Both virProcessGetStatInfo() and virProcessGetSchedInfo() are
Linux centric. Provide stubs for non-Linux platforms.
Signed-off-by: Michal Privoznik
---
src/util/virprocess.c | 24
1 file changed, 24 insertions(+)
diff --git a/src/util/virprocess.c
On a Thursday in 2022, Ani Sinha wrote:
virURIParamAppend() unconditionally returns 0. Simplify and make the return type
as void type.
Signed-off-by: Ani Sinha
---
src/util/viruri.c | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
Reviewed-by: Ján Tomko
and pushed
Jano
Signed-off-by: Andrea Bolognani
---
Pushed as trivial.
NEWS.rst | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/NEWS.rst b/NEWS.rst
index 3c06fdde0f..7ee03fde0e 100644
--- a/NEWS.rst
+++ b/NEWS.rst
@@ -42,12 +42,12 @@ v8.0.0 (unreleased)
as
On a Thursday in 2022, Peter Krempa wrote:
Until now we did 2 weird things when inserting the qemuCaps used for
individual test cases into the capability cache:
1) we inserted the same caps for all emulators
2) we always (expensively) copied them
Now when real capabilities are used we don't
On 1/4/22 7:57 AM, Daniel P. Berrangé wrote:
On Sun, Jan 02, 2022 at 09:41:37PM -0500, Laine Stump wrote:
I'm currently working on switching the backend of the network driver from
using iptables to using nftables. Due to some functionality that is not
available with nftables (the rule that
In the commit summary:
*populating
On a Thursday in 2022, Peter Krempa wrote:
Separate the code to aid further refactors.
Signed-off-by: Peter Krempa
---
tests/testutilsqemu.c | 96 ---
1 file changed, 53 insertions(+), 43 deletions(-)
Reviewed-by:
On a Thursday in 2022, Peter Krempa wrote:
Emulator binary change is needed to use the latest caps properly. The
comment is no longer needed, the expected error is recorded in the 'err'
file.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/missing-machine.err| 1 -
On a Thursday in 2022, Peter Krempa wrote:
Also ensure that the emulator and architecture are correct for
DO_TEST_CAPS_LATEST.
Signed-off-by: Peter Krempa
---
.../blkdeviotune.x86_64-latest.args | 40 +++
tests/qemuxml2argvdata/blkdeviotune.xml | 4 +-
On a Thursday in 2022, Peter Krempa wrote:
As demonstrated by the qemuxml2xmltest DO_TEST_CAPS_LATEST data based on
the 'x86-kvm-32-on-64' test case the post parse CPU selection code which
fills in the CPU into the definition does not have exactly the same
logic as we used to have when the cpu
On a Thursday in 2022, Peter Krempa wrote:
Fake capabilities are inserted twice, as in a few tests the architecture
is not present in the XML (testing filling in of the architecture).
Since we already know which architecture will be picked we don't need to
be adding the capabilities twice.
This
On a Thursday in 2022, Peter Krempa wrote:
Upcoming patches will modify how we populate the capability cache in
tests to be more sane. This also means that the emulator binary and
architecture used in the test files using real capabilities must match
what the real capabilities have.
On a Thursday in 2022, Peter Krempa wrote:
Upcoming patches will modify how we populate the capability cache in
tests to be more sane. This also means that the emulator binary and
*saner
architecture used in the test files using real capabilities must match
what the real capabilities have.
On a Thursday in 2022, Peter Krempa wrote:
None of our tests reference 'isapc', and 'pc-i440fx-2.6'.
Signed-off-by: Peter Krempa
---
tests/testutilsqemu.c | 13 +++--
1 file changed, 3 insertions(+), 10 deletions(-)
Reviewed-by: Ján Tomko
Jano
signature.asc
Description: PGP
On a Thursday in 2022, Peter Krempa wrote:
Use 'pc' instead.
Signed-off-by: Peter Krempa
---
tests/securityselinuxlabeldata/chardev.xml | 2 +-
tests/securityselinuxlabeldata/disks.xml | 2 +-
tests/securityselinuxlabeldata/kernel.xml | 2 +-
tests/securityselinuxlabeldata/nfs.xml | 2 +-
4
On a Thursday in 2022, Peter Krempa wrote:
The specific machine type is not important for the test. We can use 'pc'
instead.
Signed-off-by: Peter Krempa
---
.../qemuhotplug-console-compat-2-live+console-virtio.xml| 2 +-
.../qemuhotplug-console-compat-2-live.xml |
On a Thursday in 2022, Peter Krempa wrote:
Use the 'pc' alias for them as it's equivalent.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/fd-memory-no-numa-topology.args | 2 +-
tests/qemuxml2argvdata/fd-memory-no-numa-topology.xml | 2 +-
On a Thursday in 2022, Peter Krempa wrote:
Since introduction in fc03eb53c0bf3f654 there wasn't a qemuxml2argv
version. As we are touching the files convert them to
DO_TEST_CAPS_LATEST directly.
Signed-off-by: Peter Krempa
---
.../controller-usb-order.x86_64-latest.args | 38
On a Thursday in 2022, Peter Krempa wrote:
The qemuxml2argv invocation of some tests used DO_TEST_CAPS_LATEST while
the qemuxml2xmltest invocation uses fake caps. Unify them on
DO_TEST_CAPS_LATEST.
Signed-off-by: Peter Krempa
---
.../luks-disks.x86_64-4.1.0.args | 1 +
On a Thursday in 2022, Peter Krempa wrote:
According to commit 5568490 the test case was added to verify that
the '' element is covered by the schema. The test was not
registered for qemuxml2argvtest though. We can use 'net-server' instead
as it has the same type. On the other hand that one
On a Thursday in 2022, Peter Krempa wrote:
Use q35 instead.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/bios-nvram-secure.args | 2 +-
tests/qemuxml2argvdata/bios-nvram-secure.xml | 2 +-
tests/qemuxml2argvdata/boot-floppy-q35.args | 2 +-
On a Thursday in 2022, Peter Krempa wrote:
There's nothing special about the tests requiring to use very old
machine types. Most usage is cargo-culted from other tests. Switch all
the tests to use 'pc' instead.
Signed-off-by: Peter Krempa
---
On a Thursday in 2022, Peter Krempa wrote:
For tests with fake capabilities we fill in a bunch of machine types
which the tests might use. For now there's a random collection of
machine types which are not actually used. Purge the unused ones for
non-x86 machines.
Signed-off-by: Peter Krempa
On Thu, 6 Jan 2022, Daniel P. Berrangé wrote:
> On Thu, Jan 06, 2022 at 08:34:26PM +0530, Ani Sinha wrote:
> > virConnectOpenInternal() does not report error in all failure scenarios,
> > except
> > in some specific cases.
>
> I don't believe this is correct. My reading of the code is that it
On a Thursday in 2022, Peter Krempa wrote:
The two test cases were added to avoid regressions such as fixed in
17dff3584855e9. Nowadays the code is much simpler and any Q35 machine
will trigger the explicit FDC.
Remove the '2.11' machine type version and turn the '2.9' version into a
generic
On a Thursday in 2022, Peter Krempa wrote:
Commit c7f3a1f7870 turned qemuDomainMachineNeedsFDC() effectively into
qemuDomainIsQ35. Use it instead as it also matches the non-canonicalized
'q35'.
Signed-off-by: Peter Krempa
---
src/qemu/qemu_domain.c | 17 ++---
1 file changed, 2
virConnectOpenAuth() calls virConnectOpenInternal(). This later function
generates fine grained errors arising from various failure conditions that are
more accurate than a "catch all" broader VIR_ERR_OPERATION_FAILED error that
the callers of this function generates. Remove the broader error so
virURIParamAppend() unconditionally returns 0. Simplify and make the return type
as void type.
Signed-off-by: Ani Sinha
---
src/util/viruri.c | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/src/util/viruri.c b/src/util/viruri.c
index 3c73188a55..88bb0cc1f8 100644
---
On a Thursday in 2022, Peter Krempa wrote:
All the fd-passing setup of chardevs which this hack meant to disable
was moved to the host-preparation phase which is skipped for formatting
of non-real commandlines.
Signed-off-by: Peter Krempa
---
src/qemu/qemu_process.c | 9 ++---
1 file
Signed-off-by: Peter Krempa
Reviewed-by: Ani Sinha
Reviewed-by: Ján Tomko
---
NEWS.rst | 31 +++
1 file changed, 31 insertions(+)
diff --git a/NEWS.rst b/NEWS.rst
index 3c06fdde0f..9e0b101741 100644
--- a/NEWS.rst
+++ b/NEWS.rst
@@ -59,6 +59,37 @@ v8.0.0
When -device is configured via JSON a bug [1] is triggered in qemu were
the DEVICE_DELETED event for the removal of the device frontend is no
longer delivered to libvirt. Without the DEVICE_DELETED event we don't
remove the corresponding entries in the VM XML.
Until qemu will be fixed we must
QEMU fix was posted
https://lists.gnu.org/archive/html/qemu-devel/2022-01/msg00542.html
I'll hold off this version until just before the freeze. In case the
qemu commit gets merged we can avoid some of the churn by adding
in-progress qemu caps and updating to it.
Note that I've trimmed patch 1
On a Wednesday in 2022, Daniel P. Berrangé wrote:
On Wed, Jan 05, 2022 at 04:06:49PM +0100, Ján Tomko wrote:
Add a new attribute to control the killpriv feature:
...
Ewww, ewww, ewww.
This is a horrible element & attribute name. Even with the docs I have
little
On Thu, Jan 06, 2022 at 08:34:26PM +0530, Ani Sinha wrote:
> virConnectOpenInternal() does not report error in all failure scenarios,
> except
> in some specific cases.
I don't believe this is correct. My reading of the code is that it
is already reporting errors in all exit paths that can
On Thu, Jan 06, 2022 at 20:34:26 +0530, Ani Sinha wrote:
> virConnectOpenInternal() does not report error in all failure scenarios,
> except
> in some specific cases. This inconsistent behavior forces the caller of this
> function report a generic error for all failure modes which then hides
>
On 1/6/22 16:04, Ani Sinha wrote:
> virConnectOpenInternal() does not report error in all failure scenarios,
> except
> in some specific cases. This inconsistent behavior forces the caller of this
> function report a generic error for all failure modes which then hides
> specific
> error
virConnectOpenInternal() does not report error in all failure scenarios, except
in some specific cases. This inconsistent behavior forces the caller of this
function report a generic error for all failure modes which then hides specific
error scenarios. This change makes virConnectOpenInternal()
On 12/10/21 21:34, Praveen K Paladugu wrote:
> Signed-off-by: Vineeth Pillai
> Signed-off-by: Praveen K Paladugu
> ---
> src/ch/ch_domain.c | 25 +
> src/ch/ch_domain.h | 4 ++
> src/ch/ch_driver.c | 134 +
> 3 files changed, 163
On 12/10/21 21:34, Praveen K Paladugu wrote:
> From: Vineeth Pillai
>
> Signed-off-by: Vineeth Pillai
> Signed-off-by: Praveen K Paladugu
> ---
> src/ch/ch_domain.c | 6 ++
> src/ch/ch_domain.h | 5 +
> 2 files changed, 11 insertions(+)
Reviewed-by: Michal Privoznik
Michal
On 12/10/21 21:34, Praveen K Paladugu wrote:
> From: Vineeth Pillai
>
> Signed-off-by: Vineeth Pillai
> Signed-off-by: Praveen K Paladugu
> ---
> src/ch/ch_domain.c | 50 +-
> src/ch/ch_domain.h | 11 ++
> 2 files changed, 56 insertions(+),
On 12/10/21 21:34, Praveen K Paladugu wrote:
> From: Vineeth Pillai
>
> Add domainGetVcpuPinInfo and nodeGetCPUMap callbacks to ch driver
>
> Signed-off-by: Vineeth Pillai
> Signed-off-by: Praveen K Paladugu
> ---
> src/ch/ch_domain.h | 10 +++--
> src/ch/ch_driver.c | 51
On 12/10/21 21:34, Praveen K Paladugu wrote:
> From: Vineeth Pillai
>
> Signed-off-by: Vineeth Pillai
> Signed-off-by: Praveen K Paladugu
> ---
> src/ch/ch_conf.h| 2 +
> src/ch/ch_domain.c | 27 ++-
> src/ch/ch_domain.h | 2 +
> src/ch/ch_driver.c | 1 +
>
On 12/10/21 21:34, Praveen K Paladugu wrote:
> Refactor some cgroup management methods from qemu into hypervisor.
> These methods will be shared with ch driver for cgroup management.
>
> Signed-off-by: Praveen K Paladugu
> ---
> src/hypervisor/domain_cgroup.c | 426
On 12/10/21 21:34, Praveen K Paladugu wrote:
> Move qemuGetProcessInfo and qemuGetSchedInfo methods to util and share them
> with ch driver.
>
> Signed-off-by: Praveen K Paladugu
> ---
> src/libvirt_private.syms | 2 +
> src/qemu/qemu_driver.c | 116 ++-
>
On 1/5/22 14:31, Martin Kletzander wrote:
> Suggested-by: Erik Skultety
> Signed-off-by: Martin Kletzander
> ---
> src/util/virlog.c | 8
> 1 file changed, 4 insertions(+), 4 deletions(-)
Reviewed-by: Michal Privoznik
Michal
On Thu, 6 Jan 2022, Michal Prívozník wrote:
> On 1/6/22 13:42, Ani Sinha wrote:
> >
> >
> > On Thu, 6 Jan 2022, Michal Prívozník wrote:
> >
> >> On 1/6/22 02:33, Raphael Norwitz wrote:
> >>> Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
> >>> returns
On 1/6/22 13:42, Ani Sinha wrote:
>
>
> On Thu, 6 Jan 2022, Michal Prívozník wrote:
>
>> On 1/6/22 02:33, Raphael Norwitz wrote:
>>> Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
>>> returns VIR_ERR_OPERATION_FAILED. This change switches that error code
>>> to
On Thu, 6 Jan 2022, Michal Prívozník wrote:
> On 1/6/22 02:33, Raphael Norwitz wrote:
> > Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
> > returns VIR_ERR_OPERATION_FAILED. This change switches that error code
> > to VIR_ERR_NO_CONNECT, which is more accurate.
> >
> >
On 1/6/22 02:33, Raphael Norwitz wrote:
> Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
> returns VIR_ERR_OPERATION_FAILED. This change switches that error code
> to VIR_ERR_NO_CONNECT, which is more accurate.
>
> This should help libvirt consumers more intellegently retry
On Thu, 6 Jan 2022, Raphael Norwitz wrote:
> Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
> returns VIR_ERR_OPERATION_FAILED. This change switches that error code
> to VIR_ERR_NO_CONNECT, which is more accurate.
Hmm, I am not sure if this would be the right thing.
On Thu, 6 Jan 2022, Peter Krempa wrote:
> On Tue, Jan 04, 2022 at 17:47:43 +0530, Ani Sinha wrote:
> > DEBUG_IO and DEBUG_RAW_IO are disabled and hence the code #defined under
> > them
> > are useless. Remove them. Also remove similar useless code from
> > qemu_monitor_json.
> >
> > See also
On Tue, Jan 04, 2022 at 04:43:11 -0800, Rohit Kumar wrote:
> This change adds the domain name in the error and debug logs during
> monitor IO processing so that we may infer which VM experienced
> errors such as IO or socket hangup. This may help in debugging
> monitor IO errors.
>
>
On Tue, Jan 04, 2022 at 17:47:43 +0530, Ani Sinha wrote:
> DEBUG_IO and DEBUG_RAW_IO are disabled and hence the code #defined under them
> are useless. Remove them. Also remove similar useless code from
> qemu_monitor_json.
>
> See also 4aae00bf1287f ("qemu: monitor: Remove disabled debug
Today if virConnectOpenAuth fails, qemuMigrationSrcPerformPeer2Peer()
returns VIR_ERR_OPERATION_FAILED. This change switches that error code
to VIR_ERR_NO_CONNECT, which is more accurate.
This should help libvirt consumers more intellegently retry migrations
on intermittent connection failures.
Until now we did 2 weird things when inserting the qemuCaps used for
individual test cases into the capability cache:
1) we inserted the same caps for all emulators
2) we always (expensively) copied them
Now when real capabilities are used we don't touch them at all just
simply inser them. This
Also ensure that the emulator and architecture are correct for
DO_TEST_CAPS_LATEST.
Signed-off-by: Peter Krempa
---
.../blkdeviotune.x86_64-latest.args | 40 +++
tests/qemuxml2argvdata/blkdeviotune.xml | 4 +-
tests/qemuxml2argvtest.c | 1 +
Fake capabilities are inserted twice, as in a few tests the architecture
is not present in the XML (testing filling in of the architecture).
Since we already know which architecture will be picked we don't need to
be adding the capabilities twice.
This doesn't impact the tests as they use the
Emulator binary change is needed to use the latest caps properly. The
comment is no longer needed, the expected error is recorded in the 'err'
file.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/missing-machine.err| 1 -
.../missing-machine.x86_64-latest.err
Separate the code to aid further refactors.
Signed-off-by: Peter Krempa
---
tests/testutilsqemu.c | 96 ---
1 file changed, 53 insertions(+), 43 deletions(-)
diff --git a/tests/testutilsqemu.c b/tests/testutilsqemu.c
index bf3f919dde..94ff538382 100644
Upcoming patches will modify how we populate the capability cache in
tests to be more sane. This also means that the emulator binary and
architecture used in the test files using real capabilities must match
what the real capabilities have.
Signed-off-by: Peter Krempa
---
As demonstrated by the qemuxml2xmltest DO_TEST_CAPS_LATEST data based on
the 'x86-kvm-32-on-64' test case the post parse CPU selection code which
fills in the CPU into the definition does not have exactly the same
logic as we used to have when the cpu model was picked when formatting
the
Use 'pc' instead.
Signed-off-by: Peter Krempa
---
tests/securityselinuxlabeldata/chardev.xml | 2 +-
tests/securityselinuxlabeldata/disks.xml | 2 +-
tests/securityselinuxlabeldata/kernel.xml | 2 +-
tests/securityselinuxlabeldata/nfs.xml | 2 +-
4 files changed, 4 insertions(+), 4
For tests with fake capabilities we fill in a bunch of machine types
which the tests might use. For now there's a random collection of
machine types which are not actually used. Purge the unused ones for
non-x86 machines.
Signed-off-by: Peter Krempa
---
tests/testutilsqemu.c | 12 +---
None of our tests reference 'isapc', and 'pc-i440fx-2.6'.
Signed-off-by: Peter Krempa
---
tests/testutilsqemu.c | 13 +++--
1 file changed, 3 insertions(+), 10 deletions(-)
diff --git a/tests/testutilsqemu.c b/tests/testutilsqemu.c
index 980555b27e..bf3f919dde 100644
---
The specific machine type is not important for the test. We can use 'pc'
instead.
Signed-off-by: Peter Krempa
---
.../qemuhotplug-console-compat-2-live+console-virtio.xml| 2 +-
.../qemuhotplug-console-compat-2-live.xml | 2 +-
2 files changed, 2 insertions(+), 2
Use the 'pc' alias for them as it's equivalent.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/fd-memory-no-numa-topology.args | 2 +-
tests/qemuxml2argvdata/fd-memory-no-numa-topology.xml | 2 +-
tests/qemuxml2argvdata/fd-memory-numa-topology.args | 2 +-
The qemuxml2argv invocation of some tests used DO_TEST_CAPS_LATEST while
the qemuxml2xmltest invocation uses fake caps. Unify them on
DO_TEST_CAPS_LATEST.
Signed-off-by: Peter Krempa
---
.../luks-disks.x86_64-4.1.0.args | 1 +
tests/qemuxml2argvdata/luks-disks.xml | 5
Since introduction in fc03eb53c0bf3f654 there wasn't a qemuxml2argv
version. As we are touching the files convert them to
DO_TEST_CAPS_LATEST directly.
Signed-off-by: Peter Krempa
---
.../controller-usb-order.x86_64-latest.args | 38 +++
tests/qemuxml2argvtest.c
According to commit 5568490 the test case was added to verify that
the '' element is covered by the schema. The test was not
registered for qemuxml2argvtest though. We can use 'net-server' instead
as it has the same type. On the other hand that one was not registered
for qemuxml2xmltest.
Use q35 instead.
Signed-off-by: Peter Krempa
---
tests/qemuxml2argvdata/bios-nvram-secure.args | 2 +-
tests/qemuxml2argvdata/bios-nvram-secure.xml | 2 +-
tests/qemuxml2argvdata/boot-floppy-q35.args | 2 +-
tests/qemuxml2argvdata/boot-floppy-q35.xml
There's nothing special about the tests requiring to use very old
machine types. Most usage is cargo-culted from other tests. Switch all
the tests to use 'pc' instead.
Signed-off-by: Peter Krempa
---
...launch-security-sev-missing-platform-info.x86_64-2.12.0.args | 2 +-
Commit c7f3a1f7870 turned qemuDomainMachineNeedsFDC() effectively into
qemuDomainIsQ35. Use it instead as it also matches the non-canonicalized
'q35'.
Signed-off-by: Peter Krempa
---
src/qemu/qemu_domain.c | 17 ++---
1 file changed, 2 insertions(+), 15 deletions(-)
diff --git
The two test cases were added to avoid regressions such as fixed in
17dff3584855e9. Nowadays the code is much simpler and any Q35 machine
will trigger the explicit FDC.
Remove the '2.11' machine type version and turn the '2.9' version into a
generic q35 machine.
Signed-off-by: Peter Krempa
---
This series originally started as an attempt to remove pointless copying
of capabilitiy objects in the qemuxml2argvtest which actually took ~40%
of the runtime of the test.
Obviously few snags and places for improvements were noticed along the
way.
Peter Krempa (21):
All the fd-passing setup of chardevs which this hack meant to disable
was moved to the host-preparation phase which is skipped for formatting
of non-real commandlines.
Signed-off-by: Peter Krempa
---
src/qemu/qemu_process.c | 9 ++---
1 file changed, 2 insertions(+), 7 deletions(-)
diff
On Wed, Jan 05, 2022 at 05:21:11PM +0100, Ján Tomko wrote:
The commit that added error checking to this function
forgot to adjust the WIN32 stub.
Fixes: a873924e36b28c5b125621e35b32beb6b077bcc8
Signed-off-by: Ján Tomko
---
Pushed as a build fix.
Oops, thanks.
src/util/virdaemon.c | 16
75 matches
Mail list logo
