https://bugzilla.redhat.com/show_bug.cgi?id=872292
Libvirt should not attempt to call a QMP command that has not been
documented in qemu.git - if future qemu introduces a command by the
same name but with subtly different semantics, then libvirt will be
broken when trying to use that command.
See
> > Not sure about O_DIRECT - I'm inclined to say we should just
> > *always* use O_DIRECT - unless someone can point out a downside
> > with it ?
>
> About the only possible downside I can see is that it _might_ be
> slightly slower, as it forces libvirt to take all the I/O through
> a pipe befor
https://bugzilla.redhat.com/show_bug.cgi?id=876828
Commit 38c4a9cc introduced a regression in hot unplugging of disks
from qemu, where cgroup device ACLs were no longer being revoked
(thankfully not a security hole: cgroup ACLs only prevent open()
of the disk; so reverting the ACL prevents future
On 11/29/12 21:14, Eric Blake wrote:
This patch adds two labels and gets rid of a ton of duplicated code.
This patch also fixes some error message and swtiches most of them to
s/swtiches/switches/
proper error reporting functions.
---
src/qemu/qemu_conf.c | 194
+--
On 11/29/12 20:04, Eric Blake wrote:
This patch adds macros to help retrieve configuration values from
qemu
driver's configuration. Some configuration options are grouped
together in the process.
---
src/qemu/qemu_conf.c | 303
+--
1 file chang
> So I downgraded to a previous version that I knew worked and it did
> but
> my RTR-ADVERT messages are back. The "good" version uses a tarball
> created from git on November 19th and the "bad" version has a tarball
> created from git on November 29th.
Have you ever used 'git bisect' before? It
When I rebased my DHCPv6, etc. patches, I did that not only on to of
Laine Stump's dnsmasq-capabiliotes/bind-dynamic updates but I also
rebased to "current" git master.
Well, the problem with the large number/frequent RTR-ADVERT syslog
messages from dnsmasq that I had previously seen were fixe
> This patch adds two labels and gets rid of a ton of duplicated code.
> This patch also fixes some error message and swtiches most of them to
s/swtiches/switches/
> proper error reporting functions.
> ---
> src/qemu/qemu_conf.c | 194
> +--
> 1 f
On 11/28/2012 06:23 PM, Eric Blake wrote:
> - Original Message -
>> This new function returns true if the given address is in the range
>> of
>> any "private" or "local" networks as defined in RFC1918 (IPv4) or
>> RFC3484/RFC4193 (IPv6), otherwise they return false.
>>
>> These ranges are:
On 11/28/2012 06:32 PM, Eric Blake wrote:
>> This bug resolves CVE-2012-3411, which is described in the following
>> bugzilla report:
>>
>> https://bugzilla.redhat.com/show_bug.cgi?id=833033
>>
>> The following report is specifically for libvirt on Fedora:
>>
>> https://bugzilla.redhat.com/show
On 11/28/2012 06:13 PM, Eric Blake wrote:
>> In order to optionally take advantage of new features in dnsmasq when
>> the host's version of dnsmasq supports them, but still be able to run
>> on hosts that don't support the new features, we need to be able to
>> detect the version of dnsmasq running
This patch changes how parameters are passed to dnsmasq. Instead of
being on the command line, the parameters are put into a file (one
parameter per line) and a commandline --conf-file= specifies the
location of the file. The file is located in the same directory as
the leases file.
This also ad
This patch adds the capability for virtual guests to do IPv6
communication via a virtual network interface with no IPv6
(gateway) addresses specified. This capability currently
exists for IPv4.
This patch allows creation of a completely isolated IPv6 network.
Note that virtual guests cannot comm
These three patch files are packaged together because they serially
depend on each other. These files have been rebased to "v3" of
the dnsmasq capabilities and bind-dynamic patches.
The DHCPv6 support checks dnsmasq's version and requires a minimum of 2.64.
Also, using dnsmasq for providing the
> This patch adds macros to help retrieve configuration values from
> qemu
> driver's configuration. Some configuration options are grouped
> together in the process.
> ---
> src/qemu/qemu_conf.c | 303
> +--
> 1 file changed, 73 insertions(+), 230
> Add support for doing controlled shutdown / reboot in the LXC
> driver. The default behaviour is to try talking to /dev/initctl
> inside the container's virtual root (/proc/$INITPID/root). This
> works with sysvinit or systemd. If that file does not exist
> then send SIGTERM (for shutdown) or SIG
> The virDomainShutdownFlags and virDomainReboot APIs allow the caller
> to request the operation is implemented via either acpi button press
> or a guest agent. For containers, a couple of other methods make
> sense, a message to /dev/initctl, and direct kill(SIGTERM|HUP) of
> the container init p
Since we can't (currently) rely on the ability to provide blanket
support for all possible network changes by calling the toplevel
netdev hostside disconnect/connect functions (due to qemu only
supporting a lockstep between initialization of host side and guest
side of devices), in order to support
These two patches enable making a live change to the nwfilter of a
guest's interface via virDomainUpdateDeviceFlags (virsh
update-device).
Differences from V1:
1) add patch from Stefan Berger to do a proper comparison of the
values stored in the filterparams hashtable.
2) simplify virNWFilter
From: Stefan Berger
To detect if an interface's nwfilter has changed, we need to also
compare the filterparams, which is a hashtable of virNWFilterVarValue.
virHashEqual can do this nicely, but requires a pointer to a function
that will compare two of the items being stored in the hashes.
---
sr
> To be able todo controlled shutdown/reboot of containers an
s/todo/to do/
> API to talk to init via /dev/initctl is required. Fortunately
> this is quite straightforward to implement, and is supported
> by both sysvinit and systemd. Upstart support for /dev/initctl
> is unclear.
>
> +++ b/src
> > > When the session dies or when the system is going to be shut down
> > > we issue a virStateStop() call to instruct drivers to prepare to
> > > be stopped. This will remove any previously acquire inhibitions.
> >
> > > +#ifdef HAVE_DBUS
> > > +# include
> >
> > Again, is this necessary,
>
> > > Use the freedesktop inhibition DBus service to prevent host
> > > shutdown or session logout while any VMs are running.
> > >
> > > +#ifdef HAVE_DBUS
> > > +# include
> >
> > Do we really need this header...
>
> No, its obsolete.
Well, I _did_ see some dbus_ calls in this file:
+#ifde
> > > Currently to deal with auto-shutdown libvirtd must periodically
> > > poll all stateful drivers. Thus sucks because it requires
> > > acquiring both the driver lock and locks on every single virtual
> > > machine. Instead pass in a "inhibit" callback to
> > > virStateInitialize
> > > which dr
> On Tue, Nov 27, 2012 at 03:06:45PM -0500, Eric Blake wrote:
> > How does this interact with the libvirt-guests init script? I
> > guess
> > that if that is installed, it gets to run first before the system
> > libvirtd is shutdown, so the qemu:///system won't see any running
> > guests, and thi
- Original Message -
> REVERSE_INULL, SIZEOF_MISMATCH and some uninitialized variables.
Series now pushed to libvirt.org.
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
> If allocation of names fails, list is uninitialized.
> ---
> src/conf/snapshot_conf.c |2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
ACK. And looks like this one was introduced fairly
recently, by commit 0361917.
--
libvir-list mailing list
libvir-list@redhat.com
https://www.red
- Original Message -
> Found by coverity:
> Error: SIZEOF_MISMATCH (CWE-569):
> libvirt-0.10.2/tools/virsh-domain.c:4754: suspicious_sizeof:
> Passing
> argument "8UL /* sizeof (cpumap) */" to function
> "_vshCalloc(vshControl *, size_t, size_t, char const *, int)" and
>
> On OOM, xdr_destroy got called on xdr even though it wasn't created
> yet.
>
> Found by coverity:
> Error: UNINIT (CWE-457):
> libvirt-0.10.2/src/rpc/virnetmessage.c:214: var_decl: Declaring
> variable "xdr" without initializer.
> libvirt-0.10.2/src/rpc/virnetmessage.c:219: cond_true
> Found by coverity:
> Error: REVERSE_INULL (CWE-476):
> libvirt-0.10.2/src/conf/netdev_bandwidth_conf.c:99: deref_ptr:
> Directly dereferencing pointer "node".
> libvirt-0.10.2/src/conf/netdev_bandwidth_conf.c:107:
> check_after_deref: Null-checking "node" suggests that it may be
>
> Time values were uninitialized if the connection wasn't usable.
> ---
> tools/virsh.c | 12 ++--
> 1 files changed, 6 insertions(+), 6 deletions(-)
ACK.
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
> Found by coverity:
> Error: REVERSE_INULL (CWE-476):
> libvirt-0.10.2/src/util/processinfo.c:141: deref_ptr: Directly
> dereferencing pointer "map".
> libvirt-0.10.2/src/util/processinfo.c:142: check_after_deref:
> Null-checking "map" suggests that it may be null, but it has
>
Also remove the pointless check for NULL in auth.cephx.secret.uuid,
since this is a static array.
---
src/conf/storage_conf.c |8 +++-
1 files changed, 3 insertions(+), 5 deletions(-)
diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c
index 3fdc5b6..99c2e52 100644
--- a/src/co
On Thu, Nov 29, 2012 at 12:39:24PM +, Daniel P. Berrange wrote:
> On Thu, Nov 29, 2012 at 01:29:38PM +0100, Guido Günther wrote:
> > As of 1a50ba2cb07d8bb2aa724062889deb9efd7ad9e9 we fail to connect to the
> > monitor instead of getting an exit status != 0 from qemu itself. This
> > breaks capa
On 11/28/12 15:31, Osier Yang wrote:
> On 2012年11月28日 21:34, Ján Tomko wrote:
>> Also remove the pointles check for NULL in auth.cephx.secret.uuid,
>> since this is a static array.
>
> It's nice if there is log of coverity.
Error: FORWARD_NULL (CWE-476):
libvirt-0.10.2/src/conf/storage_conf.c:447
On Wed, Nov 28, 2012 at 3:26 PM, Daniel P. Berrange wrote:
> On Wed, Nov 28, 2012 at 02:27:53PM -0500, Dusty Mabe wrote:
>> On Wed, Nov 14, 2012 at 11:22 AM, Dusty Mabe wrote:
>> > On Thu, Nov 1, 2012 at 11:32 PM, Dusty Mabe wrote:
> Sorry for not replying before. I've been thinking about this
Time values were uninitialized if the connection wasn't usable.
---
tools/virsh.c | 12 ++--
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/tools/virsh.c b/tools/virsh.c
index 6372177..dea3f82 100644
--- a/tools/virsh.c
+++ b/tools/virsh.c
@@ -1562,20 +1562,20 @@ vshComma
If allocation of names fails, list is uninitialized.
---
src/conf/snapshot_conf.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/src/conf/snapshot_conf.c b/src/conf/snapshot_conf.c
index 06be34d..bba1bb7 100644
--- a/src/conf/snapshot_conf.c
+++ b/src/conf/snapshot_conf.
On OOM, xdr_destroy got called on xdr even though it wasn't created yet.
Found by coverity:
Error: UNINIT (CWE-457):
libvirt-0.10.2/src/rpc/virnetmessage.c:214: var_decl: Declaring
variable "xdr" without initializer.
libvirt-0.10.2/src/rpc/virnetmessage.c:219: cond_true: Condition
Found by coverity:
Error: SIZEOF_MISMATCH (CWE-569):
libvirt-0.10.2/tools/virsh-domain.c:4754: suspicious_sizeof: Passing
argument "8UL /* sizeof (cpumap) */" to function
"_vshCalloc(vshControl *, size_t, size_t, char const *, int)" and
then casting the return value to "unsigned cha
Found by coverity:
Error: REVERSE_INULL (CWE-476):
libvirt-0.10.2/src/util/processinfo.c:141: deref_ptr: Directly
dereferencing pointer "map".
libvirt-0.10.2/src/util/processinfo.c:142: check_after_deref:
Null-checking "map" suggests that it may be null, but it has already
been
Found by coverity:
Error: REVERSE_INULL (CWE-476):
libvirt-0.10.2/src/conf/netdev_bandwidth_conf.c:99: deref_ptr:
Directly dereferencing pointer "node".
libvirt-0.10.2/src/conf/netdev_bandwidth_conf.c:107:
check_after_deref: Null-checking "node" suggests that it may be
null, but
REVERSE_INULL, SIZEOF_MISMATCH and some uninitialized variables.
Ján Tomko (6):
conf: fix NULL check in virNetDevBandwidthParse
util: fix virBitmap allocation in virProcessInfoGetAffinity
virsh: use correct sizeof when allocating cpumap
virsh: do timing even for unusable connections
rpc:
> Implement the new API for sending signals to processes in a guest
> for the LXC driver. Only support sending signals to the init
> process for now, because
>
> - The kernel does not appear to expose the mapping between
>container PID numbers and host PID numbers anywhere in the
>host OS
> > Drop this hunk (rather, move it to patch 4/4). Such a check
> > should be done in the hypervisor driver, not in the entry
> > point. Otherwise, if we add a new signal later, this hunk
> > would prevent an older virsh from sending the new signal.
>
> On the other hand, you _should_ add a chec
> > Add an API for sending signals to arbitrary processes in the
> > guest OS. This is primarily useful for container based virt,
> > but can be used for machine virt too, if there is a suitable
> > guest agent,
> >
> > +
> > +if (signum > VIR_DOMAIN_PROCESS_SIGNAL_LAST) {
> > +virLib
> * tools/virsh.c: Add send-process-signal
> * tools/virsh.pod: Document new command
>
> Signed-off-by: Daniel P. Berrange
> ---
> tools/virsh-domain.c | 103
> +++
> tools/virsh.pod | 27 ++
> 2 files changed, 130 insertions(+)
> * src/remote/remote_protocol.x: message definition
> * src/remote/remote_driver.c: Register driver function
> * src/remote_protocol-structs: Test case
>
> Signed-off-by: Daniel P. Berrange
> ---
> src/remote/remote_driver.c | 1 +
> src/remote/remote_protocol.x | 10 +-
> src/remote
> Add an API for sending signals to arbitrary processes in the
> guest OS. This is primarily useful for container based virt,
> but can be used for machine virt too, if there is a suitable
> guest agent,
>
> * include/libvirt/libvirt.h.in: Add virDomainSendProcessSignal
> and virDomainProcessSig
On 11/28/2012 05:55 PM, Laine Stump wrote:
> On 11/28/2012 04:19 AM, Martin Kletzander wrote:
>> On 11/28/2012 06:08 AM, Laine Stump wrote:
>>> This resolves: https://bugzilla.redhat.com/show_bug.cgi?id=879473
>>>
>>> The name attribute is required for portgroup elements (yes, the RNG
>>> specifies
From: "Daniel P. Berrange"
* tools/virsh.c: Add send-process-signal
* tools/virsh.pod: Document new command
Signed-off-by: Daniel P. Berrange
---
tools/virsh-domain.c | 103 +++
tools/virsh.pod | 27 ++
2 files changed, 130 inse
From: "Daniel P. Berrange"
* src/remote/remote_protocol.x: message definition
* src/remote/remote_driver.c: Register driver function
* src/remote_protocol-structs: Test case
Signed-off-by: Daniel P. Berrange
---
src/remote/remote_driver.c | 1 +
src/remote/remote_protocol.x | 10 +-
From: "Daniel P. Berrange"
Add an API for sending signals to arbitrary processes in the
guest OS. This is primarily useful for container based virt,
but can be used for machine virt too, if there is a suitable
guest agent,
* include/libvirt/libvirt.h.in: Add virDomainSendProcessSignal
and virD
---
Pushed as trivial.
src/qemu/qemu_agent.c | 2 +-
src/qemu/qemu_monitor.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/qemu/qemu_agent.c b/src/qemu/qemu_agent.c
index 7062d53..893f7f2 100644
--- a/src/qemu/qemu_agent.c
+++ b/src/qemu/qemu_agent.c
@@ -242,7 +242,
From: "Daniel P. Berrange"
Implement the new API for sending signals to processes in a guest
for the LXC driver. Only support sending signals to the init
process for now, because
- The kernel does not appear to expose the mapping between
container PID numbers and host PID numbers anywhere in
A follow up to
https://www.redhat.com/archives/libvir-list/2012-November/msg01283.html
Changed in this version
- Switch pid to long long to allow process groups & 64-bit pids
- Add explicit enum entries for RT signals
- Do case insensitive parsing in virsh
- Fix version numbers
- Improve
On Thu, Nov 29, 2012 at 01:29:38PM +0100, Guido Günther wrote:
> As of 1a50ba2cb07d8bb2aa724062889deb9efd7ad9e9 we fail to connect to the
> monitor instead of getting an exit status != 0 from qemu itself. This
> breaks capabilities probing for the non QMP case.
> ---
> src/qemu/qemu_capabilities.c
As of 1a50ba2cb07d8bb2aa724062889deb9efd7ad9e9 we fail to connect to the
monitor instead of getting an exit status != 0 from qemu itself. This
breaks capabilities probing for the non QMP case.
---
src/qemu/qemu_capabilities.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git
This patch adds two labels and gets rid of a ton of duplicated code.
This patch also fixes some error message and swtiches most of them to
proper error reporting functions.
---
src/qemu/qemu_conf.c | 194 +--
1 file changed, 78 insertions(+), 116 del
Done while hacking on other series (that is not yet ready).
Peter Krempa (2):
qemu: Refactor config parameter retrieval
qemu: Refactor error reporting in qemu driver configuration parser
src/qemu/qemu_conf.c | 479 +++
1 file changed, 142 inser
This patch adds macros to help retrieve configuration values from qemu
driver's configuration. Some configuration options are grouped
together in the process.
---
src/qemu/qemu_conf.c | 303 +--
1 file changed, 73 insertions(+), 230 deletions(-)
dif
On Tue, Nov 27, 2012 at 12:44:43PM -0500, Eric Blake wrote:
> - Original Message -
> > The following is a list of random bug fixes I've identified
> > while working on LXC
>
> ACK 1 through 8, and 10. Nit in the commit message for 8/10. See
> my proposal for an alternate fix for 9.
This
On 11/29/2012 07:07 PM, Wayne Sun wrote:
v2: break down the case to small cases with separate flags
* Use setVcpusFlags API to set domain vcpus with flags
* 3 cases added, each only deal with one set flag value as in
config, live or maximum
* cases are independent on domain states, API will r
v2: break down the case to small cases with separate flags
* Use setVcpusFlags API to set domain vcpus with flags
* 3 cases added, each only deal with one set flag value as in
config, live or maximum
* cases are independent on domain states, API will report error
if not suitable for certain st
On 11/28/2012 03:58 PM, Daniel P. Berrange wrote:
Viktor Mihajlovski (3):
qemu: Wait for monitor socket even without pid
qemu: Fix QMP Capabability Probing Failure
qemu: Add QEMU version computation to QMP probing
I pushed this series now.
Daniel
Thx
--
Mit freundlichen Grüßen/
On 11/28/2012 08:10 PM, Wayne Sun wrote:
v2: break down the case to small cases with separate flags
* Use setVcpusFlags API to set domain vcpus with flags
* 3 cases added, each only deal with one set flag value as in
config, live or maximum
* cases are independent on domain states, API will r
On 29.11.2012 00:41, Eric Blake wrote:
>> This will call FITRIM within guest. The API has 4 arguments,
>> however, only 2 will be used for now (@dom and @minumum).
>> The rest two are there if in future qemu guest agent learns them.
>
> s/in future/in the future/
>
>
>> +/**
>> + * virDomainFSTr
67 matches
Mail list logo
