March 2020 - Devel - libvirt List Archives

[PATCH python] tox: Test with Python 3.6, 3.7 and 3.8
by Radostin Stoyanov 30 Mar '20

30 Mar '20

Support for Python 2.X has been dropped with commit b22e4f2. Signed-off-by: Radostin Stoyanov <rstoyanov(a)fedoraproject.org> --- tox.ini | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tox.ini b/tox.ini index e8c33f1..de683b9 100644 --- a/tox.ini +++ b/tox.ini @@ -1,5 +1,5 @@ [tox] -envlist = py26,py27,py34 +envlist = py36,py37,py38 [testenv] deps= -- 2.25.1

2 1

[libvirt PATCH v3 00/12] gitlab: expand the CI job coverage
by Daniel P. Berrangé 30 Mar '20

30 Mar '20

The main goals with this series are: - Introduce a minimal job building the website and publishing an artifact which can be deployed onto libvirt.org - Introduce a minimal job building the libvirt.pot for import into Weblate (only runs on git master branch) - Introduce a signed-off-by check. We can add a rule to block direct pushes without a SoB, but this will be needed to block merge requests without a SoB. - Expanding CI jobs to get coverage closer to Travis/Jenkins - Reducing cross-build jobs to just interesting variants, since the full set hasn't shown value in detecting failures The Linux native job coverage is now a superset of that achieved by Travis/Jenkins. For post-merge testing the full set of jobs are run on git master. This measured approx 50 minutes total duration. For pre-merge testing the job count is reduced for quicker turnaround time for developers Measured ~35 minutes total duration for a cold cache, vs ~20 minutes for warm cache. Changed in v3: - Add job for validating Signed-off-by - Add job for validating code style - Use ccache for native/cross build jobs - Eliminated the extra build stage and put all native jobs in the same place, but with filters - Keep all cross build jobs, but filter them based on branch Changed in v2: - Add more native test jobs to run on git master - Restrict git clone depth - User variable for "make" command name - Add test job to build the master pot file - Remove extra configure args for website job - Re-ordered patches to reduce repeated changes Daniel P. Berrangé (12): gitlab: add variable for make command name gitlab: restrict git history to 100 commits gitlab: create an explicit stage for cross build jobs gitlab: use CI for building website contents gitlab: reduce number of cross build jobs run by default gitlab: rename the cross build jobs gitlab: add mingw cross build CI jobs gitlab: add x86_64 native CI jobs gitlab: add job for building latest potfile gitlab: introduce use of ccache for speeding up rebuilds gitlab: introduce a check for validate DCO sign-off gitlab: add explicit early job for syntax-check .gitlab-ci.yml | 234 +++++++++++++++++++++++++++++++++++++---- scripts/require-dco.py | 96 +++++++++++++++++ 2 files changed, 307 insertions(+), 23 deletions(-) create mode 100755 scripts/require-dco.py -- 2.24.1

3 39

[libvirt PATCH] tools: explain that '^' means 'Ctrl' for console escape sequence
by Daniel P. Berrangé 30 Mar '20

30 Mar '20

Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- 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 @@ cmdRunConsole(vshControl *ctl, virDomainPtr dom, } vshPrintExtra(ctl, _("Connected to domain %s\n"), virDomainGetName(dom)); - vshPrintExtra(ctl, _("Escape character is %s\n"), priv->escapeChar); + vshPrintExtra(ctl, _("Escape character is %s"), priv->escapeChar); + if (priv->escapeChar[0] == '^') { + vshPrintExtra(ctl, " (Ctrl + %c)", priv->escapeChar[1]); + } + vshPrintExtra(ctl, "\n"); fflush(stdout); if (virshRunConsole(ctl, dom, name, flags) == 0) return true; -- 2.24.1

3 5

[libvirt PATCH] docs: Clarify semantics of model/@usable attribute in dom caps
by Jiri Denemark 30 Mar '20

30 Mar '20

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 <jdenemar(a)redhat.com> --- docs/formatdomaincaps.html.in | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/docs/formatdomaincaps.html.in b/docs/formatdomaincaps.html.in index 66e758501b..4056e0bd31 100644 --- a/docs/formatdomaincaps.html.in +++ b/docs/formatdomaincaps.html.in @@ -249,9 +249,11 @@ The <code>mode</code> element contains a list of supported CPU models, each described by a dedicated <code>model</code> element. The <code>usable</code> attribute specifies whether the model can - be used on the host. A special value <code>unknown</code> indicates - libvirt does not have enough information to provide the usability - data. + be used directly on the host. When usable='no' the corresponding model + cannot be used without disabling some features that the CPU of such + model is expected to have. A special value <code>unknown</code> + indicates libvirt does not have enough information to provide the + usability data. </dd> </dl> -- 2.26.0

2 1

[PATCH] util: virdaemon: fix compilation on mingw
by Rafael Fonseca 30 Mar '20

30 Mar '20

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 code more portable by using existing vir* wrappers. Signed-off-by: Rafael Fonseca <r4f4rfs(a)gmail.com> --- src/util/virdaemon.c | 46 +++++++++++++++++++++++++++++++++++++++----- 1 file changed, 41 insertions(+), 5 deletions(-) diff --git a/src/util/virdaemon.c b/src/util/virdaemon.c index 789adc6c50..5d92c7def7 100644 --- a/src/util/virdaemon.c +++ b/src/util/virdaemon.c @@ -20,9 +20,7 @@ #include <config.h> -#include <sys/types.h> #include <sys/stat.h> -#include <sys/wait.h> #include <fcntl.h> #include <unistd.h> #include <stdbool.h> @@ -32,9 +30,13 @@ #include "virfile.h" #include "virlog.h" #include "viralloc.h" +#include "virprocess.h" +#include "vircommand.h" #include "configmake.h" +#ifndef WIN32 + int virDaemonForkIntoBackground(const char *argv0) { @@ -42,7 +44,7 @@ virDaemonForkIntoBackground(const char *argv0) if (virPipeQuiet(statuspipe) < 0) return -1; - pid_t pid = fork(); + pid_t pid = virFork(); switch (pid) { case 0: { @@ -71,7 +73,7 @@ virDaemonForkIntoBackground(const char *argv0) if (setsid() < 0) goto cleanup; - nextpid = fork(); + nextpid = virFork(); switch (nextpid) { case 0: /* grandchild */ return statuspipe[1]; @@ -102,7 +104,7 @@ virDaemonForkIntoBackground(const char *argv0) VIR_FORCE_CLOSE(statuspipe[1]); /* We wait to make sure the first child forked successfully */ - if ((got = waitpid(pid, &exitstatus, 0)) < 0 || + if ((got = virProcessWait(pid, &exitstatus, 0)) < 0 || got != pid || exitstatus != 0) { goto error; @@ -250,3 +252,37 @@ virDaemonUnixSocketPaths(const char *sock_prefix, VIR_FREE(rundir); return ret; } + +#else /* WIN32 */ + +int virDaemonForkIntoBackground(const char *argv0 G_GNUC_UNUSED) +{ + errno = ENOTSUP; + return -1; +} + +void virDaemonSetupLogging(const char *daemon_name G_GNUC_UNUSED, + unsigned int log_level G_GNUC_UNUSED, + char *log_filters G_GNUC_UNUSED, + char *log_outputs G_GNUC_UNUSED, + bool privileged G_GNUC_UNUSED, + bool verbose G_GNUC_UNUSED, + bool godaemon G_GNUC_UNUSED) +{ + /* NOOP */ + errno = ENOTSUP; + return; +} + +int virDaemonUnixSocketPaths(const char *sock_prefix G_GNUC_UNUSED, + bool privileged G_GNUC_UNUSED, + char *unix_sock_dir G_GNUC_UNUSED, + char **sockfile G_GNUC_UNUSED, + char **rosockfile G_GNUC_UNUSED, + char **adminSockfile G_GNUC_UNUSED) +{ + errno = ENOTSUP; + return -1; +} + +#endif /* WIN32 */ -- 2.25.1

3 2

[libvirt PATCH v2 0/5] QEMU: support QEMU built without TCG
by Tobin Feldman-Fitzthum 27 Mar '20

27 Mar '20

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 is set only when the TCG is not available. This avoids some issues with backwards compatability. For a domain that was created using QEMU >= 2.10 with KVM, there is no information in the cached XML file that says whether or not TCG was enabled. Versions of QEMU older than 2.10 do not have a QMP interface for determining whether QEMU is available. Since QEMU_CAPS_TCG_DISABLED is set only when TCG is disabled, we do not have to do any extra work to infer an appropriate value in either of these cases. QEMU_CAPS_TCG_DISABLED is set via QMP command qom-list-types. Tobin Feldman-Fitzthum (5): add QEMU_CAPS_TCG_DISABLED and probe conditionally add virQEMUCapsGetVirtType convenience function add virQEMUCapsProbeQMPTCGState function to set QEMU_CAPS_TCG_DISABLED use new functions when probing Add second qom-list-types call to test data. src/qemu/qemu_capabilities.c | 81 +- src/qemu/qemu_capabilities.h | 3 + .../caps_2.10.0.aarch64.replies | 2695 ++- .../caps_2.10.0.ppc64.replies | 2799 ++- .../caps_2.10.0.s390x.replies | 1027 +- .../caps_2.10.0.x86_64.replies | 1702 +- .../caps_2.11.0.s390x.replies | 1063 +- .../caps_2.11.0.x86_64.replies | 1688 +- .../caps_2.12.0.aarch64.replies | 2960 +++- .../caps_2.12.0.ppc64.replies | 2947 +++- .../caps_2.12.0.s390x.replies | 1087 +- .../caps_2.12.0.x86_64.replies | 1827 +- .../caps_3.0.0.ppc64.replies | 2979 +++- .../caps_3.0.0.s390x.replies | 1117 +- .../caps_3.0.0.x86_64.replies | 1843 +- .../caps_3.1.0.ppc64.replies | 2999 +++- .../caps_3.1.0.x86_64.replies | 1863 +- .../caps_4.0.0.aarch64.replies | 3223 +++- .../caps_4.0.0.ppc64.replies | 3211 +++- .../caps_4.0.0.s390x.replies | 1259 +- .../caps_4.0.0.x86_64.replies | 1975 ++- .../caps_4.1.0.x86_64.replies | 2207 ++- .../caps_4.2.0.aarch64.replies | 3409 +++- .../caps_4.2.0.ppc64.replies | 3248 +++- .../caps_4.2.0.s390x.replies | 1259 +- .../caps_4.2.0.x86_64.replies | 2296 ++- .../caps_5.0.0.aarch64.replies | 3445 +++- .../caps_5.0.0.ppc64.replies | 14002 +++++++++------- .../qemucapabilitiesdata/caps_5.0.0.ppc64.xml | 1220 +- .../caps_5.0.0.x86_64.replies | 2319 ++- 30 files changed, 65236 insertions(+), 8517 deletions(-) -- 2.20.1 (Apple Git-117)

1 5

Entering freeze for libvirt-6.2.0
by Daniel Veillard 27 Mar '20

27 Mar '20

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 place: https://libvirt.org/sources/ the rc1 is also tagged in git. Please give it a try, so we can have a normal release, in those time of pandemy, take care, stay safe, thanks ! Daniel -- Daniel Veillard | Red Hat Developers Tools http://developer.redhat.com/ veillard(a)redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt PATCH 0/5] QEMU: support QEMU built without TCG
by Tobin Feldman-Fitzthum 27 Mar '20

27 Mar '20

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 available. This avoids some issues with backwards compatability. For a domain that was created using QEMU >= 2.10 with KVM, there is no information in the cached XML file that says whether or not TCG was enabled. Versions of QEMU older than 2.10 do not have a QMP interface for determining whether QEMU is available. Since QEMU_CAPS_TCG_DISABLED is set only when TCG is disabled, we do not have to do any extra work to infer an appropriate value in either of these cases. QEMU_CAPS_TCG_DISABLED is set via QMP command qom-list-types. Tobin Feldman-Fitzthum (5): add QEMU_CAPS_TCG_DISABLED and probe conditionally add virQEMUCapsGetVirtType convenience function add virQEMUCapsProbeQMPTCGState function to set QEMU_CAPS_TCG_DISABLED use new functions when probing Add second qom-list-types call to test data. src/qemu/qemu_capabilities.c | 80 +- src/qemu/qemu_capabilities.h | 3 + .../caps_2.10.0.aarch64.replies | 2695 ++- .../caps_2.10.0.ppc64.replies | 2799 ++- .../caps_2.10.0.s390x.replies | 1027 +- .../caps_2.10.0.x86_64.replies | 1702 +- .../caps_2.11.0.s390x.replies | 1063 +- .../caps_2.11.0.x86_64.replies | 1688 +- .../caps_2.12.0.aarch64.replies | 2960 +++- .../caps_2.12.0.ppc64.replies | 2947 +++- .../caps_2.12.0.s390x.replies | 1087 +- .../caps_2.12.0.x86_64.replies | 1827 +- .../caps_3.0.0.ppc64.replies | 2979 +++- .../caps_3.0.0.s390x.replies | 1117 +- .../caps_3.0.0.x86_64.replies | 1843 +- .../caps_3.1.0.ppc64.replies | 2999 +++- .../caps_3.1.0.x86_64.replies | 1863 +- .../caps_4.0.0.aarch64.replies | 3223 +++- .../caps_4.0.0.ppc64.replies | 3211 +++- .../caps_4.0.0.s390x.replies | 1259 +- .../caps_4.0.0.x86_64.replies | 1975 ++- .../caps_4.1.0.x86_64.replies | 2207 ++- .../caps_4.2.0.aarch64.replies | 3409 +++- .../caps_4.2.0.ppc64.replies | 3248 +++- .../caps_4.2.0.s390x.replies | 1259 +- .../caps_4.2.0.x86_64.replies | 2296 ++- .../caps_5.0.0.aarch64.replies | 3445 +++- .../caps_5.0.0.ppc64.replies | 14002 +++++++++------- .../qemucapabilitiesdata/caps_5.0.0.ppc64.xml | 1220 +- .../caps_5.0.0.x86_64.replies | 2319 ++- 30 files changed, 65235 insertions(+), 8517 deletions(-) -- 2.20.1 (Apple Git-117)

1 5

[libvirt PATCH 0/2] gitlab: Improve make handling
by Andrea Bolognani 27 Mar '20

27 Mar '20

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

2 4

[PATCH 0/2] Two trivial syntax-check fixes
by Michal Privoznik 27 Mar '20

27 Mar '20

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 syntax-check.mk: Drop needless _src2 variable build-aux/syntax-check.mk | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) -- 2.24.1

2 4