[libvirt PATCH v2 4/5] use new functions when probing

Probe for TCG with QMP right after probing for KVM. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 13 - 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 528bf6e504..b3ba1cb079

[libvirt PATCH v2 1/5] add QEMU_CAPS_TCG_DISABLED and probe conditionally

Only probe QEMU binary with accel=tcg if TCG is not disabled. Similarly, only add a VIR_DOMAIN_VIRT_QEMU guest if TCG is available. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 22 ++ src/qemu/qemu_capabilities.h | 1 + 2 files changed, 15

[libvirt PATCH v2 0/5] QEMU: support QEMU built without TCG

Edit: left out one line from v1. Since version 2.10, QEMU can be built without TCG. This patch adds capabillity QEMU_CAPS_TCG_DISABLED that allows libvirt to use a QEMU built without TCG. Rather than create a capability that is set whenever TCG is enabled (almost always), QEMU_CAPS_TCG_DISABLED

[libvirt PATCH v2 3/5] add virQEMUCapsProbeQMPTCGState function to set QEMU_CAPS_TCG_DISABLED

Probe for TCG with QMP right after probing for KVM. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 32 1 file changed, 32 insertions(+) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index

[libvirt PATCH v2 2/5] add virQEMUCapsGetVirtType convenience function

Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 14 ++ src/qemu/qemu_capabilities.h | 2 ++ 2 files changed, 16 insertions(+) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 61123f0e63..0b44ee0745 100644 ---

[PATCH] qemu-shim: add a --with-config option

From: Marc-André Lureau Add an option to ease setting up a VM with existing libvirt qemu.conf. Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1817776 Signed-off-by: Marc-André Lureau --- docs/manpages/virt-qemu-run.rst | 6 ++ src/libvirt_private.syms| 1 -

Entering freeze for libvirt-6.2.0

I forgot to announce plans, sorry about this, entering freeze now I hope there isn't too many pending patchestargetting this release. The plan would be to push the RC2 on Tuesday with a release on April 2nd if everything goes fine. You can find the signed tarball and source rpm at the usual

[libvirt PATCH] docs: Clarify semantics of model/@usable attribute in dom caps

The documentation could confuse people to expect that CPU models with usable='no' attribute are not usable at all on the current host. But they cannot be only used without explicitly disabling some features. Signed-off-by: Jiri Denemark --- docs/formatdomaincaps.html.in | 8 +--- 1 file

[libvirt PATCH 0/5] QEMU: support QEMU built without TCG

Since version 2.10, QEMU can be built without TCG. This patch adds capabillity QEMU_CAPS_TCG_DISABLED that allows libvirt to use a QEMU built without TCG. Rather than create a capability that is set whenever TCG is enabled (almost always), QEMU_CAPS_TCG_DISABLED is set only when the TCG is not

[libvirt PATCH 2/5] add virQEMUCapsGetVirtType convenience function

Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 14 ++ src/qemu/qemu_capabilities.h | 2 ++ 2 files changed, 16 insertions(+) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 0371ca9b03..9c946339a3 100644 ---

[libvirt PATCH 3/5] add virQEMUCapsProbeQMPTCGState function to set QEMU_CAPS_TCG_DISABLED

virQEMUCapsProbeQMPTCGState will set QEMU_CAPS_TCG_DISABLED if QEMU version is at least 2.10, KVM is enabled, and qom-list-types does not include tcg-accel. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 32 1 file changed, 32

[libvirt PATCH 4/5] use new functions when probing

Probe for TCG with QMP right after probing for KVM. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 13 - 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index c84ad0e831..410b58bf47

[libvirt PATCH 1/5] add QEMU_CAPS_TCG_DISABLED and probe conditionally

Only probe QEMU binary with accel=tcg if TCG is not disabled. Similarly, only add a VIR_DOMAIN_VIRT_QEMU guest if TCG is available. Signed-off-by: Tobin Feldman-Fitzthum --- src/qemu/qemu_capabilities.c | 21 + src/qemu/qemu_capabilities.h | 1 + 2 files changed, 14

Re: [libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence

On Fri, 2020-03-27 at 16:52 +0100, Michal Prívozník wrote: > On 27. 3. 2020 16:42, Daniel P. Berrangé wrote: > > > > +if (priv->escapeChar[0] == '^') { > > > > +vshPrintExtra(ctl, " (Ctrl + %c)", priv->escapeChar[1]); > > > > +} > > > > I'll remove the {} to keep syntax-check

[libvirt-jenkins-ci PATCH 6/7] lcitool: Use cross_commands[] for all distros

It's nicer to use, and consistent with how we're building the native part of the Dockerfile. Signed-off-by: Andrea Bolognani --- guests/lcitool | 33 +++-- 1 file changed, 19 insertions(+), 14 deletions(-) diff --git a/guests/lcitool b/guests/lcitool index

[libvirt-jenkins-ci PATCH 7/7] lcitool: Create and expose ccache wrappers

VM-based builds have used ccache by default for a very long time, and now container-based builds will too. Signed-off-by: Andrea Bolognani --- guests/lcitool | 18 ++ 1 file changed, 18 insertions(+) diff --git a/guests/lcitool b/guests/lcitool index 117e1a5..011fc07 100755 ---

[libvirt-jenkins-ci PATCH 1/7] lcitool: Improve ccache symlinks creation

Instead of having the same code snippet twice, once for Linux and once for FreeBSD, the only actual difference being the name of the compiler, include the compiler's path in the inventory among other paths and then figure out the correct name for the symlink based on that information.

[libvirt-jenkins-ci PATCH 2/7] lcitool: Configure ccache using environment variables

We're going to stop creating $HOME/.ccache soon, so we need an alternative way to configure ccache; environment variables work perfectly fine. Signed-off-by: Andrea Bolognani --- guests/playbooks/update/tasks/users.yml | 7 --- guests/playbooks/update/templates/bashrc.j2 | 1 +

[libvirt-jenkins-ci PATCH 3/7] lcitool: Create ccache wrappers outside of $HOME

Having the wrappers inside $HOME works fine when we have control over which user will run the build, as is the case for the current Jenkins-based environment, but not when the user is chosen by an external entity, as is the case for container-based builds. Signed-off-by: Andrea Bolognani ---

[libvirt-jenkins-ci PATCH 5/7] lcitool: Use commands[] for deb-based distros

It's nicer to use, and will make further changes easier. This commit is better viewed with 'git show -w'. Signed-off-by: Andrea Bolognani --- guests/lcitool | 29 ++--- 1 file changed, 14 insertions(+), 15 deletions(-) diff --git a/guests/lcitool b/guests/lcitool index

[libvirt-jenkins-ci PATCH 0/7] lcitool: Create and expose ccache wrappers

This series makes ccache work out of the box for container-based builds. Most of it is refactoring. Applies on top of https://www.redhat.com/archives/libvir-list/2020-March/msg01263.html Andrea Bolognani (7): lcitool: Improve ccache symlinks creation lcitool: Configure ccache using

[libvirt-jenkins-ci PATCH 4/7] lcitool: Refactor cross_arch handling a bit

It will make further changes easier. This commit is better viewed with 'git show -w'. Signed-off-by: Andrea Bolognani --- guests/lcitool | 34 -- 1 file changed, 16 insertions(+), 18 deletions(-) diff --git a/guests/lcitool b/guests/lcitool index

Re: On the need to move to a merge request workflow

On Fri, Mar 27, 2020 at 02:15:01PM +0100, Peter Krempa wrote: > On Thu, Mar 26, 2020 at 15:50:20 +, Daniel Berrange wrote: > > On Thu, Mar 26, 2020 at 02:10:59PM +0100, Peter Krempa wrote: > > > On Fri, Mar 06, 2020 at 11:44:07 +, Daniel Berrange wrote: > > > > We've discussed the idea of

Re: [libvirt PATCH 2/2] gitlab: Define $MAKEFLAGS in environment

On Fri, Mar 27, 2020 at 06:15:59PM +0100, Andrea Bolognani wrote: > This way we don't need to call getconf over and over, and the actual > build steps don't end up drowned in the noise. > > Signed-off-by: Andrea Bolognani > --- > .gitlab-ci.yml | 15 --- > 1 file changed, 8

[PATCH] util: virdaemon: fix compilation on mingw

The daemons are not supported on Win32 and therefore were not compiled in that platform. However, with the daemon code sharing, all the code in utils *is* compiled and it failed because `waitpid`, `fork`, and `setsid` are not available. So, as before, let's not build them on Win32 and make the

Re: [libvirt PATCH 1/2] gitlab: Add and use 'environment' template

On Fri, Mar 27, 2020 at 06:15:58PM +0100, Andrea Bolognani wrote: > This template provides a standardized place where environment > variables, which might not be static but depend on the output of > some shell command, can be defined for later use. It's not very > interesting right now, but we'll

[libvirt-jenkins-ci PATCH 3/3] lcitool: Include some paths in the generated Dockerfiles

These paths are a reflections of the contents of the container image, so it makes sense to expose them in the container's environment. This will allow the GitLab CI integration for both libvirt and libosinfo to stop setting these values themselves and start relying on them being already present

[libvirt-jenkins-ci PATCH 1/3] lcitool: Include paths in the inventory

Figuring them out at runtime is neat, but in practice they don't change frequently enough for it to be necessary; more importantly, including them in the inventory means we can use them in the Dockerfile generator in addition to the Ansible playbooks. Signed-off-by: Andrea Bolognani ---

[libvirt-jenkins-ci PATCH 2/3] lcitool: Refactor varmap generation a bit

Some values don't depend on whether the underlying OS user deb or rpm as the package format, so we can set them in common code. Signed-off-by: Andrea Bolognani --- guests/lcitool | 15 +-- 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/guests/lcitool b/guests/lcitool

[libvirt-jenkins-ci PATCH 0/3] lcitool: Include some paths in the generated Dockerfiles

Andrea Bolognani (3): lcitool: Include paths in the inventory lcitool: Refactor varmap generation a bit lcitool: Include some paths in the generated Dockerfiles guests/host_vars/libvirt-centos-7/main.yml| 9 +++ guests/host_vars/libvirt-centos-8/main.yml| 9 +++

Re: [libvirt PATCH v3 10/12] gitlab: introduce use of ccache for speeding up rebuilds

On Fri, Mar 27, 2020 at 03:59:51PM +0100, Andrea Bolognani wrote: > Anyway, what I do *not* want to see is something along the lines of > > x86-freebsd-12: > variables: > MAKE: gmake > CC: clang > > in every .gitlab-ci.yml for every project under the libvirt umbrella: > not

[libvirt PATCH 2/2] gitlab: Define $MAKEFLAGS in environment

This way we don't need to call getconf over and over, and the actual build steps don't end up drowned in the noise. Signed-off-by: Andrea Bolognani --- .gitlab-ci.yml | 15 --- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml index

[libvirt PATCH 0/2] gitlab: Improve make handling

Andrea Bolognani (2): gitlab: Add and use 'environment' template gitlab: Define $MAKEFLAGS in environment .gitlab-ci.yml | 28 1 file changed, 20 insertions(+), 8 deletions(-) -- 2.25.1

[libvirt PATCH 1/2] gitlab: Add and use 'environment' template

This template provides a standardized place where environment variables, which might not be static but depend on the output of some shell command, can be defined for later use. It's not very interesting right now, but we'll put it to good use in a second. This pattern is already used in

Re: [PATCH v2 4/4] qemu: add support for 'multidevs' option

On a Friday in 2020, Christian Schoenebeck wrote: This option prevents misbehaviours on guest if a qemu 9pfs export contains multiple devices, due to the potential file ID collisions this otherwise may cause. Signed-off-by: Christian Schoenebeck --- src/qemu/qemu_command.c | 7 +++

Re: [PATCH v2 3/4] conf: add 'multidevs' option

On a Friday in 2020, Christian Schoenebeck wrote: Introduce new 'multidevs' option for filesystem. This option prevents misbehaviours on guest if a qemu 9pfs export contains multiple devices, due to the potential file ID collisions this otherwise may cause. Signed-off-by:

Re: [PATCH v2 2/4] qemu: capabilities: add QEMU_CAPS_VIRTFS_MULTIDEVS

On a Friday in 2020, Christian Schoenebeck wrote: The QEMU 9pfs 'multidevs' option exists since QEMU 4.2, so just set this capability based on that QEMU version. Signed-off-by: Christian Schoenebeck --- src/qemu/qemu_capabilities.c | 5 + src/qemu/qemu_capabilities.h

Re: [PATCH v2 1/4] docs: virtfs: add section separators

On a Friday in 2020, Christian Schoenebeck wrote: Signed-off-by: Christian Schoenebeck --- docs/formatdomain.html.in | 7 +++ 1 file changed, 7 insertions(+) Reviewed-by: Ján Tomko and pushed Jano signature.asc Description: PGP signature

Re: [libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence

On 27. 3. 2020 16:42, Daniel P. Berrangé wrote: > On Fri, Mar 27, 2020 at 04:34:48PM +0100, Michal Prívozník wrote: >> On 27. 3. 2020 16:11, Daniel P. Berrangé wrote: >>> Signed-off-by: Daniel P. Berrangé >>> --- >>> tools/virsh-domain.c | 6 +- >>> 1 file changed, 5 insertions(+), 1

[PATCH 3/3] qemuxml2argv: Test the query portion for HTTP disks

Add a simple test case for preserving the query portion of http disk's path. Signed-off-by: Peter Krempa --- tests/qemuxml2argvdata/disk-network-http.x86_64-latest.args | 5 +++-- tests/qemuxml2argvdata/disk-network-http.xml | 2 +-

[PATCH 2/3] qemuBlockStorageSourceGetURI: Properly preserve query component

Look for the questionmark in the name and move the contents into the query portion so that we format the URI back properly. Signed-off-by: Peter Krempa --- src/qemu/qemu_block.c | 6 ++ 1 file changed, 6 insertions(+) diff --git a/src/qemu/qemu_block.c b/src/qemu/qemu_block.c index

[PATCH 0/3] Preserve query component of path for HTTP disks

Peter Krempa (3): virStorageSourceParseBackingURI: Preserve query parts of URI qemuBlockStorageSourceGetURI: Properly preserve query component qemuxml2argv: Test the query portion for HTTP disks src/qemu/qemu_block.c | 6 ++ src/util/virstoragefile.c

[PATCH 1/3] virStorageSourceParseBackingURI: Preserve query parts of URI

For non-NBD URIs we need to preserve the query part as it may be important to refer to the image. If the query doesn't start with 'socket=' concatenate it to the name of the image. Signed-off-by: Peter Krempa --- src/util/virstoragefile.c | 15 +++ tests/virstoragetest.c| 4

Re: [libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence

On Fri, Mar 27, 2020 at 04:34:48PM +0100, Michal Prívozník wrote: > On 27. 3. 2020 16:11, Daniel P. Berrangé wrote: > > Signed-off-by: Daniel P. Berrangé > > --- > > tools/virsh-domain.c | 6 +- > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > diff --git a/tools/virsh-domain.c

Re: [PATCH 2/2] syntax-check.mk: Drop needless _src2 variable

On Fri, Mar 27, 2020 at 04:29:55PM +0100, Michal Privoznik wrote: > The aim was to break a long line, but it is not long anymore. > > Signed-off-by: Michal Privoznik > --- > build-aux/syntax-check.mk | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) Reviewed-by: Daniel P. Berrangé

Re: [PATCH 1/2] syntax-check: Disallow plain fork() in src/libvirt.c

On Fri, Mar 27, 2020 at 04:29:54PM +0100, Michal Privoznik wrote: > The last usage of plain fork() was removed in v0.9.7-rc1~50, but > we forgot to update the syntax-check exemption list accordingly. > > Signed-off-by: Michal Privoznik > --- > build-aux/syntax-check.mk | 2 +- > 1 file changed,

Re: [libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence

On 27. 3. 2020 16:11, Daniel P. Berrangé wrote: > Signed-off-by: Daniel P. Berrangé > --- > tools/virsh-domain.c | 6 +- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c > index 8591e483a5..ada0189685 100644 > ---

[PATCH 0/2] Two trivial syntax-check fixes

While reviewing patches [1] I've found two problems with syntax-check which are trivial to fix. These are the patches I've pushed. 1: https://www.redhat.com/archives/libvir-list/2020-March/msg01124.html Michal Prívozník (2): syntax-check: Disallow plain fork() in src/libvirt.c

[PATCH 2/2] syntax-check.mk: Drop needless _src2 variable

The aim was to break a long line, but it is not long anymore. Signed-off-by: Michal Privoznik --- build-aux/syntax-check.mk | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/build-aux/syntax-check.mk b/build-aux/syntax-check.mk index 888e358848..db3c0cff0f 100644 ---

[PATCH 1/2] syntax-check: Disallow plain fork() in src/libvirt.c

The last usage of plain fork() was removed in v0.9.7-rc1~50, but we forgot to update the syntax-check exemption list accordingly. Signed-off-by: Michal Privoznik --- build-aux/syntax-check.mk | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/build-aux/syntax-check.mk

Re: [PATCH 0/4] Introduce shared daemon startup code

On 26. 3. 2020 16:17, Rafael Fonseca wrote: > Libvirt ships several daemons (libvirtd, virtlogd, virtlockd) that all > have similar code around general daemon startup. Right now the code is > duplicated for each daemon, but it could be shared, like in a new file > src/util/virdaemon.c. > > Rafael

Re: [PATCH 1/4] util: introduce shared daemon startup code

On 26. 3. 2020 16:18, Rafael Fonseca wrote: > Several daemons have similar code around general daemon startup code. > Let's move it into a file and share it among them. > > Signed-off-by: Rafael Fonseca > --- > src/libvirt_private.syms | 6 + > src/util/Makefile.inc.am | 2 + >

Re: [libvirt PATCH] scripts: avoid error in DCO check on empty branches

On Fri, 2020-03-27 at 14:41 +, Daniel P. Berrangé wrote: > If the DCO check is run on an empty branch (ie one which has no commits > different from master), it throws an error due to trying to interpret > the empty string as a git commit SHA. > > Signed-off-by: Daniel P. Berrangé > --- >

[libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence

Signed-off-by: Daniel P. Berrangé --- tools/virsh-domain.c | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 8591e483a5..ada0189685 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -3002,7 +3002,11 @@

Re: [libvirt PATCH v3 10/12] gitlab: introduce use of ccache for speeding up rebuilds

On Fri, 2020-03-27 at 10:47 +, Daniel P. Berrangé wrote: > On Thu, Mar 26, 2020 at 05:38:47PM +0100, Andrea Bolognani wrote: > > On Thu, 2020-03-26 at 14:05 +, Daniel P. Berrangé wrote: > > > On Thu, Mar 26, 2020 at 02:50:48PM +0100, Andrea Bolognani wrote: > > > > Ultimately I think we

[libvirt PATCH] scripts: avoid error in DCO check on empty branches

If the DCO check is run on an empty branch (ie one which has no commits different from master), it throws an error due to trying to interpret the empty string as a git commit SHA. Signed-off-by: Daniel P. Berrangé --- scripts/require-dco.py | 5 - 1 file changed, 4 insertions(+), 1

[libvirt PATCH] gitlab: don't run the DCO check on git master

The "only: branches" does not in fact exclude "master", so it must be excluded explicitly. Signed-off-by: Daniel P. Berrangé --- Pushed as a build fix since the DCO check fails on master .gitlab-ci.yml | 1 + 1 file changed, 1 insertion(+) diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml index

[libvirt PATCH 08/39] cpu_x86: Use g_auto* in x86DataToCPU

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 20 ++-- 1 file changed, 6 insertions(+), 14 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index fce7a2b8c5..d0ef66f1e3 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -771,9 +771,9 @@

[libvirt PATCH 16/39] cpu_x86: Use g_auto* in virCPUx86Compare

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 19 --- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 6758fcc170..e4e21fbed4 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1848,34 +1848,31 @@

[libvirt PATCH 28/39] cpu_x86: Use g_auto* in virCPUx86CopyMigratable

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 10 +++--- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index a65215caed..16e73b5b98 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -3113,7 +3113,7 @@

[libvirt PATCH 27/39] cpu_x86: Use g_auto* in virCPUx86ExpandFeatures

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 22 -- 1 file changed, 8 insertions(+), 14 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 3f6d889722..a65215caed 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -3059,30 +3059,29 @@ static

[libvirt PATCH 24/39] cpu_x86: Use g_auto* in virCPUx86Update

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 15 +-- 1 file changed, 5 insertions(+), 10 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 9f6027ec5c..8da8b3ada2 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2857,9 +2857,8 @@ static int

[libvirt PATCH 38/39] cputest: Add data for Intel(R) Xeon(R) Gold 6130 CPU

Skylake-Server with family 6, model 85, stepping 4, which is currently mis-detected as Cascadelake-Server CPU model. Signed-off-by: Jiri Denemark --- tests/cputest.c |1 + .../x86_64-cpuid-Xeon-Gold-6130-disabled.xml |7 +

[libvirt PATCH 14/39] cpu_x86: Use g_auto* in virCPUx86DataParse

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 23 --- 1 file changed, 8 insertions(+), 15 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index ba269df66d..6c3f9fc0be 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1636,8 +1636,8 @@

[libvirt PATCH 30/39] cpu_x86: Move and rename x86ModelHasSignature

Later in this series the function will work on a newly introduced virCPUx86Signatures structure. Let's move it to the place were all related functions will be added and rename the function as virCPUx86SignaturesMatch for easier review of the virCPUx86Signatures patch. Signed-off-by: Jiri Denemark

[libvirt PATCH 33/39] cpu_x86: Introduce virCPUx86SignatureFromCPUID

It can be used for separating family, model, and stepping numbers from a single 32b integer as reported by CPUID. Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 19 ++- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c

[libvirt PATCH 22/39] cpu_x86: Use g_auto* in virCPUx86Baseline

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 46 +- 1 file changed, 17 insertions(+), 29 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 7bd19a1676..ccbae3dabd 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@

[libvirt PATCH 07/39] cpu_x86: Use glib allocation in virCPUx86GetModels

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 10 +- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index bcb87d6f93..fce7a2b8c5 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -3123,21 +3123,13 @@

[libvirt PATCH 37/39] cputest: Add data for Intel(R) Xeon(R) Platinum 9242 CPU

Cascadelake-Server with family 6, model 85, stepping 7. Signed-off-by: Jiri Denemark --- tests/cputest.c |1 + ...6_64-cpuid-Xeon-Platinum-9242-disabled.xml |7 + ...86_64-cpuid-Xeon-Platinum-9242-enabled.xml | 10 +

[libvirt PATCH 35/39] cpu_x86: Don't check return value of x86ModelCopy

Thanks to glib allocation functions which abort on OOM the function cannot ever return NULL. Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 14 -- 1 file changed, 4 insertions(+), 10 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index ed6c1b43d3..51c98efca9

[libvirt PATCH 31/39] cpu_x86: Move and rename x86FormatSignatures

Later in this series the function will work on a newly introduced virCPUx86Signatures structure. Let's move it to the place were all related functions will be added and rename the function as virCPUx86SignaturesFormat for easier review of the virCPUx86Signatures patch. Signed-off-by: Jiri

[libvirt PATCH 26/39] cpu_x86: Use g_auto* in virCPUx86Translate

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 26 ++ 1 file changed, 10 insertions(+), 16 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 0be2cf517d..3f6d889722 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -3019,45 +3019,39 @@

[libvirt PATCH 36/39] cpu_x86: Add support for stepping part of CPU signature

CPU models defined in the cpu_map can use signature/@stepping attribute to match a limited set of stepping numbers. The value is a bitmap for bits 0..15 each corresponding to a single stepping value. For example, stepping='4-6,9' will match 4, 5, 6, and 9. Omitting the attribute is equivalent to

[libvirt PATCH 34/39] cpu_x86: Replace 32b signatures in virCPUx86Model with a struct

The CPU models in our cpu_map define their signatures using separate family and model numbers. Let's store the signatures in the same way in our runtime representation of the cpu_map. Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 149 +- 1 file

[libvirt PATCH 39/39] cpu_map: Distinguish Cascadelake-Server from Skylake-Server

The signatures of these two CPU model differ only in stepping as both report family 6 and model 85. Skylake-Server uses stepping 4 or less and Cascadelake-Server uses stepping 5..7. https://bugzilla.redhat.com/show_bug.cgi?id=1761678 Signed-off-by: Jiri Denemark ---

[libvirt PATCH 32/39] cpu_x86: Introduce virCPUx86SignaturesFree

The function will be used for freeing virCPUx86Signatures structure introduced later in this series. Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 11 +-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index

[libvirt PATCH 21/39] cpu_x86: Use g_auto* in virCPUx86GetHost

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 14 ++ 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 3b3a428ecd..7bd19a1676 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2641,18 +2641,18 @@ static int

[libvirt PATCH 29/39] cpu_x86: Move and rename x86ModelCopySignatures

Later in this series the function will work on a newly introduced virCPUx86Signatures structure. Let's move it to the place were all related functions will be added and rename the function as virCPUx86SignaturesCopy for easier review of the virCPUx86Signatures patch. Signed-off-by: Jiri Denemark

[libvirt PATCH 25/39] cpu_x86: Use g_auto* in virCPUx86UpdateLive

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 42 +++--- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 8da8b3ada2..0be2cf517d 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@

[libvirt PATCH 00/39] Distinguish Cascadelake-Server from Skylake-Server

The signatures of these two CPU model differ only in stepping as both report family 6 and model 85. Skylake-Server uses stepping 4 or less and Cascadelake-Server uses stepping 5..7. https://bugzilla.redhat.com/show_bug.cgi?id=1761678 Jiri Denemark (39): cpu_x86: Drop noTSX hint for

[libvirt PATCH 17/39] cpu_x86: Use g_auto* in x86Decode

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 26 +- 1 file changed, 9 insertions(+), 17 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index e4e21fbed4..26d872622e 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2034,15 +2034,14 @@

[libvirt PATCH 10/39] cpu_x86: Use g_auto* in x86FeatureParse

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 31 --- 1 file changed, 12 insertions(+), 19 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index a678bcde61..d2bbcbd891 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1023,13 +1023,12

[libvirt PATCH 05/39] cpu_x86: Use glib allocation for virCPUx86Model

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 31 --- 1 file changed, 8 insertions(+), 23 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index cfb8a2e4a5..4a0413db16 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -,29 +,18

[libvirt PATCH 19/39] cpu_x86: Use g_auto* in x86Encode

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 51 +++ 1 file changed, 21 insertions(+), 30 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index ce43da2bfa..431fbd8ec6 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@

[libvirt PATCH 11/39] cpu_x86: Use g_auto* in x86ModelFromCPU

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 14 +- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index d2bbcbd891..9455fd1440 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1173,7 +1173,7 @@

[libvirt PATCH 02/39] cpu_x86: Use glib allocation for virCPU{, x86}Data

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 45 ++--- 1 file changed, 22 insertions(+), 23 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index eeb16c2763..f0d9b7056a 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@

[libvirt PATCH 12/39] cpu_x86: Use g_auto* in x86ModelParse

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 23 +-- 1 file changed, 9 insertions(+), 14 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 9455fd1440..fc94c029e4 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1493,41 +1493,36 @@

[libvirt PATCH 09/39] cpu_x86: Use g_auto* in x86VendorParse

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 20 +++- 1 file changed, 7 insertions(+), 13 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index d0ef66f1e3..a678bcde61 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -849,9 +849,8 @@

[libvirt PATCH 01/39] cpu_x86: Drop noTSX hint for incompatible CPUs

The hint was introduced a long time ago when broken TSX implementation was found in Haswell and Broadwell CPUs. Since then many more CPUs with TSX were introduced and and disabled due to TAA vulnerability. Thus the hint is not very useful and I think removing it is a better choice then updating

[libvirt PATCH 20/39] cpu_x86: Use g_auto* in virCPUx86CheckFeature

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 11 +++ 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 431fbd8ec6..3b3a428ecd 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2282,21 +2282,16 @@ static int

[libvirt PATCH 06/39] cpu_x86: Use glib allocation for virCPUx86Map

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 14 +++--- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 4a0413db16..bcb87d6f93 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1566,23 +1566,24 @@

[libvirt PATCH 13/39] cpu_x86: Use g_auto* in virCPUx86LoadMap

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 10 +++--- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index fc94c029e4..ba269df66d 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -1559,18 +1559,14 @@

[libvirt PATCH 15/39] cpu_x86: Use g_auto* in x86Compute

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 64 ++- 1 file changed, 24 insertions(+), 40 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 6c3f9fc0be..6758fcc170 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@

[libvirt PATCH 04/39] cpu_x86: Use glib allocation for virCPUx86Feature

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 22 -- 1 file changed, 4 insertions(+), 18 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 6fad4af922..cfb8a2e4a5 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -893,28 +893,17 @@

[libvirt PATCH 03/39] cpu_x86: Use glib allocation for virCPUx86Vendor

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 9 - 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index f0d9b7056a..6fad4af922 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -830,9 +830,10 @@

[libvirt PATCH 18/39] cpu_x86: Use g_auto* in x86EncodePolicy

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 26d872622e..ce43da2bfa 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2168,7 +2168,7 @@ x86EncodePolicy(virCPUx86Data

[libvirt PATCH 23/39] cpu_x86: Use g_auto* in x86UpdateHostModel

Signed-off-by: Jiri Denemark --- src/cpu/cpu_x86.c | 14 +- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index ccbae3dabd..9f6027ec5c 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2821,17 +2821,16 @@ static int

Re: On the need to move to a merge request workflow

On Thu, Mar 26, 2020 at 15:50:20 +, Daniel Berrange wrote: > On Thu, Mar 26, 2020 at 02:10:59PM +0100, Peter Krempa wrote: > > On Fri, Mar 06, 2020 at 11:44:07 +, Daniel Berrange wrote: > > > We've discussed the idea of replacing our mailing list review workflow > > > with > > > a merge

Re: [libvirt PATCH v3 11/12] gitlab: introduce a check for validate DCO sign-off

On Fri, Mar 27, 2020 at 12:51:23PM +0100, Andrea Bolognani wrote: > On Fri, 2020-03-27 at 10:53 +, Daniel P. Berrangé wrote: > > On Thu, Mar 26, 2020 at 06:06:37PM +0100, Andrea Bolognani wrote: > > > On Thu, 2020-03-26 at 12:35 +, Daniel P. Berrangé wrote: > > > > +if

Re: [libvirt PATCH v3 11/12] gitlab: introduce a check for validate DCO sign-off

On Fri, 2020-03-27 at 10:53 +, Daniel P. Berrangé wrote: > On Thu, Mar 26, 2020 at 06:06:37PM +0100, Andrea Bolognani wrote: > > On Thu, 2020-03-26 at 12:35 +, Daniel P. Berrangé wrote: > > > +if "localhost" in line: > > > +print(" ❌ FAIL: bad email in %s" %

  1   2   >