On 08/08/13 08:43, John Ferlan wrote:
Update formatsecret docs to describe the various options and provide examples
in order to set up secrets for each type of secret.
---
docs/formatsecret.html.in | 156 ++
1 file changed, 145 insertions(+), 11
On 20/08/13 05:21, John Ferlan wrote:
From e31f3596893302ee1f96d2eb0cf4e006294c528c Mon Sep 17 00:00:00 2001
From: John Ferlanjfer...@redhat.com
Date: Wed, 7 Aug 2013 09:05:43 -0400
Subject: [PATCH 4/7] docs: Update the formatdomain disk examples
Add more iSCSI examples including having a
When backing files of a disk file are stored in NFS shared folder,
qemu process requires at least the read permission.
But in some rare situation, these backing files can not even be read
If so, we should treat the diskchains as broken.
This patch add a checking for this, report negative in such
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common dealloc function.
This patch also fixes a few forgotten checks of return from
Explicitly let the user know about the unknown volume name.
---
tools/virsh-pool.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c
index b8fc8d7..592b81f 100644
--- a/tools/virsh-pool.c
+++ b/tools/virsh-pool.c
@@ -1000,7 +1000,7 @@
Use the new semantics of vshStringToArray to avoid leaking the array of
volumes to be deleted. The array would be leaked in case the first
volume was found in the domain definition. Also refactor the code a bit
to sanitize naming of variables hoding arrays and dimensions of the
arrays.
Resolves:
Introduced by commit e0139e30444:
1777/* Updating pool metadata */
(40) Event var_deref_op: Dereferencing null pointer newvol.
Also see events: [assign_zero]
1778pool-def-allocation += newvol-allocation;
1779pool-def-available -= newvol-allocation;
---
Introduced by commit e0139e30444. virStorageVolDefFree free'ed the
pointers that are still used by the added volume object, this changes
it back to VIR_FREE.
---
src/storage/storage_driver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/storage/storage_driver.c
Osier Yang (2):
storage: Fix coverity warning
storage: Fix the use-after-free memory bug
src/storage/storage_driver.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--
1.8.1.4
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
Usage of vshStringToArray in virsh was problematic in a few places.
Clean it up to avoid a few memleaks. Also improve an error message.
Peter Krempa (3):
virsh: modify vshStringToArray to duplicate the elements too
virsh-pool: Improve error message in cmdPoolList
virsh: Don't leak list of
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
configured any, we set the zero in cgroup subsystem making the kernel
kill the corresponding qemu process immediately. The proper fix is to
set hard_limit iff user has
On 20/08/13 17:05, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common dealloc function.
This patch also
On 20/08/13 17:05, Peter Krempa wrote:
Explicitly let the user know about the unknown volume name.
s/volume name/pool type/,
ACK with above.
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
On 08/20/13 11:16, Osier Yang wrote:
On 20/08/13 17:05, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common
Please see notes in individual patches.
Peter Krempa (2):
virBitmapParse: Fix behavior in case of error
virBitmapParse: Don't shadow errors
po/POTFILES.in | 1 +
src/conf/domain_conf.c | 5 +
src/conf/network_conf.c | 3 ---
src/nodeinfo.c | 5 +
A few of the callers of virBitmapParse shadow the returned error.
---
Notes:
I'm kind of worried that we are making some error messages
worse compared to what they were before. If you don't like
the way this will turn out, I'll prepare a inverse fix
where all callers will be fixed
Re-arrange the code so that the returned bitmap is always initialized to
NULL even on early failures and return an error message as some callers
are already expecting it.
---
Notes:
Version 2:
Was already ACKed in v1, but:
* fixed bracing of arguments of the _() macro
* added
Hi.
I tried this on Fedora 19, using libvirt 1.0.5 and also tested with
git v1.1.1-maint as well as git master on Ubuntu 12.04.
How to reproduce:
python -
import libvirt as l
c = l.virConnect(test:///default)
v = c.virStream()
-
Here's my GDB session with git master:
Hello!
QEMU is switching the default QCOW2 version from v2 (compat=0.10) to v3
(compat=1.1) [1]
Currently, libvirt only specifies the compat=0.10 option if it was explicitly
requested (to avoid parsing qemu-img help output [2]) and assumes the format
to be v2 when it calls qemu-img without the
Hi
Any comments?
Thanks
-Original Message-
From: libvir-list-boun...@redhat.com
[mailto:libvir-list-boun...@redhat.com]
On Behalf Of Chen HanXiao
Sent: Wednesday, August 14, 2013 9:30 AM
To: 'Daniel P. Berrange'
Cc: libvir-list@redhat.com
Subject: Re: [libvirt] [PATCH
On Mon, Aug 19, 2013 at 08:19:56PM +0200, Giuseppe Scrivano wrote:
---
I have started working on:
https://bugzilla.redhat.com/show_bug.cgi?id=916786
+/**
+ * virConnectGetCPUMapDesc:
+ *
+ * @conn: virConnect connection
+ *
+ * Get the content of the cpu_map.xml file used by the
On Mon, Aug 19, 2013 at 08:07:42PM +0200, Guido Günther wrote:
On Mon, Aug 19, 2013 at 11:37:24AM -0600, Eric Blake wrote:
On 08/19/2013 11:28 AM, Guido Günther wrote:
and use it when available
---
configure.ac | 1 +
m4/virt-linker-no-indirect.m4 | 30
On 20/08/13 17:05, Peter Krempa wrote:
Use the new semantics of vshStringToArray to avoid leaking the array of
volumes to be deleted. The array would be leaked in case the first
volume was found in the domain definition. Also refactor the code a bit
to sanitize naming of variables hoding arrays
On Tue, Aug 20, 2013 at 11:33:36AM +0200, Ján Tomko wrote:
Hello!
QEMU is switching the default QCOW2 version from v2 (compat=0.10) to v3
(compat=1.1) [1]
Currently, libvirt only specifies the compat=0.10 option if it was explicitly
requested (to avoid parsing qemu-img help output [2])
On 20/08/13 17:20, Peter Krempa wrote:
On 08/20/13 11:16, Osier Yang wrote:
On 20/08/13 17:05, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with
On 20/08/13 17:10, Michal Privoznik wrote:
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
configured any, we set the zero in cgroup subsystem making the kernel
kill the corresponding qemu process immediately. The
On 08/20/13 11:58, Osier Yang wrote:
On 20/08/13 17:05, Peter Krempa wrote:
Use the new semantics of vshStringToArray to avoid leaking the array of
volumes to be deleted. The array would be leaked in case the first
volume was found in the domain definition. Also refactor the code a bit
to
On 08/20/13 12:08, Osier Yang wrote:
On 20/08/13 17:10, Michal Privoznik wrote:
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
s/default,/default/
configured any, we set the zero in cgroup subsystem making the
On 08/20/13 12:01, Osier Yang wrote:
On 20/08/13 17:20, Peter Krempa wrote:
On 08/20/13 11:16, Osier Yang wrote:
On 20/08/13 17:05, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
On 08/20/2013 11:10 AM, Michal Privoznik wrote:
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
configured any, we set the zero in cgroup subsystem making the kernel
kill the corresponding qemu process immediately.
On 08/19/2013 11:09 PM, John Ferlan wrote:
On 08/16/2013 12:34 PM, Ján Tomko wrote:
On 08/08/2013 02:43 AM, John Ferlan wrote:
VIR_SECRET_GET_VALUE_INTERNAL_CALL);
if (!secret_value) {
-
On Mon, Aug 12, 2013 at 10:19:47PM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=924153
Commit 904e05a2 (v0.9.9) added a per-disk seclabel element with
an attribute relabel='no' in order to try and minimize the
impact of shutdown delays when an NFS server disappears.
If there's no hard_limit set and domain uses VFIO we still must lock the
guest memory (prerequisite from qemu). Hence, we should compute the
amount to be locked from max_baloon.
---
src/qemu/qemu_command.c | 15 +--
1 file changed, 13 insertions(+), 2 deletions(-)
diff --git
On 20.08.2013 14:11, Ján Tomko wrote:
On 08/20/2013 11:10 AM, Michal Privoznik wrote:
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
configured any, we set the zero in cgroup subsystem making the kernel
kill the
On 08/20/2013 03:43 AM, Osier Yang wrote:
On 20/08/13 05:21, John Ferlan wrote:
From e31f3596893302ee1f96d2eb0cf4e006294c528c Mon Sep 17 00:00:00 2001
From: John Ferlan jfer...@redhat.com
Date: Wed, 7 Aug 2013 09:05:43 -0400
Subject: [PATCH 4/7] docs: Update the formatdomain disk examples
On 08/16/13 14:41, Peter Krempa wrote:
On 08/16/13 12:39, Daniel P. Berrange wrote:
On Fri, Aug 16, 2013 at 12:32:06PM +0200, Peter Krempa wrote:
The bitmap parsing code might cause a crash of the application using it.
Fix it and add tests so that it doesn't happen again.
Peter Krempa (3):
On 08/20/2013 03:09 PM, Michal Privoznik wrote:
On 20.08.2013 14:11, Ján Tomko wrote:
On 08/20/2013 11:10 AM, Michal Privoznik wrote:
Since 16bcb3 we have a regression. The hard_limit is set
unconditionally. By default, the limit is zero. Hence, if user hasn't
configured any, we set the zero
On 08/20/2013 07:06 AM, Michal Privoznik wrote:
If there's no hard_limit set and domain uses VFIO we still must lock the
guest memory (prerequisite from qemu). Hence, we should compute the
amount to be locked from max_baloon.
s/baloon/balloon/
---
src/qemu/qemu_command.c | 15
On Thu, Aug 01, 2013 at 09:28:10PM +0200, Jiri Denemark wrote:
Otherwise defining variables that hold callbacks pointers is ugly and
several places have to be changed when new parameters are added.
---
src/qemu/qemu_monitor.c | 6 +-
src/qemu/qemu_monitor.h | 171
On Thu, Aug 01, 2013 at 09:28:11PM +0200, Jiri Denemark wrote:
We will have to pass a mock-up of the driver when testing monitor
events.
---
src/qemu/qemu_capabilities.c | 5 ++-
src/qemu/qemu_monitor.c | 24 +++
src/qemu/qemu_monitor.h | 69 --
On Thu, Aug 01, 2013 at 09:28:12PM +0200, Jiri Denemark wrote:
---
src/qemu/qemu_driver.c | 114 --
src/qemu/qemu_hotplug.c | 142
src/qemu/qemu_hotplug.h | 16 ++
3 files changed, 134 insertions(+),
On Thu, Aug 01, 2013 at 09:28:14PM +0200, Jiri Denemark wrote:
Each test case label now contains more data useful to identify the test.
---
tests/qemuhotplugtest.c | 29 -
1 file changed, 12 insertions(+), 17 deletions(-)
ACK
Daniel
--
|: http://berrange.com
On Thu, Aug 01, 2013 at 09:28:13PM +0200, Jiri Denemark wrote:
---
src/qemu/qemu_driver.c | 52 --
src/qemu/qemu_hotplug.c | 67
-
src/qemu/qemu_hotplug.h | 9 +++
3 files changed, 64 insertions(+),
On Thu, Aug 01, 2013 at 09:28:15PM +0200, Jiri Denemark wrote:
We need to make sure a device is properly added/removed (or not) to a
domain definition to check that a hotplug API did not lie to us.
---
tests/qemuhotplugtest.c| 67 +--
On Thu, Aug 01, 2013 at 09:28:16PM +0200, Jiri Denemark wrote:
---
tests/qemuhotplugtest.c | 15 +--
1 file changed, 9 insertions(+), 6 deletions(-)
ACK
Daniel
--
|: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org -o-
On Thu, Aug 01, 2013 at 09:28:17PM +0200, Jiri Denemark wrote:
This is a generic XML usable for hotplugging various types of devices.
---
.../qemuxml2argv-hotplug-base.args | 7
.../qemuxml2argvdata/qemuxml2argv-hotplug-base.xml | 38
++
On Thu, Aug 01, 2013 at 09:28:21PM +0200, Jiri Denemark wrote:
---
src/Makefile.am | 1 +
src/qemu/qemu_process.c | 5 +++--
src/qemu/qemu_process.h | 2 +-
src/qemu/qemu_processpriv.h | 37 +
4 files changed, 42 insertions(+), 3
On 20/08/13 21:10, John Ferlan wrote:
On 08/20/2013 03:43 AM, Osier Yang wrote:
On 20/08/13 05:21, John Ferlan wrote:
From e31f3596893302ee1f96d2eb0cf4e006294c528c Mon Sep 17 00:00:00 2001
From: John Ferlan jfer...@redhat.com
Date: Wed, 7 Aug 2013 09:05:43 -0400
Subject: [PATCH 4/7] docs:
On Thu, Aug 01, 2013 at 09:28:19PM +0200, Jiri Denemark wrote:
Some tests need the monitor to operate on an already existing VM object
rather than on a new mock-up the monitor test normally creates.
---
tests/qemuhotplugtest.c | 2 +-
tests/qemumonitorjsontest.c | 26
On Thu, Aug 01, 2013 at 09:28:18PM +0200, Jiri Denemark wrote:
diff --git a/tests/qemuhotplugtest.c b/tests/qemuhotplugtest.c
index 4712334..bb047fe 100644
--- a/tests/qemuhotplugtest.c
+++ b/tests/qemuhotplugtest.c
@@ -313,12 +327,22 @@ mymain(void)
driver.config =
On Thu, Aug 01, 2013 at 09:28:20PM +0200, Jiri Denemark wrote:
The driver is then passed to monitor event handlers.
---
tests/qemuhotplugtest.c | 2 +-
tests/qemumonitortestutils.c | 5 +++--
tests/qemumonitortestutils.h | 6 --
3 files changed, 8 insertions(+), 5 deletions(-)
ACK
On Thu, Aug 01, 2013 at 09:28:23PM +0200, Jiri Denemark wrote:
---
tests/qemuhotplugtest.c | 29 +++--
tests/qemumonitortestutils.c | 2 ++
2 files changed, 25 insertions(+), 6 deletions(-)
ACK
Daniel
--
|: http://berrange.com -o-
On Thu, Aug 01, 2013 at 09:28:22PM +0200, Jiri Denemark wrote:
We don't want tests to wait 5 seconds for an event which we know will
never come.
---
src/Makefile.am | 1 +
src/qemu/qemu_hotplug.c | 10 ++
src/qemu/qemu_hotplugpriv.h | 31
On 08/20/2013 03:06 PM, Michal Privoznik wrote:
If there's no hard_limit set and domain uses VFIO we still must lock the
guest memory (prerequisite from qemu). Hence, we should compute the
amount to be locked from max_baloon.
You forgot about the bookkeepers, didn't you?
---
On Thu, Aug 01, 2013 at 09:28:24PM +0200, Jiri Denemark wrote:
---
tests/qemuhotplugtest.c | 23 +++
1 file changed, 23 insertions(+)
ACK
Daniel
--
|: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org -o-
On Thu, Aug 01, 2013 at 09:28:26PM +0200, Jiri Denemark wrote:
---
tests/qemuhotplugtest.c| 18 +
.../qemuhotplugtestdata/qemuhotplug-disk-scsi.xml | 7
.../qemuhotplug-hotplug-base+disk-scsi.xml | 46
++
3 files
On Thu, Aug 01, 2013 at 09:28:25PM +0200, Jiri Denemark wrote:
---
tests/qemuhotplugtest.c| 17
tests/qemuhotplugtestdata/qemuhotplug-disk-usb.xml | 7
.../qemuhotplug-hotplug-base+disk-usb.xml | 45
++
3 files
Explicitly let the user know about the unknown pool type.
---
Notes:
Version 2:
* fixed commit message
* ACK'd in v1 (would cause conflict if reordering ...)
tools/virsh-pool.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/virsh-pool.c
Use the new semantics of vshStringToArray to avoid leaking the array of
volumes to be deleted. The array would be leaked in case the first
volume was found in the domain definition. Also refactor the code a bit
to sanitize naming of variables hoding arrays and dimensions of the
arrays.
Resolves:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common dealloc function.
This patch also fixes a few forgotten checks of return from
On 08/20/2013 02:09 AM, Osier Yang wrote:
On 08/08/13 08:43, John Ferlan wrote:
Update formatsecret docs to describe the various options and provide
examples
in order to set up secrets for each type of secret.
---
docs/formatsecret.html.in | 156
Fixed after review from Osier, please see individual patches.
Peter Krempa (3):
virsh: modify vshStringToArray to duplicate the elements too
virsh-pool: Improve error message in cmdPoolList
virsh: Don't leak list of volumes when undefining domain with storage
tools/virsh-domain.c | 134
To try to stay on the end of month release schedule, I am suggesting
to enter the freeze for 1.1.2 next Tuesday, i.e. the 27th Aug, and
then shoot for a release on the morning of Monday 2nd of Sep.
So unless I hear disagreements, that's the plan :-)
Daniel
--
Daniel Veillard | Open
On 08/20/2013 05:08 AM, Osier Yang wrote:
Introduced by commit e0139e30444:
1777 /* Updating pool metadata */
(40) Event var_deref_op: Dereferencing null pointer newvol.
Also see events: [assign_zero]
1778 pool-def-allocation += newvol-allocation;
1779
On 08/20/2013 05:08 AM, Osier Yang wrote:
Introduced by commit e0139e30444. virStorageVolDefFree free'ed the
pointers that are still used by the added volume object, this changes
it back to VIR_FREE.
---
src/storage/storage_driver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Introduced by commit e0139e30444:
1777/* Updating pool metadata */
(40) Event var_deref_op: Dereferencing null pointer newvol.
Also see events: [assign_zero]
1778pool-def-allocation += newvol-allocation;
1779pool-def-available -= newvol-allocation;
---
On 08/20/2013 09:28 AM, Osier Yang wrote:
Introduced by commit e0139e30444:
1777 /* Updating pool metadata */
(40) Event var_deref_op: Dereferencing null pointer newvol.
Also see events: [assign_zero]
1778 pool-def-allocation += newvol-allocation;
1779
On 20/08/13 22:15, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common dealloc function.
This patch also
On 08/20/2013 08:15 AM, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
common dealloc function.
This patch
On 20/08/13 22:15, Peter Krempa wrote:
Use the new semantics of vshStringToArray to avoid leaking the array of
volumes to be deleted. The array would be leaked in case the first
volume was found in the domain definition. Also refactor the code a bit
to sanitize naming of variables hoding arrays
On 20/08/13 23:42, Eric Blake wrote:
On 08/20/2013 08:15 AM, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
On 08/20/13 17:42, Eric Blake wrote:
On 08/20/2013 08:15 AM, Peter Krempa wrote:
At a slightly larger memory expense allow stealing of items from the
string array returned from vshStringToArray and turn the result into a
string list compatible with virStringSplit. This will allow to use the
On 08/20/2013 09:45 AM, Peter Krempa wrote:
+++ b/tools/virsh.h
@@ -37,6 +37,7 @@
# include virerror.h
# include virthread.h
# include virnetdevbandwidth.h
+# include virstring.h
Is this change necessary?
It's to import virStringFreeList to virsh as it's used to free the
string
On 08/20/13 17:56, Eric Blake wrote:
On 08/20/2013 09:45 AM, Peter Krempa wrote:
+++ b/tools/virsh.h
@@ -37,6 +37,7 @@
# include virerror.h
# include virthread.h
# include virnetdevbandwidth.h
+# include virstring.h
Is this change necessary?
It's to import virStringFreeList to
On 08/20/2013 07:04 AM, Daniel P. Berrange wrote:
On Mon, Aug 12, 2013 at 10:19:47PM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=924153
The solution is to separate the semantics of a chain that must
not be labeled (which the user can set even on persistent domains)
On 08/16/2013 05:46 PM, Jim Fehlig wrote:
From: Dario Faggioli dario.faggi...@citrix.com
...snip...
+
+ cleanup:
+if (ret != 0) {
+for (i = 0; i nr_nodes; i++)
+VIR_FREE(cpus[i]);
+virCapabilitiesFreeNUMAInfo(caps);
+}
+
Coverity got grumpy with
On 08/20/2013 11:01 AM, John Ferlan wrote:
The following resolves Coverity's complaint and keeps things safer:
-for (i = 0; i nr_nodes; i++)
+for (i = 0; cpus i nr_nodes; i++)
Pre-approved ACK if you want to work that into a formal patch.
--
Eric Blake eblake
If we reached cleanup: prior to allocating cpus, it was possible that
'nr_nodes' had a value, but cpus was NULL leading to a possible NULL
deref. Add a 'cpus' as an end condition to for loop
---
Will push based upon pre-approved ACK:
On 08/07/2013 08:43 PM, John Ferlan wrote:
These patches address a couple of issues I ran into while doing extra
testing on the iSCSI chap authentication patches:
https://www.redhat.com/archives/libvir-list/2013-July/msg01378.html
The first 2 or 3 patches should be candidiates to be put
On 07/15/2013 04:13 AM, Guannan Ren wrote:
On 07/13/2013 01:36 AM, Michal Privoznik wrote:
On 17.05.2013 08:30, Guannan Ren wrote:
s/dictionay/dictionary/ in $SUBJ
Currently, when there is no blockjob, dom.blockJobInfo('vda')
still reports error because it didn't distinguish return value 0
by adding it to AM_LDFLAGS instead of every linking rule and
by avoiding a forked grep.
---
Daniel kind of nacked the AM_LDFLAGS part already but I think it's a
reasonable cleanup. We should rather use AM_LDFLAGS everywhere which
(we currently don't and which would be another cleanup). Or are
On 08/20/2013 08:39 AM, Guido Günther wrote:
by adding it to AM_LDFLAGS instead of every linking rule and
by avoiding a forked grep.
---
Daniel kind of nacked the AM_LDFLAGS part already but I think it's a
reasonable cleanup. We should rather use AM_LDFLAGS everywhere which
(we currently
and use it when available
---
configure.ac | 1 +
daemon/Makefile.am| 1 +
m4/virt-linker-no-indirect.m4 | 30 ++
src/Makefile.am | 1 +
tests/Makefile.am | 1 +
tools/Makefile.am | 1 +
6 files
On Mon, Aug 19, 2013 at 01:36:36PM -0600, Eric Blake wrote:
On 08/19/2013 12:07 PM, Guido Günther wrote:
+NO_INDIRECT_LDFLAGS=
+`$LD --help 21 | grep -- --no-copy-dt-needed-entries /dev/null`
\
Doesn't do what you think (it tries to execute the output of grep -
which is
Reformat the description to more cleanly delineate the attributes
for a disk element.
---
Similar to the recently changed source attribute.
docs/formatdomain.html.in | 121 +++---
1 file changed, 71 insertions(+), 50 deletions(-)
diff --git
On 08/20/2013 08:40 AM, Guido Günther wrote:
and use it when available
---
+AC_DEFUN([LIBVIRT_LINKER_NO_INDIRECT],[
+AC_MSG_CHECKING([for how to avoid indirect lib deps])
+
+NO_INDIRECT_LDFLAGS=
+`$LD --help 21 | grep -- --no-copy-dt-needed-entries /dev/null` \
+
On 08/20/2013 01:54 AM, Guannan Ren wrote:
When backing files of a disk file are stored in NFS shared folder,
qemu process requires at least the read permission.
But in some rare situation, these backing files can not even be read
If so, we should treat the diskchains as broken.
This patch
Original series + some fixes/improvements.
Tomas Meszaros (9):
virsh: C99 style for info_domfstrim and opts_lxc_enter_namespace
virsh: Add vshCmdCompleter and vshOptCompleter
virsh: Improve readline generators and readline completion
virsh: Add vshDomainCompleter
virsh: Add
Change info_domfstrim and opts_lxc_enter_namespace initialization style
to C99.
---
tools/virsh-domain.c | 21 -
1 file changed, 16 insertions(+), 5 deletions(-)
diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
index b29f934..5d4913d 100644
--- a/tools/virsh-domain.c
* vshMalloc is now used in vshDetermineCommandName
* vshStrdup instead of vshMalloc + snprintf
* joined if's in vshReadlineOptionsCompletionGenerator
---
tools/virsh.c | 395 ++
1 file changed, 373 insertions(+), 22 deletions(-)
diff --git
* global variable __my_conn renamed to vshConn
* @name is now const char *
* label cleanup renamed to error
---
tools/virsh.c | 53 +
tools/virsh.h | 2 ++
2 files changed, 55 insertions(+)
diff --git a/tools/virsh.c b/tools/virsh.c
index
vshRebootShutdownModeCompleter returns available modes
for reboot/shutdown commands.
---
tools/virsh.c | 28
tools/virsh.h | 1 +
2 files changed, 29 insertions(+)
diff --git a/tools/virsh.c b/tools/virsh.c
index 85d74ad..965b141 100644
--- a/tools/virsh.c
+++
* label cleanup renamed to error
* vshSuspendTargetCompleter added to opts_dom_pm_suspend
---
tools/virsh-domain.c | 3 ++-
tools/virsh.c| 28
tools/virsh.h| 1 +
3 files changed, 31 insertions(+), 1 deletion(-)
diff --git a/tools/virsh-domain.c
Add .completer and .completer_flags to snapshotCmds.
Provides domain completion for most of the snapshotCmds commands.
---
tools/virsh-snapshot.c | 45 -
1 file changed, 36 insertions(+), 9 deletions(-)
diff --git a/tools/virsh-snapshot.c
Add .completer and .completer_flags to domMonitoringCmds.
Provides domain completion for domMonitoringCmds commands.
---
tools/virsh-domain-monitor.c | 32 +---
1 file changed, 25 insertions(+), 7 deletions(-)
diff --git a/tools/virsh-domain-monitor.c
completer and completer_flags added to the _vshCmdDef and _vshCmdOptDef
structures so it will be possible for completion generators to
conveniently call completer functions with desired flags.
---
tools/virsh.h | 7 +++
1 file changed, 7 insertions(+)
diff --git a/tools/virsh.h
* opts_dom_pm_suspend completer initialization for target removed
(moved to another patch)
---
tools/virsh-domain.c | 230 +++
1 file changed, 179 insertions(+), 51 deletions(-)
diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
index
On 08/20/2013 02:02 PM, Tomas Meszaros wrote:
Your commit message should give a brief summary of your change and why
it is important (Add vshDomainCompleter is the what, but you didn't
spell out the why - something along the lines of future patches will
use this to improve readline completion).
On 08/20/2013 02:02 PM, Tomas Meszaros wrote:
Change info_domfstrim and opts_lxc_enter_namespace initialization style
to C99.
---
tools/virsh-domain.c | 21 -
1 file changed, 16 insertions(+), 5 deletions(-)
ACK. I wonder if it is worth adding a 'make syntax-check' test
On 08/20/2013 02:30 PM, Eric Blake wrote:
I'm not a fan of this. We worked hard to get some commands (like 'virsh
echo') to NOT need a connection, and this would be undoing that work. I
think we should NOT autoconnect UNTIL we reach the point that completion
is being attempted on something
1 - 100 of 106 matches
Mail list logo