th
On Fri, Mar 20, 2015 at 09:51:23AM -0600, Eric Blake wrote:
On 03/20/2015 06:21 AM, Martin Kletzander wrote:
Wikipedia's list of common misspellings [1] has a machine-readable
version. This patch fixes those misspellings mentioned in the list
which don't have multiple right variants (as e.g. "a
On Mon, Mar 23, 2015 at 11:33:50AM +0800, Yanbing Du wrote:
>
>
> On 03/20/2015 10:39 PM, Pavel Hrdina wrote:
> > From: Luyao Huang
> >
> > We will ignore --maximum option when only use setvcpus with
> > this option, like this (this error is another issue):
> >
> > # virsh setvcpus test3 --max
On 21.03.2015 01:06, Jayashree Deshpande wrote:
> ---
> libvirt-designer/libvirt-designer-domain.c | 194
> ---
> libvirt-designer/libvirt-designer-domain.h | 41 --
> libvirt-designer/libvirt-designer-internal.c | 3 +-
> libvirt-designer/libvirt-designer-inter
On 20.03.2015 20:23, Noel Burton-Krahn wrote:
> Hi Michal,
>
> I think issuing a libvirt migrate to a host where the network disks
> don't already exist would be a prequisite failure. Libvirt can never
> copy a network disk, but it shouldn't fail trying to migrate an existing
> domain that contai
53 matches
Mail list logo