On Thu, Oct 12, 2017 at 14:57:36 -0500, Eric Blake wrote:
> On 10/12/2017 02:07 PM, Peter Krempa wrote:
> > The backing store indexes were not bound to the storage sources in any
> > way. To allow us to bind a given alias to a given storage source we need
> > to save the index in virStorageSource.
On Thu, Oct 12, 2017 at 15:34:28 -0500, Eric Blake wrote:
> On 10/12/2017 02:07 PM, Peter Krempa wrote:
> > Express a properly terminated backing chain by putting a
> > virStorageSource of type VIR_STORAGE_TYPE_NONE in the chain. The newly
> > used helpers simplify this greatly.
> >
> > The change
directory /var/lib alway is Persistence directory, but in redhat system,
/var/run is memory directory.
our running domain xml is saved at /var/run/libvirt/qemu. so if we cold reset
system,
the /var/run/libvirt/qemu directory is clear, but
/var/lib/libvirt/qemu/domain-*** is saved. so there
have
On 10/12/2017 08:48 AM, Erik Skultety wrote:
> On Tue, Oct 10, 2017 at 04:20:06PM -0400, John Ferlan wrote:
>> Rather than separate callbacks for the NumOfNetworks, GetNames, and
>> Export functions, let's combine them all into one multi-purpose callback.
>>
>> Signed-off-by: John Ferlan
>> ---
On Thu, Oct 12, 2017 at 05:10:43PM +0200, Jiri Denemark wrote:
>On Mon, Oct 09, 2017 at 10:00:03 +0800, Chao Fan wrote:
>> The command "info migrate" of qemu outputs the dirty-pages-rate during
>> migration, but page size is different in different architectures. So
>> page size should be output to
On Fri, Oct 13, 2017 at 09:13:07AM +0800, Chao Fan wrote:
>On Thu, Oct 12, 2017 at 05:10:43PM +0200, Jiri Denemark wrote:
>>On Mon, Oct 09, 2017 at 10:00:03 +0800, Chao Fan wrote:
>>> The command "info migrate" of qemu outputs the dirty-pages-rate during
>>> migration, but page size is different in
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> xsaveopt is artificially removed from the host to test disabled feature
> which is only included in QEMU's version of the CPU model.
>
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c| 1 +
> ...x86_64-c
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c | 2 +-
> tests/cputestdata/x86_64-cpuid-Xeon-E3-1245.json | 299
> +++
> 2 files changed, 300 insertions(+), 1 deletion(-)
>
There's ano
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> arat is now enabled even if the hardware does not support it.
>
> Signed-off-by: Jiri Denemark
> ---
> .../x86_64-cpuid-Core-i7-2600-disabled.xml | 1 -
> .../x86_64-cpuid-Core-i7-2600-enabled.xml | 1 +
> .../cputestdata/x86_
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> xsaves is supported by current QEMU/KVM on this CPU.
Current as of when? Does this cause migration issues to have one
version support and another not? What about save/restore type operations?
I'm not against this, just curious. Not sure I ever w
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> This CPU was incorrectly detected as SandyBridge before because the
> number of additional elements was the same for both
> SandyBridge and Westmere CPU models, but SandyBridge is newer (the CPU
> signature does not help here because it doesn't matc
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c| 1 +
> .../x86_64-cpuid-Xeon-E7-4830-disabled.xml | 5 +
> .../x86_64-cpuid-Xeon-E7-4830-enabled.xml | 8 +
> .../x86_64-cpuid-Xeon-E7-
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> The unavailable features do not make any difference in this case,
> because this is a SandyBridge CPU which has an empty list of unavailable
> features.
>
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c |
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Gather query-cpu-definitions results and use them for testing CPU model
> usability blockers in CPUID to virCPUDef translation.
>
> Signed-off-by: Jiri Denemark
> ---
> src/qemu/qemu_capabilities.c| 2 +-
> src/qemu/qemu_capspriv.h|
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> To make the code reusable by other tests.
>
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c | 97
> +++--
> 1 file changed, 60 insertions(+), 37 deletions(-)
>
> diff --git a/tests/cpute
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> When testing cpuDecode for computing guest CPU definition from CPUID
> data (the CPU definition reported by domain capabilities), we need to
> use CPU models (and their usability blockers) from QEMU if they are
> available to cpuDecode in the same wa
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Signed-off-by: Jiri Denemark
> ---
> src/libvirt_private.syms | 1 +
> 1 file changed, 1 insertion(+)
>
This should either be merged earlier when changing the function
arguments or in reality in the "next" patch which actually uses it.
Reviewed-
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Add a helper that will retrun true if a virStorageSource has backing
s/retrun/return/
> store. This will also aid in refactoring of the code further down.
> ---
The commit message mentions adding only one helper, but doesn't name it.
I like to menti
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Express a properly terminated backing chain by putting a
> virStorageSource of type VIR_STORAGE_TYPE_NONE in the chain. The newly
> used helpers simplify this greatly.
>
> The change fixes a bug as formatting an incomplete backing chain and
> parsing i
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> If the actual result does not match our expectation, the tests would
> not correctly show the difference if a CPU feature is disabled in the
> expected result and the actual result does not mention it at all. The
> test could complain about an unrela
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Various version of json_reformat use different number of spaces for
> indenting. Let's use a simple python reformatter to gain full control
> over the formatting for consistent results.
>
> Signed-off-by: Jiri Denemark
> ---
> tests/cputestdata/cp
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Storage driver uses virStorageSource only partially to store it's
> configuration but fully when parsing backing files of storage volumes.
> This patch sets the 'type' field to a value other than
> VIR_STORAGE_TYPE_NONE so that further patches can add a
@title, consider:
Replace bool with cpuTestCPUIDJson enum
It's just the "json bool" that looks strange.
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> We will soon need to handle more than two values.
>
> Signed-off-by: Jiri Denemark
> ---
> tests/cputest.c | 79
> ++
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> chain->path may be NULL e.g. for NDB drives, so the check needs to avoid
s/NDB/NBD/ ?
> dereferencing the path in such case
> ---
> src/util/virstoragefile.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
Reviewed-by: Eric Blake
>
> diff
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Set the type so that the iterators will work after upcoming
> modification.
> ---
> tests/virstoragetest.c | 1 +
> 1 file changed, 1 insertion(+)
Could almost be squashed to the previous patch; but I'm also fine with
it separate.
Reviewed-by: Eric B
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> The backing store indexes were not bound to the storage sources in any
> way. To allow us to bind a given alias to a given storage source we need
> to save the index in virStorageSource. The backing store ids are now
> generated when detecting the backi
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> We can now use the backing store ID directly rather than counting the
> number of images seen while looking up images.
> ---
> src/util/virstoragefile.c | 18 +++---
> 1 file changed, 3 insertions(+), 15 deletions(-)
>
I guess if we've al
Hi,
On Thu, Oct 12, 2017 at 05:25:34PM +0100, Daniel P. Berrange wrote:
> On Thu, Oct 12, 2017 at 05:53:21PM +0200, Guido Günther wrote:
> > Domains fail to start without fuse like
> >
> > error: internal error: guest failed to start: fuse: device not found, try
> > 'modprobe fuse' first
> >
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> When decoding CPUID data to virCPUDef we need to be careful about using
> a CPU model which cannot be directly used on the current host. Normally,
> libvirt would notice the features which prevent the model from being
> usable and it would disable th
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Signed-off-by: Jiri Denemark
> ---
> src/cpu/cpu_x86.c | 61
> ---
> 1 file changed, 31 insertions(+), 30 deletions(-)
>
Reviewed-by: John Ferlan
John
--
libvir-list mailing list
libvir-lis
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> If a given CPU model is supported by the hypervisor, we want to know
> more about it, e.g., what features may block its usage on the current
> host and such details are stored in the virDomainCapsCPUModelsPtr list
> which virCPUModelIsAllowed uses to
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Use VIR_STEAL_PTR to remove conditional cleanup.
> ---
> src/conf/domain_conf.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
Reviewed-by: Eric Blake
>
> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
> index 54be9028d.
From: Wim ten Have
Add support for describing NUMA distances in a domain's
XML description.
Below is an example of a 4 node setup:
From: Wim ten Have
Test a bidirectional xen-xl domxml to and from native for numa
support administration as brought under this patch series.
Signed-off-by: Wim ten Have
---
.../test-fullvirt-vnuma-autocomplete.cfg | 26 +++
.../test-fullvirt-vnuma-autocomplete.xml | 85
From: Wim ten Have
This patch generates a NUMA distance-aware libxl description from the
information extracted from a NUMA distance-aware libvirt XML file.
By default, if no NUMA node distance information is supplied in the
libvirt XML file, this patch uses the distances 10 for local and 20
for
From: Wim ten Have
This patch converts NUMA configurations between the Xen libxl
configuration file format and libvirt's XML format.
XML HVM domain on a 4 node (2 cores/socket) configuration:
From: Wim ten Have
This patch extends guest domain administration adding support to
advertise node sibling distances when configuring NUMA guests also
referred to as vNUMA (Virtual NUMA).
NUMA (Non-Uniform Memory Access), a method of configuring a cluster
of nodes within a single multiprocessing
Add a helper that will retrun true if a virStorageSource has backing
store. This will also aid in refactoring of the code further down.
---
src/conf/domain_conf.c| 2 +-
src/conf/storage_conf.c | 5 +++--
src/libvirt_private.syms | 2 ++
src/qemu/qemu_
The backing store indexes were not bound to the storage sources in any
way. To allow us to bind a given alias to a given storage source we need
to save the index in virStorageSource. The backing store ids are now
generated when detecting the backing chain.
Since we don't re-detect the backing chai
Express a properly terminated backing chain by putting a
virStorageSource of type VIR_STORAGE_TYPE_NONE in the chain. The newly
used helpers simplify this greatly.
The change fixes a bug as formatting an incomplete backing chain and
parsing it back would end up in expressing a terminated chain sin
The file object is needed when formatting the command line, but it makes
nesting of the objects less easy for use with blockdev. Separate the
wrapping into the 'file' object into a helper used specifically for disk
sources in the old code path.
---
src/qemu/qemu_block.c | 14 +++---
src/
Set the type so that the iterators will work after upcoming
modification.
---
tests/virstoragetest.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/virstoragetest.c b/tests/virstoragetest.c
index ffebd4dc1..ad4514871 100644
--- a/tests/virstoragetest.c
+++ b/tests/virstoragetest.c
@@ -5
On 10/12/2017 02:07 PM, Peter Krempa wrote:
> Index will remain an internal property even if we allow backing store
> parsing from the XML, so we need to allow backing store without it in
> the schema.
> ---
> docs/schemas/domaincommon.rng | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git
See individual patches please.
Peter Krempa (3):
qemu: command: Separate wrapping of disk backend props to 'file'
object
qemu: block: Add support for file/block/dir storage to JSON disk src
generator
util: storagefile: Track whether a virStorageSource was auto-detected
src/qemu/qem
When formatting an inactive or migratable XML we will need to suppress
backing chain members which were detected from the disk to keep
semantics straight. This means we need to record, whether a
virStorageSource originates from autodetection.
---
src/util/virstoragefile.c | 3 +++
src/util/virstor
qemuBlockStorageSourceGetBackendProps now is able to format the JSON
definition for regular storage too.
---
src/qemu/qemu_block.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/qemu/qemu_block.c b/src/qemu/qemu_block.c
index 4c0f675ca..75a8d6200 100644
--- a/src/qemu/qemu_block.c
+
Storage driver uses virStorageSource only partially to store it's
configuration but fully when parsing backing files of storage volumes.
This patch sets the 'type' field to a value other than
VIR_STORAGE_TYPE_NONE so that further patches can add a terminator
element to backing chains without breaki
chain->path may be NULL e.g. for NDB drives, so the check needs to avoid
dereferencing the path in such case
---
src/util/virstoragefile.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/util/virstoragefile.c b/src/util/virstoragefile.c
index 9568a5068..818a679de 100644
---
We can now use the backing store ID directly rather than counting the
number of images seen while looking up images.
---
src/util/virstoragefile.c | 18 +++---
1 file changed, 3 insertions(+), 15 deletions(-)
diff --git a/src/util/virstoragefile.c b/src/util/virstoragefile.c
index 2b9
Use VIR_STEAL_PTR to remove conditional cleanup.
---
src/conf/domain_conf.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index 54be9028d..f7c9160b4 100644
--- a/src/conf/domain_conf.c
+++ b/src/conf/domain_conf.c
@@ -8315,
Index will remain an internal property even if we allow backing store
parsing from the XML, so we need to allow backing store without it in
the schema.
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.
This series fixes and improves storage of backing store indexes and of
the completness of the backing chain in the XML.
Peter Krempa (9):
conf: domain: Simplify return from backing store parser
conf: Make backing store index optional
util: storage: Store backing chain index in virStorageSour
On 10/12/2017 04:37 AM, Wim ten Have wrote:
On Fri, 6 Oct 2017 08:49:46 -0600
Jim Fehlig wrote:
On 09/08/2017 08:47 AM, Wim Ten Have wrote:
From: Wim ten Have
Add libvirtd NUMA cell domain administration functionality to
describe underlying cell id sibling distances in full fashion
when con
On Thu, Oct 12, 2017 at 05:53:21PM +0200, Guido Günther wrote:
> Domains fail to start without fuse like
>
> error: internal error: guest failed to start: fuse: device not found, try
> 'modprobe fuse' first
> Failure in libvirt_lxc startup: no error
>
> so check for it too.
>
> References:
Domains fail to start without fuse like
error: internal error: guest failed to start: fuse: device not found, try
'modprobe fuse' first
Failure in libvirt_lxc startup: no error
so check for it too.
References:
https://ci.debian.net/data/autopkgtest/unstable/amd64/libv/libvirt/20171012_1059
On Mon, Oct 09, 2017 at 10:00:03 +0800, Chao Fan wrote:
> The command "info migrate" of qemu outputs the dirty-pages-rate during
> migration, but page size is different in different architectures. So
> page size should be output to calculate dirty pages in bytes.
>
> Page size is already implement
On Mon, Sep 18, 2017 at 22:27:52 +0530, Nitesh Konkar wrote:
> libvirtd throws unhandled signal 11 on ppc while running
> virsh cpu-compare with missing model tag in the xml. This
> patch errors out in such situation.
>
> Signed-off-by: Nitesh Konkar
> ---
> src/cpu/cpu_ppc64.c | 6 ++
> 1 f
On Mon, Oct 09, 2017 at 02:57:38PM -0400, John Ferlan wrote:
> Let's use the ObjectRWLockable for the various list lock mgmt.
> Only time need Write lock will be for Add/Remove logic.
>
> Signed-off-by: John Ferlan
> ---
Reviewed-by: Erik Skultety
--
libvir-list mailing list
libvir-list@redhat.
On Thu, Oct 12, 2017 at 03:46:25PM +0200, Erik Skultety wrote:
> On Tue, Oct 10, 2017 at 09:54:28AM -0400, John Ferlan wrote:
> > Rather than separate functions for NumOfDevices, GetNames, and
> > ListExport - let's converge the code into one handler.
> >
> > Signed-off-by: John Ferlan
> > ---
>
>
On Mon, Oct 09, 2017 at 02:57:39PM -0400, John Ferlan wrote:
> Rather than separate functions for NumOfSecrets, GetUUIDs, and
> ListExport - let's converge the code into one handler.
>
> Signed-off-by: John Ferlan
> ---
NACK to the idea, linking [1] to avoid repeating the comments.
[1] https://w
When migration fails, QEMU may provide a description of the error in
the reply to query-migrate QMP command. We can fetch this error and use
it instead of the generic "unexpectedly failed" message.
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_driver.c | 3 ++-
src/qemu/qemu_migration.c
On Tue, Oct 10, 2017 at 09:54:28AM -0400, John Ferlan wrote:
> Rather than separate functions for NumOfDevices, GetNames, and
> ListExport - let's converge the code into one handler.
>
> Signed-off-by: John Ferlan
> ---
NACK, for the very same reasons stated in [1], IOW we really should not make
On Tue, Oct 10, 2017 at 09:54:27AM -0400, John Ferlan wrote:
> It only returns 0 or 1 anyway, let's be realistic
>
> Signed-off-by: John Ferlan
For patches 1-2:
Reviewed-by: Erik Skultety
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
On Tue, Oct 10, 2017 at 04:20:06PM -0400, John Ferlan wrote:
> Rather than separate callbacks for the NumOfNetworks, GetNames, and
> Export functions, let's combine them all into one multi-purpose callback.
>
> Signed-off-by: John Ferlan
> ---
NACK to the idea of consolidating all of them into a
On Mon, Oct 09, 2017 at 10:16:48 +0200, Marc Hartmayer wrote:
> On Thu, Oct 05, 2017 at 02:11 PM +0200, Jiri Denemark
> wrote:
> > But it's going to be a bit complicated because we ask QEMU what the
> > host CPU is and the interface we used would need to be enhanced to
> > give us different resul
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> Signed-off-by: Jiri Denemark
> ---
> src/qemu/qemu_capabilities.c | 51
> ++--
> 1 file changed, 49 insertions(+), 2 deletions(-)
>
> diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> query-cpu-definitions QMP command returns a list of unavailable features
> which prevent CPU models from being usable on the current host. So far
> we only checked whether the list was empty to mark CPU models as
> (un)usable. This patch parses all u
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> The "preferred" parameter is not used by any caller of cpuDecode
> anymore. It's only used internally in cpu_x86 to implement cpuBaseline.
>
> Signed-off-by: Jiri Denemark
> ---
> src/cpu/cpu.c | 15 ---
> src/cpu/cpu.
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> All APIs which expect a list of CPU models supported by hypervisors were
> switched from char **models and int models to just accept a pointer to
> virDomainCapsCPUModels object stored in domain capabilities. This avoids
> the need to transform virDo
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> When a hypervisor marks a CPU model as unusable on the current host, it
> may also give us a list of features which prevent the model from being
> usable. Storing this list in virDomainCapsCPUModel will help the CPU
> driver with creating a host-mode
On 10/04/2017 10:58 AM, Jiri Denemark wrote:
> The API makes a deep copy of a NULL-terminated string list.
>
> Signed-off-by: Jiri Denemark
> ---
> src/util/virstring.c | 37 +
> src/util/virstring.h | 3 +++
> 2 files changed, 40 insertions(+)
>
> diff --
On Wed, Oct 11, 2017 at 12:11 PM +0200, Jiri Denemark
wrote:
> Signed-off-by: Jiri Denemark
> ---
> src/conf/cpu_conf.c | 40 +++-
> src/conf/cpu_conf.h | 4
> src/libvirt_private.syms | 1 +
> 3 files changed, 32 insertions(+), 13 deletions(
On Tue, Oct 10, 2017 at 04:20:05PM -0400, John Ferlan wrote:
> Let's use the RWObjectLockable for the various list lock mgmt.
> Only time need Write lock will be for Add, Remove, and Prune logic.
>
> Signed-off-by: John Ferlan
Reviewed-by: Erik Skultety
--
libvir-list mailing list
libvir-list@r
On Fri, 6 Oct 2017 08:49:46 -0600
Jim Fehlig wrote:
> On 09/08/2017 08:47 AM, Wim Ten Have wrote:
> > From: Wim ten Have
> >
> > Add libvirtd NUMA cell domain administration functionality to
> > describe underlying cell id sibling distances in full fashion
> > when configuring HVM guests.
>
On Wed, Oct 11, 2017 at 12:11 PM +0200, Jiri Denemark
wrote:
> The host CPU definition from host capabilities may contain features
> unknown to QEMU. Thus whenever we want to use this CPU definition, we
> have to filter the unknown features.
>
> https://bugzilla.redhat.com/show_bug.cgi?id=1495171
On Wed, Oct 11, 2017 at 19:42:36 +0200, Marc Hartmayer wrote:
> On Wed, Oct 11, 2017 at 12:11 PM +0200, Jiri Denemark
> wrote:
> > The new function is called qemuProcessRefreshCPU.
> >
> > Signed-off-by: Jiri Denemark
> > ---
> > src/qemu/qemu_process.c | 66
> > +++
On Wed, Oct 11, 2017 at 05:03:05PM -0400, John Ferlan wrote:
> Need to free @groups in the parent on success similar to other
> APIs (virFile*) which use virGetGroupList and virFork.
>
> Reported by Coverity.
>
> Signed-off-by: John Ferlan
> ---
> src/util/vircommand.c | 1 +
> 1 file changed, 1
77 matches
Mail list logo