On 12/02/2015 04:15 AM, Daniel P. Berrange wrote:
> Yep, we need to agree which min platform we're targetting. It is nice
> to see RHEL7 rebasing GTK, as that makes life easier. We should not
> only consider RHEL though. It would be desirable if someone can provide
> a list of GTK versions in curre
Eduardo Lima (Etrunko) wrote:
> On 12/02/2015 11:25 PM, Jim Fehlig wrote:
>> On 12/02/2015 04:15 AM, Daniel P. Berrange wrote:
>>> Yep, we need to agree which min platform we're targetting. It is nice
>>> to see RHEL7 rebasing GTK, as that makes life easier. We s
FYI, I added the attached patch to SUSE's virt-manager package. I thought it one
of those distro-specific patches to maintain forever, but in the end decided it
was worth a shot to upstream since the existing regex includes "gerd's firmware
repo" :-).
Regards,
Jim
Support paths to SUSE OVMF firmwa
Extend the domcapabilities regex to include SUSE's OVMF
file naming convention.
Signed-off-by: Jim Fehlig
---
virtinst/domcapabilities.py | 1 +
1 file changed, 1 insertion(+)
diff --git a/virtinst/domcapabilities.py b/virtinst/domcapabilities.py
index 86bdfe8..f074b4e 100644
--- a/vir
On 01/27/2016 05:37 PM, Jim Fehlig wrote:
> FYI, I added the attached patch to SUSE's virt-manager package. I thought it
> one
> of those distro-specific patches to maintain forever, but in the end decided
> it
> was worth a shot to upstream since the existing regex includes
Cole Robinson wrote:
> On 01/27/2016 09:51 PM, Jim Fehlig wrote:
>> Extend the domcapabilities regex to include SUSE's OVMF
>> file naming convention.
>>
>> Signed-off-by: Jim Fehlig
>> ---
>> virtinst/domcapabilities.py | 1 +
>> 1 file c
Richard W.M. Jones wrote:
> On Thu, Jan 28, 2016 at 11:27:14AM -0500, Cole Robinson wrote:
>> On 01/27/2016 09:51 PM, Jim Fehlig wrote:
>>> Extend the domcapabilities regex to include SUSE's OVMF
>>> file naming convention.
>>>
>>>
On 02/04/2016 10:48 AM, Richard W.M. Jones wrote:
> On Thu, Jan 28, 2016 at 02:47:24PM -0700, Jim Fehlig wrote:
>> Richard W.M. Jones wrote:
>>> What's the full path to the OVMF files on SUSE?
>> /usr/share/qemu/ovmf-x86_64-{code,vars}.bin is the path to the "p
Cole Robinson wrote:
> On 03/01/2016 11:21 AM, Clair Grant wrote:
>> Virt-install on kvm provides boot loader options for UEFI rather than BIOS.
>> There does not seem to be a similar option in virt-install on xen. Is there
>> another means for being UEFI-based on xen?
>>
>
> Personally I have n
Cole Robinson wrote:
> On 03/01/2016 01:18 PM, Jim Fehlig wrote:
>> Cole Robinson wrote:
>>> On 03/01/2016 11:21 AM, Clair Grant wrote:
>>>> Virt-install on kvm provides boot loader options for UEFI rather than
>>>> BIOS. There does not seem to be a simil
Hi All,
I had once again nearly forgot about vhostmd whenI received a mail from
Michael-John Turner asking about packaging it for Debian. He was looking for the
canonical source, which was on the now defunct gitorious. Sadly, I never moved
the code togithub, gitlab, other hosting service that may
Cole Robinson wrote:
> On 03/09/2016 05:58 PM, Jim Fehlig wrote:
>> Hi All,
>>
>> I had once again nearly forgot about vhostmd whenI received a mail from
>> Michael-John Turner asking about packaging it for Debian. He was looking for
>> the
>> canonic
On 06/15/2016 06:42 AM, Pavel Grunt wrote:
> Hi,
>
> On Tue, 2016-06-14 at 23:40 +0200, Fabiano Fidêncio wrote:
>> These small series has as objective start "preparing the ground" for
>> the work Sagar Huge has been done on re-design/re-write virt-viewer's
>> UI in a way to have it looking as somet
On 06/15/2016 12:56 PM, Charles Arnold wrote:
On 6/15/2016 at 12:13 PM, Cole Robinson wrote:
>> On 06/15/2016 11:40 AM, Charles Arnold wrote:
>> On 6/15/2016 at 09:21 AM, Pavel Hrdina wrote:
On Wed, Jun 15, 2016 at 08:31:42AM -0600, Charles Arnold wrote:
> Tell libvirt what ma
t/capabilities.py| 7 ++-
> 5 files changed, 17 insertions(+), 6 deletions(-)
FWIW, I tested these patches on one of my Xen systems.
Tested-by: Jim Fehlig
Regards,
Jim
___
virt-tools-list mailing list
virt-tools-list@redhat.com
https://www.redhat.com/mailman/listinfo/virt-tools-list
g so.
Regards,
Jim
>From 5314d51a91f57542d97f04aa589e79d13352ce22 Mon Sep 17 00:00:00 2001
From: Jim Fehlig
Date: Wed, 31 Aug 2016 11:26:24 -0600
Subject: [PATCH 1/2] virt-manager: use virDomainMigrate3 API
libvirt has supported the migration V3 protocol for many years now.
A nice feature of
On 09/02/2016 04:38 PM, Cole Robinson wrote:
> On 08/31/2016 02:04 PM, Jim Fehlig wrote:
>> Hi All,
>>
>> I received a bug report (apologies for no link, but it's not publicly
>> accessible) about migration failing with virt-manager between Xen hosts. It's
>
pty scheme. It's not clear what the 'xenmigr' scheme
is used for. 'xenmigr' is not referenced by any files in the Xen
code-base, including old branches with xend.
Drop setting scheme to 'xenmigr' when creating the Xen migration URI.
Signed-off-by: Jim Fehlig
---
Lin Ma wrote:
>
>
Cole Robinson 2017/3/18 星期六 上午 1:10 >>>
>>On 03/17/2017 04:02 AM, Lin Ma wrote:
>>> Create a network with a device pool containing all the VFs of an SR-IOV
>>> device.
>>>
>>> Signed-off-by: Lin Ma
>>> ---
>>> ui/createnet.ui | 92 ++
On a host system with keyboard configured to en-US, it was noticed
that virt-install created install XML with keymap='de'. The host
system did not have /etc/vconsole.conf, so /etc/sysconfig/keyboard
was the next file to check, which contained the following
# Path:Hardware/Keyboard
## Descr
On 05/22/2018 05:14 AM, Trapp, Michael wrote:
On 17.05.18, 14:23, "Daniel P. Berrangé" wrote:
On Thu, May 17, 2018 at 12:13:58PM +, Trapp, Michael wrote:
Hi
I would like to add virtio based communication to vhostmd.
The current vhostmd implementation writes the metric data of all VMs and
Some of the constructs used in the configure script and Makefiles
are a bit outdated, e.g. use of '$(mkinstalldirs)' instead of
'$(MKDIR_P)'. Change these old constructs to their modern counterparts.
Signed-off-by: Jim Fehlig
---
Makefile.am | 15 +++
configure
vhostmd is long overdue for a systemd service file. Add one, and
while at it support specifying the type of init system in the
configure script.
Signed-off-by: Jim Fehlig
---
Makefile.am | 36 +---
configure.ac| 24
Patch 1 is a bit of cleanup, patch 2 provides the service file.
Jim Fehlig (2):
Modernize build files
Add systemd service file
Makefile.am | 45 +
configure.ac| 26 --
vhostmd.service | 14 ++
3 files
support for xenstore as a transport since I know of a
few users of that feature. I've been pinged about its current broken state
that is fixed by patch 3.
Jim Fehlig (3):
Remove xenctrl support
Remove support for libxenstat
Fix compilation with xenstore
configure.ac
The xenctrl interface has changed over the years, causing
the associated code in vhostmd to no longer compile. Instead
of trying to get this low-level xen interface working again,
simply remove it from vhostmd. libvirt or custom "actions"
should be used to gather metrics.
Signed-o
The xenstore transport is worth keeping, so update it to
work with "modern" Xen. By modern I mean something like 4.4
or newer.
Signed-off-by: Jim Fehlig
---
configure.ac | 2 +-
libmetrics/libmetrics.c | 2 +-
vhostmd/xenstore-update.c | 2 +-
3 files changed, 3 insert
Old, crufty code that no longer compiles and should be
removed from the sources. libvirt or custom "actions"
should be used to gather metrics.
Signed-off-by: Jim Fehlig
---
configure.ac | 12 -
include/metric.h | 4 --
vhostmd/Makefile.am | 5 --
vhostmd
On 06/07/2018 09:42 AM, Trapp, Michael wrote:
I fully understand the requirements regarding coding style and the way the changes should be split and discussed in detail,
but may I ask you to provide some general feedback about the basic concept and
if this an option at all before we start di
On 06/07/2018 10:00 AM, Richard W.M. Jones wrote:
On Wed, Jun 06, 2018 at 01:46:53PM -0600, Jim Fehlig wrote:
Patch 1 is a bit of cleanup, patch 2 provides the service file.
This all looks good, ACK.
Thanks. I've pushed this and the other series.
Now I guess the question is whethe
On 06/08/2018 02:30 AM, Daniel P. Berrangé wrote:
On Thu, Jun 07, 2018 at 04:23:32PM -0600, Jim Fehlig wrote:
On 06/07/2018 09:42 AM, Trapp, Michael wrote:
I fully understand the requirements regarding coding style and the way the
changes should be split and discussed in detail,
but may I ask
On 06/08/2018 05:58 AM, Trapp, Michael wrote:
On 08.06.18, 10:31, "Daniel P. Berrangé" wrote:
On Thu, Jun 07, 2018 at 04:23:32PM -0600, Jim Fehlig wrote:
> On 06/07/2018 09:42 AM, Trapp, Michael wrote:
> > I fully understand the requirements regarding codin
cript=check'.
Signed-off-by: Jim Fehlig
---
configure.ac | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/configure.ac b/configure.ac
index c6b660e..9414440 100644
--- a/configure.ac
+++ b/configure.ac
@@ -59,10 +59,12 @@ AC_ARG_WITH([init_script],
[with_init_s
Side note: The 'v2' should be in the subject prefix, not the subject. E.g. 'git
send-email --subject-prefix="vhostmd PATCH V2" ...'.
On 06/19/2018 07:33 AM, Michael Trapp wrote:
vhostmd has no signal handler for SIGPIPE and a restart of libvirtd results in a
stopped vhostmd. The root cause seem
Although the upstream vhostmd config file serves primarily as an
example, it should at least work and not unconditionally call
commands that no longer exist.
While at it, improve the comments describing the element.
Signed-off-by: Jim Fehlig
---
vhostmd.xml | 95
Fix the short option for help, add help to the help output, and
add help to the vhostmd man page.
Signed-off-by: Jim Fehlig
---
docs/man/vhostmd.8 | 3 +++
vhostmd/vhostmd.c | 7 ---
2 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/docs/man/vhostmd.8 b/docs/man/vhostmd.8
On 06/21/2018 07:03 AM, Michael Trapp wrote:
vhostmd has no signal handler for SIGPIPE and a restart of libvirtd results in
a stopped vhostmd. The root cause seems to be a UDS socket between vhostmd and
libvirtd which is closed by a libvirtd restart.
In addition to the signal handler the connecti
I added a few words to the commit message.
On 06/21/2018 07:19 AM, Michael Trapp wrote:
---
vhostmd/virt-util.c | 8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/vhostmd/virt-util.c b/vhostmd/virt-util.c
index 1c31305..b2b9670 100644
--- a/vhostmd/virt-util.c
+++ b/v
On 07/06/2018 07:02 AM, Richard W.M. Jones wrote:
On Wed, Jun 20, 2018 at 04:40:58PM -0600, Jim Fehlig wrote:
Fix the short option for help, add help to the help output, and
add help to the vhostmd man page.
Signed-off-by: Jim Fehlig
This one's gone upstream already I think, but ACK a
On 07/06/2018 07:01 AM, Richard W.M. Jones wrote:
On Wed, Jun 20, 2018 at 04:04:45PM -0600, Jim Fehlig wrote:
Although the upstream vhostmd config file serves primarily as an
example, it should at least work and not unconditionally call
commands that no longer exist.
While at it, improve the
insertions(+), 1 deletion(-)
Reviewed-by: Jim Fehlig
Regards,
Jim
diff --git a/include/util.h b/include/util.h
index 3a21e09..c0bd19a 100644
--- a/include/util.h
+++ b/include/util.h
@@ -42,6 +42,7 @@ typedef enum {
typedef struct _vu_buffer {
unsigned int size;
unsign
On 8/30/18 4:11 AM, Michael Trapp wrote:
At the vhostmd side virtio channels are Unix domain sockets from QEMU
which are created during a VM start and removed when the VM is stopped.
Basically this implementation
- monitors a directory for new virtio channels (channel names contain VM UUID)
- fo
On 8/30/18 4:11 AM, Michael Trapp wrote:
Activate the virtio code and extend DTD/XML to support a virtio section
for virtio related configuration and a new transport type virtio.
---
README | 56 +--
vhostmd.changes | 5 +
On 8/30/18 4:11 AM, Michael Trapp wrote:
virtio functions are available in libmetrics
and in libserialclient (with minimal dependencies).
---
We would like to have the serialclient code available in a library without
additional dependencies, because linking against libmetrics requires libxml2.
On 10/1/18 1:27 AM, Trapp, Michael wrote:
Was on SLES11 with gcc-4.3 and actually the defaults from the project, I will
move to a SLES15 and fix the compile flags and the resulting errors.
Can you please share your GNU indent settings that I can fix the
whitespace/formatting issue.
I haven't
On 10/2/18 9:27 AM, Cole Robinson wrote:
On 10/02/2018 10:53 AM, Jim Fehlig wrote:
On 10/2/18 5:13 AM, Pavel Hrdina wrote:
On Mon, Oct 01, 2018 at 02:28:09PM -0400, Cole Robinson wrote:
On 09/28/2018 12:54 AM, Jim Fehlig wrote:
I've attempted to use virt-manager to create a new VM that
On 10/5/18 3:27 PM, Cole Robinson wrote:
On 10/02/2018 01:09 PM, Jim Fehlig wrote:
On 10/2/18 9:27 AM, Cole Robinson wrote:
On 10/02/2018 10:53 AM, Jim Fehlig wrote:
On 10/2/18 5:13 AM, Pavel Hrdina wrote:
On Mon, Oct 01, 2018 at 02:28:09PM -0400, Cole Robinson wrote:
On 09/28/2018 12:54 AM
uses the method to avoid actions like
setting unix file permissions on the volume, which doesn't make
sense for a network-based volume.
Signed-off-by: Jim Fehlig
---
virtinst/devices/disk.py | 4
virtinst/diskbackend.py | 18 +-
2 files changed, 21 insertions(+),
On 10/12/18 9:57 AM, Cole Robinson wrote:
On 10/09/2018 04:30 PM, Jim Fehlig wrote:
When creating a new VM and selecting a volume from a network-based
storage pool such as rbd, the volume is not recognized as network-based
and is treated as a volume from a directory storage pool.
This patch
On 10/5/18 8:34 AM, Michael Trapp wrote:
At the vhostmd side virtio channels are Unix domain sockets from QEMU
which are created during a VM start and removed when the VM is stopped.
Basically this implementation
- monitors a directory for new virtio channels (channel names contain VM UUID)
- fo
On 10/5/18 8:34 AM, Michael Trapp wrote:
Activate the virtio code and extend DTD/XML to support a virtio section
for virtio related configuration and a new transport type virtio.
---
README | 56 -
vhostmd.changes | 5
vhostmd.dtd
cated memory is initialized there
should be a memset.
---
include/util.h | 1 +
vhostmd/util.c | 4 +++-
2 files changed, 4 insertions(+), 1 deletion(-)
Reviewed-by: Jim Fehlig
Regards,
Jim
diff --git a/include/util.h b/include/util.h
index 3a21e09..c0bd19a 100644
--- a/include/util.h
al seconds.
Fix the delay between updates, that
'runtime of metrics update' + 'sleep time' matches the update_interval value.
---
vhostmd/vhostmd.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
Reviewed-by: Jim Fehlig
Regards,
Jim
diff --git a/vhostmd
On 10/5/18 8:34 AM, Michael Trapp wrote:
virtio functions are available in libmetrics
and in libserialclient (with minimal dependencies).
---
We would like to have the serialclient code available in a library without
additional dependencies, because linking against libmetrics requires libxml2.
The formatting in the usage function of vm-dump-metrics used a
lot of spaces for whitespace. Replace some occurrences with tabs
and remove others that were causing ugly formatting of help.
Signed-off-by: Jim Fehlig
---
vm-dump-metrics/main.c | 24 +---
1 file changed, 9
The use of WITH_XENSTORE in vm-dump-metrics is clumbsy in some
cases and flat out wrong in others. E.g. if WITH_XENSTORE is not
defined then vbd transport is excluded as an option for retrieving
the metrics. Improve use of WITH_XENSTORE and fix the broken logic.
Signed-off-by: Jim Fehlig
---
vm
A few fixes for vm-dump-metrics helper program. See patches for details
Jim Fehlig (2):
vm-dump-metrics: Fix formating or usage function
vm-dump-metrics: Fix WITH_XENSTORE logic
vm-dump-metrics/main.c | 56 +++---
1 file changed, 25 insertions(+), 31
On 10/30/18 3:41 AM, Trapp, Michael wrote:
On 24.10.18, 00:09, Jim Fehlig wrote:
The use of WITH_XENSTORE in vm-dump-metrics is clumbsy in some
cases and flat out wrong in others. E.g. if WITH_XENSTORE is not
defined then vbd transport is excluded as an option for retrieving
On 11/12/18 8:13 AM, Michael Trapp wrote:
Activate the virtio code and extend DTD/XML to support a virtio section
for virtio related configuration and a new transport type virtio.
---
README | 84 +
vhostmd.changes | 5 +++
vhostmd
On 11/12/18 8:12 AM, Michael Trapp wrote:
At the vhostmd side virtio channels are Unix domain sockets from QEMU
which are created during a VM start and removed when the VM is stopped.
Basically this implementation
- monitors a directory for new virtio channels (channel names contain VM UUID)
- f
On 11/15/18 5:13 AM, Trapp, Michael wrote:
On 15.11.18, 00:35, "Jim Fehlig" wrote:
On 11/12/18 8:12 AM, Michael Trapp wrote:
> At the vhostmd side virtio channels are Unix domain sockets from QEMU
> which are created during a VM start and removed when t
On 11/27/18 7:07 AM, Michael Trapp wrote:
virtio functions are available in libmetrics
and in libserialclient (with minimal dependencies).
---
We would like to have the serialclient code available in a library without
additional dependencies, because linking against libmetrics requires libxml2.
On 11/27/18 7:06 AM, Michael Trapp wrote:
Activate the virtio code and extend DTD/XML to support a virtio section
for virtio related configuration and a new transport type virtio.
---
README | 78 +
vhostmd.changes | 5 +++
vhostmd
On 11/27/18 7:06 AM, Michael Trapp wrote:
Add virtio serial channels as transport method for VM and host metrics.
A serial port at the VM side provides access to host metrics and
metrics of the reading VM, access to metrics of other VMs is not supported.
Virtio transport can be activated in the
vhostmd provides a "metrics communication channel" between a host and its hosted
virtual machines, allowing limited introspection of host resource usage from
within virtual machines.
It's been nearly 7 years since Rich Jones bumped the version to 0.5, so it is
time for a new 1.0 release! I tag
On 11/27/18 7:07 AM, Michael Trapp wrote:
virtio functions are available in libmetrics
and in libserialclient (with minimal dependencies).
---
We would like to have the serialclient code available in a library without
additional dependencies, because linking against libmetrics requires libxml2.
serialclient.h | 30 --
5 files changed, 133 insertions(+), 222 deletions(-)
delete mode 100644 libmetrics/libserialclient.c
delete mode 100644 libmetrics/libserialclient.h
As far as this patch goes:
Reviewed-by: Jim Fehlig
Regards,
Jim
diff --git a/libmetrics/Makefile.am b/li
On 12/7/18 11:12 AM, Michael Trapp wrote:
---
libmetrics/libmetrics.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libmetrics/libmetrics.c b/libmetrics/libmetrics.c
index 0a5e3ee..36c7f5f 100644
--- a/libmetrics/libmetrics.c
+++ b/libmetrics/libmetrics.c
@@ -681,8 +6
On 1/28/19 8:38 AM, Michael Trapp wrote:
We can't rely on d_type == DT_DIR because on some filesystems
d_type is not supported. A check for domain-ID-NAME/org.github.vhostmd.1
should be sufficient.
From the readdir(3) man page it looks like xfs is one of the unsupported
filesystems.
---
v
No need for the period in subject.
On 2/7/19 8:41 AM, Michael Trapp wrote:
The metrics disk can be opened with the first get_metrics() request.
---
This implements option 1 from the previously discussed solutions for
the change of the libmetrics constructor.
The virtio transport does not fill t
Same comment about period at end of subject. Also, I don't think the error
messages are broken, they are just being changed to include the function name. I
adjusted the subject and will push this patch.
Regards,
Jim
On 2/7/19 8:41 AM, Michael Trapp wrote:
---
libmetrics/libmetrics.c | 14 ++
On 2/7/19 8:41 AM, Michael Trapp wrote:
A timeout in the send request part of dump_virtio_metrics()
should prevent the send from looping forever.
Did you experience that, or just being defensive?
The timeout value is fixed to a reasonable value of 5 seconds.
I agree that 30 seconds is a bit
libvirt commit 09eb1ae0 added support for a new 'xenbus' controller
type. Add support for the controller in virtinst, including support
for the maxGrantFrames attribute.
Signed-off-by: Jim Fehlig
---
tests/cli-test-xml/compare/virt-install-many-devices.xml | 1 +
tests/cli-test-xml/co
vhostmd provides a "metrics communication channel" between a host and its hosted
virtual machines, allowing limited introspection of host resource usage from
within virtual machines.
After collecting a some bug fixes following the 1.0 release it is time for a 1.1
release. Thanks to everyone wh
requirement for virtio transport in the schema. With the
introduction of multiple transports perhaps the others shoud be optional
as well, but requiring virtio is clearly a regression.
Signed-off-by: Jim Fehlig
---
vhostmd.dtd | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a
On 11/25/19 11:26 AM, Michael Trapp wrote:
> From: d032747
You'll need to fix this or remove it. In addition, please provide an example of
the compiler warning in the commit message. Without this patch I'm not seeing
any warnings with '-Wall -Werror'.
>
> ---
> vhostmd/virtio.c | 24 +++
On 11/25/19 11:32 AM, Michael Trapp wrote:
> From: d032747
>
> Mark both global sections, disk and virtio, as optional DTD elements.
> Based on that, the configured transport can be either 'disk only' or
> 'virtio only' or 'disk and virtio'.
> But a configuration with disabled disk and disabled v
On 11/25/19 11:36 AM, Michael Trapp wrote:
> From: d032747
>
>>From a security perspective virtio should be the preferred transport
> and therefore the default transport order is virtio, disk[, xenstore]
This makes it sound as though we are changing the default transport order, when
in fact we
>From coverity scan
vhostmd-1.1/vhostmd/util.c:415:14: warning: Potential leak of memory pointed to
by 'cp'
return(NULL);
Signed-off-by: Jim Fehlig
---
vhostmd/util.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/vhostmd/util.c b/vhostmd/util.c
, NULL,
553|-> XML_PARSE_NOENT | XML_PARSE_NONET |
554|XML_PARSE_NOWARNING);
555| if (!xml) {
Signed-off-by: Jim Fehlig
---
vhostmd/vhostmd.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/vhostmd/vhostmd.c b/vhostmd/vhostmd.c
index 3d1d53e
>From coverity scan
vhostmd-1.1/libmetrics/libmetrics.c:817: invalid_type: Argument "5L" to format
specifier "%u" was expected to have type "unsigned int" but has type "long".
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 4 ++--
1 file cha
and only setting it to success after the function
bodies have successfully executed.
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 16
1 file changed, 4 insertions(+), 12 deletions(-)
diff --git a/libmetrics/libmetrics.c b/libmetrics/libmetrics.c
index a5582e7..49b38ea 10
_end: va_end was not called for
"argptr".
224| buf->use += count;
225| buf->content[buf->use] = '\0';
226|-> }
227|
228| /*
Signed-off-by: Jim Fehlig
---
vhostmd/util.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/vh
read_mdisk(mdisk);
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libmetrics/libmetrics.c b/libmetrics/libmetrics.c
index 2819f80..756645c 100644
--- a/libmetrics/libmetrics.c
+++ b/libmetrics/libmetrics.c
@@ -601,7 +601,7 @
called again. Protect agains these potential double-frees by setting
the mdisk contents NULL after freeing them.
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 24
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git a/libmetrics/libmetrics.c b/libm
sk->pctxt, mdisk->buffer,
411| mdisk->length, "mdisk.xml", NULL,
412|->XML_PARSE_NOENT | XML_PARSE_NONET |
413| XML_PARSE_NOWARNING);
414| if (!mdisk->doc) {
It should be safe to remove the option.
Signed-off-by: Jim Fehlig
uot;//metrics/metric[name='%s'][@context='%s']", pmdef->name,
pmdef->context);
^~~~~~~~~~~
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 15 +++
1 file change
d-1.1/libmetrics/libmetrics.c:900: noescape: Resource "fp" is not freed
or pointed-to in "fwrite".
vhostmd-1.1/libmetrics/libmetrics.c:909: leaked_storage: Variable "fp" going
out of scope leaks the storage it points to.
907| free(response);
908|
s/transport", ctxt)".
vhostmd-1.1/vhostmd/vhostmd.c:474: leaked_storage: Variable "obj" going out of
scope leaks the storage it points to.
472| #else
473| vu_log (VHOSTMD_ERR, "No support for xenstore transport in
this vhostmd");
474|->
>From coverity scan
vhostmd-1.1/libmetrics/libmetrics.c:613:18: warning: Potential leak of memory
pointed to by 'pmdef.context'
errno = ENOMEM;
Signed-off-by: Jim Fehlig
---
libmetrics/libmetrics.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git
|->close(fd);
846| return -1;
847| }
Signed-off-by: Jim Fehlig
---
vhostmd/vhostmd.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/vhostmd/vhostmd.c b/vhostmd/vhostmd.c
index 1600a87..1395bc5 100644
--- a/vhostmd/vhostmd.c
+++ b/vhostmd/vhostmd.c
@@ -
>From coverity scan
vhostmd-1.1/vhostmd/util.c:421: returned_null: "strstr" returns "NULL" (checked
6 out of 7 times).
Signed-off-by: Jim Fehlig
---
vhostmd/util.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/vhostmd/util.c b/vhost
lared with attribute warn_unused_result [-Wunused-result]
ftruncate(fd, mdisk_size);
^
819|
820| /* truncate to a possible new size */
821|->ftruncate(fd, mdisk_size);
822|
823| /* zero fill metrics data */
Signed-off-by: Jim Fehlig
---
quot;, *dest, str);
Signed-off-by: Jim Fehlig
---
vhostmd/util.c| 3 ++-
vhostmd/vhostmd.c | 3 ++-
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/vhostmd/util.c b/vhostmd/util.c
index 317dbc6..d7ce3fc 100644
--- a/vhostmd/util.c
+++ b/vhostmd/util.c
@@ -488,7 +488,8 @
e destination string unterminated.
192| address.sun_family = AF_LOCAL;
193|
194|-> strncpy(address.sun_path, c->uds_name, SUN_PATH_LEN);
195|
196| if ((c->fd = socket(AF_LOCAL, SOCK_STREAM, 0)) == -1)
Signed-off-by: Jim Fehlig
---
vhostmd/virtio.c | 2 +-
1 file change
vu_log(VHOSTMD_ERR, "Failed to start virtio thread '%s'\n",
958| strerror(rc));
959|-> return -1;
960| }
961| }
Signed-off-by: Jim Fehlig
---
vhostmd/vhostmd.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-
-fixes
Jim Fehlig (18):
libmetrics: Set pointer NULL after free
libmetrics: Initialize local variable 'ret' to failure
libmetrics: Check return value of asprintf
libmetrics: Remove unsafe XML_PARSE_NOENT option
libmetrics: Ensure libmetrics mutex is unlocked in error paths
On 1/16/20 1:48 AM, Richard W.M. Jones wrote:
>
> All looks sensible to me, ACK series.
Thanks, pushed.
Regards,
Jim
On 7/19/23 09:50, Daniel P. Berrangé wrote:
On Wed, Jul 19, 2023 at 10:12:07AM +, Sergio Gelato wrote:
Does this project share the view expressed in
https://github.com/systemd/systemd/issues/28113#issuecomment-1621986461
that "when it comes to systemd's PoV xen dom0 is not a VM. i.e. the VM
or failed to connect to xend
> 2013-05-18 05:44:47.636+0000: 1304: error : xenDaemonOpen_tcp:700 :
> unable to connect to 'localhost:8000': Connection refused
> 2013-05-18 05:45:13.647+: 1289: error : virNetSocketReadWire:1362
> : Cannot recv data: Connection reset by peer
1 - 100 of 113 matches
Mail list logo