date:20131218

When test-locking is enabled, CFLAGS have -Dbool=char, which makes
pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
double negation fixes the build breaker.

Signed-off-by: Martin Kletzander mklet...@redhat.com
---

Notes:
I still can't build with '--enable-test-locking' and I'm not sure this
is the proper way to fix it.  Also, the code wasn't touched for
months, so I'm not pushing it as a build-breaker.

 src/util/virfile.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/util/virfile.c b/src/util/virfile.c
index 631cd06..88674ec 100644
--- a/src/util/virfile.c
+++ b/src/util/virfile.c
@@ -262,7 +262,7 @@ virFileWrapperFdNew(int *fd, const char *name, unsigned int 
flags)
 }

 VIR_FORCE_CLOSE(*fd);
-*fd = pipefd[output];
+*fd = pipefd[!!output];
 return ret;

 error:
-- 
1.8.5.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] qemu: fix typo PCi = PCI

On 16.12.2013 17:31, Martin Kletzander wrote:
 Signed-off-by: Martin Kletzander mklet...@redhat.com
 ---
  src/qemu/qemu_process.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)
 
 diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c
 index bd9546e..d0fde54 100644
 --- a/src/qemu/qemu_process.c
 +++ b/src/qemu/qemu_process.c
 @@ -3829,7 +3829,7 @@ int qemuProcessStart(virConnectPtr conn,
   * Normally PCI addresses are assigned in the virDomainCreate
   * or virDomainDefine methods. We might still need to assign
   * some here to cope with the question of upgrades. Regardless
 - * we also need to populate the PCi address set cache for later
 + * we also need to populate the PCI address set cache for later
   * use in hotplug
   */
  if (virQEMUCapsGet(priv-qemuCaps, QEMU_CAPS_DEVICE)) {
 @@ -4560,7 +4560,7 @@ int qemuProcessAttach(virConnectPtr conn 
 ATTRIBUTE_UNUSED,
   * Normally PCI addresses are assigned in the virDomainCreate
   * or virDomainDefine methods. We might still need to assign
   * some here to cope with the question of upgrades. Regardless
 - * we also need to populate the PCi address set cache for later
 + * we also need to populate the PCI address set cache for later
   * use in hotplug
   */
  if (virQEMUCapsGet(priv-qemuCaps, QEMU_CAPS_DEVICE)) {
 

ACK

Michal

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] libxl: libxl_get_max_cpus returning a libxl error from 4.4 onward

On mer, 2013-12-18 at 07:39 +0100, Martin Kletzander wrote:
 On Tue, Dec 17, 2013 at 06:43:25PM +0100, Dario Faggioli wrote:
  diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
  index 692c3b7..a31b094 100644
  --- a/src/libxl/libxl_driver.c
  +++ b/src/libxl/libxl_driver.c
  @@ -1101,9 +1101,11 @@ libxlConnectGetMaxVcpus(virConnectPtr conn, const 
  char *type ATTRIBUTE_UNUSED)
 
   cfg = libxlDriverConfigGet(driver);
   ret = libxl_get_max_cpus(cfg-ctx);
  -/* libxl_get_max_cpus() will return 0 if there were any failures,
  -   e.g. xc_physinfo() failing */
  -if (ret == 0)
  +/* On failure, libxl_get_max_cpus() will return ERROR_FAIL from Xen 4.4
  + * onward, but it ever returning 0 is obviously wrong too (and it is
  + * what happens, on failure, on Xen 4.3 and earlier). Therefore, a 
  'less
  + * or equal' is the catchall we want. */
  +if (ret = 0)
   ret = -1;
 
   virObjectUnref(cfg);
 
 
 Good to change that, thanks, but the call to:
 virBitmapNew(libxl_get_max_nodes(priv-ctx)) // in the same file
 should be fixed as well, especially since virBitmapNew takes an
 unsigned argument.  Could you modify that bit as well?
 
Right! I checkd for other uses of *_max_cpus() and forgot to do the same
for *_max_nodes()! :-/

I'll do that, thanks for looking at the patch.

Regards,
Dario

-- 
This happens because I choose it to happen! (Raistlin Majere)
-
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems RD Ltd., Cambridge (UK)



signature.asc
Description: This is a digitally signed message part
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] lxc: simplify command when connecting lxc multi console

On Wed, Dec 18, 2013 at 04:54:27PM +0800, Chen Hanxiao wrote:
 From: Chen Hanxiao chenhanx...@cn.fujitsu.com
 
 If we config more than one consoles for container,
 we always named them as consoleN.
 We had to type a complex option --devname consoleN.
 This patch enables option --devname N
 to be equal to  --devname consoleN.
 
 Signed-off-by: Chen Hanxiao chenhanx...@cn.fujitsu.com
 ---
  src/lxc/lxc_driver.c | 12 +++-
  1 file changed, 11 insertions(+), 1 deletion(-)

NACK, I don't see any compelling reason todo this.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] lxc: simplify command when connecting lxc multi console

On Wed, Dec 18, 2013 at 09:52:05AM +, Daniel P. Berrange wrote:
 On Wed, Dec 18, 2013 at 04:54:27PM +0800, Chen Hanxiao wrote:
  From: Chen Hanxiao chenhanx...@cn.fujitsu.com
  
  If we config more than one consoles for container,
  we always named them as consoleN.
  We had to type a complex option --devname consoleN.
  This patch enables option --devname N
  to be equal to  --devname consoleN.
  
  Signed-off-by: Chen Hanxiao chenhanx...@cn.fujitsu.com
  ---
   src/lxc/lxc_driver.c | 12 +++-
   1 file changed, 11 insertions(+), 1 deletion(-)
 
 NACK, I don't see any compelling reason todo this.

I should say though if we want simplification in virsh code that
would be acceptablee

   virsh console -n 2 myguestname

would say connect to the 2nd console in the XML. virsh would
get the XML description and find the 2nd console and extract
the alias string from it, and then call the libvirt API with
that string.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] libvirt live modify

Hi

Can we modify a Virtual machine while its running, I want to update the
discription and title

Br.

Umar
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCHv3] qemu: ask for -enable-fips when FIPS is required

On 12/17/13 19:36, Eric Blake wrote:
 On a system that is enforcing FIPS, most libraries honor the
 current mode by default.  Qemu, on the other hand, refused to
 honor FIPS mode unless you add the '-enable-fips' command
 line option; worse, this option is not discoverable via QMP,
 and is only present on binaries built for Linux.  So, if we
 detect FIPS mode, then we unconditionally ask for FIPS; either
 qemu is new enough to have the option and then correctly
 cripple insecure VNC passwords, or it is so old that we are
 correctly avoiding a FIPS violation by preventing qemu from
 starting.  Meanwhile, if we don't detect FIPS mode, then
 omitting the argument is safe whether the qemu has the option
 (but it would do nothing because FIPS is disabled) or whether
 qemu lacks the option (including in the case where we are not
 running on Linux).
 
 The testsuite was a bit interesting: we don't want our test
 to depend on whether it is being run in FIPS mode, so I had
 to tweak things to set the capability bit outside of our
 normal interaction with capability parsing.
 
 This fixes https://bugzilla.redhat.com/show_bug.cgi?id=1035474
 
 * src/qemu/qemu_capabilities.h (QEMU_CAPS_ENABLE_FIPS): New bit.
 * src/qemu/qemu_capabilities.c (virQEMUCapsInitQMP): Conditionally
 set capability according to detection of FIPS mode.
 * src/qemu/qemu_command.c (qemuBuildCommandLine): Use it.
 * tests/qemucapabilitiestest.c (testQemuCaps): Conditionally set
 capability to test expected output.
 * tests/qemucapabilitiesdata/caps_1.2.2-1.caps: Update list.
 * tests/qemucapabilitiesdata/caps_1.6.0-1.caps: Likewise.
 
 Signed-off-by: Eric Blake ebl...@redhat.com
 ---
 
 v3: use virFileReadAll correctly
 
  src/qemu/qemu_capabilities.c | 27 ++-
  src/qemu/qemu_capabilities.h |  1 +
  src/qemu/qemu_command.c  |  2 ++
  tests/qemucapabilitiesdata/caps_1.2.2-1.caps |  1 +
  tests/qemucapabilitiesdata/caps_1.6.0-1.caps |  1 +
  tests/qemucapabilitiestest.c | 20 +++-
  6 files changed, 46 insertions(+), 6 deletions(-)

ACK. I verified that the detection works correctly when the file is
present and contains the expected data.

Peter




signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

On 12/18/13 11:41, Umar Draz wrote:
 Hi
 
 Can we modify a Virtual machine while its running, I want to update the
 discription and title

You can use the virDomainSetMetadata and virDomainGetMetadata APIs.

http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

Or the virsh desc command from the commandline.

Peter



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Fix crash in virsystemdtest with dbus 1.7.6

2013-12-18 Thread Cédric Bosdonnat

D-bus introduced some changes in its locking code. Overriding the init
function skips the new locking init and thus crashes later in libvirt
test. Removing the function makes the test pass again.
---
 tests/virsystemdmock.c | 5 -
 1 file changed, 5 deletions(-)

diff --git a/tests/virsystemdmock.c b/tests/virsystemdmock.c
index b8fc031..b3b82b4 100644
--- a/tests/virsystemdmock.c
+++ b/tests/virsystemdmock.c
@@ -31,11 +31,6 @@ void dbus_connection_set_change_sigpipe(dbus_bool_t 
will_modify_sigpipe ATTRIBUT
 {
 }
 
-dbus_bool_t dbus_threads_init_default(void)
-{
-return 1;
-}
-
 DBusConnection *dbus_bus_get(DBusBusType type ATTRIBUTE_UNUSED,
  DBusError *error ATTRIBUTE_UNUSED)
 {
-- 
1.8.4.4

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Peter,

Thanks for the response, Would you please help how I can use
*virDomainSetMetadata* with PHP or any url for help?

Br.

Umar


On Wed, Dec 18, 2013 at 4:11 PM, Peter Krempa pkre...@redhat.com wrote:

 On 12/18/13 11:41, Umar Draz wrote:
  Hi
 
  Can we modify a Virtual machine while its running, I want to update the
  discription and title

 You can use the virDomainSetMetadata and virDomainGetMetadata APIs.

 http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

 Or the virsh desc command from the commandline.

 Peter




-- 
Umar Draz
Network Architect
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Block info query: Add flag to allow failure if not active

2013-12-18 Thread John Ferlan

Currently the qemuDomainGetBlockInfo will return allocation == physical
for most backing stores. For a qcow2 block backed device it's possible
to return the highest lv extent allocated from qemu for an active guest.
That is a value where allocation != physical and one would hope be less.
However, if the guest is not running, then the code falls back to returning
allocation == physical. This turns out to be problematic for rhev which
monitors the size of the backing store. During a migration, before the
VM has been started on the target and while it is deemed inactive on the
source, allocation is returned as physical triggering the code to extend
the file unnecessarily.

Thus, this patch will check a new flag to the call to determine whether
a failure should be returned to indicate to the caller that the code is
unable to fetch the allocation value. This will allow the caller to decide
how to proceed.

Note that I chose not to return a failure as a more general rule mostly
because of backwards compatability and current expectations reasons.
Changing a default return action could be an unexpected action. Although
returning allocation == physical perhaps is incorrect, it is how the
code has been functioning. Having a 'virsh domblkinfo' return a failure
for some inactive domains while succeeding for others could be confusing.
A flag forces the caller to decide how to operate.

As it turns out, there seems to be quite a bit of history regarding this.
Details can be found in the BZ:

https://bugzilla.redhat.com/show_bug.cgi?id=1040507
Signed-off-by: John Ferlan jfer...@redhat.com
---
NOTE:
Trying to keep this as simple as possible since it needs to be backported.
If someone has a better idea for a flag name - I'm open to suggestions...

 include/libvirt/libvirt.h.in | 23 ---
 src/libvirt.c| 11 ++-
 src/qemu/qemu_driver.c   | 33 +++--
 3 files changed, 57 insertions(+), 10 deletions(-)

diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in
index 6f79c49..420642b 100644
--- a/include/libvirt/libvirt.h.in
+++ b/include/libvirt/libvirt.h.in
@@ -2095,7 +2095,8 @@ int virDomainBlockResize 
(virDomainPtr dom,
 
 /** virDomainBlockInfo:
  *
- * This struct provides information about the size of a block device backing 
store
+ * This struct provides information about the size of a block device
+ * backing store
  *
  * Examples:
  *
@@ -2108,13 +2109,29 @@ int virDomainBlockResize 
(virDomainPtr dom,
  *
  *  - qcow2 file in filesystem
  *   * capacity: logical size from qcow2 header
- *   * allocation, physical: logical size of the file / highest qcow 
extent (identical)
+ *   * allocation, physical: logical size of the file /
+ *   highest qcow extent (identical)
  *
  *  - qcow2 file in a block device
  *   * capacity: logical size from qcow2 header
- *   * allocation: highest qcow extent written
+ *   * allocation: highest qcow extent written for an active domain
  *   * physical: size of the block device container
  */
+
+/**
+ * virDomainGetBlockInfoFlags:
+ *
+ * Flags available for virDomainGetBlockInfo().
+ */
+typedef enum {
+VIR_DOMAIN_BLOCK_CHECK_ACTIVE = 1  0, /* If not active, then return
+ * failure when attempting to
+ * fetch data from an inactive
+ * domain, such as allocation
+ * from a qcow2 block device
+ */
+} virDomainGetBlockInfoFlags;
+
 typedef struct _virDomainBlockInfo virDomainBlockInfo;
 typedef virDomainBlockInfo *virDomainBlockInfoPtr;
 struct _virDomainBlockInfo {
diff --git a/src/libvirt.c b/src/libvirt.c
index d15d617..2159eaf 100644
--- a/src/libvirt.c
+++ b/src/libvirt.c
@@ -9027,7 +9027,7 @@ error:
  * @domain: a domain object
  * @disk: path to the block device, or device shorthand
  * @info: pointer to a virDomainBlockInfo structure allocated by the user
- * @flags: extra flags; not used yet, so callers should always pass 0
+ * @flags: bitwise-OR of supported virDomainGetBlockInfoFlags
  *
  * Extract information about a domain's block device.
  *
@@ -9038,6 +9038,15 @@ error:
  * can be found by calling virDomainGetXMLDesc() and inspecting
  * elements within //domain/devices/disk.
  *
+ * If the VIR_DOMAIN_BLOCK_CHECK_ACTIVE flag is set, then for any block
+ * information data that could differ depending on whether the domain was
+ * active or not, the hypervisor will return a failure when the domain
+ * is deemed to be inactive or unable to provide the data, such as may be
+ * the case during a migration. For example, the 'allocation' value for
+ * a qcow2 block device backed storage can return the highest extent written.
+ * However, this can only be determined if the

Re: [libvirt] libvirt live modify

On 18.12.2013 12:27, Umar Draz wrote:
 Hi Peter,
 
 Thanks for the response, Would you please help how I can
 use *virDomainSetMetadata* with PHP or any url for help?
 
 Br.
 
 Umar

[CC'ing Michal Novotny - the libvirt-php maintainer]

I don't think you can invoke those APIs via PHP. But I expect them to be
able to add easily. Michal?

Michal

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Michal and Umar,
it's correct I can't invoke this APIs via PHP yet. Please note that
libvirt-php is not automatically generated when doing libvirt build and
it's a separate project. It's rather easy to implement it to the
libvirt-php so if you want to implement it, feel free to implement or if
you can wait a day or two until I have some time I can implement it (the
project is not officially supported and it's done in the spare time mainly).

I'd just like to ask you whether there are some more APIs you'd like me
to implement when I'll be implementing it. Are there any other APIs
you'd like?

Thanks,
Michal

On 12/18/2013 12:59 PM, Michal Privoznik wrote:
 On 18.12.2013 12:27, Umar Draz wrote:
 Hi Peter,

 Thanks for the response, Would you please help how I can
 use *virDomainSetMetadata* with PHP or any url for help?

 Br.

 Umar
 [CC'ing Michal Novotny - the libvirt-php maintainer]

 I don't think you can invoke those APIs via PHP. But I expect them to be
 able to add easily. Michal?

 Michal

-- 
Michal Novotny minov...@redhat.com, RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

HI Michal,

Ok , I will wait. I hope you will get sometime to do this.

Br.

Umar


On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny minov...@redhat.com wrote:

 Hi Michal and Umar,
 it's correct I can't invoke this APIs via PHP yet. Please note that
 libvirt-php is not automatically generated when doing libvirt build and
 it's a separate project. It's rather easy to implement it to the
 libvirt-php so if you want to implement it, feel free to implement or if
 you can wait a day or two until I have some time I can implement it (the
 project is not officially supported and it's done in the spare time
 mainly).

 I'd just like to ask you whether there are some more APIs you'd like me
 to implement when I'll be implementing it. Are there any other APIs
 you'd like?

 Thanks,
 Michal

 On 12/18/2013 12:59 PM, Michal Privoznik wrote:
  On 18.12.2013 12:27, Umar Draz wrote:
  Hi Peter,
 
  Thanks for the response, Would you please help how I can
  use *virDomainSetMetadata* with PHP or any url for help?
 
  Br.
 
  Umar
  [CC'ing Michal Novotny - the libvirt-php maintainer]
 
  I don't think you can invoke those APIs via PHP. But I expect them to be
  able to add easily. Michal?
 
  Michal

 --
 Michal Novotny minov...@redhat.com, RHCE, Red Hat
 Virtualization | libvirt-php bindings | php-virt-control.org




-- 
Umar Draz
Network Architect
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Umar,
that's fine. Give one or two days for this please. Maybe I'll implement
earlier but I'm a bit overloaded after my PTO so I can't promise it will
be earlier.

Also, what APIs would you like to implement?

I'm thinking of implementing these two:

http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

Would you like any other APIs?

Thanks,
Michal

On 12/18/2013 01:15 PM, Umar Draz wrote:
 HI Michal,

 Ok , I will wait. I hope you will get sometime to do this.

 Br.

 Umar


 On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny minov...@redhat.com
 mailto:minov...@redhat.com wrote:

 Hi Michal and Umar,
 it's correct I can't invoke this APIs via PHP yet. Please note that
 libvirt-php is not automatically generated when doing libvirt
 build and
 it's a separate project. It's rather easy to implement it to the
 libvirt-php so if you want to implement it, feel free to implement
 or if
 you can wait a day or two until I have some time I can implement
 it (the
 project is not officially supported and it's done in the spare
 time mainly).

 I'd just like to ask you whether there are some more APIs you'd
 like me
 to implement when I'll be implementing it. Are there any other APIs
 you'd like?

 Thanks,
 Michal

 On 12/18/2013 12:59 PM, Michal Privoznik wrote:
  On 18.12.2013 12:27, Umar Draz wrote:
  Hi Peter,
 
  Thanks for the response, Would you please help how I can
  use *virDomainSetMetadata* with PHP or any url for help?
 
  Br.
 
  Umar
  [CC'ing Michal Novotny - the libvirt-php maintainer]
 
  I don't think you can invoke those APIs via PHP. But I expect
 them to be
  able to add easily. Michal?
 
  Michal

 --
 Michal Novotny minov...@redhat.com mailto:minov...@redhat.com,
 RHCE, Red Hat
 Virtualization | libvirt-php bindings | php-virt-control.org
 http://php-virt-control.org




 -- 
 Umar Draz
 Network Architect

-- 
Michal Novotny minov...@redhat.com, RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Fix race leading to crash when setting up dbus watches

Currently the virDBusAddWatch does

  virEventAddHandle(fd, flags,
virDBusWatchCallback,
watch, NULL);
  dbus_watch_set_data(watch, info, virDBusWatchFree);

Unfortunately this is racy - since the event loop is in a
different thread, the virDBusWatchCallback method may be
run before we get to calling dbus_watch_set_data. We must
reverse the order of these calls

See https://bugzilla.redhat.com/show_bug.cgi?id=885445

Signed-off-by: Daniel P. Berrange berra...@redhat.com
---
 src/util/virdbus.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/util/virdbus.c b/src/util/virdbus.c
index 4e4c267..a0cbbfe 100644
--- a/src/util/virdbus.c
+++ b/src/util/virdbus.c
@@ -238,15 +238,15 @@ static dbus_bool_t virDBusAddWatch(DBusWatch *watch,
 # else
 fd = dbus_watch_get_fd(watch);
 # endif
+dbus_watch_set_data(watch, info, virDBusWatchFree);
 info-bus = (DBusConnection *)data;
 info-watch = virEventAddHandle(fd, flags,
 virDBusWatchCallback,
 watch, NULL);
 if (info-watch  0) {
-VIR_FREE(info);
+dbus_watch_set_data(watch, NULL, NULL);
 return 0;
 }
-dbus_watch_set_data(watch, info, virDBusWatchFree);
 
 return 1;
 }
-- 
1.8.4.2

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Rename virDomainGetRootFilesystem to virDomainGetFilesystemForTarget

The virDomainGetRootFilesystem method can be generalized to allow
any filesystem path to be obtained.

While doing this, start a new test case for purpose of testing various
helper methods in the domain_conf.{c,h} files, such as this one.

Signed-off-by: Daniel P. Berrange berra...@redhat.com
---
 .gitignore |   1 +
 src/conf/domain_conf.c |   5 +-
 src/conf/domain_conf.h |   3 +-
 src/libvirt_private.syms   |   2 +-
 src/lxc/lxc_container.c|   2 +-
 src/lxc/lxc_process.c  |   2 +-
 tests/Makefile.am  |   6 ++
 tests/domainconfdata/getfilesystem.xml |  28 
 tests/domainconftest.c | 120 +
 tests/testutils.c  |  57 
 tests/testutils.h  |   5 ++
 11 files changed, 225 insertions(+), 6 deletions(-)
 create mode 100644 tests/domainconfdata/getfilesystem.xml
 create mode 100644 tests/domainconftest.c

diff --git a/.gitignore b/.gitignore
index d5a6cf5..cbb3e8a 100644
--- a/.gitignore
+++ b/.gitignore
@@ -147,6 +147,7 @@
 /tests/commandtest
 /tests/conftest
 /tests/cputest
+/tests/domainconftest
 /tests/domainsnapshotxml2xmltest
 /tests/esxutilstest
 /tests/eventtest
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index 0079234..dd36026 100644
--- a/src/conf/domain_conf.c
+++ b/src/conf/domain_conf.c
@@ -17827,12 +17827,13 @@ virDiskNameToBusDeviceIndex(virDomainDiskDefPtr disk,
 }
 
 virDomainFSDefPtr
-virDomainGetRootFilesystem(virDomainDefPtr def)
+virDomainGetFilesystemForTarget(virDomainDefPtr def,
+const char *path)
 {
 size_t i;
 
 for (i = 0; i  def-nfss; i++) {
-if (STREQ(def-fss[i]-dst, /))
+if (STREQ(def-fss[i]-dst, path))
 return def-fss[i];
 }
 
diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h
index 647d115..ea7f1a1 100644
--- a/src/conf/domain_conf.h
+++ b/src/conf/domain_conf.h
@@ -2541,7 +2541,8 @@ int virDiskNameToBusDeviceIndex(virDomainDiskDefPtr disk,
 int *busIdx,
 int *devIdx);
 
-virDomainFSDefPtr virDomainGetRootFilesystem(virDomainDefPtr def);
+virDomainFSDefPtr virDomainGetFilesystemForTarget(virDomainDefPtr def,
+  const char *path);
 int virDomainFSIndexByName(virDomainDefPtr def, const char *name);
 int virDomainVideoDefaultType(const virDomainDef *def);
 int virDomainVideoDefaultRAM(const virDomainDef *def, int type);
diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms
index 2dbb8f8..a870e9c 100644
--- a/src/libvirt_private.syms
+++ b/src/libvirt_private.syms
@@ -225,7 +225,7 @@ virDomainFSTypeFromString;
 virDomainFSTypeToString;
 virDomainFSWrpolicyTypeFromString;
 virDomainFSWrpolicyTypeToString;
-virDomainGetRootFilesystem;
+virDomainGetFilesystemForTarget;
 virDomainGraphicsAuthConnectedTypeFromString;
 virDomainGraphicsAuthConnectedTypeToString;
 virDomainGraphicsDefFree;
diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c
index c6bdc8c..c32b085 100644
--- a/src/lxc/lxc_container.c
+++ b/src/lxc/lxc_container.c
@@ -1823,7 +1823,7 @@ static int lxcContainerChild(void *data)
 if (lxcContainerSetID(vmDef)  0)
 goto cleanup;
 
-root = virDomainGetRootFilesystem(vmDef);
+root = virDomainGetFilesystemForTarget(vmDef, /);
 
 if (argv-nttyPaths) {
 const char *tty = argv-ttyPaths[0];
diff --git a/src/lxc/lxc_process.c b/src/lxc/lxc_process.c
index cc9c1a2..f0190a0 100644
--- a/src/lxc/lxc_process.c
+++ b/src/lxc/lxc_process.c
@@ -934,7 +934,7 @@ virLXCProcessReadLogOutput(virDomainObjPtr vm,
 static int
 virLXCProcessEnsureRootFS(virDomainObjPtr vm)
 {
-virDomainFSDefPtr root = virDomainGetRootFilesystem(vm-def);
+virDomainFSDefPtr root = virDomainGetFilesystemForTarget(vm-def, /);
 
 if (root)
 return 0;
diff --git a/tests/Makefile.am b/tests/Makefile.am
index 568b7a0..90fb3a0 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -65,6 +65,7 @@ EXTRA_DIST =  \
commanddata \
confdata \
cputestdata \
+   domainconfdata \
domainschemadata \
domainschematest \
domainsnapshotschematest \
@@ -141,6 +142,7 @@ test_programs = virshtest sockettest \
 virportallocatortest \
sysinfotest \
virstoragetest \
+   domainconftest \
$(NULL)
 
 if WITH_REMOTE
@@ -893,6 +895,10 @@ sysinfotest_SOURCES = \
sysinfotest.c testutils.h testutils.c
 sysinfotest_LDADD = $(LDADDS)
 
+domainconftest_SOURCES = \
+   domainconftest.c testutils.h testutils.c
+domainconftest_LDADD = $(LDADDS)
+
 fdstreamtest_SOURCES = \
fdstreamtest.c testutils.h testutils.c
 fdstreamtest_LDADD = $(LDADDS)
diff --git a/tests/domainconfdata/getfilesystem.xml 
b/tests/domainconfdata/getfilesystem.xml
new

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Ian Campbell

On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:
  
  Might this libxl fix be relevant:
  commit 5420f26507fc5c9853eb1076401a8658d72669da
  Author: Jim Fehlig jfeh...@suse.com
  Date:   Fri Jan 11 12:22:26 2013 +
  
  libxl: Set vfb and vkb devid if not done so by the caller
  
  Other devices set a sensible devid if the caller has not done 
  so.
  Do the same for vfb and vkb.  While at it, factor out the 
  common code
  used to determine a sensible devid, so it can be used by other
  libxl__device_*_add functions.
  
  Signed-off-by: Jim Fehlig jfeh...@suse.com
  Acked-by: Ian Campbell ian.campb...@citrix.com
  Committed-by: Ian Campbell ian.campb...@citrix.com
  
  and a follow up in dfeccbeaa. Although the comment implies that nic's
  were already correctly assigning a devid if the caller specified -1, so
  I don't know why it doesn't work for you :-(
 
 Ok, yes, the commit above indeed changes libxl__device_nic_add to call
 libxl__device_nextid for the devid... Just how is this actually called.
 Maybe not sufficient but git grep libxl__device_nic_add in the xen code only
 shows the definition and a declaration in libxl_internal.h to me...

I have a feeling a macro might be involved...

Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
add the eventual function names in comments to provide grep fodder

Ian.

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Michal,

Thanks for your efforts and help. If I need any other API I will definitely
ask you.

Br.

Umar


On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny minov...@redhat.com wrote:

 Hi Umar,
 that's fine. Give one or two days for this please. Maybe I'll implement
 earlier but I'm a bit overloaded after my PTO so I can't promise it will
 be earlier.

 Also, what APIs would you like to implement?

 I'm thinking of implementing these two:

 http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
 http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

 Would you like any other APIs?

 Thanks,
 Michal

 On 12/18/2013 01:15 PM, Umar Draz wrote:
  HI Michal,
 
  Ok , I will wait. I hope you will get sometime to do this.
 
  Br.
 
  Umar
 
 
  On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny minov...@redhat.com
  mailto:minov...@redhat.com wrote:
 
  Hi Michal and Umar,
  it's correct I can't invoke this APIs via PHP yet. Please note that
  libvirt-php is not automatically generated when doing libvirt
  build and
  it's a separate project. It's rather easy to implement it to the
  libvirt-php so if you want to implement it, feel free to implement
  or if
  you can wait a day or two until I have some time I can implement
  it (the
  project is not officially supported and it's done in the spare
  time mainly).
 
  I'd just like to ask you whether there are some more APIs you'd
  like me
  to implement when I'll be implementing it. Are there any other APIs
  you'd like?
 
  Thanks,
  Michal
 
  On 12/18/2013 12:59 PM, Michal Privoznik wrote:
   On 18.12.2013 12:27, Umar Draz wrote:
   Hi Peter,
  
   Thanks for the response, Would you please help how I can
   use *virDomainSetMetadata* with PHP or any url for help?
  
   Br.
  
   Umar
   [CC'ing Michal Novotny - the libvirt-php maintainer]
  
   I don't think you can invoke those APIs via PHP. But I expect
  them to be
   able to add easily. Michal?
  
   Michal
 
  --
  Michal Novotny minov...@redhat.com mailto:minov...@redhat.com,
  RHCE, Red Hat
  Virtualization | libvirt-php bindings | php-virt-control.org
  http://php-virt-control.org
 
 
 
 
  --
  Umar Draz
  Network Architect

 --
 Michal Novotny minov...@redhat.com, RHCE, Red Hat
 Virtualization | libvirt-php bindings | php-virt-control.org




-- 
Umar Draz
Network Architect
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] lxc: simplify command when connecting lxc multi console

2013-12-18 Thread Chen Hanxiao

 -Original Message-
 From: Daniel P. Berrange [mailto:berra...@redhat.com]
 Sent: Wednesday, December 18, 2013 5:55 PM
 To: Chen Hanxiao
 Cc: libvir-list@redhat.com
 Subject: Re: [libvirt] [PATCH] lxc: simplify command when connecting lxc multi
 console

 On Wed, Dec 18, 2013 at 09:52:05AM +, Daniel P. Berrange wrote:
  On Wed, Dec 18, 2013 at 04:54:27PM +0800, Chen Hanxiao wrote:
   From: Chen Hanxiao chenhanx...@cn.fujitsu.com

   If we config more than one consoles for container,
   we always named them as consoleN.
   We had to type a complex option --devname consoleN.
   This patch enables option --devname N
   to be equal to  --devname consoleN.

   Signed-off-by: Chen Hanxiao chenhanx...@cn.fujitsu.com
   ---
src/lxc/lxc_driver.c | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)

  NACK, I don't see any compelling reason todo this.

 I should say though if we want simplification in virsh code that
 would be acceptablee

virsh console -n 2 myguestname

 would say connect to the 2nd console in the XML. virsh would
 get the XML description and find the 2nd console and extract
 the alias string from it, and then call the libvirt API with
 that string.

Thanks, I see.

 Daniel
 --
 |: http://berrange.com  -o-
 http://www.flickr.com/photos/dberrange/ :|
 |: http://libvirt.org  -o-
 http://virt-manager.org :|
 |: http://autobuild.org   -o-
 http://search.cpan.org/~danberr/ :|
 |: http://entangle-photo.org   -o-
 http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] python pypi package - backport older version?

2013-12-18 Thread Jason Antman

Hello,

First off, thanks for the wonderful project, and thanks even more for
all of the excellent work on the Python bindings lately.

I'm working on a small project to collect data on kvm hosts/guests via a
python collector client and display the data in a web interface. It
works beautifully when using my various distros' OS packages for
libvirt-python, but I've hit a bit of a snag when trying to do automated
testing. I'm quite happy to see libvirt-python packaged as a python
package and on pypi, but the version there is 1.2.0, which requires
libvirt = 0.9.11.

Unfortunately, I was planning to test my python app on travis-ci, which
runs Ubuntu Precise (12.04 LTS), and the latest version of libvirt for
that is 0.9.8 (package python-libvirt 0.9.8-2ubuntu17.13).

If I were to submit a patch, would there be any interest in me
backporting the setup.py/packaging work to an older version of the
python bindings - say 0.9.8 - and getting it published on pypi?

Thanks,
Jason Antman
CMGdigital
Atlanta, GA

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] configure: make --with-test-suite work

On 12/18/2013 12:03 AM, Martin Kletzander wrote:
 Our option '--with-test-suite' could have never worked since it was
 defined as AC_ARG_ENABLE([with-test-suite], ...), thus working only as
 '--enable-with-test-suite', but documented in configure.ac as
 AC_HELP_STRING([--with-test-suite], ...).
 In my optinion, the help string is as it should be, but the option is

s/optinion/opinion/

 wrong.
 
 Signed-off-by: Martin Kletzander mklet...@redhat.com
 ---
  configure.ac | 16 
  1 file changed, 8 insertions(+), 8 deletions(-)

Indeed, you could use './configure --disable-with-test-suite' to change
the behavior, but that just looks wrong.  It's been broken since the
introduction in commit 3a2fc27.

ACK.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On 12/18/2013 02:35 AM, Martin Kletzander wrote:
 When test-locking is enabled, CFLAGS have -Dbool=char, which makes

Huh?  bool should never be defined to char in CFLAGS; it should only be
replaced if gnulib thinks the compiler is too old.

 pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
 double negation fixes the build breaker.

I don't get how this could possibly make a difference.  The 'output'
variable is declared bool, and then only ever assigned 'false' or
'true', which if bool is replaced by gnulib are still guaranteed to be 0
and 1, so using it directly as an array index is still safe.

 
 Signed-off-by: Martin Kletzander mklet...@redhat.com
 ---
 
 Notes:
 I still can't build with '--enable-test-locking' and I'm not sure this
 is the proper way to fix it.  Also, the code wasn't touched for
 months, so I'm not pushing it as a build-breaker.

NACK.  I need more details about the actual failure you are seeing, but
this is not the right patch.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Stefan Bader

On 18.12.2013 13:27, Ian Campbell wrote:
 On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:

 Might this libxl fix be relevant:
 commit 5420f26507fc5c9853eb1076401a8658d72669da
 Author: Jim Fehlig jfeh...@suse.com
 Date:   Fri Jan 11 12:22:26 2013 +
 
 libxl: Set vfb and vkb devid if not done so by the caller
 
 Other devices set a sensible devid if the caller has not done 
 so.
 Do the same for vfb and vkb.  While at it, factor out the 
 common code
 used to determine a sensible devid, so it can be used by other
 libxl__device_*_add functions.
 
 Signed-off-by: Jim Fehlig jfeh...@suse.com
 Acked-by: Ian Campbell ian.campb...@citrix.com
 Committed-by: Ian Campbell ian.campb...@citrix.com
 
 and a follow up in dfeccbeaa. Although the comment implies that nic's
 were already correctly assigning a devid if the caller specified -1, so
 I don't know why it doesn't work for you :-(

 Ok, yes, the commit above indeed changes libxl__device_nic_add to call
 libxl__device_nextid for the devid... Just how is this actually called.
 Maybe not sufficient but git grep libxl__device_nic_add in the xen code 
 only
 shows the definition and a declaration in libxl_internal.h to me...
 
 I have a feeling a macro might be involved...
 
 Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
 add the eventual function names in comments to provide grep fodder

Oh duh, yeah. So in DEFINE_DEVICE_ADD a libxl_device_nic_add is created which
calls to libxl__device_nic_add. When I look for the single _ version I find a
call from xl_cmdimpl.c and its public declaration in libxl.h.
So I guess the bug is that libvirt in the libxl driver never seems to do so. Ok,
thanks a lot for digging the out the DEFINE. As nice those are to create similar
functions from template, grep and tags fail to be helpful with them.

-Stefan
 
 Ian.
 
 --
 libvir-list mailing list
 libvir-list@redhat.com
 https://www.redhat.com/mailman/listinfo/libvir-list
 




signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] configure: make --with-test-suite work

On Wed, Dec 18, 2013 at 06:05:15AM -0700, Eric Blake wrote:
 On 12/18/2013 12:03 AM, Martin Kletzander wrote:
  Our option '--with-test-suite' could have never worked since it was
  defined as AC_ARG_ENABLE([with-test-suite], ...), thus working only as
  '--enable-with-test-suite', but documented in configure.ac as
  AC_HELP_STRING([--with-test-suite], ...).
  In my optinion, the help string is as it should be, but the option is

 s/optinion/opinion/

  wrong.
 
  Signed-off-by: Martin Kletzander mklet...@redhat.com
  ---
   configure.ac | 16 
   1 file changed, 8 insertions(+), 8 deletions(-)

 Indeed, you could use './configure --disable-with-test-suite' to change
 the behavior, but that just looks wrong.  It's been broken since the
 introduction in commit 3a2fc27.

 ACK.


I fixed the typo, added the commit info and pushed, thanks.

Martin


signature.asc
Description: Digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On 12/18/2013 06:09 AM, Eric Blake wrote:
 On 12/18/2013 02:35 AM, Martin Kletzander wrote:
 When test-locking is enabled, CFLAGS have -Dbool=char, which makes
 
 Huh?  bool should never be defined to char in CFLAGS; it should only be
 replaced if gnulib thinks the compiler is too old.

I actually looked at configure.ac, and wow, we really ARE doing
-Dbool=char when trying to use CIL.  Yuck.  I think --with-test-locking
has bit-rotted for some time now, and is going to need a LOT of work
before it can be considered usable.  That option was added in commit
6962a2d, in 2009, which predates my time on libvirt, and I don't ever
recall using it myself.  I'm not familiar enough with CIL to speculate
why it doesn't work with C99 bool (or maybe that was a limitation back
in 2009 but has since been fixed).  Dan?

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Fix crash in virsystemdtest with dbus 1.7.6

On 12/18/2013 04:14 AM, Cédric Bosdonnat wrote:
 D-bus introduced some changes in its locking code. Overriding the init
 function skips the new locking init and thus crashes later in libvirt
 test. Removing the function makes the test pass again.
 ---
  tests/virsystemdmock.c | 5 -
  1 file changed, 5 deletions(-)

Ouch.  We added the override because it was needed on older platforms.
Have you tested that this doesn't break the test on RHEL 5?

 
 diff --git a/tests/virsystemdmock.c b/tests/virsystemdmock.c
 index b8fc031..b3b82b4 100644
 --- a/tests/virsystemdmock.c
 +++ b/tests/virsystemdmock.c
 @@ -31,11 +31,6 @@ void dbus_connection_set_change_sigpipe(dbus_bool_t 
 will_modify_sigpipe ATTRIBUT
  {
  }
  
 -dbus_bool_t dbus_threads_init_default(void)
 -{
 -return 1;
 -}
 -
  DBusConnection *dbus_bus_get(DBusBusType type ATTRIBUTE_UNUSED,
   DBusError *error ATTRIBUTE_UNUSED)
  {
 

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On Wed, Dec 18, 2013 at 06:09:40AM -0700, Eric Blake wrote:
 On 12/18/2013 02:35 AM, Martin Kletzander wrote:
  When test-locking is enabled, CFLAGS have -Dbool=char, which makes

 Huh?  bool should never be defined to char in CFLAGS; it should only be
 replaced if gnulib thinks the compiler is too old.


Gnulib doesn't do that, it's our configure.ac thing (but I must say I
don't know whether that's copied from somewhere or not) introduced in
6962a2dd7e:

AC_ARG_ENABLE([test-locking],
  [AS_HELP_STRING([--enable-test-locking],
[thread locking tests using CIL @:@default=no@:@])],
[case ${enableval} in
   yes|no) ;;
   *)  AC_MSG_ERROR([bad value ${enableval} for test-locking option]) ;;
 esac],
  [enableval=no])
enable_locking=$enableval

if test $enable_locking = yes; then
  LOCK_CHECKING_CFLAGS=-Dbool=char -D_Bool=char -save-temps
  AC_SUBST([LOCK_CHECKING_CFLAGS])
fi
AM_CONDITIONAL([WITH_CIL],[test $enable_locking = yes])

  pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
  double negation fixes the build breaker.

 I don't get how this could possibly make a difference.  The 'output'
 variable is declared bool, and then only ever assigned 'false' or
 'true', which if bool is replaced by gnulib are still guaranteed to be 0
 and 1, so using it directly as an array index is still safe.


But the thing is that due to -Dbool=char the variable is declared
char, not bool.

 
  Signed-off-by: Martin Kletzander mklet...@redhat.com
  ---
 
  Notes:
  I still can't build with '--enable-test-locking' and I'm not sure this
  is the proper way to fix it.  Also, the code wasn't touched for
  months, so I'm not pushing it as a build-breaker.

 NACK.  I need more details about the actual failure you are seeing, but
 this is not the right patch.

 --
 Eric Blake   eblake redhat com+1-919-301-3266
 Libvirt virtualization library http://libvirt.org



signature.asc
Description: Digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Fix race leading to crash when setting up dbus watches

On 12/18/2013 05:21 AM, Daniel P. Berrange wrote:
 Currently the virDBusAddWatch does
 
   virEventAddHandle(fd, flags,
 virDBusWatchCallback,
 watch, NULL);
   dbus_watch_set_data(watch, info, virDBusWatchFree);
 
 Unfortunately this is racy - since the event loop is in a
 different thread, the virDBusWatchCallback method may be
 run before we get to calling dbus_watch_set_data. We must
 reverse the order of these calls
 
 See https://bugzilla.redhat.com/show_bug.cgi?id=885445
 
 Signed-off-by: Daniel P. Berrange berra...@redhat.com
 ---
  src/util/virdbus.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)

ACK.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Ian Campbell

On Wed, 2013-12-18 at 14:12 +0100, Stefan Bader wrote:
 On 18.12.2013 13:27, Ian Campbell wrote:
  On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:
 
  Might this libxl fix be relevant:
  commit 5420f26507fc5c9853eb1076401a8658d72669da
  Author: Jim Fehlig jfeh...@suse.com
  Date:   Fri Jan 11 12:22:26 2013 +
  
  libxl: Set vfb and vkb devid if not done so by the caller
  
  Other devices set a sensible devid if the caller has not done 
  so.
  Do the same for vfb and vkb.  While at it, factor out the 
  common code
  used to determine a sensible devid, so it can be used by other
  libxl__device_*_add functions.
  
  Signed-off-by: Jim Fehlig jfeh...@suse.com
  Acked-by: Ian Campbell ian.campb...@citrix.com
  Committed-by: Ian Campbell ian.campb...@citrix.com
  
  and a follow up in dfeccbeaa. Although the comment implies that nic's
  were already correctly assigning a devid if the caller specified -1, so
  I don't know why it doesn't work for you :-(
 
  Ok, yes, the commit above indeed changes libxl__device_nic_add to call
  libxl__device_nextid for the devid... Just how is this actually called.
  Maybe not sufficient but git grep libxl__device_nic_add in the xen code 
  only
  shows the definition and a declaration in libxl_internal.h to me...
  
  I have a feeling a macro might be involved...
  
  Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
  add the eventual function names in comments to provide grep fodder
 
 Oh duh, yeah. So in DEFINE_DEVICE_ADD a libxl_device_nic_add is created which
 calls to libxl__device_nic_add. When I look for the single _ version I find a
 call from xl_cmdimpl.c and its public declaration in libxl.h.
 So I guess the bug is that libvirt in the libxl driver never seems to do so

The macro creates libxl__add_nics which adds the nics from the
libxl_domain_config-nics array. I don't think libvirt needs to call
libxl_device_nic_add manually unless it is hotplugging a new nic at
runtime.

 . Ok,
 thanks a lot for digging the out the DEFINE. As nice those are to create 
 similar
 functions from template, grep and tags fail to be helpful with them.
 
 -Stefan
  
  Ian.
  
  --
  libvir-list mailing list
  libvir-list@redhat.com
  https://www.redhat.com/mailman/listinfo/libvir-list
  
 
 


--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On 12/18/2013 06:25 AM, Martin Kletzander wrote:

 pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
 double negation fixes the build breaker.

 I don't get how this could possibly make a difference.  The 'output'
 variable is declared bool, and then only ever assigned 'false' or
 'true', which if bool is replaced by gnulib are still guaranteed to be 0
 and 1, so using it directly as an array index is still safe.

 
 But the thing is that due to -Dbool=char the variable is declared
 char, not bool.

But if the variable contains exactly 0 or 1, it still makes no
difference as an array index whether it was typed bool or char.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] python pypi package - backport older version?

On Wed, Dec 18, 2013 at 07:51:00AM -0500, Jason Antman wrote:
 Hello,
 
 First off, thanks for the wonderful project, and thanks even more for
 all of the excellent work on the Python bindings lately.
 
 I'm working on a small project to collect data on kvm hosts/guests via a
 python collector client and display the data in a web interface. It
 works beautifully when using my various distros' OS packages for
 libvirt-python, but I've hit a bit of a snag when trying to do automated
 testing. I'm quite happy to see libvirt-python packaged as a python
 package and on pypi, but the version there is 1.2.0, which requires
 libvirt = 0.9.11.
 
 Unfortunately, I was planning to test my python app on travis-ci, which
 runs Ubuntu Precise (12.04 LTS), and the latest version of libvirt for
 that is 0.9.8 (package python-libvirt 0.9.8-2ubuntu17.13).
 
 If I were to submit a patch, would there be any interest in me
 backporting the setup.py/packaging work to an older version of the
 python bindings - say 0.9.8 - and getting it published on pypi?

It is not practical to support any version older than 0.9.11.
The code generator requires that the API XML description files
be installed by libvirt and it did not do that prior to the
0.9.11 release

For older distros just use the libvirt-python APIs that were
bundled with that distros' version fo libvirt.

Regards,
Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Stefan Bader

On 18.12.2013 14:28, Ian Campbell wrote:
 On Wed, 2013-12-18 at 14:12 +0100, Stefan Bader wrote:
 On 18.12.2013 13:27, Ian Campbell wrote:
 On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:

 Might this libxl fix be relevant:
 commit 5420f26507fc5c9853eb1076401a8658d72669da
 Author: Jim Fehlig jfeh...@suse.com
 Date:   Fri Jan 11 12:22:26 2013 +
 
 libxl: Set vfb and vkb devid if not done so by the caller
 
 Other devices set a sensible devid if the caller has not done 
 so.
 Do the same for vfb and vkb.  While at it, factor out the 
 common code
 used to determine a sensible devid, so it can be used by other
 libxl__device_*_add functions.
 
 Signed-off-by: Jim Fehlig jfeh...@suse.com
 Acked-by: Ian Campbell ian.campb...@citrix.com
 Committed-by: Ian Campbell ian.campb...@citrix.com
 
 and a follow up in dfeccbeaa. Although the comment implies that nic's
 were already correctly assigning a devid if the caller specified -1, so
 I don't know why it doesn't work for you :-(

 Ok, yes, the commit above indeed changes libxl__device_nic_add to call
 libxl__device_nextid for the devid... Just how is this actually called.
 Maybe not sufficient but git grep libxl__device_nic_add in the xen code 
 only
 shows the definition and a declaration in libxl_internal.h to me...

 I have a feeling a macro might be involved...

 Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
 add the eventual function names in comments to provide grep fodder

 Oh duh, yeah. So in DEFINE_DEVICE_ADD a libxl_device_nic_add is created which
 calls to libxl__device_nic_add. When I look for the single _ version I find a
 call from xl_cmdimpl.c and its public declaration in libxl.h.
 So I guess the bug is that libvirt in the libxl driver never seems to do so
 
 The macro creates libxl__add_nics which adds the nics from the
 libxl_domain_config-nics array. I don't think libvirt needs to call
 libxl_device_nic_add manually unless it is hotplugging a new nic at
 runtime.

Bah, this is another one, DEFINE_DEVICES_ADD in libxl_devices.c, that does that.
That seems to be used from a few places. That needs more research to understand
the call paths. Probably simpler starting from the entry which looks to be
libxl_domain_create_new from libvirt's side.

 
 . Ok,
 thanks a lot for digging the out the DEFINE. As nice those are to create 
 similar
 functions from template, grep and tags fail to be helpful with them.

 -Stefan

 Ian.

 --
 libvir-list mailing list
 libvir-list@redhat.com
 https://www.redhat.com/mailman/listinfo/libvir-list



 
 




signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 1/2] virsh: man: Mention that volumes need to be in storage pool for undefine

https://bugzilla.redhat.com/show_bug.cgi?id=105

When undefining a VM with storage the man page doesn't explicitly
mention that the volumes need to be a part of the storage pool otherwise
it won't work.
---
 tools/virsh.pod | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/virsh.pod b/tools/virsh.pod
index 677931f..59213c6 100644
--- a/tools/virsh.pod
+++ b/tools/virsh.pod
@@ -1794,7 +1794,8 @@ on inactive domains. Volume deletion is only attempted 
after the domain is
 undefined; if not all of the requested volumes could be deleted, the
 error message indicates what still remains behind. If a volume path is not
 found in the domain definition, it's treated as if the volume was successfully
-deleted.
+deleted. Only volumes managed by libvirt in storage pools can be removed this
+way.
 (See Bdomblklist for list of target names associated to a domain).
 Example: --storage vda,/path/to/storage.img

-- 
1.8.5.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 0/2] undefine with storage fixes

Two fixes related to the virsh undefine --storage command.

Peter Krempa (2):
  virsh: man: Mention that volumes need to be in storage pool for
undefine
  virsh: domain: Fix undefine with storage of 'volume' disks

 tools/virsh-domain.c | 36 ++--
 tools/virsh.pod  |  3 ++-
 2 files changed, 36 insertions(+), 3 deletions(-)

-- 
1.8.5.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 2/2] virsh: domain: Fix undefine with storage of 'volume' disks

The undefine code that removes the storage along with the VM didn't take
into acount the existence of 'volume' type disks. Add the functionality.
---
 tools/virsh-domain.c | 36 ++--
 1 file changed, 34 insertions(+), 2 deletions(-)

diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
index 8b80e1e..760dca5 100644
--- a/tools/virsh-domain.c
+++ b/tools/virsh-domain.c
@@ -2939,6 +2939,7 @@ cmdUndefine(vshControl *ctl, const vshCmd *cmd)
 int nvol_nodes;
 char *source = NULL;
 char *target = NULL;
+char *pool = NULL;
 size_t i;
 size_t j;

@@ -3048,6 +3049,7 @@ cmdUndefine(vshControl *ctl, const vshCmd *cmd)
 vshUndefineVolume vol;
 VIR_FREE(source);
 VIR_FREE(target);
+VIR_FREE(pool);

 /* get volume source and target paths */
 if (!(target = virXPathString(string(./target/@dev), ctxt)))
@@ -3057,9 +3059,12 @@ cmdUndefine(vshControl *ctl, const vshCmd *cmd)
   ./source/@file|
   ./source/@dir|
   ./source/@name|
-  ./source/@dev), ctxt)))
+  ./source/@dev|
+  ./source/@volume), ctxt)))
 continue;

+pool = virXPathString(string(./source/@pool), ctxt);
+
 /* lookup if volume was selected by user */
 if (vol_list) {
 bool found = false;
@@ -3075,7 +3080,33 @@ cmdUndefine(vshControl *ctl, const vshCmd *cmd)
 continue;
 }

-if (!(vol.vol = virStorageVolLookupByPath(ctl-conn, source))) {
+if (pool) {
+virStoragePoolPtr storagepool = NULL;
+
+if (!source) {
+vshPrint(ctl,
+ _(Missing storage volume name for disk '%s'),
+ target);
+continue;
+}
+
+if (!(storagepool = virStoragePoolLookupByName(ctl-conn,
+   pool))) {
+vshPrint(ctl,
+ _(Storage pool '%s' for volume '%s' not found.),
+ pool, target);
+vshResetLibvirtError();
+continue;
+}
+
+vol.vol = virStorageVolLookupByName(storagepool, source);
+virStoragePoolFree(storagepool);
+
+} else {
+   vol.vol = virStorageVolLookupByPath(ctl-conn, source);
+}
+
+if (!vol.vol) {
 vshPrint(ctl,
  _(Storage volume '%s'(%s) is not managed by libvirt. 

Remove it manually.\n), target, source);
@@ -3190,6 +3221,7 @@ out:
 cleanup:
 VIR_FREE(source);
 VIR_FREE(target);
+VIR_FREE(pool);
 for (i = 0; i  nvols; i++) {
 VIR_FREE(vols[i].source);
 VIR_FREE(vols[i].target);
-- 
1.8.5.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] python pypi package - backport older version?

On Wed, Dec 18, 2013 at 01:56:46PM +, Daniel P. Berrange wrote:
 On Wed, Dec 18, 2013 at 07:51:00AM -0500, Jason Antman wrote:
  Hello,
  
  First off, thanks for the wonderful project, and thanks even more for
  all of the excellent work on the Python bindings lately.
  
  I'm working on a small project to collect data on kvm hosts/guests via a
  python collector client and display the data in a web interface. It
  works beautifully when using my various distros' OS packages for
  libvirt-python, but I've hit a bit of a snag when trying to do automated
  testing. I'm quite happy to see libvirt-python packaged as a python
  package and on pypi, but the version there is 1.2.0, which requires
  libvirt = 0.9.11.
  
  Unfortunately, I was planning to test my python app on travis-ci, which
  runs Ubuntu Precise (12.04 LTS), and the latest version of libvirt for
  that is 0.9.8 (package python-libvirt 0.9.8-2ubuntu17.13).
  
  If I were to submit a patch, would there be any interest in me
  backporting the setup.py/packaging work to an older version of the
  python bindings - say 0.9.8 - and getting it published on pypi?
 
 It is not practical to support any version older than 0.9.11.
 The code generator requires that the API XML description files
 be installed by libvirt and it did not do that prior to the
 0.9.11 release
 
 For older distros just use the libvirt-python APIs that were
 bundled with that distros' version fo libvirt.

BTW Canonical provide an add-on  Cloud Archive repository for
the LTS release which includes updated versions of OpenStack,
QEMU/KVM and libvirt. So if you need the updated libvirt it might
be an option to use that repo.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Fix crash in virsystemdtest with dbus 1.7.6

On Wed, Dec 18, 2013 at 06:25:07AM -0700, Eric Blake wrote:
 On 12/18/2013 04:14 AM, Cédric Bosdonnat wrote:
  D-bus introduced some changes in its locking code. Overriding the init
  function skips the new locking init and thus crashes later in libvirt
  test. Removing the function makes the test pass again.
  ---
   tests/virsystemdmock.c | 5 -
   1 file changed, 5 deletions(-)
 
 Ouch.  We added the override because it was needed on older platforms.
 Have you tested that this doesn't break the test on RHEL 5?

Did we ? I saw this method was added right at the first commit of this
code.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] add --nocow option to vol-create and vol-clone

On Thu, Dec 05, 2013 at 11:16:42AM +, Daniel P. Berrange wrote:
 On Thu, Dec 05, 2013 at 06:35:12PM +0800, Chunyan Liu wrote:
  Btrfs has terrible performance when hosting VM images, even more when the 
  guest
  in those VM are also using btrfs as file system. One way to mitigate this 
  bad
  performance is to turn off COW attributes on VM files (since having copy on
  write for this kind of data is not useful).
  
  According to 'chattr' manpage, NOCOW could be set to new or empty file only 
  on
  btrfs, so this patch tries to add a --nocow option to vol-create functions 
  and
  vol-clone function, so that users could have a chance to set NOCOW to a new
  volume if that happens to create on a btrfs like file system.
 
 What effect / impact does setting this flag have from a functional
 POV ?  Why would we not just unconditonally enable it on btrfs so
 it was fast out of the box ? I'm loathe to add a btrfs-specific
 flag to our public API.

Quoting from the qemu-devel thread on the same subject:

   When the NOCOW attribute is set on a file, reflink copying (aka
   file-level snapshots) do not work:
  
   $ cp --reflink test.img test-snapshot.img
  
   This produces EINVAL.
  
   It is a regression if qemu-img create suddenly starts breaking this
   standard btrfs feature for existing users.

So as with QEMU, I don't think libvirt can do something which could
break existing users of brtfs in this way. So this would have to be
an opt-in of some kind.

We already have a way to express features for storage volumes in
the XML description. We could use this to express a 'nocow' feature.
This is preferrable to an API flag, since this would let a user query
XML for an existing volume to discover if it had 'nocow' or not.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Umar,
please see commit
http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
.

The functions mentioned below have been implemented.

Michal

On 12/18/2013 01:29 PM, Umar Draz wrote:
Hi Michal,

Thanks for your efforts and help. If I need any other API I will
definitely ask you.

Br.

Umar

On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny minov...@redhat.com
mailto:minov...@redhat.com wrote:

Hi Umar,
that's fine. Give one or two days for this please. Maybe I'll
implement
earlier but I'm a bit overloaded after my PTO so I can't promise
it will
be earlier.

Also, what APIs would you like to implement?

I'm thinking of implementing these two:

http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

Would you like any other APIs?

Thanks,
Michal

On 12/18/2013 01:15 PM, Umar Draz wrote:
HI Michal,

Ok , I will wait. I hope you will get sometime to do this.

Br.

Umar

On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:

Hi Michal and Umar,
it's correct I can't invoke this APIs via PHP yet. Please
note that
libvirt-php is not automatically generated when doing libvirt
build and
it's a separate project. It's rather easy to implement it to the
libvirt-php so if you want to implement it, feel free to
implement
or if
you can wait a day or two until I have some time I can implement
it (the
project is not officially supported and it's done in the spare
time mainly).

I'd just like to ask you whether there are some more APIs you'd
like me
to implement when I'll be implementing it. Are there any
other APIs
you'd like?

Thanks,
Michal

On 12/18/2013 12:59 PM, Michal Privoznik wrote:
On 18.12.2013 12:27, Umar Draz wrote:
Hi Peter,

Thanks for the response, Would you please help how I can
use *virDomainSetMetadata* with PHP or any url for help?

Br.

Umar
[CC'ing Michal Novotny - the libvirt-php maintainer]

I don't think you can invoke those APIs via PHP. But I expect
them to be
able to add easily. Michal?

Michal

--
Michal Novotny minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com,
RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org
http://php-virt-control.org
http://php-virt-control.org

--
Umar Draz
Network Architect

--
Michal Novotny minov...@redhat.com mailto:minov...@redhat.com,
RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org
http://php-virt-control.org

--
Umar Draz
Network Architect

--
Michal Novotny minov...@redhat.com, RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Block info query: Add flag to allow failure if not active

On Wed, Dec 18, 2013 at 06:58:10AM -0500, John Ferlan wrote:
 Currently the qemuDomainGetBlockInfo will return allocation == physical
 for most backing stores. For a qcow2 block backed device it's possible
 to return the highest lv extent allocated from qemu for an active guest.
 That is a value where allocation != physical and one would hope be less.
 However, if the guest is not running, then the code falls back to returning
 allocation == physical. This turns out to be problematic for rhev which
 monitors the size of the backing store. During a migration, before the
 VM has been started on the target and while it is deemed inactive on the
 source, allocation is returned as physical triggering the code to extend
 the file unnecessarily.
 
 Thus, this patch will check a new flag to the call to determine whether
 a failure should be returned to indicate to the caller that the code is
 unable to fetch the allocation value. This will allow the caller to decide
 how to proceed.
 
 Note that I chose not to return a failure as a more general rule mostly
 because of backwards compatability and current expectations reasons.
 Changing a default return action could be an unexpected action. Although
 returning allocation == physical perhaps is incorrect, it is how the
 code has been functioning. Having a 'virsh domblkinfo' return a failure
 for some inactive domains while succeeding for others could be confusing.
 A flag forces the caller to decide how to operate.
 
 As it turns out, there seems to be quite a bit of history regarding this.
 Details can be found in the BZ:
 
 https://bugzilla.redhat.com/show_bug.cgi?id=1040507
 Signed-off-by: John Ferlan jfer...@redhat.com
 ---
 NOTE:
 Trying to keep this as simple as possible since it needs to be backported.
 If someone has a better idea for a flag name - I'm open to suggestions...
 
  include/libvirt/libvirt.h.in | 23 ---
  src/libvirt.c| 11 ++-
  src/qemu/qemu_driver.c   | 33 +++--
  3 files changed, 57 insertions(+), 10 deletions(-)
 
 diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in
 index 6f79c49..420642b 100644
 --- a/include/libvirt/libvirt.h.in
 +++ b/include/libvirt/libvirt.h.in
 @@ -2095,7 +2095,8 @@ int virDomainBlockResize 
 (virDomainPtr dom,
  
  /** virDomainBlockInfo:
   *
 - * This struct provides information about the size of a block device backing 
 store
 + * This struct provides information about the size of a block device
 + * backing store
   *
   * Examples:
   *
 @@ -2108,13 +2109,29 @@ int virDomainBlockResize 
 (virDomainPtr dom,
   *
   *  - qcow2 file in filesystem
   *   * capacity: logical size from qcow2 header
 - *   * allocation, physical: logical size of the file / highest qcow 
 extent (identical)
 + *   * allocation, physical: logical size of the file /
 + *   highest qcow extent (identical)
   *
   *  - qcow2 file in a block device
   *   * capacity: logical size from qcow2 header
 - *   * allocation: highest qcow extent written
 + *   * allocation: highest qcow extent written for an active domain
   *   * physical: size of the block device container
   */
 +
 +/**
 + * virDomainGetBlockInfoFlags:
 + *
 + * Flags available for virDomainGetBlockInfo().
 + */
 +typedef enum {
 +VIR_DOMAIN_BLOCK_CHECK_ACTIVE = 1  0, /* If not active, then return
 + * failure when attempting to
 + * fetch data from an inactive
 + * domain, such as allocation
 + * from a qcow2 block device
 + */
 +} virDomainGetBlockInfoFlags;

I'm not convinced this flag is desirable. Can't apps just
check whether the guest is running themselves, or indeed
something like RHEV surely knows what its own VM is doing
already.

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] Implementation deficiency in virInitctlSetRunLevel

2013-12-18 Thread Reco

 Hello, list.

I was pointed here by maintainer of libvirt package in Debian, Guido
Günther. For the sake of completeness, the original bug report can be
viewed at this link:

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=732394

To sum up the bug report, current implementation of
virInitctlSetRunLevel function (src/util/virinitctl.c) lacks any sanity
checks before writing to container's /dev/initctl. In the absence of
such checks, libvirtd can be easily tricked to write runlevel check
request to an arbitrary main hosts' file (including
hosts' /run/initctl, as described in the bug report). All it takes is
one symlink in place of containers' /dev/initctl.

I've checked current libvirtd's git, and it seems to me that the
problem is still here.

Attached to this letter is a patch which tries to mitigate the issue by
checking whenever container's /dev/initctl is a pipe actually.

Sincerely yours, Reco

PS I'm not subscribed to this list, in case of further questions please
CC me.
--- a/src/util/virinitctl.c 2013-12-18 11:13:10.078432196 +0400
+++ b/src/util/virinitctl.c 2013-12-18 11:26:50.0 +0400
@@ -24,7 +24,10 @@
 #include config.h

 #include sys/param.h
+#include sys/types.h
+#include sys/stat.h
 #include fcntl.h
+#include unistd.h

 #include internal.h
 #include virinitctl.h
@@ -122,6 +125,7 @@
 int fd = -1;
 char *path = NULL;
 int ret = -1;
+struct stat attrs;

 memset(req, 0, sizeof(req));

@@ -139,7 +143,10 @@
 return -1;
 }

-if ((fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
+if (lstat(path, attrs) == -1)
+goto cleanup;
+
+if ((attrs.st_mode  S_IFIFO)  (fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
 if (errno == ENOENT) {
 ret = 0;
 goto cleanup
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH v2] libxl: libxl_get_max_cpus returning a libxl error from 4.4 onward

Starting from commit 2e82c18c in Xen (will be included in Xen 4.4)
both libxl_get_max_cpus() and libxl_get_max_ndoes() start returning
a proper libxl error code, in case of failure. This patch fixes
this in the libxl driver.

Note that, although it is now basically impossible for them to return
0, that would, theoretically, still be wrong. Also, checking that the
returned value is '= 0' makes the code correct for both Xen 4.4 and
Xen 4.3 (and 4.2), and that is why we go for it (rather than
just ' 0').

Signed-off-by: Dario Faggioli dario.faggi...@citrix.com
Cc: Jim Fehlig jfeh...@suse.com
Cc: Ian Jackson ian.jack...@eu.citrix.com
Cc: Martin Kletzander mklet...@redhat.com
---
Changes from v1:
 * taking care of libxl_get_max_nodes() too
---
 src/libxl/libxl_driver.c |   15 +++
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
index 692c3b7..29aa6c7 100644
--- a/src/libxl/libxl_driver.c
+++ b/src/libxl/libxl_driver.c
@@ -1101,9 +1101,11 @@ libxlConnectGetMaxVcpus(virConnectPtr conn, const char 
*type ATTRIBUTE_UNUSED)
 
 cfg = libxlDriverConfigGet(driver);
 ret = libxl_get_max_cpus(cfg-ctx);
-/* libxl_get_max_cpus() will return 0 if there were any failures,
-   e.g. xc_physinfo() failing */
-if (ret == 0)
+/* On failure, libxl_get_max_cpus() will return ERROR_FAIL from Xen 4.4
+ * onward, but it ever returning 0 is obviously wrong too (and it is
+ * what happens, on failure, on Xen 4.3 and earlier). Therefore, a 'less
+ * or equal' is the catchall we want. */
+if (ret = 0)
 ret = -1;
 
 virObjectUnref(cfg);
@@ -3980,6 +3982,7 @@ libxlDomainGetNumaParameters(virDomainPtr dom,
 
 for (i = 0; i  LIBXL_NUMA_NPARAM  i  *nparams; i++) {
 virMemoryParameterPtr param = params[i];
+int numnodes;
 
 switch (i) {
 case 0:
@@ -3998,8 +4001,12 @@ libxlDomainGetNumaParameters(virDomainPtr dom,
 /* Node affinity */
 
 /* Let's allocate both libxl and libvirt bitmaps */
+numnodes = libxl_get_max_nodes(priv-ctx);
+if (numnodes = 0)
+goto cleanup;
+
 if (libxl_node_bitmap_alloc(priv-ctx, nodemap, 0) ||
-!(nodes = virBitmapNew(libxl_get_max_nodes(priv-ctx {
+!(nodes = virBitmapNew(numnodes))) {
 virReportOOMError();
 goto cleanup;
 }

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Michal,

Thanks, Now what I need here? I already have installed libvirt-php-0.4.8,
do I need to copy

libvirt-php.chttp://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
libvirt-php.http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0h
in the src directory and reconfigure the php-libvirt?

Br.

Umar

On Wed, Dec 18, 2013 at 7:31 PM, Michal Novotny minov...@redhat.com wrote:

Hi Umar,
please see commit

http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
.

The functions mentioned below have been implemented.

Michal

On 12/18/2013 01:29 PM, Umar Draz wrote:
Hi Michal,

Thanks for your efforts and help. If I need any other API I will
definitely ask you.

Br.

Umar

On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny minov...@redhat.com
mailto:minov...@redhat.com wrote:

Hi Umar,
that's fine. Give one or two days for this please. Maybe I'll
implement
earlier but I'm a bit overloaded after my PTO so I can't promise
it will
be earlier.

Also, what APIs would you like to implement?

I'm thinking of implementing these two:

http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

Would you like any other APIs?

Thanks,
Michal

On 12/18/2013 01:15 PM, Umar Draz wrote:
HI Michal,

Ok , I will wait. I hope you will get sometime to do this.

Br.

Umar

On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:

Hi Michal and Umar,
it's correct I can't invoke this APIs via PHP yet. Please
note that
libvirt-php is not automatically generated when doing libvirt
build and
it's a separate project. It's rather easy to implement it to
the
libvirt-php so if you want to implement it, feel free to
implement
or if
you can wait a day or two until I have some time I can
implement
it (the
project is not officially supported and it's done in the spare
time mainly).

I'd just like to ask you whether there are some more APIs you'd
like me
to implement when I'll be implementing it. Are there any
other APIs
you'd like?

Thanks,
Michal

On 12/18/2013 12:59 PM, Michal Privoznik wrote:
On 18.12.2013 12:27, Umar Draz wrote:
Hi Peter,

Thanks for the response, Would you please help how I can
use *virDomainSetMetadata* with PHP or any url for help?

Br.

Umar
[CC'ing Michal Novotny - the libvirt-php maintainer]

I don't think you can invoke those APIs via PHP. But I expect
them to be
able to add easily. Michal?

Michal

--
Umar Draz
Network Architect

--
Michal Novotny minov...@redhat.com mailto:minov...@redhat.com,
RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org
http://php-virt-control.org

--
Umar Draz
Network Architect

--
Michal Novotny minov...@redhat.com, RHCE, Red Hat
Virtualization | libvirt-php bindings | php-virt-control.org

--
Umar Draz
Network Architect
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libxl: cannot connect to PV console

2013-12-18 Thread Ian Jackson

Jim Fehlig writes (Re: libxl: cannot connect to PV console):
 Dario Faggioli wrote:
  Hey Jim,
 
  I'm on libvirt.git's trunk and I see the following:
  [xen@ghoul3 libvirt.git]$ sudo ./tools/virsh list
   IdName   State
  
   3 debian_32  running
   4 fedora20_64running
 
  [xen@ghoul3 libvirt.git]$ sudo ./tools/virsh console debian_32
  Connected to domain debian_32
  Escape character is ^]
  error: internal error: cannot find character device (null)
 
  [xen@ghoul3 libvirt.git]$ sudo ./tools/virsh console fedora20_64
  Connected to domain fedora20_64
  Escape character is ^]
  error: internal error: cannot find character device (null)

 
 Looking at libxlDomainOpenConsole() in src/libxl/libxl_driver.c, it
 currently only supports a serial console.  Do you have one defined in
 these domains?  E.g.

If this is the root cause, it's a pretty bad error message.  Should I
or Dario take a look at the libvirt code to see if we can improve it ?

Ian.

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Block info query: Add flag to allow failure if not active

2013-12-18 Thread John Ferlan



On 12/18/2013 09:35 AM, Daniel P. Berrange wrote:
 On Wed, Dec 18, 2013 at 06:58:10AM -0500, John Ferlan wrote:

 
 I'm not convinced this flag is desirable. Can't apps just
 check whether the guest is running themselves, or indeed
 something like RHEV surely knows what its own VM is doing
 already.

I do agree with you and that's been my argument in the referenced BZ;
however, as I understand it - they have a thread that continually polls
for blockInfo information and a separate thread that handles the
migration. The blockInfo thread doesn't have the state information.

I contend it's just as simple to add a check about the domain state and
to get/check the reason value as well. That is - I think the blockInfo
thread should be more aware of state. Of course, the return argument is
that libvirt shouldn't return different answers on consecutive fetches
where the first fetch is when the guest is active and the second is
when it's not.

To be honest this feels a bit hackish and it requires a deliberate
change to their code to not only provide the flag, but handle the error.
Since they'd already be changing their code to handle this, then why not
change to check state. There is a downside to their checking state
though and that's a timing condition where their check could indicate
the guest was active, but the next clock tick it becomes inactive.
Thus, their code would be in the same position of getting what is felt
is the wrong answer.

I figure opening it up to community discussion will help decide one
way or another the best approach.

I'm all for adjusting the documentation and leaving the code the way it
is now as well.

John

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Umar,
you have to clone it from git repository and compile it and install. You
can simply do it by:

git clone git://libvirt.org/libvirt-php.git
./autogen.sh
make
sudo make install

And then you have to restart your Apache webserver.

For example how to use it, see the test in
test-domain-create-get-metadata.phpt file (it's normal PHP script,
however I use phpt extension as PHP test but it's standard PHP code).

Michal

On 12/18/2013 03:45 PM, Umar Draz wrote:
Hi Michal,

Thanks, Now what I need here? I already have
installed libvirt-php-0.4.8, do I need to copy

libvirt-php.c
http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
libvirt-php.
http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0h
in the src directory and reconfigure the php-libvirt?

Br.

Umar

On Wed, Dec 18, 2013 at 7:31 PM, Michal Novotny minov...@redhat.com
mailto:minov...@redhat.com wrote:

Hi Umar,
please see commit

http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
.

The functions mentioned below have been implemented.

Michal

On 12/18/2013 01:29 PM, Umar Draz wrote:
Hi Michal,

Thanks for your efforts and help. If I need any other API I will
definitely ask you.

Br.

Umar

On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny
minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:

Hi Umar,
that's fine. Give one or two days for this please. Maybe I'll
implement
earlier but I'm a bit overloaded after my PTO so I can't promise
it will
be earlier.

Also, what APIs would you like to implement?

I'm thinking of implementing these two:

http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata

http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata

Would you like any other APIs?

Thanks,
Michal

On 12/18/2013 01:15 PM, Umar Draz wrote:
HI Michal,

Ok , I will wait. I hope you will get sometime to do this.

Br.

Umar

On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:

Hi Michal and Umar,
it's correct I can't invoke this APIs via PHP yet. Please
note that
libvirt-php is not automatically generated when doing
libvirt
build and
it's a separate project. It's rather easy to implement
it to the
libvirt-php so if you want to implement it, feel free to
implement
or if
you can wait a day or two until I have some time I can
implement
it (the
project is not officially supported and it's done in
the spare
time mainly).

I'd just like to ask you whether there are some more
APIs you'd
like me
to implement when I'll be implementing it. Are there any
other APIs
you'd like?

Thanks,
Michal

On 12/18/2013 12:59 PM, Michal Privoznik wrote:
On 18.12.2013 12:27, Umar Draz wrote:
Hi Peter,

Thanks for the response, Would you please help how
I can
use *virDomainSetMetadata* with PHP or any url for
help?

Br.

Umar
[CC'ing Michal Novotny - the libvirt-php maintainer]

I don't think you can invoke those APIs via PHP. But
I expect
them to be
able to add easily. Michal?

Michal

--
Michal Novotny minov...@redhat.com
mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com,
RHCE, Red Hat
Virtualization | libvirt-php bindings |
php-virt-control.org http://php-virt-control.org
http://php-virt-control.org
http://php-virt-control.org

Re: [libvirt] [PATCH] Block info query: Add flag to allow failure if not active

On Wed, Dec 18, 2013 at 09:49:46AM -0500, John Ferlan wrote:
 
 
 On 12/18/2013 09:35 AM, Daniel P. Berrange wrote:
  On Wed, Dec 18, 2013 at 06:58:10AM -0500, John Ferlan wrote:
 
  
  I'm not convinced this flag is desirable. Can't apps just
  check whether the guest is running themselves, or indeed
  something like RHEV surely knows what its own VM is doing
  already.
 
 I do agree with you and that's been my argument in the referenced BZ;
 however, as I understand it - they have a thread that continually polls
 for blockInfo information and a separate thread that handles the
 migration. The blockInfo thread doesn't have the state information.

But it must have a virDomainPtr instance, so it can use
virDomainGetState() or virDomainIsActive() if it cares
about this.

 I contend it's just as simple to add a check about the domain state and
 to get/check the reason value as well. That is - I think the blockInfo
 thread should be more aware of state. Of course, the return argument is
 that libvirt shouldn't return different answers on consecutive fetches
 where the first fetch is when the guest is active and the second is
 when it's not.

Sure, I agree that libvirt isn't ideal here - our hands are tied by
the fact that QEMU doesn't make this data available to us offline.
If we changed anything on the libvirt side, then I'd want to see us
do a proper fix to get the data but that'd require qemu help;

 
 To be honest this feels a bit hackish and it requires a deliberate
 change to their code to not only provide the flag, but handle the error.
 Since they'd already be changing their code to handle this, then why not
 change to check state. There is a downside to their checking state
 though and that's a timing condition where their check could indicate
 the guest was active, but the next clock tick it becomes inactive.
 Thus, their code would be in the same position of getting what is felt
 is the wrong answer.

If they check the state after querying the data, then yes they could
get bogus data.  If they check the state before querying the info,
then the race condition scenario does not lead to bad behaviour - they
would be simply be discarding the allocation value unncessarily
which is exactly the same situation they'd be in if the libvirt API
itself checked this too, and the guest shutoff after libvirt checked
its state. So I don't see a need for this libvirt change here.

 I figure opening it up to community discussion will help decide one
 way or another the best approach.
 
 I'm all for adjusting the documentation and leaving the code the way it
 is now as well.

Updating docs would be my recommendation.

Regards,
Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Stefan Bader

On 18.12.2013 14:28, Ian Campbell wrote:
 On Wed, 2013-12-18 at 14:12 +0100, Stefan Bader wrote:
 On 18.12.2013 13:27, Ian Campbell wrote:
 On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:

 Might this libxl fix be relevant:
 commit 5420f26507fc5c9853eb1076401a8658d72669da
 Author: Jim Fehlig jfeh...@suse.com
 Date:   Fri Jan 11 12:22:26 2013 +
 
 libxl: Set vfb and vkb devid if not done so by the caller
 
 Other devices set a sensible devid if the caller has not done 
 so.
 Do the same for vfb and vkb.  While at it, factor out the 
 common code
 used to determine a sensible devid, so it can be used by other
 libxl__device_*_add functions.
 
 Signed-off-by: Jim Fehlig jfeh...@suse.com
 Acked-by: Ian Campbell ian.campb...@citrix.com
 Committed-by: Ian Campbell ian.campb...@citrix.com
 
 and a follow up in dfeccbeaa. Although the comment implies that nic's
 were already correctly assigning a devid if the caller specified -1, so
 I don't know why it doesn't work for you :-(

 Ok, yes, the commit above indeed changes libxl__device_nic_add to call
 libxl__device_nextid for the devid... Just how is this actually called.
 Maybe not sufficient but git grep libxl__device_nic_add in the xen code 
 only
 shows the definition and a declaration in libxl_internal.h to me...

 I have a feeling a macro might be involved...

 Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
 add the eventual function names in comments to provide grep fodder

 Oh duh, yeah. So in DEFINE_DEVICE_ADD a libxl_device_nic_add is created which
 calls to libxl__device_nic_add. When I look for the single _ version I find a
 call from xl_cmdimpl.c and its public declaration in libxl.h.
 So I guess the bug is that libvirt in the libxl driver never seems to do so
 
 The macro creates libxl__add_nics which adds the nics from the
 libxl_domain_config-nics array. I don't think libvirt needs to call
 libxl_device_nic_add manually unless it is hotplugging a new nic at
 runtime.
 

Hm, so I think this is the path:

libxl_domain_create_new
- do_domain_create
   - initiate_domain_create
  - libxl__bootloader_run (HVM domain, skipping bootloader)
  - domcreate_bootloader_done
 - domcreate_rebuild_done
- domcreate_launch_dm
   - libxl__spawn_local_dm
 - domcreate_devmodel_started

In libxl__spawn_local_dm, there is the following loop:

for (i = 0; i  d_config-num_nics; i++) {
/* We have to init the nic here, because we still haven't
 * called libxl_device_nic_add at this point, but qemu needs
 * the nic information to be complete.
 */
ret = libxl__device_nic_setdefault(gc, d_config-nics[i], domid);
if (ret)
goto error_out;
}

So I think when starting the dm, the devid just is not set as setdefault does
not seem to do so. I would be done in the later domcreate_devmodel_started
callback but that is too late for the generated qemu arguments.

-Stefan



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCHv3] qemu: ask for -enable-fips when FIPS is required

On 12/18/2013 03:59 AM, Peter Krempa wrote:
 On 12/17/13 19:36, Eric Blake wrote:
 On a system that is enforcing FIPS, most libraries honor the
 current mode by default.  Qemu, on the other hand, refused to
 honor FIPS mode unless you add the '-enable-fips' command
 line option; worse, this option is not discoverable via QMP,
 and is only present on binaries built for Linux.  So, if we
 detect FIPS mode, then we unconditionally ask for FIPS; either
 qemu is new enough to have the option and then correctly
 cripple insecure VNC passwords, or it is so old that we are
 correctly avoiding a FIPS violation by preventing qemu from
 starting.  Meanwhile, if we don't detect FIPS mode, then
 omitting the argument is safe whether the qemu has the option
 (but it would do nothing because FIPS is disabled) or whether
 qemu lacks the option (including in the case where we are not
 running on Linux).

 The testsuite was a bit interesting: we don't want our test
 to depend on whether it is being run in FIPS mode, so I had
 to tweak things to set the capability bit outside of our
 normal interaction with capability parsing.

 This fixes https://bugzilla.redhat.com/show_bug.cgi?id=1035474


 ACK. I verified that the detection works correctly when the file is
 present and contains the expected data.

Thanks; pushed.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH 1/2] virsh: man: Mention that volumes need to be in storage pool for undefine

On 12/18/2013 07:10 AM, Peter Krempa wrote:
 https://bugzilla.redhat.com/show_bug.cgi?id=105
 
 When undefining a VM with storage the man page doesn't explicitly
 mention that the volumes need to be a part of the storage pool otherwise
 it won't work.
 ---
  tools/virsh.pod | 3 ++-
  1 file changed, 2 insertions(+), 1 deletion(-)

ACK.  (Although this is yet another place where libvirt auto-creating
temporary storage pools for any disk referenced by a domain might be
nice...)

 
 diff --git a/tools/virsh.pod b/tools/virsh.pod
 index 677931f..59213c6 100644
 --- a/tools/virsh.pod
 +++ b/tools/virsh.pod
 @@ -1794,7 +1794,8 @@ on inactive domains. Volume deletion is only attempted 
 after the domain is
  undefined; if not all of the requested volumes could be deleted, the
  error message indicates what still remains behind. If a volume path is not
  found in the domain definition, it's treated as if the volume was 
 successfully
 -deleted.
 +deleted. Only volumes managed by libvirt in storage pools can be removed this
 +way.
  (See Bdomblklist for list of target names associated to a domain).
  Example: --storage vda,/path/to/storage.img
 

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH 2/2] virsh: domain: Fix undefine with storage of 'volume' disks

On 12/18/2013 07:10 AM, Peter Krempa wrote:
 The undefine code that removes the storage along with the VM didn't take
 into acount the existence of 'volume' type disks. Add the functionality.

s/acount/account/

 ---
  tools/virsh-domain.c | 36 ++--
  1 file changed, 34 insertions(+), 2 deletions(-)
 

ACK.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libvirt live modify

Hi Michal

I tried

but on autogen.sh is failed

automake: error: global options already processed
automake: Please contact bug-autom...@gnu.org.
 at /usr/share/automake-1.14/Automake/Channels.pm line 662, GEN0 line 77.
Automake::Channels::msg('automake', '', 'global options already
processed') called at /usr/share/automake-1.14/Automake/ChannelDefs.pm line
212
Automake::ChannelDefs::prog_error('global options already
processed') called at /usr/share/automake-1.14/Automake/Options.pm line 421
Automake::Options::process_global_option_list('HASH(0x246df60)')
called at /usr/bin/automake line 5331
Automake::scan_autoconf_traces('configure.ac') called at
/usr/bin/automake line 5431
Automake::scan_autoconf_files() called at /usr/bin/automake line
8259
autoreconf: automake failed with exit status: 29


Br.

Umar


On Wed, Dec 18, 2013 at 7:50 PM, Michal Novotny minov...@redhat.com wrote:

 Hi Umar,
 you have to clone it from git repository and compile it and install. You
 can simply do it by:

 git clone git://libvirt.org/libvirt-php.git
 ./autogen.sh
 make
 sudo make install

 And then you have to restart your Apache webserver.

 For example how to use it, see the test in
 test-domain-create-get-metadata.phpt file (it's normal PHP script,
 however I use phpt extension as PHP test but it's standard PHP code).

 Michal

 On 12/18/2013 03:45 PM, Umar Draz wrote:
  Hi Michal,
 
  Thanks, Now what I need here? I already have
  installed libvirt-php-0.4.8, do I need to copy
 
  libvirt-php.c
  
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
 
  libvirt-php.
  
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
 h
  in the src directory and reconfigure the php-libvirt?
 
  Br.
 
  Umar
 
 
  On Wed, Dec 18, 2013 at 7:31 PM, Michal Novotny minov...@redhat.com
  mailto:minov...@redhat.com wrote:
 
  Hi Umar,
  please see commit
 
 http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
  .
 
  The functions mentioned below have been implemented.
 
  Michal
 
  On 12/18/2013 01:29 PM, Umar Draz wrote:
   Hi Michal,
  
   Thanks for your efforts and help. If I need any other API I will
   definitely ask you.
  
   Br.
  
   Umar
  
  
   On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny
  minov...@redhat.com mailto:minov...@redhat.com
   mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
  
   Hi Umar,
   that's fine. Give one or two days for this please. Maybe I'll
   implement
   earlier but I'm a bit overloaded after my PTO so I can't
 promise
   it will
   be earlier.
  
   Also, what APIs would you like to implement?
  
   I'm thinking of implementing these two:
  
  
  http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
  
  http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata
  
   Would you like any other APIs?
  
   Thanks,
   Michal
  
   On 12/18/2013 01:15 PM, Umar Draz wrote:
HI Michal,
   
Ok , I will wait. I hope you will get sometime to do this.
   
Br.
   
Umar
   
   
On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
   minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
   
Hi Michal and Umar,
it's correct I can't invoke this APIs via PHP yet. Please
   note that
libvirt-php is not automatically generated when doing
  libvirt
build and
it's a separate project. It's rather easy to implement
  it to the
libvirt-php so if you want to implement it, feel free to
   implement
or if
you can wait a day or two until I have some time I can
  implement
it (the
project is not officially supported and it's done in
  the spare
time mainly).
   
I'd just like to ask you whether there are some more
  APIs you'd
like me
to implement when I'll be implementing it. Are there any
   other APIs
you'd like?
   
Thanks,
Michal
   
On 12/18/2013 12:59 PM, Michal Privoznik wrote:
 On 18.12.2013 12:27, Umar Draz wrote:

Re: [libvirt] libvirt live modify

Hi Umar,
then you don't have working automake setup. Please google how to solve
it, I apologize but I don't really have much time to solve it right now.

Thanks for understanding,
Michal

On 12/18/2013 04:21 PM, Umar Draz wrote:
 Hi Michal

 I tried

 but on autogen.sh is failed

 automake: error: global options already processed
 automake: Please contact bug-autom...@gnu.org
 mailto:bug-autom...@gnu.org.
  at /usr/share/automake-1.14/Automake/Channels.pm line 662, GEN0
 line 77.
 Automake::Channels::msg('automake', '', 'global options
 already processed') called at
 /usr/share/automake-1.14/Automake/ChannelDefs.pm line 212
 Automake::ChannelDefs::prog_error('global options already
 processed') called at /usr/share/automake-1.14/Automake/Options.pm
 line 421

 Automake::Options::process_global_option_list('HASH(0x246df60)')
 called at /usr/bin/automake line 5331
 Automake::scan_autoconf_traces('configure.ac
 http://configure.ac') called at /usr/bin/automake line 5431
 Automake::scan_autoconf_files() called at /usr/bin/automake
 line 8259
 autoreconf: automake failed with exit status: 29


 Br.

 Umar


 On Wed, Dec 18, 2013 at 7:50 PM, Michal Novotny minov...@redhat.com
 mailto:minov...@redhat.com wrote:

 Hi Umar,
 you have to clone it from git repository and compile it and
 install. You
 can simply do it by:

 git clone git://libvirt.org/libvirt-php.git
 http://libvirt.org/libvirt-php.git
 ./autogen.sh
 make
 sudo make install

 And then you have to restart your Apache webserver.

 For example how to use it, see the test in
 test-domain-create-get-metadata.phpt file (it's normal PHP script,
 however I use phpt extension as PHP test but it's standard PHP code).

 Michal

 On 12/18/2013 03:45 PM, Umar Draz wrote:
  Hi Michal,
 
  Thanks, Now what I need here? I already have
  installed libvirt-php-0.4.8, do I need to copy
 
  libvirt-php.c
 
 
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
  libvirt-php.
 
 
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0h
  in the src directory and reconfigure the php-libvirt?
 
  Br.
 
  Umar
 
 
  On Wed, Dec 18, 2013 at 7:31 PM, Michal Novotny
 minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
 
  Hi Umar,
  please see commit
 
 
 http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
  .
 
  The functions mentioned below have been implemented.
 
  Michal
 
  On 12/18/2013 01:29 PM, Umar Draz wrote:
   Hi Michal,
  
   Thanks for your efforts and help. If I need any other API
 I will
   definitely ask you.
  
   Br.
  
   Umar
  
  
   On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny
  minov...@redhat.com mailto:minov...@redhat.com
 mailto:minov...@redhat.com mailto:minov...@redhat.com
   mailto:minov...@redhat.com mailto:minov...@redhat.com
 mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
  
   Hi Umar,
   that's fine. Give one or two days for this please.
 Maybe I'll
   implement
   earlier but I'm a bit overloaded after my PTO so I
 can't promise
   it will
   be earlier.
  
   Also, what APIs would you like to implement?
  
   I'm thinking of implementing these two:
  
  
 
 http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
  
 
 http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata
  
   Would you like any other APIs?
  
   Thanks,
   Michal
  
   On 12/18/2013 01:15 PM, Umar Draz wrote:
HI Michal,
   
Ok , I will wait. I hope you will get sometime to do
 this.
   
Br.
   
Umar
   
   
On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
   minov...@redhat.com mailto:minov...@redhat.com
 mailto:minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com
 mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com
 mailto:minov...@redhat.com mailto:minov...@redhat.com
 mailto:minov...@redhat.com
  mailto:minov...@redhat.com

Re: [libvirt] libvirt live modify

Hi Michal,

No worries, I will

But again Thanks

Br.

Umar


On Wed, Dec 18, 2013 at 8:25 PM, Michal Novotny minov...@redhat.com wrote:

 Hi Umar,
 then you don't have working automake setup. Please google how to solve
 it, I apologize but I don't really have much time to solve it right now.

 Thanks for understanding,
 Michal

 On 12/18/2013 04:21 PM, Umar Draz wrote:
  Hi Michal
 
  I tried
 
  but on autogen.sh is failed
 
  automake: error: global options already processed
  automake: Please contact bug-autom...@gnu.org
  mailto:bug-autom...@gnu.org.
   at /usr/share/automake-1.14/Automake/Channels.pm line 662, GEN0
  line 77.
  Automake::Channels::msg('automake', '', 'global options
  already processed') called at
  /usr/share/automake-1.14/Automake/ChannelDefs.pm line 212
  Automake::ChannelDefs::prog_error('global options already
  processed') called at /usr/share/automake-1.14/Automake/Options.pm
  line 421
 
  Automake::Options::process_global_option_list('HASH(0x246df60)')
  called at /usr/bin/automake line 5331
  Automake::scan_autoconf_traces('configure.ac
  http://configure.ac') called at /usr/bin/automake line 5431
  Automake::scan_autoconf_files() called at /usr/bin/automake
  line 8259
  autoreconf: automake failed with exit status: 29
 
 
  Br.
 
  Umar
 
 
  On Wed, Dec 18, 2013 at 7:50 PM, Michal Novotny minov...@redhat.com
  mailto:minov...@redhat.com wrote:
 
  Hi Umar,
  you have to clone it from git repository and compile it and
  install. You
  can simply do it by:
 
  git clone git://libvirt.org/libvirt-php.git
  http://libvirt.org/libvirt-php.git
  ./autogen.sh
  make
  sudo make install
 
  And then you have to restart your Apache webserver.
 
  For example how to use it, see the test in
  test-domain-create-get-metadata.phpt file (it's normal PHP script,
  however I use phpt extension as PHP test but it's standard PHP code).
 
  Michal
 
  On 12/18/2013 03:45 PM, Umar Draz wrote:
   Hi Michal,
  
   Thanks, Now what I need here? I already have
   installed libvirt-php-0.4.8, do I need to copy
  
   libvirt-php.c
  
  
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
 
   libvirt-php.
  
  
 http://libvirt.org/git/?p=libvirt-php.git;a=blob;f=src/libvirt-php.c;h=7a2a3c0651e882abf538473ff013cb68e499be43;hb=e3e783413d8f001ee0a266edbac949c78616fde0
 h
   in the src directory and reconfigure the php-libvirt?
  
   Br.
  
   Umar
  
  
   On Wed, Dec 18, 2013 at 7:31 PM, Michal Novotny
  minov...@redhat.com mailto:minov...@redhat.com
   mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
  
   Hi Umar,
   please see commit
  
 
 http://libvirt.org/git/?p=libvirt-php.git;a=commit;h=e3e783413d8f001ee0a266edbac949c78616fde0
   .
  
   The functions mentioned below have been implemented.
  
   Michal
  
   On 12/18/2013 01:29 PM, Umar Draz wrote:
Hi Michal,
   
Thanks for your efforts and help. If I need any other API
  I will
definitely ask you.
   
Br.
   
Umar
   
   
On Wed, Dec 18, 2013 at 5:19 PM, Michal Novotny
   minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com
mailto:minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com wrote:
   
Hi Umar,
that's fine. Give one or two days for this please.
  Maybe I'll
implement
earlier but I'm a bit overloaded after my PTO so I
  can't promise
it will
be earlier.
   
Also, what APIs would you like to implement?
   
I'm thinking of implementing these two:
   
   
  
  http://libvirt.org/html/libvirt-libvirt.html#virDomainGetMetadata
   
  
  http://libvirt.org/html/libvirt-libvirt.html#virDomainSetMetadata
   
Would you like any other APIs?
   
Thanks,
Michal
   
On 12/18/2013 01:15 PM, Umar Draz wrote:
 HI Michal,

 Ok , I will wait. I hope you will get sometime to do
  this.

 Br.

 Umar


 On Wed, Dec 18, 2013 at 5:06 PM, Michal Novotny
minov...@redhat.com mailto:minov...@redhat.com
  mailto:minov...@redhat.com mailto:minov...@redhat.com
   mailto:minov...@redhat.com

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On Wed, Dec 18, 2013 at 06:09:40AM -0700, Eric Blake wrote:
 On 12/18/2013 02:35 AM, Martin Kletzander wrote:
  When test-locking is enabled, CFLAGS have -Dbool=char, which makes
 
 Huh?  bool should never be defined to char in CFLAGS; it should only be
 replaced if gnulib thinks the compiler is too old.
 
  pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
  double negation fixes the build breaker.
 
 I don't get how this could possibly make a difference.  The 'output'
 variable is declared bool, and then only ever assigned 'false' or
 'true', which if bool is replaced by gnulib are still guaranteed to be 0
 and 1, so using it directly as an array index is still safe.
 
  
  Signed-off-by: Martin Kletzander mklet...@redhat.com
  ---
  
  Notes:
  I still can't build with '--enable-test-locking' and I'm not sure this
  is the proper way to fix it.  Also, the code wasn't touched for
  months, so I'm not pushing it as a build-breaker.
 
 NACK.  I need more details about the actual failure you are seeing, but
 this is not the right patch.

The reason for this is that CIL didn't understand the bool type.
GNULIB's automagic detection didn't help here because the build
is still done by gcc which supports it just fine. We use -save-temps
and then CIL analyses those intermediate files and threw up.

I'll see if it is still a problem on F20 CIL.


Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH 2/2] virsh: domain: Fix undefine with storage of 'volume' disks

On 12/18/13 16:18, Eric Blake wrote:
 On 12/18/2013 07:10 AM, Peter Krempa wrote:
 The undefine code that removes the storage along with the VM didn't take
 into acount the existence of 'volume' type disks. Add the functionality.
 
 s/acount/account/
 
 ---
  tools/virsh-domain.c | 36 ++--
  1 file changed, 34 insertions(+), 2 deletions(-)

 
 ACK.

Pushed; Thanks.

Peter




signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] build: fix building with '--enable-test-locking'

On Wed, Dec 18, 2013 at 06:34:43AM -0700, Eric Blake wrote:
 On 12/18/2013 06:25 AM, Martin Kletzander wrote:

  pipefd[bool] fail (obviously).  Forcing the subscript to be bool by
  double negation fixes the build breaker.
 
  I don't get how this could possibly make a difference.  The 'output'
  variable is declared bool, and then only ever assigned 'false' or
  'true', which if bool is replaced by gnulib are still guaranteed to be 0
  and 1, so using it directly as an array index is still safe.
 
 
  But the thing is that due to -Dbool=char the variable is declared
  char, not bool.

 But if the variable contains exactly 0 or 1, it still makes no
 difference as an array index whether it was typed bool or char.


Yep, but '--enable-werror' is sad aboot it.  Anyways, the code for
test-locking is way too old and I don't need it in for anything right
now, so I'm ok with keeping everything as-is.  Thanks for having a
look at it and sorry for wasting your time.

Martin


signature.asc
Description: Digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Remove redefinition of bool type when --enable-test-locking

Old versions of CIL did not understand the 'bool' data type,
but at least 1.7.3 does now cope. We can remove the old hack
which redefined bool and no longer compiles successfully.

Signed-off-by: Daniel P. Berrange berra...@redhat.com
---
 configure.ac | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/configure.ac b/configure.ac
index 5446634..675f1fc 100644
--- a/configure.ac
+++ b/configure.ac
@@ -2129,7 +2129,7 @@ AC_ARG_ENABLE([test-locking],
 enable_locking=$enableval
 
 if test $enable_locking = yes; then
-  LOCK_CHECKING_CFLAGS=-Dbool=char -D_Bool=char -save-temps
+  LOCK_CHECKING_CFLAGS=-save-temps
   AC_SUBST([LOCK_CHECKING_CFLAGS])
 fi
 AM_CONDITIONAL([WITH_CIL],[test $enable_locking = yes])
-- 
1.8.4.2

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] Remove stray semicolon after pragma macros

The VIR_WARNINGS_NO_CAST_ALIGN / VIR_WARNINGS_RESET should
not have any trailing ';' since they are pragmas. The use
of a ';' results in an empty statement which confuses CIL.

Signed-off-by: Daniel P. Berrange berra...@redhat.com
---
 src/nwfilter/nwfilter_dhcpsnoop.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/nwfilter/nwfilter_dhcpsnoop.c 
b/src/nwfilter/nwfilter_dhcpsnoop.c
index e8fcfef..a96a790 100644
--- a/src/nwfilter/nwfilter_dhcpsnoop.c
+++ b/src/nwfilter/nwfilter_dhcpsnoop.c
@@ -981,9 +981,9 @@ virNWFilterSnoopDHCPDecode(virNWFilterSnoopReqPtr req,
 /* go through the protocol headers */
 switch (ntohs(pep-eh_type)) {
 case ETHERTYPE_IP:
-VIR_WARNINGS_NO_CAST_ALIGN;
+VIR_WARNINGS_NO_CAST_ALIGN
 pip = (struct iphdr *) pep-eh_data;
-VIR_WARNINGS_RESET;
+VIR_WARNINGS_RESET
 len -= offsetof(virNWFilterSnoopEthHdr, eh_data);
 break;
 default:
-- 
1.8.4.2

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] libxl: cannot connect to PV console

On mer, 2013-12-18 at 14:45 +, Ian Jackson wrote:
 Jim Fehlig writes (Re: libxl: cannot connect to PV console):
  Dario Faggioli wrote:
   [xen@ghoul3 libvirt.git]$ sudo ./tools/virsh console fedora20_64
   Connected to domain fedora20_64
   Escape character is ^]
   error: internal error: cannot find character device (null)
 
  
  Looking at libxlDomainOpenConsole() in src/libxl/libxl_driver.c, it
  currently only supports a serial console.  Do you have one defined in
  these domains?  E.g.
 
 If this is the root cause, it's a pretty bad error message.

Indeed.

 Should I
 or Dario take a look at the libvirt code to see if we can improve it ?

Looks like it's this piece of code, in src/libxl/libxl_driver.c:

  libxlDomainOpenConsole(..)
...
if (vm-def-nserials)
chr = vm-def-serials[0];

if (!chr) {
virReportError(VIR_ERR_INTERNAL_ERROR,
   _(cannot find character device %s),
   NULLSTR(dev_name));
goto cleanup;
}
...
  }

I guess I can send a patch to turn that message into something like
cannot find any serial console define for the domain, either always or
in the specific case where vm-def-serials[0] is NULL, as I feel like
the character device (null) part is the worst and most misleading.

At the same time, I'd be interested in figuring out why this does not
work / stopped working.

I'll investigate more.

Regards,
Dario

-- 
This happens because I choose it to happen! (Raistlin Majere)
-
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems RD Ltd., Cambridge (UK)



signature.asc
Description: This is a digitally signed message part
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH v2] libxl: libxl_get_max_cpus returning a libxl error from 4.4 onward

On Wed, Dec 18, 2013 at 03:39:12PM +0100, Dario Faggioli wrote:
 Starting from commit 2e82c18c in Xen (will be included in Xen 4.4)
 both libxl_get_max_cpus() and libxl_get_max_ndoes() start returning

s/ndoes/nodes/

 a proper libxl error code, in case of failure. This patch fixes
 this in the libxl driver.

 Note that, although it is now basically impossible for them to return
 0, that would, theoretically, still be wrong. Also, checking that the
 returned value is '= 0' makes the code correct for both Xen 4.4 and
 Xen 4.3 (and 4.2), and that is why we go for it (rather than
 just ' 0').

 Signed-off-by: Dario Faggioli dario.faggi...@citrix.com
 Cc: Jim Fehlig jfeh...@suse.com
 Cc: Ian Jackson ian.jack...@eu.citrix.com
 Cc: Martin Kletzander mklet...@redhat.com
 ---
 Changes from v1:
  * taking care of libxl_get_max_nodes() too
 ---
  src/libxl/libxl_driver.c |   15 +++
  1 file changed, 11 insertions(+), 4 deletions(-)


ACK, fixed  pushed.

Thanks,
Martin


signature.asc
Description: Digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH v2] libxl: libxl_get_max_cpus returning a libxl error from 4.4 onward

On mer, 2013-12-18 at 17:40 +0100, Martin Kletzander wrote:
 On Wed, Dec 18, 2013 at 03:39:12PM +0100, Dario Faggioli wrote:
  Starting from commit 2e82c18c in Xen (will be included in Xen 4.4)
  both libxl_get_max_cpus() and libxl_get_max_ndoes() start returning
 
 s/ndoes/nodes/
 
Ouch! :-(

  ---
  Changes from v1:
   * taking care of libxl_get_max_nodes() too
  ---
   src/libxl/libxl_driver.c |   15 +++
   1 file changed, 11 insertions(+), 4 deletions(-)
 
 
 ACK, fixed  pushed.
 
Cool. Thanks,
Dario

-- 
This happens because I choose it to happen! (Raistlin Majere)
-
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems RD Ltd., Cambridge (UK)



signature.asc
Description: This is a digitally signed message part
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Remove redefinition of bool type when --enable-test-locking

On Wed, Dec 18, 2013 at 04:16:17PM +, Daniel P. Berrange wrote:
 Old versions of CIL did not understand the 'bool' data type,
 but at least 1.7.3 does now cope. We can remove the old hack
 which redefined bool and no longer compiles successfully.

 Signed-off-by: Daniel P. Berrange berra...@redhat.com
 ---
  configure.ac | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

 diff --git a/configure.ac b/configure.ac
 index 5446634..675f1fc 100644
 --- a/configure.ac
 +++ b/configure.ac
 @@ -2129,7 +2129,7 @@ AC_ARG_ENABLE([test-locking],
  enable_locking=$enableval

  if test $enable_locking = yes; then
 -  LOCK_CHECKING_CFLAGS=-Dbool=char -D_Bool=char -save-temps
 +  LOCK_CHECKING_CFLAGS=-save-temps
AC_SUBST([LOCK_CHECKING_CFLAGS])
  fi
  AM_CONDITIONAL([WITH_CIL],[test $enable_locking = yes])
 --
 1.8.4.2


ACK,

Martin


signature.asc
Description: Digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] python pypi package - backport older version?

2013-12-18 Thread Jason Antman

On 12/18/2013 09:15 AM, Daniel P. Berrange wrote:
 On Wed, Dec 18, 2013 at 01:56:46PM +, Daniel P. Berrange wrote:
 On Wed, Dec 18, 2013 at 07:51:00AM -0500, Jason Antman wrote:
 Hello,

 First off, thanks for the wonderful project, and thanks even more for
 all of the excellent work on the Python bindings lately.

 I'm working on a small project to collect data on kvm hosts/guests via a
 python collector client and display the data in a web interface. It
 works beautifully when using my various distros' OS packages for
 libvirt-python, but I've hit a bit of a snag when trying to do automated
 testing. I'm quite happy to see libvirt-python packaged as a python
 package and on pypi, but the version there is 1.2.0, which requires
 libvirt = 0.9.11.

 Unfortunately, I was planning to test my python app on travis-ci, which
 runs Ubuntu Precise (12.04 LTS), and the latest version of libvirt for
 that is 0.9.8 (package python-libvirt 0.9.8-2ubuntu17.13).

 If I were to submit a patch, would there be any interest in me
 backporting the setup.py/packaging work to an older version of the
 python bindings - say 0.9.8 - and getting it published on pypi?
 It is not practical to support any version older than 0.9.11.
 The code generator requires that the API XML description files
 be installed by libvirt and it did not do that prior to the
 0.9.11 release

 For older distros just use the libvirt-python APIs that were
 bundled with that distros' version fo libvirt.
 BTW Canonical provide an add-on  Cloud Archive repository for
 the LTS release which includes updated versions of OpenStack,
 QEMU/KVM and libvirt. So if you need the updated libvirt it might
 be an option to use that repo.

 Daniel
Daniel,

Ok, thanks for the heads-up and info. I'll see if I can get that working
on travis-ci sometime this week, thanks for the tip. I guess in the
interim I'll just roll a VM of a more recent distro to test on.

-Jason

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 0/3] libxl: improve vcpu pinning

Basically, by making it possible to both query and change the vcpu to pcpu
pinning of a (persistent) domain, even when it is not running.

That happens by providing the implementation of virDomainGetVcpuPinInfo() and
virDomainPinVcpuFlags() within he libxl driver, which is what happens in the
first two patches.  The third patch is something that can also be seen as a
bugfix, and that's why I kept it separated from the second one, for easier
review (although, the 'bug' does not really manifests, until
virDomainPinVcpuFlags is implemented in patch 2).

The logic is a lot similar to what happens in the QEMU driver.

The patches are available in the following git branch:

 git://xenbits.xen.org/people/dariof/libvirt.git libxl/VcpuPinX

Thanks and Regards,
Dario

---

Dario Faggioli (3):
  libxl: implement virDomainGetVcpuPinInfo
  libxl: implement virDomainPinVcpuFlags
  libxl: correctly handle affinity reset in virDomainPinVcpu[Flags]


 src/libxl/libxl_driver.c |  171 --
 1 file changed, 150 insertions(+), 21 deletions(-)

-- 
This happens because I choose it to happen! (Raistlin Majere)
-
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems RD Ltd., Cambridge (UK)

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 3/3] libxl: correctly handle affinity reset in virDomainPinVcpu[Flags]

By actually removing the vcpupin element (from within the
cputune section) from the XML, rather than jus update it with
a fully set vcpu affinity mask.

Signed-off-by: Dario Faggioli dario.faggi...@citrix.com
Cc: Jim Fehlig jfeh...@suse.com
Cc: Ian Jackson ian.jack...@eu.citrix.com
---
 src/libxl/libxl_driver.c |   18 ++
 1 file changed, 18 insertions(+)

diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
index 4079434..9a0898a 100644
--- a/src/libxl/libxl_driver.c
+++ b/src/libxl/libxl_driver.c
@@ -2363,6 +2363,7 @@ libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int 
vcpu,
 libxlDriverPrivatePtr driver = dom-conn-privateData;
 libxlDriverConfigPtr cfg = libxlDriverConfigGet(driver);
 virDomainDefPtr targetDef = NULL;
+virBitmapPtr pcpumap = NULL;
 virDomainObjPtr vm;
 int ret = -1;
 
@@ -2391,6 +2392,10 @@ libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int 
vcpu,
 
 sa_assert(targetDef);
 
+pcpumap = virBitmapNewData(cpumap, maplen);
+if (!pcpumap)
+goto cleanup;
+
 if (flags  VIR_DOMAIN_AFFECT_LIVE) {
 libxl_bitmap map = { .size = maplen, .map = cpumap };
 libxlDomainObjPrivatePtr priv;
@@ -2404,6 +2409,17 @@ libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int 
vcpu,
 }
 }
 
+/* full bitmap means reset the settings (if any). */
+if (virBitmapIsAllSet(pcpumap)) {
+if (virDomainVcpuPinDel(targetDef, vcpu)  0) {
+virReportError(VIR_ERR_INTERNAL_ERROR,
+   _(Failed to delete vcpupin xml for vcpu '%d'),
+   vcpu);
+goto cleanup;
+}
+goto out;
+}
+
 if (!targetDef-cputune.vcpupin) {
 if (VIR_ALLOC(targetDef-cputune.vcpupin)  0)
 goto cleanup;
@@ -2419,6 +2435,7 @@ libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int 
vcpu,
 goto cleanup;
 }
 
+out:
 ret = 0;
 
 if (flags  VIR_DOMAIN_AFFECT_LIVE) {
@@ -2430,6 +2447,7 @@ libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int 
vcpu,
 cleanup:
 if (vm)
 virObjectUnlock(vm);
+virBitmapFree(pcpumap);
 virObjectUnref(cfg);
 return ret;
 }

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 1/3] libxl: implement virDomainGetVcpuPinInfo

So that it is possible to query vcpu related information of
a persistent but not running domain, like it is for the QEMU
driver.

In fact, before this patch, we have:
 # virsh list --all
  IdName   State
 
  5 debian_32  running
  - fedora20_64shut off
 # virsh vcpuinfo fedora20_64
 error: this function is not supported by the connection driver: 
virDomainGetVcpuPinInfo

After (same situation as above, i.e., fedora20_64 not running):
 # virsh vcpuinfo fedora20_64
 VCPU:   0
 CPU:N/A
 State:  N/A
 CPU timeN/A
 CPU Affinity:   

 VCPU:   1
 CPU:N/A
 State:  N/A
 CPU timeN/A
 CPU Affinity:   

Signed-off-by: Dario Faggioli dario.faggi...@citrix.com
Cc: Jim Fehlig jfeh...@suse.com
Cc: Ian Jackson ian.jack...@eu.citrix.com
---
 src/libxl/libxl_driver.c |   83 ++
 1 file changed, 83 insertions(+)

diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
index 692c3b7..750d4ec 100644
--- a/src/libxl/libxl_driver.c
+++ b/src/libxl/libxl_driver.c
@@ -2415,6 +2415,88 @@ cleanup:
 return ret;
 }
 
+static int
+libxlDomainGetVcpuPinInfo(virDomainPtr dom, int ncpumaps,
+  unsigned char *cpumaps, int maplen,
+  unsigned int flags)
+{
+libxlDriverPrivatePtr driver = dom-conn-privateData;
+libxlDriverConfigPtr cfg = libxlDriverConfigGet(driver);
+virDomainObjPtr vm = NULL;
+virDomainDefPtr targetDef = NULL;
+virDomainVcpuPinDefPtr *vcpupin_list;
+virBitmapPtr cpumask = NULL;
+int maxcpu, hostcpus, vcpu, pcpu, n, ret = -1;
+unsigned char *cpumap;
+bool pinned;
+
+virCheckFlags(VIR_DOMAIN_AFFECT_LIVE |
+  VIR_DOMAIN_AFFECT_CONFIG, -1);
+
+if (!(vm = libxlDomObjFromDomain(dom)))
+goto cleanup;
+
+if (virDomainGetVcpuPinInfoEnsureACL(dom-conn, vm-def)  0)
+goto cleanup;
+
+if (virDomainLiveConfigHelperMethod(cfg-caps, driver-xmlopt, vm,
+flags, targetDef)  0)
+goto cleanup;
+
+if (flags  VIR_DOMAIN_AFFECT_LIVE) {
+targetDef = vm-def;
+}
+
+sa_assert(targetDef);
+
+/* Clamp to actual number of vcpus */
+if (ncpumaps  targetDef-vcpus)
+ncpumaps = targetDef-vcpus;
+
+if (!cpumaps || ncpumaps  1) {
+virReportError(VIR_ERR_INVALID_ARG, %s,
+   _(cannot return affinity via a NULL pointer));
+goto cleanup;
+}
+
+/* we use cfg-ctx, as vm-privateData-ctx may be NULL if VM is down */
+if ((hostcpus = libxl_get_max_cpus(cfg-ctx))  0)
+goto cleanup;
+
+maxcpu = maplen * 8;
+if (maxcpu  hostcpus)
+maxcpu = hostcpus;
+
+/* initialize cpumaps */
+memset(cpumaps, 0xff, maplen * ncpumaps);
+if (maxcpu % 8) {
+for (vcpu = 0; vcpu  ncpumaps; vcpu++) {
+cpumap = VIR_GET_CPUMAP(cpumaps, maplen, vcpu);
+cpumap[maplen - 1] = (1  maxcpu % 8) - 1;
+}
+}
+
+/* if vcpupin setting exists, there may be unused pcpus */
+for (n = 0; n  targetDef-cputune.nvcpupin; n++) {
+vcpupin_list = targetDef-cputune.vcpupin;
+vcpu = vcpupin_list[n]-vcpuid;
+cpumask = vcpupin_list[n]-cpumask;
+cpumap = VIR_GET_CPUMAP(cpumaps, maplen, vcpu);
+for (pcpu = 0; pcpu  maxcpu; pcpu++) {
+if (virBitmapGetBit(cpumask, pcpu, pinned)  0)
+goto cleanup;
+if (!pinned)
+VIR_UNUSE_CPU(cpumap, pcpu);
+}
+}
+ret = ncpumaps;
+
+cleanup:
+if (vm)
+virObjectUnlock(vm);
+virObjectUnref(cfg);
+return ret;
+}
 
 static int
 libxlDomainGetVcpus(virDomainPtr dom, virVcpuInfoPtr info, int maxinfo,
@@ -4237,6 +4319,7 @@ static virDriver libxlDriver = {
 .domainGetVcpusFlags = libxlDomainGetVcpusFlags, /* 0.9.0 */
 .domainPinVcpu = libxlDomainPinVcpu, /* 0.9.0 */
 .domainGetVcpus = libxlDomainGetVcpus, /* 0.9.0 */
+.domainGetVcpuPinInfo = libxlDomainGetVcpuPinInfo, /* 1.2.1 */
 .domainGetXMLDesc = libxlDomainGetXMLDesc, /* 0.9.0 */
 .connectDomainXMLFromNative = libxlConnectDomainXMLFromNative, /* 0.9.0 */
 .connectDomainXMLToNative = libxlConnectDomainXMLToNative, /* 0.9.0 */

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 2/3] libxl: implement virDomainPinVcpuFlags

And use it to implement libxlDomainPinVcpu(), similarly to what
happens in the QEMU driver. This way, it is possible to both
query and change the vcpu affinity of a persistent but not
running domain.

In face, before this patch, we have:
 # virsh list --all
  IdName   State
 
  5 debian_32  running
  - fedora20_64shut off
 # virsh vcpupin fedora20_64 0 2-4 --current
 error: this function is not supported by the connection driver: 
virDomainPinVcpuFlags

After (same situation as above):
 # virsh vcpupin  fedora20_64 0 2-4 --current
 # virsh vcpupin  fedora20_64 0
 VCPU: CPU Affinity
 --
0: 2-4

Signed-off-by: Dario Faggioli dario.faggi...@citrix.com
Cc: Jim Fehlig jfeh...@suse.com
Cc: Ian Jackson ian.jack...@eu.citrix.com
---
 src/libxl/libxl_driver.c |   72 --
 1 file changed, 50 insertions(+), 22 deletions(-)

diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
index 750d4ec..4079434 100644
--- a/src/libxl/libxl_driver.c
+++ b/src/libxl/libxl_driver.c
@@ -2356,45 +2356,61 @@ cleanup:
 }
 
 static int
-libxlDomainPinVcpu(virDomainPtr dom, unsigned int vcpu, unsigned char *cpumap,
-   int maplen)
+libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int vcpu,
+unsigned char *cpumap, int maplen,
+unsigned int flags)
 {
 libxlDriverPrivatePtr driver = dom-conn-privateData;
 libxlDriverConfigPtr cfg = libxlDriverConfigGet(driver);
-libxlDomainObjPrivatePtr priv;
+virDomainDefPtr targetDef = NULL;
 virDomainObjPtr vm;
 int ret = -1;
-libxl_bitmap map;
+
+virCheckFlags(VIR_DOMAIN_AFFECT_LIVE |
+  VIR_DOMAIN_AFFECT_CONFIG, -1);
 
 if (!(vm = libxlDomObjFromDomain(dom)))
 goto cleanup;
 
-if (virDomainPinVcpuEnsureACL(dom-conn, vm-def)  0)
+if (virDomainPinVcpuFlagsEnsureACL(dom-conn, vm-def, flags)  0)
 goto cleanup;
 
-if (!virDomainObjIsActive(vm)) {
+if ((flags  VIR_DOMAIN_AFFECT_LIVE)  !virDomainObjIsActive(vm)) {
 virReportError(VIR_ERR_OPERATION_INVALID, %s,
-   _(cannot pin vcpus on an inactive domain));
+   _(domain is inactive));
 goto cleanup;
 }
 
-priv = vm-privateData;
-
-map.size = maplen;
-map.map = cpumap;
-if (libxl_set_vcpuaffinity(priv-ctx, dom-id, vcpu, map) != 0) {
-virReportError(VIR_ERR_INTERNAL_ERROR,
-   _(Failed to pin vcpu '%d' with libxenlight), vcpu);
+if (virDomainLiveConfigHelperMethod(cfg-caps, driver-xmlopt, vm,
+flags, targetDef)  0)
 goto cleanup;
+
+if (flags  VIR_DOMAIN_AFFECT_LIVE) {
+targetDef = vm-def;
 }
 
-if (!vm-def-cputune.vcpupin) {
-if (VIR_ALLOC(vm-def-cputune.vcpupin)  0)
+sa_assert(targetDef);
+
+if (flags  VIR_DOMAIN_AFFECT_LIVE) {
+libxl_bitmap map = { .size = maplen, .map = cpumap };
+libxlDomainObjPrivatePtr priv;
+
+priv = vm-privateData;
+if (libxl_set_vcpuaffinity(priv-ctx, dom-id, vcpu, map) != 0) {
+virReportError(VIR_ERR_INTERNAL_ERROR,
+   _(Failed to pin vcpu '%d' with libxenlight),
+   vcpu);
 goto cleanup;
-vm-def-cputune.nvcpupin = 0;
+}
 }
-if (virDomainVcpuPinAdd(vm-def-cputune.vcpupin,
-vm-def-cputune.nvcpupin,
+
+if (!targetDef-cputune.vcpupin) {
+if (VIR_ALLOC(targetDef-cputune.vcpupin)  0)
+goto cleanup;
+targetDef-cputune.nvcpupin = 0;
+}
+if (virDomainVcpuPinAdd(targetDef-cputune.vcpupin,
+targetDef-cputune.nvcpupin,
 cpumap,
 maplen,
 vcpu)  0) {
@@ -2403,11 +2419,14 @@ libxlDomainPinVcpu(virDomainPtr dom, unsigned int vcpu, 
unsigned char *cpumap,
 goto cleanup;
 }
 
-if (virDomainSaveStatus(driver-xmlopt, cfg-stateDir, vm)  0)
-goto cleanup;
-
 ret = 0;
 
+if (flags  VIR_DOMAIN_AFFECT_LIVE) {
+ret = virDomainSaveStatus(driver-xmlopt, cfg-stateDir, vm);
+} else if (flags  VIR_DOMAIN_AFFECT_CONFIG) {
+ret = virDomainSaveConfig(cfg-configDir, targetDef);
+}
+
 cleanup:
 if (vm)
 virObjectUnlock(vm);
@@ -2416,6 +2435,14 @@ cleanup:
 }
 
 static int
+libxlDomainPinVcpu(virDomainPtr dom, unsigned int vcpu, unsigned char *cpumap,
+   int maplen)
+{
+return libxlDomainPinVcpuFlags(dom, vcpu, cpumap, maplen,
+   VIR_DOMAIN_AFFECT_LIVE);
+}
+
+static int
 libxlDomainGetVcpuPinInfo(virDomainPtr dom, int ncpumaps,
   unsigned char *cpumaps, int maplen,

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

On 18.12.2013 15:33, Reco wrote:
  Hello, list.
 
 I was pointed here by maintainer of libvirt package in Debian, Guido
 Günther. For the sake of completeness, the original bug report can be
 viewed at this link:
 
 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=732394
 
 To sum up the bug report, current implementation of
 virInitctlSetRunLevel function (src/util/virinitctl.c) lacks any sanity
 checks before writing to container's /dev/initctl. In the absence of
 such checks, libvirtd can be easily tricked to write runlevel check
 request to an arbitrary main hosts' file (including
 hosts' /run/initctl, as described in the bug report). All it takes is
 one symlink in place of containers' /dev/initctl.
 
 I've checked current libvirtd's git, and it seems to me that the
 problem is still here.
 
 Attached to this letter is a patch which tries to mitigate the issue by
 checking whenever container's /dev/initctl is a pipe actually.
 
 Sincerely yours, Reco
 
 PS I'm not subscribed to this list, in case of further questions please
 CC me.
 
 
 
 --
 libvir-list mailing list
 libvir-list@redhat.com
 https://www.redhat.com/mailman/listinfo/libvir-list
 

[Pasting the patch here]

 --- a/src/util/virinitctl.c 2013-12-18 11:13:10.078432196 +0400
 +++ b/src/util/virinitctl.c 2013-12-18 11:26:50.0 +0400
 @@ -24,7 +24,10 @@
  #include config.h
 
  #include sys/param.h
 +#include sys/types.h
 +#include sys/stat.h
  #include fcntl.h
 +#include unistd.h
 
  #include internal.h
  #include virinitctl.h
 @@ -122,6 +125,7 @@
  int fd = -1;
  char *path = NULL;
  int ret = -1;
 +struct stat attrs;
 
  memset(req, 0, sizeof(req));
 
 @@ -139,7 +143,10 @@
  return -1;
  }
 
 -if ((fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
 +if (lstat(path, attrs) == -1)
 +goto cleanup;

While the part above looks okay, I think we should report error if the 
/dev/initctl is not a pipe. Moreover, with your approach, if it's really not a 
pipe, we don't call open() and proceed to safewrite(-1, ...) at line 153 which 
will report EBADF error.

 +
 +if ((attrs.st_mode  S_IFIFO)  (fd = open(path, 
 O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
  if (errno == ENOENT) {
  ret = 0;
  goto cleanup

Missing ';' at EOL ^^^ - I guess you missed that during copy-paste.

Looking forward to v2.
BTW: consider using git when sending patches. I couldn't really apply this 
patch via 'git am'.

http://libvirt.org/hacking.html#patches

Michal

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

On Wed, Dec 18, 2013 at 06:33:21PM +0400, Reco wrote:
  Hello, list.
 
 I was pointed here by maintainer of libvirt package in Debian, Guido
 Günther. For the sake of completeness, the original bug report can be
 viewed at this link:
 
 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=732394
 
 To sum up the bug report, current implementation of
 virInitctlSetRunLevel function (src/util/virinitctl.c) lacks any sanity
 checks before writing to container's /dev/initctl. In the absence of
 such checks, libvirtd can be easily tricked to write runlevel check
 request to an arbitrary main hosts' file (including
 hosts' /run/initctl, as described in the bug report). All it takes is
 one symlink in place of containers' /dev/initctl.
 
 I've checked current libvirtd's git, and it seems to me that the
 problem is still here.
 
 Attached to this letter is a patch which tries to mitigate the issue by
 checking whenever container's /dev/initctl is a pipe actually.
 
 Sincerely yours, Reco
 
 PS I'm not subscribed to this list, in case of further questions please
 CC me.

 --- a/src/util/virinitctl.c 2013-12-18 11:13:10.078432196 +0400
 +++ b/src/util/virinitctl.c 2013-12-18 11:26:50.0 +0400
 @@ -24,7 +24,10 @@
  #include config.h
 
  #include sys/param.h
 +#include sys/types.h
 +#include sys/stat.h
  #include fcntl.h
 +#include unistd.h
 
  #include internal.h
  #include virinitctl.h
 @@ -122,6 +125,7 @@
  int fd = -1;
  char *path = NULL;
  int ret = -1;
 +struct stat attrs;
 
  memset(req, 0, sizeof(req));
 
 @@ -139,7 +143,10 @@
  return -1;
  }
 
 -if ((fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
 +if (lstat(path, attrs) == -1)
 +goto cleanup;
 +
 +if ((attrs.st_mode  S_IFIFO)  (fd = open(path, 
 O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
  if (errno == ENOENT) {
  ret = 0;
  goto cleanup

Hmm, using lstat sets up a race condition though. I would suggest we
use O_NOFOLLOW with open() but that only works for the final component
of the path - so if /dev is a symlink in the guest it'll still cause
problems.

There are also a few other places where we use /proc/$PID/root/dev for
hotplug where we mknod. If the guest setup a bad /dev symlink it could
cause us problems. 

I think we may actually have to instead rely on forking a child which
does setns(/proc/$PID/ns/mnt) to make the changes safely in the container
namespace.

Regards,
Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Fix crash in virsystemdtest with dbus 1.7.6

On Wed, Dec 18, 2013 at 06:25:07AM -0700, Eric Blake wrote:
 On 12/18/2013 04:14 AM, Cédric Bosdonnat wrote:
  D-bus introduced some changes in its locking code. Overriding the init
  function skips the new locking init and thus crashes later in libvirt
  test. Removing the function makes the test pass again.
  ---
   tests/virsystemdmock.c | 5 -
   1 file changed, 5 deletions(-)
 
 Ouch.  We added the override because it was needed on older platforms.
 Have you tested that this doesn't break the test on RHEL 5?

I tested RHEL5.10 and it had no negative impact on the test

Daniel
-- 
|: http://berrange.com  -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org  -o- http://virt-manager.org :|
|: http://autobuild.org   -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org   -o-   http://live.gnome.org/gtk-vnc :|

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH 1/2] virsh: man: Mention that volumes need to be in storage pool for undefine

On 12/18/13 16:16, Eric Blake wrote:
 On 12/18/2013 07:10 AM, Peter Krempa wrote:
 https://bugzilla.redhat.com/show_bug.cgi?id=105

 When undefining a VM with storage the man page doesn't explicitly
 mention that the volumes need to be a part of the storage pool otherwise
 it won't work.
 ---
  tools/virsh.pod | 3 ++-
  1 file changed, 2 insertions(+), 1 deletion(-)
 
 ACK.  (Although this is yet another place where libvirt auto-creating
 temporary storage pools for any disk referenced by a domain might be
 nice...)

I'm not quite sure if we want to expose the temporary storage pool
functionality outside of libvirtd. Let's see how the review of my
gluster series will end up where we can discuss that.

Peter

P.S.: Pushed.




signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH] Remove stray semicolon after pragma macros

On 12/18/2013 09:17 AM, Daniel P. Berrange wrote:
 The VIR_WARNINGS_NO_CAST_ALIGN / VIR_WARNINGS_RESET should
 not have any trailing ';' since they are pragmas. The use
 of a ';' results in an empty statement which confuses CIL.
 
 Signed-off-by: Daniel P. Berrange berra...@redhat.com
 ---
  src/nwfilter/nwfilter_dhcpsnoop.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)

ACK.

 
 diff --git a/src/nwfilter/nwfilter_dhcpsnoop.c 
 b/src/nwfilter/nwfilter_dhcpsnoop.c
 index e8fcfef..a96a790 100644
 --- a/src/nwfilter/nwfilter_dhcpsnoop.c
 +++ b/src/nwfilter/nwfilter_dhcpsnoop.c
 @@ -981,9 +981,9 @@ virNWFilterSnoopDHCPDecode(virNWFilterSnoopReqPtr req,
  /* go through the protocol headers */
  switch (ntohs(pep-eh_type)) {
  case ETHERTYPE_IP:
 -VIR_WARNINGS_NO_CAST_ALIGN;
 +VIR_WARNINGS_NO_CAST_ALIGN
  pip = (struct iphdr *) pep-eh_data;
 -VIR_WARNINGS_RESET;
 +VIR_WARNINGS_RESET
  len -= offsetof(virNWFilterSnoopEthHdr, eh_data);
  break;
  default:
 

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] lxcDomainShutdownFlags: Cleanup @flags usage

Currently, the @flags usage is a bit unclear at first sight to say the
least. There's no need for such unclear code especially when we can
borrow the working code from qemuDomainShutdownFlags().

In addition, this fixes one bug too. If user requested both
VIR_DOMAIN_SHUTDOWN_INITCTL and VIR_DOMAIN_SHUTDOWN_SIGNAL at the same
time, he is basically saying: 'Use the force Luke! If initctl fails try
sending a signal.' But with the current code we don't do that. If
initctl fails for some reason (e.g. inability to write to /dev/initctl)
we don't try sending any signal but fail immediately. To make things
worse, making a domain shutdown with bare _SIGNAL was working by blind
chance of a @rc variable being placed at correct place on the stack so
its initial value was zero.

Signed-off-by: Michal Privoznik mpriv...@redhat.com
---
 src/lxc/lxc_driver.c | 26 +-
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c
index c499182..a563b70 100644
--- a/src/lxc/lxc_driver.c
+++ b/src/lxc/lxc_driver.c
@@ -2594,7 +2594,8 @@ lxcDomainShutdownFlags(virDomainPtr dom,
 virDomainObjPtr vm;
 char *vroot = NULL;
 int ret = -1;
-int rc;
+int rc = 0;
+bool useInitctl = false, initctlRequested, signalRequested;
 
 virCheckFlags(VIR_DOMAIN_SHUTDOWN_INITCTL |
   VIR_DOMAIN_SHUTDOWN_SIGNAL, -1);
@@ -2623,25 +2624,24 @@ lxcDomainShutdownFlags(virDomainPtr dom,
 (unsigned long long)priv-initpid)  0)
 goto cleanup;
 
-if (flags == 0 ||
-(flags  VIR_DOMAIN_SHUTDOWN_INITCTL)) {
-if ((rc = virInitctlSetRunLevel(VIR_INITCTL_RUNLEVEL_POWEROFF,
-vroot))  0) {
+initctlRequested = flags  VIR_DOMAIN_SHUTDOWN_INITCTL;
+signalRequested = flags  VIR_DOMAIN_SHUTDOWN_SIGNAL;
+
+if (initctlRequested || !flags)
+useInitctl = true;
+
+if (useInitctl) {
+rc = virInitctlSetRunLevel(VIR_INITCTL_RUNLEVEL_POWEROFF, vroot);
+if (rc  0  !signalRequested)
 goto cleanup;
-}
-if (rc == 0  flags != 0 
-((flags  ~VIR_DOMAIN_SHUTDOWN_INITCTL) == 0)) {
+if (rc == 0  !signalRequested) {
 virReportError(VIR_ERR_OPERATION_UNSUPPORTED, %s,
_(Container does not provide an initctl pipe));
 goto cleanup;
 }
-} else {
-rc = 0;
 }
 
-if (rc == 0 
-(flags == 0 ||
- (flags  VIR_DOMAIN_SHUTDOWN_SIGNAL))) {
+if (rc == 0  !useInitctl) {
 if (kill(priv-initpid, SIGTERM)  0 
 errno != ESRCH) {
 virReportSystemError(errno,
-- 
1.8.5.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

2013-12-18 Thread Reco

 Hi.

On Wed, 18 Dec 2013 18:15:17 +0100
Michal Privoznik mpriv...@redhat.com wrote:

 While the part above looks okay, I think we should report error if the 
 /dev/initctl is not a pipe. Moreover, with your approach, if it's really not 
 a pipe, we don't call open() and proceed to safewrite(-1, ...) at line 153 
 which will report EBADF error.

I've adjusted the patch to your suggestion.


 Missing ';' at EOL ^^^ - I guess you missed that during copy-paste.

OOPS. My mistake.

 
 Looking forward to v2.

Attached. I've tried following Daniel suggestions too.


 BTW: consider using git when sending patches. I couldn't really apply this 
 patch via 'git am'.

I cannot figure out how to force 'git send-email' to reply on a mail.
It seems to insist to send a new one.

Reco
commit be722f5c49a6ceab5af7d4c5e672c0b70f969efd
Author: Reco recovery...@gmail.com
Date:   Wed Dec 18 23:44:17 2013 +0400

Check whenever init control is a pipe

diff --git a/src/util/virinitctl.c b/src/util/virinitctl.c
index 64bc23a..2a2ec27 100644
--- a/src/util/virinitctl.c
+++ b/src/util/virinitctl.c
@@ -24,7 +24,10 @@
 #include config.h
 
 #include sys/param.h
+#include sys/types.h
+#include dirent.h
 #include fcntl.h
+#include string.h
 
 #include internal.h
 #include virinitctl.h
@@ -46,11 +49,21 @@
  *  Copyright (C) 1995-2004 Miquel van Smoorenburg
  */
 
+#ifdef _DIRENT_HAVE_D_TYPE
+# if defined(__FreeBSD_kernel__)
+#  define VIR_INITCTL_DIR   /etc
+#  define VIR_INITCTL_FIFO  .initctl
+# else
+#  define VIR_INITCTL_DIR   /dev
+#  define VIR_INITCTL_FIFO  initctl
+# endif
+#else
 # if defined(__FreeBSD_kernel__)
 #  define VIR_INITCTL_FIFO  /etc/.initctl
 # else
 #  define VIR_INITCTL_FIFO  /dev/initctl
 # endif
+#endif
 
 # define VIR_INITCTL_MAGIC 0x03091969
 # define VIR_INITCTL_CMD_START  0
@@ -123,6 +136,13 @@ int virInitctlSetRunLevel(virInitctlRunLevel level,
 char *path = NULL;
 int ret = -1;
 
+#ifdef _DIRENT_HAVE_D_TYPE
+int dfd = -1;
+DIR *dev;
+char *dpath = NULL;
+struct dirent* dentry;
+#endif
+
 memset(req, 0, sizeof(req));
 
 req.magic = VIR_INITCTL_MAGIC;
@@ -131,6 +151,63 @@ int virInitctlSetRunLevel(virInitctlRunLevel level,
 /* Yes it is an 'int' field, but wants a numeric character. Go figure */
 req.runlevel = '0' + level;
 
+#ifdef _DIRENT_HAVE_D_TYPE
+if (vroot) {
+if (virAsprintf(dpath, %s/%s, vroot, VIR_INITCTL_DIR)  0)
+return -1;
+} else {
+if (VIR_STRDUP(dpath, VIR_INITCTL_DIR)  0)
+return -1;
+}
+
+if ((dfd = open(dpath,
+O_RDONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY|O_DIRECTORY|O_NOFOLLOW))  0) {
+virReportSystemError(errno,
+ _(%s is not a directory),
+ dpath);
+ret = -1;
+goto cleanup;
+}
+
+if ((dev = fdopendir(dfd)) == NULL) {
+virReportSystemError(errno,
+ _(Failed to open directory %s),
+ dpath);
+ret = -1;
+goto cleanup;
+}
+
+while ((dentry = readdir(dev))) {
+if (strcmp(dentry-d_name, VIR_INITCTL_FIFO) != 0)
+continue;
+
+if (dentry-d_type != DT_FIFO)
+{
+virReportSystemError(errno,
+ _(Init control %s/%s is not a pipe),
+ dpath, VIR_INITCTL_FIFO);
+closedir(dev);
+ret = -1;
+goto cleanup;
+}
+
+if (virAsprintf(path, %s/%s, dpath, VIR_INITCTL_FIFO)  0)
+{
+closedir(dev);
+ret = -1;
+goto cleanup;
+}
+
+break;
+}
+
+closedir(dev);
+
+if (! path) {
+ret = 0;
+goto cleanup;
+}
+#else
 if (vroot) {
 if (virAsprintf(path, %s/%s, vroot, VIR_INITCTL_FIFO)  0)
 return -1;
@@ -138,8 +215,9 @@ int virInitctlSetRunLevel(virInitctlRunLevel level,
 if (VIR_STRDUP(path, VIR_INITCTL_FIFO)  0)
 return -1;
 }
+#endif
 
-if ((fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY))  0) {
+if ((fd = open(path, O_WRONLY|O_NONBLOCK|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW))  0) {
 if (errno == ENOENT) {
 ret = 0;
 goto cleanup;
@@ -160,6 +238,10 @@ int virInitctlSetRunLevel(virInitctlRunLevel level,
 ret = 1;
 
 cleanup:
+#ifdef _DIRENT_HAVE_D_TYPE
+VIR_FREE(dpath);
+VIR_FORCE_CLOSE(dfd);
+#endif
 VIR_FREE(path);
 VIR_FORCE_CLOSE(fd);
 return ret;
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

On 12/18/2013 12:47 PM, Reco wrote:
 
 BTW: consider using git when sending patches. I couldn't really apply this 
 patch via 'git am'.
 
 I cannot figure out how to force 'git send-email' to reply on a mail.

git send-email --in-reply-to='messag...@example.com'

for the correct value of messageid.  But generally posting a new thread
is better than threading your v2 to your v1 anyways, as top-level
threads are easier to spot than a patch nested deep in reply to another
message.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] lvm storage pool destroy problem

Hi I am trying to stop the storage pool through php-libvirt api which is a
LVM volume group.

when I tried to stop it I got error

error: internal error Child process (/sbin/vgchange -aln mylvmgrp)
unexpected exit status 5:   Can't deactivate volume group mylvmgrp with 4
open logical volume(s)

Then I tried throggh virsh command but the result remain same.

*root@kvm02:~# virsh pool-destroy vms*
error: Failed to destroy pool vms
error: internal error Child process (/sbin/vgchange -aln mylvmgrp)
unexpected exit status 5:   Can't deactivate volume group mylvmgrp with 4
open logical volume(s)

The same error also occured if I use direct vgchange -aln on my kvm host,
but then I tried


*vgchange -aly mylvmgrp*

this time its worked

  14 logical volume(s) in volume group mylvmgrp now active

1) Now would you please help me? is this libvirt issue? or LVM?
2) Is this possible virsh pool-destroy pool also run the vgchange -aly
instead of -aln

Br.

Umar
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

2013-12-18 Thread Reco

 Hi.

On Wed, 18 Dec 2013 13:42:25 -0700
Eric Blake ebl...@redhat.com wrote:

 git send-email --in-reply-to='messag...@example.com'
 
 for the correct value of messageid.  But generally posting a new thread
 is better than threading your v2 to your v1 anyways, as top-level
 threads are easier to spot than a patch nested deep in reply to another
 message.

Thanks. Should I resend the patch the proper way again?

Reco

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel

On 12/18/2013 02:06 PM, Reco wrote:
  Hi.
 
 On Wed, 18 Dec 2013 13:42:25 -0700
 Eric Blake ebl...@redhat.com wrote:
 
 git send-email --in-reply-to='messag...@example.com'

 for the correct value of messageid.  But generally posting a new thread
 is better than threading your v2 to your v1 anyways, as top-level
 threads are easier to spot than a patch nested deep in reply to another
 message.
 
 Thanks. Should I resend the patch the proper way again?
 
 Reco

We also prefer that patches have authorship attributed to a legal name,
rather than a nickname.  But rather than resend your patch, I think we
should pay attention to Dan's advice - that any solution we come up with
that doesn't involve forking a child process, then resetting that child
into the correct namespace, is prone to races and therefore does not
solve the problem at hand (namely, that symlinks created in the guest
namespace must NOT be followed in the parent in host namespace).

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [Xen-devel] Setting devid for emulated NICs (Xen 4.3.1 / libvirt 1.2.0) using libxl driver

2013-12-18 Thread Jim Fehlig

Stefan Bader wrote:
On 18.12.2013 14:28, Ian Campbell wrote:

On Wed, 2013-12-18 at 14:12 +0100, Stefan Bader wrote:

On 18.12.2013 13:27, Ian Campbell wrote:

On Tue, 2013-12-17 at 18:32 +0100, Stefan Bader wrote:

Might this libxl fix be relevant:
commit 5420f26507fc5c9853eb1076401a8658d72669da
Author: Jim Fehlig jfeh...@suse.com
Date: Fri Jan 11 12:22:26 2013 +

libxl: Set vfb and vkb devid if not done so by the caller

Other devices set a sensible devid if the caller has not
done so.
Do the same for vfb and vkb. While at it, factor out the
common code
used to determine a sensible devid, so it can be used by
other
libxl__device_*_add functions.

Signed-off-by: Jim Fehlig jfeh...@suse.com
Acked-by: Ian Campbell ian.campb...@citrix.com
Committed-by: Ian Campbell ian.campb...@citrix.com

and a follow up in dfeccbeaa. Although the comment implies that nic's
were already correctly assigning a devid if the caller specified -1, so
I don't know why it doesn't work for you :-(

Ok, yes, the commit above indeed changes libxl__device_nic_add to call
libxl__device_nextid for the devid... Just how is this actually called.
Maybe not sufficient but git grep libxl__device_nic_add in the xen code
only
shows the definition and a declaration in libxl_internal.h to me...

I have a feeling a macro might be involved...

Here we go, look for DEFINE_DEVICE_REMOVE in libxl.c. We should really
add the eventual function names in comments to provide grep fodder

Oh duh, yeah. So in DEFINE_DEVICE_ADD a libxl_device_nic_add is created
which
calls to libxl__device_nic_add. When I look for the single _ version I find
a
call from xl_cmdimpl.c and its public declaration in libxl.h.
So I guess the bug is that libvirt in the libxl driver never seems to do so

The macro creates libxl__add_nics which adds the nics from the
libxl_domain_config-nics array. I don't think libvirt needs to call
libxl_device_nic_add manually unless it is hotplugging a new nic at
runtime.

Hm, so I think this is the path:

libxl_domain_create_new
- do_domain_create
- initiate_domain_create
- libxl__bootloader_run (HVM domain, skipping bootloader)
- domcreate_bootloader_done
- domcreate_rebuild_done
- domcreate_launch_dm
- libxl__spawn_local_dm
- domcreate_devmodel_started

In libxl__spawn_local_dm, there is the following loop:

for (i = 0; i d_config-num_nics; i++) {
/* We have to init the nic here, because we still haven't
* called libxl_device_nic_add at this point, but qemu needs
* the nic information to be complete.
*/
ret = libxl__device_nic_setdefault(gc, d_config-nics[i], domid);
if (ret)
goto error_out;
}

So I think when starting the dm, the devid just is not set as setdefault does
not seem to do so. I would be done in the later domcreate_devmodel_started
callback but that is too late for the generated qemu arguments.

Sorry for jumping in late...

I stumbled across this problem just before openSUSE13.1 released and did
a quick fix in libvirt

https://build.opensuse.org/package/view_file/Virtualization:openSUSE13.1/libvirt/libxl-hvm-nic.patch?expand=1

I removed setting the NIC devid in the libxl driver a while back to be
consistent with other devices

http://libvirt.org/git/?p=libvirt.git;a=commit;h=ba64b97134a6129a48684f22f31be92c3b6eef96

The quick fix was to essentially revert the above commit until I could
investigate further. Thank you for now having done that investigation
:). Can the devid assignment logic be moved from
libxl__device_nic_add() to libxl__device_nic_setdefault()?

Regards,
Jim

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [v7 2/6] change qemu driver to use hostdev common library

2013-12-18 Thread Jim Fehlig

Jim Fehlig wrote:
 Chunyan Liu wrote:
   
 Change qemu driver to use hostdev common library instead of APIs in
 qemu_hostdev.[ch]
   
 

 This patch causes 'make check' to fail qemuxml2xmltest and
 qemuargv2xmltest tests.  Also, there's a segfault when running
 qemuhotplugtest.

 ~/upstream/libvirt/tests libtool --mode=execute gdb qemuhotplugtest
 (gdb) r
 ...
 Program received signal SIGSEGV, Segmentation fault.
 0x77a6ed64 in virLockManagerPluginGetDriver (plugin=0x0) at
 locking/lock_manager.c:291
 291return plugin-driver;
 (gdb) bt
 #0  0x77a6ed64 in virLockManagerPluginGetDriver (plugin=0x0) at
 locking/lock_manager.c:291
 #1  0x77a6fc3f in virDomainLockManagerNew (plugin=0x0,
 uri=0x4fb743 qemu:///session,
 dom=0x7464f0, withResources=false) at locking/domain_lock.c:132
 #2  0x77a702a9 in virDomainLockDiskAttach (plugin=0x0,
 uri=0x4fb743 qemu:///session,
 dom=0x7464f0, disk=0x75f240) at locking/domain_lock.c:256
 #3  0x00456010 in qemuDomainAttachVirtioDiskDevice (conn=0x0,
 driver=0x72eba0 driver,
 vm=0x7464f0, disk=0x75f240) at qemu/qemu_hotplug.c:257
 #4  0x00457d06 in qemuDomainAttachDeviceDiskLive (conn=0x0,
 driver=0x72eba0 driver,
 vm=0x7464f0, dev=0x75eab0) at qemu/qemu_hotplug.c:777
 #5  0x0041d1a4 in testQemuHotplugAttach (vm=0x7464f0,
 dev=0x75eab0) at qemuhotplugtest.c:109
 #6  0x0041d83c in testQemuHotplug (data=0x7fffdc90) at
 qemuhotplugtest.c:279
 #7  0x0041ef7b in virtTestRun (title=0x4eea60 hotplug-base
 ATTACH disk-virtio,
 body=0x41d46e testQemuHotplug, data=0x7fffdc90) at testutils.c:138
 #8  0x0041e195 in mymain () at qemuhotplugtest.c:430
 #9  0x00420291 in virtTestMain (argc=1, argv=0x7fffe258,
 func=0x41da09 mymain)
 at testutils.c:593
 #10 0x0041ebc5 in main (argc=1, argv=0x7fffe258) at
 qemuhotplugtest.c:487
   


While chatting on IRC yesterday, Chunyan mentioned she was not seeing
this segfault. After applying the latest updates to my test setup, I'm
no longer seeing it either.

Regards,
Jim

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [v8 2/6] add a unit test for hostdev common library

Add unit test for hostdev common library. Current tests are based on virpcimock.

Signed-off-by: Chunyan Liu cy...@suse.com
---
 tests/Makefile.am  |5 +
 tests/virhostdevtest.c |  481 
 tests/virpcimock.c |   45 +-
 3 files changed, 530 insertions(+), 1 deletions(-)
 create mode 100644 tests/virhostdevtest.c

diff --git a/tests/Makefile.am b/tests/Makefile.am
index 568b7a0..3e66d8c 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -141,6 +141,7 @@ test_programs = virshtest sockettest \
 virportallocatortest \
sysinfotest \
virstoragetest \
+   virhostdevtest \
$(NULL)
 
 if WITH_REMOTE
@@ -754,6 +755,10 @@ vircgroupmock_la_CFLAGS = $(AM_CFLAGS)
 vircgroupmock_la_LDFLAGS = -module -avoid-version \
 -rpath /evil/libtool/hack/to/force/shared/lib/creation
 
+virhostdevtest_SOURCES = \
+   virhostdevtest.c testutils.h testutils.c
+virhostdevtest_LDADD = $(LDADDS)
+
 virpcitest_SOURCES = \
virpcitest.c testutils.h testutils.c
 virpcitest_LDADD = $(LDADDS)
diff --git a/tests/virhostdevtest.c b/tests/virhostdevtest.c
new file mode 100644
index 000..5b45548
--- /dev/null
+++ b/tests/virhostdevtest.c
@@ -0,0 +1,481 @@
+/*
+ * Copyright (C) 2013 SUSE LINUX Products GmbH, Nuernberg, Germany.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library.  If not, see
+ * http://www.gnu.org/licenses/.
+ *
+ * Author: Chunyan Liu cy...@suse.com
+ */
+
+#include config.h
+
+#include testutils.h
+
+#ifdef __linux__
+
+# include stdlib.h
+# include stdio.h
+# include sys/types.h
+# include sys/stat.h
+# include fcntl.h
+# include virhostdev.h
+
+# define VIR_FROM_THIS VIR_FROM_NONE
+
+# define CHECK_LIST_COUNT(list, cnt)\
+if ((count = virPCIDeviceListCount(list)) != cnt) { \
+virReportError(VIR_ERR_INTERNAL_ERROR,  \
+   Unexpected count of items in  #list : %d,\
+   expecting %zu, count, (size_t) cnt);   \
+goto cleanup;   \
+}
+
+#ifdef DEBUG
+#define DPRINTF(fmt, ...) printf(fmt, ## __VA_ARGS__)
+#else
+#define DPRINTF(fmt, ...) do { } while (0)
+#endif
+
+#define TEST_STATE_DIR abs_builddir /hostdevmgr
+static const char *drv_name = test_driver;
+static const char *dom_name = test_domain;
+static const unsigned char *uuid =
+(unsigned char *)(f92360b0-2541-8791-fb32-d1f838811541);
+static int nhostdevs = 3;
+static virDomainHostdevDefPtr hostdevs[] = {NULL, NULL, NULL};
+static virPCIDevicePtr dev[] = {NULL, NULL, NULL};
+static virHostdevManagerPtr mgr = NULL;
+
+static void
+myCleanup(void)
+{
+size_t i;
+for (i = 0; i  nhostdevs; i++) {
+ virPCIDeviceFree(dev[i]);
+ virDomainHostdevDefFree(hostdevs[i]);
+}
+
+if (mgr) {
+virObjectUnref(mgr-activePciHostdevs);
+virObjectUnref(mgr-inactivePciHostdevs);
+virObjectUnref(mgr-activeUsbHostdevs);
+VIR_FREE(mgr-stateDir);
+VIR_FREE(mgr);
+}
+}
+
+static int
+myInit(void)
+{
+size_t i;
+
+for (i = 0; i  nhostdevs; i++) {
+virDomainHostdevSubsys subsys;
+hostdevs[i] = virDomainHostdevDefAlloc();
+if (!hostdevs[i])
+goto cleanup;
+hostdevs[i]-mode = VIR_DOMAIN_HOSTDEV_MODE_SUBSYS;
+subsys.type = VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI;
+subsys.u.pci.addr.domain = 0;
+subsys.u.pci.addr.bus = 0;
+subsys.u.pci.addr.slot = i + 1;
+subsys.u.pci.addr.function = 0;
+subsys.u.pci.backend = VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM;
+hostdevs[i]-source.subsys = subsys;
+}
+
+for (i = 0; i  nhostdevs; i++) {
+if (!(dev[i] = virPCIDeviceNew(0, 0, i + 1, 0)) ||
+virPCIDeviceSetStubDriver(dev[i], pci-stub)  0)
+goto cleanup;
+}
+
+if (VIR_ALLOC(mgr)  0)
+goto cleanup;
+if ((mgr-activePciHostdevs = virPCIDeviceListNew()) == NULL)
+goto cleanup;
+if ((mgr-activeUsbHostdevs = virUSBDeviceListNew()) == NULL)
+goto cleanup;
+if ((mgr-inactivePciHostdevs = virPCIDeviceListNew()) == NULL)
+goto cleanup;
+if ((mgr-activeScsiHostdevs = virSCSIDeviceListNew()) == NULL)
+goto cleanup;
+if (VIR_STRDUP(mgr-stateDir,

[libvirt] [v8 5/6] add hostdev pci backend type for xen

Add VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN. For legacy xen, it will use pciback as
stub driver.

Signed-off-by: Chunyan Liu cy...@suse.com
---
 docs/schemas/domaincommon.rng |1 +
 src/conf/domain_conf.c|3 ++-
 src/conf/domain_conf.h|1 +
 src/libxl/libxl_domain.c  |9 +
 src/qemu/qemu_command.c   |3 +--
 src/qemu/qemu_hotplug.c   |4 +---
 src/util/virhostdev.c |   12 
 7 files changed, 27 insertions(+), 6 deletions(-)

diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng
index be32c6b..ebbeaca 100644
--- a/docs/schemas/domaincommon.rng
+++ b/docs/schemas/domaincommon.rng
@@ -3318,6 +3318,7 @@
 choice
   valuekvm/value
   valuevfio/value
+  valuexen/value
 /choice
   /attribute
   empty/
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index 0079234..9475b25 100644
--- a/src/conf/domain_conf.c
+++ b/src/conf/domain_conf.c
@@ -617,7 +617,8 @@ VIR_ENUM_IMPL(virDomainHostdevSubsysPciBackend,
   VIR_DOMAIN_HOSTDEV_PCI_BACKEND_TYPE_LAST,
   default,
   kvm,
-  vfio)
+  vfio,
+  xen)
 
 VIR_ENUM_IMPL(virDomainHostdevCaps, VIR_DOMAIN_HOSTDEV_CAPS_TYPE_LAST,
   storage,
diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h
index 647d115..5068b09 100644
--- a/src/conf/domain_conf.h
+++ b/src/conf/domain_conf.h
@@ -414,6 +414,7 @@ typedef enum {
 VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT, /* detect automaticaly, prefer 
VFIO */
 VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM,/* force legacy kvm style */
 VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO,   /* force vfio */
+VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN,/* force legacy xen style, use 
pciback */
 
 VIR_DOMAIN_HOSTDEV_PCI_BACKEND_TYPE_LAST
 } virDomainHostdevSubsysPciBackendType;
diff --git a/src/libxl/libxl_domain.c b/src/libxl/libxl_domain.c
index 68009db..32e798b 100644
--- a/src/libxl/libxl_domain.c
+++ b/src/libxl/libxl_domain.c
@@ -393,6 +393,15 @@ libxlDomainDeviceDefPostParse(virDomainDeviceDefPtr dev,
 STRNEQ(def-os.type, hvm))
 dev-data.chr-targetType = VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_XEN;
 
+if (dev-type == VIR_DOMAIN_DEVICE_HOSTDEV) {
+virDomainHostdevDefPtr hostdev = dev-data.hostdev;
+
+if (hostdev-mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS 
+hostdev-source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI 

+hostdev-source.subsys.u.pci.backend == 
VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT)
+hostdev-source.subsys.u.pci.backend = 
VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN;
+}
+
 return 0;
 }
 
diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c
index d50c9aa..6802376 100644
--- a/src/qemu/qemu_command.c
+++ b/src/qemu/qemu_command.c
@@ -5538,8 +5538,7 @@ qemuBuildPCIHostdevDevStr(virDomainDefPtr def,
 virBufferAddLit(buf, vfio-pci);
 break;
 
-case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT:
-case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_TYPE_LAST:
+default:
 virReportError(VIR_ERR_INTERNAL_ERROR,
_(invalid PCI passthrough type '%s'),
virDomainHostdevSubsysPciBackendTypeToString(backend));
diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c
index 9ea95b7..3d15d9f 100644
--- a/src/qemu/qemu_hotplug.c
+++ b/src/qemu/qemu_hotplug.c
@@ -1193,9 +1193,7 @@ qemuDomainAttachHostPciDevice(virQEMUDriverPtr driver,
 
 break;
 
-case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT:
-case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM:
-case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_TYPE_LAST:
+default:
 break;
 }
 
diff --git a/src/util/virhostdev.c b/src/util/virhostdev.c
index 63690b7..1b0354f 100644
--- a/src/util/virhostdev.c
+++ b/src/util/virhostdev.c
@@ -373,6 +373,13 @@ virHostdevGetPciHostDeviceList(virDomainHostdevDefPtr 
*hostdevs, int nhostdevs)
 }
 break;
 
+case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN :
+if (virPCIDeviceSetStubDriver(dev, pciback)  0) {
+virObjectUnref(list);
+return NULL;
+}
+break;
+
 case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM :
 case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT:
 if (virPCIDeviceSetStubDriver(dev, pci-stub)  0) {
@@ -1191,6 +1198,11 @@ virHostdevUpdateActivePciHostdevs(virHostdevManagerPtr 
mgr,
 goto cleanup;
 break;
 
+case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN :
+if (virPCIDeviceSetStubDriver(dev, pciback)  0)
+goto cleanup;
+break;
+
 case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM :
 case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT:
 if (virPCIDeviceSetStubDriver(dev, pci-stub)  0)
-- 
1.6.0.2

--
libvir-list mailing list
libvir-list@redhat.com

[libvirt] [v8 0/6]Write separate module for hostdev passthrough

These patches implements a separate module for hostdev passthrough so that it
could be shared by different drivers and can maintain a global state of a host
device.

patch 1/6: extract hostdev passthrough function from qemu_hostdev.c and make it
reusable by multiple drivers.
patch 2/6: add a unit test for hostdev common library.
patch 3/6: switch qemu driver to use the common library instead of its own
hostdev passthrough APIs.
patch 4/6: switch lxc driver to use the common library instead of its own
hostdev passthrough APIs.
patch 5/6: add a hostdev pci backend type for xen usage.
patch 6/6: add pci passthrough to libxl driver.

---
Changes:
  * fix v7 comments
  * reorganize order of patches
  * rebase to latest source code

Chunyan Liu (6):
  add hostdev passthrough common library
  add a unit test for hostdev common library
  change qemu driver to use hostdev common library
  change lxc driver to use hostdev common library
  add hostdev pci backend type for xen
  add pci passthrough to libxl driver

 docs/schemas/domaincommon.rng  |1 +
 po/POTFILES.in |3 +-
 src/Makefile.am|3 +-
 src/conf/domain_conf.c |3 +-
 src/conf/domain_conf.h |1 +
 src/libvirt_private.syms   |   21 +
 src/libxl/libxl_conf.c |   63 +
 src/libxl/libxl_conf.h |4 +
 src/libxl/libxl_domain.c   |9 +
 src/libxl/libxl_driver.c   |  448 +-
 src/lxc/lxc_conf.h |4 -
 src/lxc/lxc_driver.c   |   47 +-
 src/lxc/lxc_hostdev.c  |  413 -
 src/lxc/lxc_hostdev.h  |   43 -
 src/lxc/lxc_process.c  |   24 +-
 src/qemu/qemu_command.c|4 +-
 src/qemu/qemu_conf.h   |9 +-
 src/qemu/qemu_domain.c |   22 +
 src/qemu/qemu_driver.c |   81 +-
 src/qemu/qemu_hostdev.c| 1454 -
 src/qemu/qemu_hostdev.h|   76 -
 src/qemu/qemu_hotplug.c|  136 +-
 src/qemu/qemu_process.c|   40 +-
 src/util/virhostdev.c  | 1703 
 src/util/virhostdev.h  |  134 ++
 src/util/virpci.c  |   30 +-
 src/util/virpci.h  |9 +-
 src/util/virscsi.c |   28 +-
 src/util/virscsi.h |8 +-
 src/util/virusb.c  |   29 +-
 src/util/virusb.h  |8 +-
 tests/Makefile.am  |5 +
 .../qemuxml2argv-hostdev-pci-address.xml   |1 +
 .../qemuxml2argvdata/qemuxml2argv-net-hostdev.xml  |1 +
 tests/qemuxml2argvdata/qemuxml2argv-pci-rom.xml|2 +
 tests/virhostdevtest.c |  481 ++
 tests/virpcimock.c |   45 +-
 37 files changed, 3181 insertions(+), 2212 deletions(-)
 delete mode 100644 src/lxc/lxc_hostdev.c
 delete mode 100644 src/lxc/lxc_hostdev.h
 delete mode 100644 src/qemu/qemu_hostdev.c
 delete mode 100644 src/qemu/qemu_hostdev.h
 create mode 100644 src/util/virhostdev.c
 create mode 100644 src/util/virhostdev.h
 create mode 100644 tests/virhostdevtest.c

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [v8 6/6] add pci passthrough to libxl driver

Add pci passthrough to libxl driver, support attach-device, detach-device and
start a vm with pci hostdev specified.

Signed-off-by: Chunyan Liu cy...@suse.com
---
 src/libxl/libxl_conf.c   |   63 +++
 src/libxl/libxl_conf.h   |4 +
 src/libxl/libxl_driver.c |  448 +-
 3 files changed, 514 insertions(+), 1 deletions(-)

diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c
index aaeb00e..7791e4a 100644
--- a/src/libxl/libxl_conf.c
+++ b/src/libxl/libxl_conf.c
@@ -1126,6 +1126,66 @@ libxlDriverConfigGet(libxlDriverPrivatePtr driver)
 return cfg;
 }
 
+int
+libxlMakePci(virDomainHostdevDefPtr hostdev, libxl_device_pci *pcidev)
+{
+if (hostdev-mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS)
+return -1;
+if (hostdev-source.subsys.type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI)
+return -1;
+
+pcidev-domain = hostdev-source.subsys.u.pci.addr.domain;
+pcidev-bus = hostdev-source.subsys.u.pci.addr.bus;
+pcidev-dev = hostdev-source.subsys.u.pci.addr.slot;
+pcidev-func = hostdev-source.subsys.u.pci.addr.function;
+
+return 0;
+}
+
+static int
+libxlMakePciList(virDomainDefPtr def, libxl_domain_config *d_config)
+{
+virDomainHostdevDefPtr *l_hostdevs = def-hostdevs;
+size_t nhostdevs = def-nhostdevs;
+size_t npcidevs = 0;
+libxl_device_pci *x_pcidevs;
+size_t i, j;
+
+if (nhostdevs == 0)
+return 0;
+
+if (VIR_ALLOC_N(x_pcidevs, nhostdevs)  0)
+return -1;
+
+for (i = 0, j = 0; i  nhostdevs; i++) {
+if (l_hostdevs[i]-mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS)
+continue;
+if (l_hostdevs[i]-source.subsys.type != 
VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI)
+continue;
+
+libxl_device_pci_init(x_pcidevs[j]);
+
+if (libxlMakePci(l_hostdevs[i], x_pcidevs[j])  0)
+goto error;
+
+npcidevs++;
+j++;
+}
+
+VIR_SHRINK_N(x_pcidevs, nhostdevs, nhostdevs - npcidevs);
+d_config-pcidevs = x_pcidevs;
+d_config-num_pcidevs = npcidevs;
+
+return 0;
+
+error:
+for (i = 0; i  npcidevs; i++)
+libxl_device_pci_dispose(x_pcidevs[i]);
+
+VIR_FREE(x_pcidevs);
+return -1;
+}
+
 virCapsPtr
 libxlMakeCapabilities(libxl_ctx *ctx)
 {
@@ -1174,6 +1234,9 @@ libxlBuildDomainConfig(libxlDriverPrivatePtr driver,
 if (libxlMakeVfbList(driver, def, d_config)  0)
 return -1;
 
+if (libxlMakePciList(def, d_config)  0)
+return -1;
+
 d_config-on_reboot = def-onReboot;
 d_config-on_poweroff = def-onPoweroff;
 d_config-on_crash = def-onCrash;
diff --git a/src/libxl/libxl_conf.h b/src/libxl/libxl_conf.h
index ffa93bd..958a65b 100644
--- a/src/libxl/libxl_conf.h
+++ b/src/libxl/libxl_conf.h
@@ -38,6 +38,7 @@
 # include virchrdev.h
 
 
+# define LIBXL_DRIVER_NAME xenlight
 # define LIBXL_VNC_PORT_MIN  5900
 # define LIBXL_VNC_PORT_MAX  65535
 
@@ -148,6 +149,9 @@ libxlMakeVfb(libxlDriverPrivatePtr driver,
  virDomainGraphicsDefPtr l_vfb, libxl_device_vfb *x_vfb);
 
 int
+libxlMakePci(virDomainHostdevDefPtr hostdev, libxl_device_pci *pcidev);
+
+int
 libxlBuildDomainConfig(libxlDriverPrivatePtr driver,
virDomainObjPtr vm, libxl_domain_config *d_config);
 
diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
index 29aa6c7..7d4cc36 100644
--- a/src/libxl/libxl_driver.c
+++ b/src/libxl/libxl_driver.c
@@ -52,6 +52,7 @@
 #include virsysinfo.h
 #include viraccessapicheck.h
 #include viratomic.h
+#include virhostdev.h
 
 #define VIR_FROM_THIS VIR_FROM_LIBXL
 
@@ -328,6 +329,7 @@ libxlVmReap(libxlDriverPrivatePtr driver,
 virDomainShutoffReason reason)
 {
 libxlDomainObjPrivatePtr priv = vm-privateData;
+virHostdevManagerPtr hostdev_mgr;
 
 if (libxl_domain_destroy(priv-ctx, vm-def-id, NULL)  0) {
 virReportError(VIR_ERR_INTERNAL_ERROR,
@@ -335,6 +337,11 @@ libxlVmReap(libxlDriverPrivatePtr driver,
 return -1;
 }
 
+hostdev_mgr = virHostdevManagerGetDefault();
+if (hostdev_mgr != NULL)
+virHostdevReAttachDomainHostdevs(hostdev_mgr, LIBXL_DRIVER_NAME,
+ vm-def, VIR_SP_PCI_HOSTDEV);
+
 libxlVmCleanup(driver, vm, reason);
 return 0;
 }
@@ -558,6 +565,7 @@ libxlVmStart(libxlDriverPrivatePtr driver, virDomainObjPtr 
vm,
 #ifdef LIBXL_HAVE_DOMAIN_CREATE_RESTORE_PARAMS
 libxl_domain_restore_params params;
 #endif
+virHostdevManagerPtr hostdev_mgr;
 
 if (libxlDomainObjPrivateInitCtx(vm)  0)
 goto error;
@@ -616,6 +624,13 @@ libxlVmStart(libxlDriverPrivatePtr driver, virDomainObjPtr 
vm,
 goto error;
 }
 
+VIR_DEBUG(Preparing host PCI devices);
+hostdev_mgr = virHostdevManagerGetDefault();
+if (hostdev_mgr == NULL ||
+virHostdevPrepareDomainHostdevs(hostdev_mgr, LIBXL_DRIVER_NAME,
+vm-def, VIR_SP_PCI_HOSTDEV)  0)
+goto error;
+
 /* use as

[libvirt] [v8 4/6] change lxc driver to use hostdev common library

Change lxc driver to use hostdev common library instead of APIs in
lxc_hostdev.[ch]

Signed-off-by: Chunyan Liu cy...@suse.com
---
 po/POTFILES.in|1 -
 src/Makefile.am   |1 -
 src/lxc/lxc_conf.h|4 -
 src/lxc/lxc_driver.c  |   47 ---
 src/lxc/lxc_hostdev.c |  416 -
 src/lxc/lxc_hostdev.h |   43 -
 src/lxc/lxc_process.c |   24 +++-
 7 files changed, 48 insertions(+), 488 deletions(-)
 delete mode 100644 src/lxc/lxc_hostdev.c
 delete mode 100644 src/lxc/lxc_hostdev.h

diff --git a/po/POTFILES.in b/po/POTFILES.in
index fcd3af9..f999445 100644
--- a/po/POTFILES.in
+++ b/po/POTFILES.in
@@ -62,7 +62,6 @@ src/locking/lock_manager.c
 src/locking/sanlock_helper.c
 src/lxc/lxc_cgroup.c
 src/lxc/lxc_fuse.c
-src/lxc/lxc_hostdev.c
 src/lxc/lxc_container.c
 src/lxc/lxc_conf.c
 src/lxc/lxc_controller.c
diff --git a/src/Makefile.am b/src/Makefile.am
index 8b4a25c..96563dc 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -619,7 +619,6 @@ LXC_DRIVER_SOURCES =
\
lxc/lxc_container.c lxc/lxc_container.h \
lxc/lxc_cgroup.c lxc/lxc_cgroup.h   \
lxc/lxc_domain.c lxc/lxc_domain.h   \
-   lxc/lxc_hostdev.c lxc/lxc_hostdev.h \
lxc/lxc_monitor.c lxc/lxc_monitor.h \
lxc/lxc_process.c lxc/lxc_process.h \
lxc/lxc_fuse.c lxc/lxc_fuse.h   \
diff --git a/src/lxc/lxc_conf.h b/src/lxc/lxc_conf.h
index f6cbfc9..5940b16 100644
--- a/src/lxc/lxc_conf.h
+++ b/src/lxc/lxc_conf.h
@@ -92,10 +92,6 @@ struct _virLXCDriver {
 /* Immutable pointer, self-locking APIs */
 virDomainObjListPtr domains;
 
-/* Immutable pointer. Requires lock to be held before
- * calling APIs. */
-virUSBDeviceListPtr activeUsbHostdevs;
-
 /* Immutable pointer, self-locking APIs */
 virObjectEventStatePtr domainEventState;
 
diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c
index c499182..2824c96 100644
--- a/src/lxc/lxc_driver.c
+++ b/src/lxc/lxc_driver.c
@@ -70,6 +70,7 @@
 #include virstring.h
 #include viraccessapicheck.h
 #include viraccessapichecklxc.h
+#include virhostdev.h
 
 #define VIR_FROM_THIS VIR_FROM_LXC
 
@@ -1417,9 +1418,6 @@ static int lxcStateInitialize(bool privileged,
 if (!(lxc_driver-securityManager = lxcSecurityInit(cfg)))
 goto cleanup;
 
-if ((lxc_driver-activeUsbHostdevs = virUSBDeviceListNew()) == NULL)
-goto cleanup;
-
 if ((virLXCDriverGetCapabilities(lxc_driver, true)) == NULL)
 goto cleanup;
 
@@ -1534,7 +1532,6 @@ static int lxcStateCleanup(void)
 
 virSysinfoDefFree(lxc_driver-hostsysinfo);
 
-virObjectUnref(lxc_driver-activeUsbHostdevs);
 virObjectUnref(lxc_driver-caps);
 virObjectUnref(lxc_driver-securityManager);
 virObjectUnref(lxc_driver-xmlopt);
@@ -3172,6 +3169,7 @@ lxcDomainAttachDeviceHostdevSubsysUSBLive(virLXCDriverPtr 
driver,
 mode_t mode;
 bool created = false;
 virUSBDevicePtr usb = NULL;
+virHostdevManagerPtr hostdev_mgr;
 
 if (virDomainHostdevFind(vm-def, def, NULL) = 0) {
 virReportError(VIR_ERR_OPERATION_FAILED, %s,
@@ -3179,6 +3177,14 @@ 
lxcDomainAttachDeviceHostdevSubsysUSBLive(virLXCDriverPtr driver,
 return -1;
 }
 
+hostdev_mgr = virHostdevManagerGetDefault();
+if (hostdev_mgr == NULL ||
+virHostdevPrepareUsbHostdevs(hostdev_mgr,
+ LXC_DRIVER_NAME,
+ vm-def-name,
+ def, 1, 0)  0)
+return -1;
+
 if (virAsprintf(vroot, /proc/%llu/root,
 (unsigned long long)priv-initpid)  0)
 goto cleanup;
@@ -3254,6 +3260,11 @@ 
lxcDomainAttachDeviceHostdevSubsysUSBLive(virLXCDriverPtr driver,
 ret = 0;
 
 cleanup:
+virHostdevReAttachUsbHostdevs(hostdev_mgr,
+  LXC_DRIVER_NAME,
+  vm-def-name,
+  def,
+  1);
 virDomainAuditHostdev(vm, def, attach, ret == 0);
 if (ret  0  created)
 unlink(dstfile);
@@ -3716,8 +3727,7 @@ cleanup:
 
 
 static int
-lxcDomainDetachDeviceHostdevUSBLive(virLXCDriverPtr driver,
-virDomainObjPtr vm,
+lxcDomainDetachDeviceHostdevUSBLive(virDomainObjPtr vm,
 virDomainDeviceDefPtr dev)
 {
 virLXCDomainObjPrivatePtr priv = vm-privateData;
@@ -3726,6 +3736,7 @@ lxcDomainDetachDeviceHostdevUSBLive(virLXCDriverPtr 
driver,
 char *dst = NULL;
 char *vroot = NULL;
 virUSBDevicePtr usb = NULL;
+virHostdevManagerPtr hostdev_mgr;
 
 if ((idx = virDomainHostdevFind(vm-def,
 dev-data.hostdev,
@@ -3770,9 +3781,10 @@

Re: [libvirt] [PATCH RFC 02/27] qemu: snapshot: Avoid libvirtd crash when qemu crashes while snapshotting

On 12/16/2013 09:32 AM, Peter Krempa wrote:
 We shouldn't access the domain definition while we are in the monitor
 section as the domain is unlocked. Additionaly after we exit from the

s/Additionaly/Additionally/

 monitor we need to check if the VM is still alive. Not doing so resulted
 into crash if qemu exits while attempting to do a external VM snapshot.

s/into/in a/
s/a external/an external/

 ---
  src/qemu/qemu_driver.c | 46 +++---
  1 file changed, 35 insertions(+), 11 deletions(-)
 

ACK.  This one can be applied in spite of the RFC.  And patch 1 was
indeed helpful in testing this.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 02/27] qemu: snapshot: Avoid libvirtd crash when qemu crashes while snapshotting

On 12/18/2013 08:42 PM, Eric Blake wrote:
 On 12/16/2013 09:32 AM, Peter Krempa wrote:
 We shouldn't access the domain definition while we are in the monitor
 section as the domain is unlocked. Additionaly after we exit from the
 
 s/Additionaly/Additionally/
 
 monitor we need to check if the VM is still alive. Not doing so resulted
 into crash if qemu exits while attempting to do a external VM snapshot.
 
 s/into/in a/
 s/a external/an external/
 
 ---
  src/qemu/qemu_driver.c | 46 +++---
  1 file changed, 35 insertions(+), 11 deletions(-)

 
 ACK.  This one can be applied in spite of the RFC.  And patch 1 was
 indeed helpful in testing this.

Spoke too soon.  After trying to redefine and restart my domain, I got
the dreaded state change lock error, which means there's still a locking
discrepancy in your patch.

2013-12-19 04:09:14.351+: 10084: error : qemuMonitorIO:656 :
internal error: End of file from monitor
2013-12-19 04:10:44.000+: 10100: warning :
qemuDomainObjBeginJobInternal:1093 : Cannot start job (modify, none) for
domain f18-live; current job is (none, snapshot) owned by (0, 10099)
2013-12-19 04:10:44.000+: 10100: error :
qemuDomainObjBeginJobInternal:1097 : Timed out during operation: cannot
acquire state change lock

It's too late for me to try and find it tonight, but I'll try again in
the morning if you haven't found it first.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 03/27] Revert DO NOT APPLY UPSTREAM: Reproducer for disk snapshot crash

On 12/16/2013 09:32 AM, Peter Krempa wrote:
 This reverts commit 96313c60b91197578b313717db154b6bac029c9e.

Also not to be applied upstream :)

 ---
  src/qemu/qemu_driver.c | 3 ---
  1 file changed, 3 deletions(-)
 
 diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
 index ab20dfb..2140e1b 100644
 --- a/src/qemu/qemu_driver.c
 +++ b/src/qemu/qemu_driver.c
 @@ -12226,9 +12226,6 @@ 
 qemuDomainSnapshotCreateSingleDiskActive(virQEMUDriverPtr driver,
  goto cleanup;
  }
 
 -kill(vm-pid, 9);
 -sleep(1);
 -
  /* create the actual snapshot */
  if (snap-format)
  formatStr = virStorageFileFormatTypeToString(snap-format);
 

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 05/27] storage: gluster: Properly fix missing slashes in volume paths

On 12/16/2013 09:32 AM, Peter Krempa wrote:
 Add a slash to the volume key so that the URI that is generated from it
 will be formated properly. The extra slash doesn't hurt in the 'key'
 attribute as we just need it to be unique.
 ---
  src/storage/storage_backend_gluster.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
 
 diff --git a/src/storage/storage_backend_gluster.c 
 b/src/storage/storage_backend_gluster.c
 index 1be9034..36e99e9 100644
 --- a/src/storage/storage_backend_gluster.c
 +++ b/src/storage/storage_backend_gluster.c
 @@ -222,7 +222,7 @@ 
 virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
 
  if (VIR_STRDUP(vol-name, name)  0)
  goto cleanup;
 -if (virAsprintf(vol-key, %s%s%s, state-volname, state-dir,
 +if (virAsprintf(vol-key, /%s%s%s, state-volname, state-dir,

This does not match the documentation.  I'd rather fix the problem with
the URI generation (in patch 4/27) and leave the key unchanged, if only
because we already released 1.2.0 with documentation that the key does
not have a leading slash.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 04/27] Revert storage: fix omitted slash in gluster volume URI

On 12/16/2013 09:32 AM, Peter Krempa wrote:
 The fix for the missing slash at the beginning of the volume URI prints
 to a bad variable causing a memory leak and another one if the variable
 will be fixed. Revert the commit instead as we can just add a slash to
 the key attribute.
 
 The fix also caused a regression, where the path of the volume didn't
 include the actual volume name and thus was invalid.
 
 This reverts commit 6cd60b687acd04ea1538690b7d80f2809e0e29d4.
 ---
  src/storage/storage_backend_gluster.c | 5 +
  1 file changed, 1 insertion(+), 4 deletions(-)
 
 diff --git a/src/storage/storage_backend_gluster.c 
 b/src/storage/storage_backend_gluster.c
 index 622526b..1be9034 100644
 --- a/src/storage/storage_backend_gluster.c
 +++ b/src/storage/storage_backend_gluster.c
 @@ -227,10 +227,7 @@ 
 virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
  goto cleanup;
 
  tmp = state-uri-path;
 -if (virAsprintf(vol-key, %s%s, state-uri-path, name)  0) {

Uggh.  I see what I did: I swapped vol-key and state-uri-path, and
didn't run my patch through valgrind (where I would have seen the problem).

 -state-uri-path = tmp;
 -goto cleanup;
 -}
 +state-uri-path = vol-key;

Rather than reverting this patch, I'd prefer to fix the string creation
to use the intended order:

diff --git i/src/storage/storage_backend_gluster.c
w/src/storage/storage_backend_gluster.c
index 622526b..aab70ba 100644
--- i/src/storage/storage_backend_gluster.c
+++ w/src/storage/storage_backend_gluster.c
@@ -227,7 +227,7 @@
virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
 goto cleanup;

 tmp = state-uri-path;
-if (virAsprintf(vol-key, %s%s, state-uri-path, name)  0) {
+if (virAsprintf(state-uri-path, %s%s, vol-key, name)  0) {
 state-uri-path = tmp;
 goto cleanup;
 }


-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 04/27] Revert storage: fix omitted slash in gluster volume URI

On 12/18/2013 09:30 PM, Eric Blake wrote:

 
 Rather than reverting this patch, I'd prefer to fix the string creation
 to use the intended order:
 
 diff --git i/src/storage/storage_backend_gluster.c
 w/src/storage/storage_backend_gluster.c
 index 622526b..aab70ba 100644
 --- i/src/storage/storage_backend_gluster.c
 +++ w/src/storage/storage_backend_gluster.c
 @@ -227,7 +227,7 @@
 virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
  goto cleanup;
 
  tmp = state-uri-path;
 -if (virAsprintf(vol-key, %s%s, state-uri-path, name)  0) {
 +if (virAsprintf(state-uri-path, %s%s, vol-key, name)  0) {

Except that's not right, either.  I'll post what I actually tested in a
separate thread.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH] storage: fix bogus target in gluster volume xml

Commit 6cd60b6 was flat out broken - it tried to print into the
wrong variable.  My testing was obviously too cursory (did the
name get a slash added?); valgrind would have caught the error.
Thankfully it didn't hit any release.

Reported by Peter Krempa.

* src/storage/storage_backend_gluster.c
(virStorageBackendGlusterRefreshVol): Fix bogus code.

Signed-off-by: Eric Blake ebl...@redhat.com
---
 src/storage/storage_backend_gluster.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/storage/storage_backend_gluster.c 
b/src/storage/storage_backend_gluster.c
index 622526b..9aa692e 100644
--- a/src/storage/storage_backend_gluster.c
+++ b/src/storage/storage_backend_gluster.c
@@ -227,7 +227,7 @@ 
virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
 goto cleanup;

 tmp = state-uri-path;
-if (virAsprintf(vol-key, %s%s, state-uri-path, name)  0) {
+if (virAsprintf(state-uri-path, /%s, vol-key)  0) {
 state-uri-path = tmp;
 goto cleanup;
 }
-- 
1.8.4.2

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

[libvirt] [PATCH 1/1] virsh nodecpustats erroneously returns stats for offline cpus on Linux

2013-12-18 Thread Pradipta Kr. Banerjee

From: Pradipta Kr. Banerjee bpra...@in.ibm.com

virsh nodecpustats erroneously returns stats for offline cpus on Linux

To retrieve node cpu statistics on Linux system, the
linuxNodeGetCPUstats function performs a minimal match of the input
cpuid with the cpuid read from /proc/cpustat. On systems with larger
cpus this can result in erroneous data.
For eg if /proc/stat has similar data
 cpu  3645648 485 361475 3691194868 176595 40645 69850 0 1526172 0
 cpu0 494045 62 42763 138516619 16107 2339 2431 0 248345 0
 cpu4 402646 59 34745 138617562 17738 1413 1196 0 193948 0
 cpu8 318662 32 32119 138715131 7903 927 982 0 115429 0
 cpu12 247004 33 30465 138791438 5251 759 913 0 51938 0

and cpu 1,2 are offline, then

'virsh nodecpustats 1' displays data for cpu12

whereas virsh nodecpustats 2 correctly throws the following error

error: Unable to get node cpu stats
error: Invalid cpuNum in linuxNodeGetCPUStats

This patch fixes the above mentioned problem

Signed-off-by: Pradipta Kr. Banerjee bpra...@in.ibm.com
---
 src/nodeinfo.c | 13 +
 1 file changed, 13 insertions(+)

diff --git a/src/nodeinfo.c b/src/nodeinfo.c
index 1838547..c9e5ff1 100644
--- a/src/nodeinfo.c
+++ b/src/nodeinfo.c
@@ -614,6 +614,8 @@ linuxNodeGetCPUStats(FILE *procstat,
 unsigned long long usr, ni, sys, idle, iowait;
 unsigned long long irq, softirq, steal, guest, guest_nice;
 char cpu_header[3 + INT_BUFSIZE_BOUND(cpuNum)];
+char cpu_header_read[3 + INT_BUFSIZE_BOUND(cpuNum)];
+char cpu_header_fmt[8];

 if ((*nparams) == 0) {
 /* Current number of cpu stats supported by linux */
@@ -631,8 +633,11 @@ linuxNodeGetCPUStats(FILE *procstat,

 if (cpuNum == VIR_NODE_CPU_STATS_ALL_CPUS) {
 strcpy(cpu_header, cpu);
+snprintf(cpu_header_fmt, sizeof(cpu_header_fmt), %%%ds, 3);
 } else {
 snprintf(cpu_header, sizeof(cpu_header), cpu%d, cpuNum);
+snprintf(cpu_header_fmt, sizeof(cpu_header_fmt), %%%ds,
+ (int)(3 + INT_BUFSIZE_BOUND(cpuNum)));
 }

 while (fgets(line, sizeof(line), procstat) != NULL) {
@@ -649,6 +654,14 @@ linuxNodeGetCPUStats(FILE *procstat,
 continue;
 }

+/*
+ * Process with stats gathering only if the cpuid from /proc/stat
+ * matches exactly with the input cpuid
+*/
+sscanf(buf, cpu_header_fmt, cpu_header_read);
+if (STRNEQ(cpu_header, cpu_header_read))
+continue;
+
 for (i = 0; i  *nparams; i++) {
 virNodeCPUStatsPtr param = params[i];

--
1.8.3.1

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Re: [libvirt] [PATCH RFC 02/27] qemu: snapshot: Avoid libvirtd crash when qemu crashes while snapshotting