Peter Krempa [Mon, 3 Aug 2020 10:09:21 +0000 (12:09 +0200)]
kbase: debuglogs: Include more data to attach to bugreports
Promote the 'What to attach?' section to a first level heading and
request also the XML config of a VM, coredump backtrace if something
crashed and ask to not tear down the environment for the possibility to
ask for additional data.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Kashyap Chamarthy <kchamart@redhat.com>
Peter Krempa [Mon, 3 Aug 2020 14:38:31 +0000 (16:38 +0200)]
docs: build: Use proper path to file in 'contribute' link
NEWS.rst is based in the root of the repository and 'hvsupport.html'
doesn't have a backing file which can be edited since it's fully
generated. Our 'contribute -> edit this page' link on the bottom of the
page is wrong in those cases.
Fix it by adding the contribute section only when there's a source and
base the 'source' of a html file in the root of the repository.
Along with that we need to modify the scripts/meson-html-gen.py script
to accept optional 'pagesrc' and the XSL template to skip the
'contribute' section when we don't have a source.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
virLoginShellGetShellArgv was not dereferencing the pointer
to the string list containing the shell parameters from the
config file, thus setting some random number as shargvlen.
Signed-off-by: Ján Tomko <jtomko@redhat.com> Fixes: 740e4d705284ba0598258291c2656f50075486f0 Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Ján Tomko [Sun, 2 Aug 2020 21:35:16 +0000 (23:35 +0200)]
tools: virt-admin: do not leak daemon-log settings
The commands daemon-log-filters and daemon-log-outputs
are used both for getting and setting the variables.
But the getter receives an allocated string, which
we do not free.
Use separate variables for the getter and the setter
to get rid of the memory leak and to stop casting
away the const.
Signed-off-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Ján Tomko [Sun, 2 Aug 2020 20:52:51 +0000 (22:52 +0200)]
vz: remove redundant NULL pointer check
The 'dom' pointer is already dereferenced earlier.
src/vz/vz_sdk.c:249:24: warning: Either the condition 'if(dom)'
is redundant or there is possible null pointer dereference:
dom. [nullPointerRedundantCheck]
Signed-off-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
It was created to get rid of conditional compilation in the resctrl code and
make it usable anywhere else. However this is not something that is going to be
used in other places because it is not portable and resctrl is just very
specific in this regard. And there is no reason why there could not be a
preprocessor conditional in the resctrl code. Also the interface of
virFileFlock() was very ambiguous which lead to some issues.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Reviewed-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
That's the way it should've been all the time. It was originally the case, but
then the rework to virFileFlock() made the function ambiguous when it was
created in commit 5a0a5f7fb5f5, and due to that it was misused in commit 657ddeff2313 and since then the lock being taken was shared rather than
exclusive.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Reviewed-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Wang Xin [Fri, 24 Jul 2020 03:34:12 +0000 (11:34 +0800)]
conf: allow shmem name change in migration
The shmem 'name' specifies the shared memory path in '/dev/shm/',
however, we may need to change it to avoid filename conflict
when VM migrate to other host. This patch remove shmem name
consistency check.
Signed-off-by: Wang Xin <wangxinxin.wang@huawei.com> Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Wang Xin [Fri, 24 Jul 2020 03:34:11 +0000 (11:34 +0800)]
qemu: add support for shmem-{plain, doorbell} role
Role(master or peer) controls how the domain behaves on migration.
For more details about migration with ivshmem, see
https://git.qemu.org/?p=qemu.git;a=blob_plain;f=docs/system/ivshmem.rst;hb=HEAD
It's a optional attribute in libvirt, and qemu will choose default
role for ivshmem device if the user is not specified.
With device property 'role', the value can be 'master' or 'peer'.
- 'master' (means 'master=on' in qemu), the guest will copy
the shared memory on migration to the destination host.
- 'peer' (means 'master=off' in qemu), the migration is disabled.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Signed-off-by: Yang Hang <yanghang44@huawei.com> Signed-off-by: Wang Xin <wangxinxin.wang@huawei.com> Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Pavel Hrdina [Thu, 16 Jul 2020 08:26:02 +0000 (10:26 +0200)]
meson: .gitlab-ci.yml: switch armv7l build to Debian 10
On Debian Sid there are broken headers for cross compilation. There is
no <asm/kvm.h> header unconditionally included by <linux/kvm.h>.
With autotools the build works correctly because the header detection
tries to compile simple code that includes the header but with Meson
the detection works differently. We check <linux/kvm.h> which is present
so it should be safe to include it.
We could workaround it in libvirt by checking for presence of
<asm/kvm.h> as well but I don't believe that's a correct approach.
IMHO it should be fixed in Debian.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com>
Pavel Hrdina [Wed, 1 Jul 2020 13:35:41 +0000 (15:35 +0200)]
meson: drop remaining bits from Makefile.am
There is no alternative to DISTCHECK_CONFIGURE_FLAGS in Meson. We could
enable -Werror unconditionally but that would affect all users. It is
mainly used in our CI environment so we can run meson with --werror to
enable it in configure time which will be picked up by meson dist as
well.
XZ_OPT is not relevant with meson since it uses shutil.make_archive
which uses lzma python module to do the compression.
Introducing rpm and srpm rules should be possible with Meson but we
don't even use them in our CI and it's easy to run 'ninja dist'
followed by rpmbuild manually.
As for the ci rules that is simply not possible in the same way, we
would have to have a lot of aliases for all the ci rules and that would
be just insane.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com>
Pavel Hrdina [Thu, 25 Jun 2020 21:29:43 +0000 (23:29 +0200)]
meson: add syntax-check
This changes the approach used with autotools where it was separate make
target. With meson it will be part of the `meson test` target but can be
disabled using --no-suite syntax-check or we can run only syntax-check
by using --suite syntax-check.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com>
Pavel Hrdina [Tue, 28 Jul 2020 12:29:32 +0000 (14:29 +0200)]
meson: tests: add file access test setup
We need to modify check-file-access.py to be usable as wrapper for
libvirt tests. This way we can run the tests using this command:
meson test --setup access
which will run all tests using check-file-access.py as a wrapper.
With autotools all file access are written into single file for all
tests and compared once the whole test suite is done.
With Meson we will compare the file access after every single test
because it is used as wrapper now. That requires writing the file
access into separate files for every single test as they are executed
in parallel.
Since the wrapper is used for all tests in Meson including tests outside
of tests directory we have to check for presence of the output file.
We should also cleanup after ourselves.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com>
projects / thirdparty / libvirt.git / log
