- Devel - libvirt List Archives

[libvirt] [dbus PATCH] connect: remove unused domain variables
by Ján Tomko 18 Apr '18

18 Apr '18

Neither virDomainRestoreFlags nor virDomainSaveImageDefineXML take a virDomain argument. Remove the unused domain variables. Introduced by commits 1ce556b and 2597d41 respectively. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- src/connect.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/src/connect.c b/src/connect.c index 8faaeaf..3a563bb 100644 --- a/src/connect.c +++ b/src/connect.c @@ -357,7 +357,6 @@ virtDBusConnectDomainRestoreFlags(GVariant *inArgs, GError **error) { virtDBusConnect *connect = userData; - g_autoptr(virDomain) domain = NULL; const gchar *from; const gchar *xml; guint flags; @@ -383,7 +382,6 @@ virtDBusConnectDomainSaveImageDefineXML(GVariant *inArgs, GError **error) { virtDBusConnect *connect = userData; - g_autoptr(virDomain) domain = NULL; const gchar *file; const gchar *xml; guint flags; -- 2.16.1

2 1

[libvirt] [PATCH 0/6] Various disk-related fixes/cleanups
by Peter Krempa 18 Apr '18

18 Apr '18

noticed while doing blockdev work but didn't turn out to be relevant. Peter Krempa (6): conf: Make argument of virDomainDiskGetDriver const qemu: Move disk driver name validation into disk validation callback conf, qemu: Replace access to disk driver name with accessors conf: Punt 'driverName' back to disk definition structure conf: Disk 'shared' state is not guest ABI qemu: driver: Fix error message in qemuDomainBlockCommit src/conf/domain_conf.c | 20 ++++++++++---------- src/conf/domain_conf.h | 4 +++- src/qemu/qemu_command.c | 12 ------------ src/qemu/qemu_domain.c | 9 +++++++++ src/qemu/qemu_driver.c | 2 +- src/qemu/qemu_hotplug.c | 7 ------- src/qemu/qemu_parse_command.c | 2 +- src/util/virstoragefile.c | 6 ------ src/util/virstoragefile.h | 1 - 9 files changed, 24 insertions(+), 39 deletions(-) -- 2.16.2

2 17

[libvirt] [PATCH] conf: Reindent virDomainObjCheckActive
by Jiri Denemark 18 Apr '18

18 Apr '18

Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- Pushed as trivial. src/conf/domain_conf.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index ae2135263e..5f1af911b7 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -6001,9 +6001,9 @@ int virDomainObjCheckActive(virDomainObjPtr dom) { if (!virDomainObjIsActive(dom)) { - virReportError(VIR_ERR_OPERATION_INVALID, - "%s", _("domain is not running")); - return -1; + virReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is not running")); + return -1; } return 0; } -- 2.17.0

1 0

[libvirt] [PATCH v2 0/2] qemu: cpu: fix "full" CPU to include all "reported" CPU features
by Nikolay Shirokovskiy 18 Apr '18

18 Apr '18

diff from v1: - add test data for CPU for which issue was triggered Patch 1 adds tests for CPU on which problem was detected. However we don't test <cpu mode='host-model' check='partial'> case which have issue. Patch 2 is actual fix. Nikolay Shirokovskiy (2): cputest: New test for Intel Core i5-650 qemu: cpu: fix "full" CPU to include all "reported" CPU features src/qemu/qemu_capabilities.c | 12 +- tests/cputest.c | 1 + .../x86_64-cpuid-Core-i5-650-disabled.xml | 5 + .../x86_64-cpuid-Core-i5-650-enabled.xml | 7 + .../cputestdata/x86_64-cpuid-Core-i5-650-guest.xml | 24 + .../cputestdata/x86_64-cpuid-Core-i5-650-host.xml | 25 + .../cputestdata/x86_64-cpuid-Core-i5-650-json.xml | 12 + tests/cputestdata/x86_64-cpuid-Core-i5-650.json | 1068 ++++++++++++++++++++ tests/cputestdata/x86_64-cpuid-Core-i5-650.xml | 30 + 9 files changed, 1181 insertions(+), 3 deletions(-) create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650-disabled.xml create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650-enabled.xml create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650-guest.xml create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650-host.xml create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650-json.xml create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650.json create mode 100644 tests/cputestdata/x86_64-cpuid-Core-i5-650.xml -- 1.8.3.1

2 5

[libvirt] [PATCH] qemuDomainObjPrivateDataClear: Don't leak @migParams
by Michal Privoznik 18 Apr '18

18 Apr '18

Allocated in qemuMigrationParamsNew() we need to free priv->job.migParams when no longer needed. ==8061== 234 (192 direct, 42 indirect) bytes in 1 blocks are definitely lost in loss record 112 of 123 ==8061== at 0x4C2CF26: calloc (vg_replace_malloc.c:711) ==8061== by 0x5325D05: virAlloc (viralloc.c:144) ==8061== by 0x1984F9: qemuMigrationParamsNew (qemu_migration_params.c:218) ==8061== by 0x19A352: qemuMigrationParamsParse (qemu_migration_params.c:1185) ==8061== by 0x1604D8: qemuDomainObjPrivateXMLParseJob (qemu_domain.c:2390) ==8061== by 0x160AE9: qemuDomainObjPrivateXMLParse (qemu_domain.c:2517) ==8061== by 0x5419EAE: virDomainObjParseXML (domain_conf.c:20442) ==8061== by 0x541A25E: virDomainObjParseNode (domain_conf.c:20555) ==8061== by 0x541A2FC: virDomainObjParseFile (domain_conf.c:20574) ==8061== by 0x13607D: testCompareStatusXMLToXMLFiles (qemuxml2xmltest.c:75) ==8061== by 0x14F3E8: virTestRun (testutils.c:180) ==8061== by 0x14DCD0: mymain (qemuxml2xmltest.c:1200) Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_domain.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 672f08b442..47595e244a 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -1895,6 +1895,8 @@ qemuDomainObjPrivateDataClear(qemuDomainObjPrivatePtr priv) virBitmapFree(priv->migrationCaps); priv->migrationCaps = NULL; + + qemuMigrationParamsFree(priv->job.migParams); } -- 2.16.1

2 1

[libvirt] [PATCH] news: Xen: announce support for setting CPU features
by Jim Fehlig 18 Apr '18

18 Apr '18

Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- docs/news.xml | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/docs/news.xml b/docs/news.xml index a5c489151..dec3f134c 100644 --- a/docs/news.xml +++ b/docs/news.xml @@ -44,6 +44,18 @@ add this controller when traditional PCI devices are in use. </description> </change> + <change> + <summary> + Xen: Support setting CPU features for host-passthrough model + </summary> + <description> + The CPU model presented to Xen HVM domains is equivalent to libvirt's + host-passthrough model, although individual features can be enabled + and disabled via the cpuid setting. The libvirt libxl driver now + supports enabling and disabling individual features of the + host-passthrough CPU model. + </description> + </change> </section> <section title="Removed features"> <change> -- 2.16.3

2 1

[libvirt] [PATCH v2 0/9] Add function that raises error if domain is not active
by Clementine Hayat 18 Apr '18

18 Apr '18

This is my GSOC patch contribution. This change was suggested on BiteSizedTasks in the libvirt wiki[1]. in libvirt there is lots of occurences of this same pattern: if (!virDomainObjIsActive(vm)) { virReportError(VIR_ERR_OPERATION_INVALID, "%s", _("domain is not running")); goto out; } This series replace these calls with a new function that check if the domain is active and log directly the error. This allows to remove almost 300 lines of code in the code base. [1] https://wiki.libvirt.org/page/BiteSizedTasks#Add_function_that_raises_error… Changes since v2: * renamed virDomainObjCheckIsActive into virDomainObjCheckActive * add the remaining occurences Clementine Hayat (9): Add function that raises error if domain is not active qemu: start using virDomainObjCheckActive test: start using virDomainObjCheckActive libxl: start using virDomainObjCheckActive bhyve: start using virDomainObjCheckActive lxc: start using virDomainObjCheckActive openvz: start using virDomainObjCheckActive uml: start using virDomainObjCheckActive vz: start using virDomainObjCheckActive src/bhyve/bhyve_driver.c | 20 +-- src/conf/domain_conf.c | 11 ++ src/conf/domain_conf.h | 2 + src/libvirt_private.syms | 1 + src/libxl/libxl_driver.c | 97 +++---------- src/lxc/lxc_driver.c | 60 ++------ src/openvz/openvz_driver.c | 20 +-- src/qemu/qemu_domain.c | 5 +- src/qemu/qemu_driver.c | 271 ++++++++----------------------------- src/test/test_driver.c | 35 +---- src/uml/uml_driver.c | 5 +- src/vz/vz_driver.c | 5 +- 12 files changed, 120 insertions(+), 412 deletions(-) -- 2.17.0

2 19

[libvirt] [PATCH v3 0/5] virClassNew rework
by Michal Privoznik 18 Apr '18

18 Apr '18

2 12

[libvirt] [PATCH 0/2] Two patches to fix the build
by Michal Privoznik 18 Apr '18

18 Apr '18

After 446d09149802677 the build was broken. I've pushed these under build breaker rule. And they're trivial too. Michal Privoznik (2): virmocklibxl.c: Fix include of virfile.h tests: Compile and link virmocklibxl with libxml tests/Makefile.am | 2 +- tests/virmocklibxl.c | 3 ++- 2 files changed, 3 insertions(+), 2 deletions(-) -- 2.16.1

1 2

[libvirt] [PATCH 0/7] Addressing some Coverity complaints...
by John Ferlan 18 Apr '18

18 Apr '18

My current pile is 48 patches - most of those are known false positives or have been posted before and either rejected or ignored. In any case, the following is a list of more recent adjustments from my coverity tree. Seems we've largely moved away from resource leaks and into return value checking issues. Some good, some not so good. John Ferlan (7): conf: Add error checking to virDomainSnapshotDiskDefFormat remote: Fix usage of ATTRIBUTE_FALLTHROUGH tests: Return failure if log not fopen'd tests: Don't call virNetServerClientClose without valid client tests: Add checks for possible errors qemu: Fix possible memory leak in migration param processing qemu: Check for missing 'return' in qemuMonitorJSONCheckReply src/conf/snapshot_conf.c | 16 ++++++++++------ src/qemu/qemu_migration_params.c | 1 + src/qemu/qemu_monitor_json.c | 7 ++++++- src/remote/remote_driver.c | 2 +- tests/commandhelper.c | 2 +- tests/testutilsqemuschema.c | 40 +++++++++++++++++++++++++++++----------- tests/virnetserverclienttest.c | 3 ++- 7 files changed, 50 insertions(+), 21 deletions(-) -- 2.13.6

2 14

[libvirt] [PATCH v7 0/9] Add setting CPU features (CPUID) with libxenlight driver.
by Marek Marczykowski-Górecki 18 Apr '18

18 Apr '18

Add support for CPUID setting based on <cpu> element. Since libxl format support only adjusting specific bits over host CPU, only mode='host-passthrough' is supported - other values are rejected (including default 'custom'). This will break some configurations working before (bare <cpu> element with for example NUMA configuration), but libxl driver never supported full 'custom' mode - it was silently ignored, which might lead to some unexpected effects. Since mode='host-passthrough' is now necessary to specify CPU options, do not enable nested HVM feature by mere presence of this element, require also enabling it in libxl.conf. Nested HVM is still in "preview" state, so better be explicit here. v2 of this patch series: https://www.redhat.com/archives/libvir-list/2017-July/msg00050.html v3 of this patch series: https://www.redhat.com/archives/libvir-list/2017-December/msg00314.html v4 of this patch series: https://www.redhat.com/archives/libvir-list/2018-February/msg00504.html v5 of this patch series: https://www.redhat.com/archives/libvir-list/2018-March/msg00796.html v6 of this patch series: https://www.redhat.com/archives/libvir-list/2018-March/msg01310.html Marek Marczykowski-Górecki (9): libxl: fix libxlDriverConfigDispose for partially constructed object libxl: pass driver config to libxlMakeDomBuildInfo libxl: warn about ignored CPU mode=custom libxl: do not enable nested HVM unless global nested_hvm option enabled xenconfig: do not override def->cpu if already set elsewhere libxl: add support for CPUID features policy tests: check CPU features handling in libxl driver xenconfig: add CPUID handling to domXML <-> xl.cfg conversion tests: add test case for CPUID in xenconfig driver src/libxl/libvirtd_libxl.aug | 2 +- src/libxl/libxl.conf | 8 +- src/libxl/libxl_conf.c | 66 +++- src/libxl/libxl_conf.h | 6 +- src/libxl/libxl_domain.c | 2 +- src/libxl/test_libvirtd_libxl.aug.in | 1 +- src/xenconfig/xen_xl.c | 236 ++++++++++++++-- src/xenconfig/xen_xl.h | 2 +- tests/libxlxml2domconfigdata/fullvirt-cpuid.json | 60 ++++- tests/libxlxml2domconfigdata/fullvirt-cpuid.xml | 34 ++- tests/libxlxml2domconfigtest.c | 27 +- tests/virmocklibxl.c | 31 ++- tests/xlconfigdata/test-fullvirt-cpuid.cfg | 25 ++- tests/xlconfigdata/test-fullvirt-cpuid.xml | 36 ++- tests/xlconfigtest.c | 1 +- 15 files changed, 492 insertions(+), 45 deletions(-) create mode 100644 tests/libxlxml2domconfigdata/fullvirt-cpuid.json create mode 100644 tests/libxlxml2domconfigdata/fullvirt-cpuid.xml create mode 100644 tests/xlconfigdata/test-fullvirt-cpuid.cfg create mode 100644 tests/xlconfigdata/test-fullvirt-cpuid.xml base-commit: dffe584aa4194b0667924632e9e1ae12c5520956 -- git-series 0.9.1

2 15

[libvirt] [PATCH v2 0/2] Fix vmware driver for VMware Fusion 10
by Rainer Müller 17 Apr '18

17 Apr '18

These changes were required to get VMware Fusion working at all. I tested with VMware Fusion 10.1.1 on macOS 10.12 Sierra. I am not sure whether calling virCapabilitiesInitCaches() makes sense at all on macOS. This function looks highly specific to Linux as it relies on /sys/devices/system, so it should probably be wrapped with an appropriate #ifdef. I do not feel proficient enough with the libvirt internals to make this change, though. v1 was here: https://www.redhat.com/archives/libvir-list/2018-April/msg00087.html diff to v1: - Fix a possible segfault when vmrun is not found in PATH. Rainer Müller (2): vmware: Fix initialization of VMware Fusion vmware: Failures in cache info init are non-fatal src/vmware/vmware_conf.c | 7 ++++++- src/vmware/vmware_driver.c | 11 ++++++++++- 2 files changed, 16 insertions(+), 2 deletions(-) -- 2.17.0

2 3

[libvirt] [jenkins-ci PATCH v3 0/3] Mostly drop CentOS 6
by Andrea Bolognani 17 Apr '18

17 Apr '18

This applies on top of https://www.redhat.com/archives/libvir-list/2018-April/msg01526.html Changes from [v2]: * several patches have been pushed already; * define the libvirt-master-build-website job using the newly variants-enabled generic-build-job template; * add 'mingw_machines' variable along with 'all_machines' and 'rpm_machines'. Changes from [v1]: * instead of dropping CentOS 6 altogether, simply stop building pretty much all projects on it; * introduce a new job type specifically for whatever little stuff we need to ensure still works on CentOS 6; * the end result is still a nice cleanup thanks to not repeating the list of machines all over the place; moreover, the message about our support of the OS is very clear because it disappears from all but one page on the CI web interface. [v1] https://www.redhat.com/archives/libvir-list/2018-April/msg00453.html [v2] https://www.redhat.com/archives/libvir-list/2018-April/msg01051.html Andrea Bolognani (3): jobs: Introduce machine groups guests: Introduce libvirt+website project projects: Add libvirt-master-build-website job guests/host_vars/libvirt-centos-6/main.yml | 3 +++ guests/vars/projects/libvirt+website.yml | 5 ++++ jobs/defaults.yaml | 16 ++++++++++++ projects/libosinfo.yaml | 16 ++---------- projects/libvirt-cim.yaml | 6 +---- projects/libvirt-dbus.yaml | 6 +---- projects/libvirt-glib.yaml | 22 +++------------- projects/libvirt-go-xml.yaml | 10 +------ projects/libvirt-go.yaml | 10 +------ projects/libvirt-perl.yaml | 16 ++---------- projects/libvirt-python.yaml | 16 ++---------- projects/libvirt.yaml | 42 ++++++++++++++++-------------- projects/osinfo-db-tools.yaml | 16 ++---------- projects/osinfo-db.yaml | 16 ++---------- projects/virt-viewer.yaml | 22 +++------------- 15 files changed, 69 insertions(+), 153 deletions(-) create mode 100644 guests/vars/projects/libvirt+website.yml -- 2.14.3

2 5

[libvirt] [jenkins-ci PATCH 0/3] Introduce variants and use them for MinGW
by Andrea Bolognani 17 Apr '18

17 Apr '18

Andrea Bolognani (3): jobs: Introduce variants jobs: Tweak autotools-mingw-job template jobs: Drop autotools-mingw-job jobs/autotools.yaml | 94 ++++------------------------------------------ jobs/defaults.yaml | 17 +++++++++ jobs/generic.yaml | 16 ++++---- jobs/go.yaml | 8 ++-- jobs/perl-makemaker.yaml | 12 +++--- jobs/perl-modulebuild.yaml | 12 +++--- jobs/python-distutils.yaml | 12 +++--- projects/libvirt-glib.yaml | 12 +++++- projects/libvirt.yaml | 12 +++++- projects/virt-viewer.yaml | 12 +++++- 10 files changed, 88 insertions(+), 119 deletions(-) -- 2.14.3

2 6

[libvirt] [PATCH v2 0/5] virClassNew rework
by Michal Privoznik 17 Apr '18

17 Apr '18

2 9

[libvirt] [PATCH] domain: avoid slash characters to the new domain name.
by Julio Faracco 17 Apr '18

17 Apr '18

The 'domrename' command needs to check if the new domain name contains the slash character. This character is not accepted by libvirt XML definition because it is an invalid char (see Cole's commit b1fc6a7b7). This commit enhace the 'domrename' command adding this check. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1333232 Signed-off-by: Julio Faracco <jcfaracco(a)gmail.com> --- src/internal.h | 9 +++++++++ src/libvirt-domain.c | 1 + 2 files changed, 10 insertions(+) diff --git a/src/internal.h b/src/internal.h index 1760e3b69..e976b1caa 100644 --- a/src/internal.h +++ b/src/internal.h @@ -484,6 +484,15 @@ goto label; \ } \ } while (0) +# define virCheckNonSlashGoto(argname, label) \ + do { \ + if (strchr(argname, '/')) { \ + virReportInvalidArg(ctl, \ + _("name %s cannot contain '/'"), \ + argname); \ + goto label; \ + } \ + } while (0) diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c index 63d2ae23d..47bc59b11 100644 --- a/src/libvirt-domain.c +++ b/src/libvirt-domain.c @@ -8536,6 +8536,7 @@ virDomainRename(virDomainPtr dom, virResetLastError(); virCheckDomainReturn(dom, -1); virCheckNonEmptyStringArgGoto(new_name, error); + virCheckNonSlashGoto(new_name, error); virCheckReadOnlyGoto(dom->conn->flags, error); if (dom->conn->driver->domainRename) { -- 2.14.1

2 2

[libvirt] [PATCHv2 0/4] qemu: enable sandbox whitelist by default
by Ján Tomko 17 Apr '18

17 Apr '18

v1: https://www.redhat.com/archives/libvir-list/2018-March/msg01965.html https://bugzilla.redhat.com/show_bug.cgi?id=1492597 v2: * also deny resource control * split out and refactor the command line building * be explicit about denying the obsolete syscalls Ján Tomko (4): Introduce QEMU_CAPS_SECCOMP_BLACKLIST Introduce qemuBuildSeccompSandboxCommandLine Refactor qemuBuildSeccompSandboxCommandLine qemu: deny privilege elevation and spawn in seccomp src/qemu/qemu.conf | 7 ++-- src/qemu/qemu_capabilities.c | 2 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 46 +++++++++++++++++----- tests/qemucapabilitiesdata/caps_2.11.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 1 + tests/qemuxml2argvdata/minimal-sandbox.args | 29 ++++++++++++++ tests/qemuxml2argvdata/minimal-sandbox.xml | 34 ++++++++++++++++ tests/qemuxml2argvtest.c | 11 ++++++ 12 files changed, 123 insertions(+), 12 deletions(-) create mode 100644 tests/qemuxml2argvdata/minimal-sandbox.args create mode 100644 tests/qemuxml2argvdata/minimal-sandbox.xml -- 2.16.1

3 14

[libvirt] [jenkins-ci PATCH v2 0/6] Mostly drop CentOS 6
by Andrea Bolognani 17 Apr '18

17 Apr '18

Changes from [v1]: * instead of dropping CentOS 6 altogether, simply stop building pretty much all projects on it; * introduce a new job type specifically for whatever little stuff we need to ensure still works on CentOS 6; * the end result is still a nice cleanup thanks to not repeating the list of machines all over the place; moreover, the message about our support of the OS is very clear because it disappears from all but one page on the CI web interface. [v1] https://www.redhat.com/archives/libvir-list/2018-April/msg00453.html Andrea Bolognani (6): jobs: Introduce autotools-website-job guests: Introduce libvirt+website project projects: Add libvirt-master-website job projects: Stop building on CentOS 6 guests: Drop projects from CentOS 6 jobs: Introduce {all_machines} and {rpm_machines} guests/host_vars/libvirt-centos-6/main.yml | 5 +-- guests/vars/projects/libvirt+website.yml | 5 +++ jobs/autotools.yaml | 67 ++++++++++++++++++++++++++++++ jobs/defaults.yaml | 14 +++++++ projects/libosinfo.yaml | 16 +------ projects/libvirt-cim.yaml | 7 +--- projects/libvirt-glib.yaml | 16 +------ projects/libvirt-go-xml.yaml | 10 +---- projects/libvirt-go.yaml | 10 +---- projects/libvirt-perl.yaml | 17 +------- projects/libvirt-python.yaml | 18 +------- projects/libvirt.yaml | 34 +++++++-------- projects/osinfo-db-tools.yaml | 16 +------ projects/osinfo-db.yaml | 16 +------ projects/virt-viewer.yaml | 16 +------ 15 files changed, 119 insertions(+), 148 deletions(-) create mode 100644 guests/vars/projects/libvirt+website.yml -- 2.14.3

2 26

[libvirt] [PATCH v2 0/5] Enable memory-backend-file.discard-data
by Michal Privoznik 17 Apr '18

17 Apr '18

diff to v1: - Merged two patches into one (5/7 and 6/7 from the original series) - reworked last patch - 1/7 from the original series is pushed now Michal Privoznik (5): qemu_capabilities: s/ObjectProps/DeviceProps/g qemu_capabilities: Separate out device props fetching qemu_monitor: Introduce qemuMonitorGetObjectProps qemu_capabilities: Introduce QEMU_CAPS_OBJECT_MEMORY_FILE_DISCARD qemu: Enable memory-backend-file.discard-data whenever possible src/qemu/qemu_capabilities.c | 267 +++++---- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 7 + src/qemu/qemu_monitor.c | 13 + src/qemu/qemu_monitor.h | 3 + src/qemu/qemu_monitor_json.c | 62 ++- src/qemu/qemu_monitor_json.h | 4 + .../qemucapabilitiesdata/caps_2.1.1.x86_64.replies | 20 +- tests/qemucapabilitiesdata/caps_2.1.1.x86_64.xml | 2 +- .../caps_2.10.0.aarch64.replies | 30 +- tests/qemucapabilitiesdata/caps_2.10.0.aarch64.xml | 2 +- .../qemucapabilitiesdata/caps_2.10.0.ppc64.replies | 28 +- tests/qemucapabilitiesdata/caps_2.10.0.ppc64.xml | 2 +- .../qemucapabilitiesdata/caps_2.10.0.s390x.replies | 32 +- tests/qemucapabilitiesdata/caps_2.10.0.s390x.xml | 2 +- .../caps_2.10.0.x86_64.replies | 540 +++++++++--------- tests/qemucapabilitiesdata/caps_2.10.0.x86_64.xml | 2 +- .../qemucapabilitiesdata/caps_2.11.0.s390x.replies | 26 +- tests/qemucapabilitiesdata/caps_2.11.0.s390x.xml | 2 +- .../caps_2.12.0.aarch64.replies | 24 +- tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 2 +- .../qemucapabilitiesdata/caps_2.12.0.ppc64.replies | 22 +- tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 2 +- .../qemucapabilitiesdata/caps_2.12.0.s390x.replies | 24 +- .../caps_2.12.0.x86_64.replies | 612 +++++++++++---------- tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 3 +- .../qemucapabilitiesdata/caps_2.4.0.x86_64.replies | 20 +- tests/qemucapabilitiesdata/caps_2.4.0.x86_64.xml | 2 +- .../qemucapabilitiesdata/caps_2.5.0.x86_64.replies | 22 +- tests/qemucapabilitiesdata/caps_2.5.0.x86_64.xml | 2 +- .../caps_2.6.0.aarch64.replies | 30 +- tests/qemucapabilitiesdata/caps_2.6.0.aarch64.xml | 2 +- .../qemucapabilitiesdata/caps_2.6.0.ppc64.replies | 28 +- tests/qemucapabilitiesdata/caps_2.6.0.ppc64.xml | 2 +- .../qemucapabilitiesdata/caps_2.6.0.x86_64.replies | 22 +- tests/qemucapabilitiesdata/caps_2.6.0.x86_64.xml | 2 +- .../qemucapabilitiesdata/caps_2.7.0.s390x.replies | 28 +- tests/qemucapabilitiesdata/caps_2.7.0.s390x.xml | 2 +- .../qemucapabilitiesdata/caps_2.7.0.x86_64.replies | 22 +- tests/qemucapabilitiesdata/caps_2.7.0.x86_64.xml | 2 +- .../qemucapabilitiesdata/caps_2.8.0.s390x.replies | 32 +- tests/qemucapabilitiesdata/caps_2.8.0.s390x.xml | 2 +- .../qemucapabilitiesdata/caps_2.8.0.x86_64.replies | 22 +- tests/qemucapabilitiesdata/caps_2.8.0.x86_64.xml | 2 +- .../qemucapabilitiesdata/caps_2.9.0.ppc64.replies | 28 +- tests/qemucapabilitiesdata/caps_2.9.0.ppc64.xml | 2 +- .../qemucapabilitiesdata/caps_2.9.0.s390x.replies | 32 +- tests/qemucapabilitiesdata/caps_2.9.0.s390x.xml | 2 +- .../qemucapabilitiesdata/caps_2.9.0.x86_64.replies | 536 +++++++++--------- tests/qemucapabilitiesdata/caps_2.9.0.x86_64.xml | 2 +- 50 files changed, 1452 insertions(+), 1128 deletions(-) -- 2.16.1

4 9

[libvirt] [PATCH 0/6] qemu: Add support for 'write-cache' disk frontend argument (blockdev-add saga)
by Peter Krempa 17 Apr '18

17 Apr '18

'write-cache' is one of the booleans the 'cache' attribute of -drive translates to. Since we are getting rid of -drive this needs to be implemented in a different way. Peter Krempa (6): tests: utils: Allow parsing test capability file without virCaps tests: qemuxm2argv: Add infrastructure for testing with real qemuCaps qemu: domain: Add helper for translating disk cachemode to qemu flags qemu: caps: Add capability for 'write-cache' parameter of disk frontends qemu: Format 'write-cache' parameter for disk frontends tests: qemuxml2argv: Test formatting of 'write-cache' parameter src/qemu/qemu_capabilities.c | 5 ++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 27 ++++++++ src/qemu/qemu_domain.c | 75 ++++++++++++++++++++++ src/qemu/qemu_domain.h | 6 ++ tests/qemucapabilitiesdata/caps_2.10.0.aarch64.xml | 1 + tests/qemucapabilitiesdata/caps_2.10.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.10.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.10.0.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.11.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.7.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.7.0.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.8.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.8.0.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.9.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.9.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.9.0.x86_64.xml | 1 + tests/qemuxml2argvdata/disk-drive-write-cache.args | 45 +++++++++++++ tests/qemuxml2argvdata/disk-drive-write-cache.xml | 45 +++++++++++++ tests/qemuxml2argvtest.c | 32 +++++++++ tests/testutilsqemu.c | 21 ++++-- tests/testutilsqemu.h | 3 + 26 files changed, 270 insertions(+), 6 deletions(-) create mode 100644 tests/qemuxml2argvdata/disk-drive-write-cache.args create mode 100644 tests/qemuxml2argvdata/disk-drive-write-cache.xml -- 2.16.2

3 24

[libvirt] [jenkins-ci PATCH] guests: Force use of nodebug kernel on Rawhide
by Andrea Bolognani 17 Apr '18

17 Apr '18

The nodebug kernel provides much better performance than the standard Rawhide kernel, so we want to use it; however, it can sometimes lag behind a bit, which might cause it to be replaced with the regular one when performing a system update. To work around the issue, perform the update in two steps: first with all repositories enabled but excluding the kernel packages, then only the kernel packages but with all repositories except fedora-rawhide-kernel-nodebug disabled. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- guests/tasks/base.yml | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/guests/tasks/base.yml b/guests/tasks/base.yml index 616db0a..70d5abc 100644 --- a/guests/tasks/base.yml +++ b/guests/tasks/base.yml @@ -104,6 +104,24 @@ state: latest when: - package_format == 'rpm' + - not ( os_name == 'Fedora' and + os_version == 'Rawhide' ) + +- name: Update installed packages + command: dnf update --refresh --exclude 'kernel*' -y + args: + warn: no + when: + - os_name == 'Fedora' + - os_version == 'Rawhide' + +- name: Update installed packages + command: dnf update --disablerepo='*' --enablerepo=fedora-rawhide-kernel-nodebug 'kernel*' -y + args: + warn: no + when: + - os_name == 'Fedora' + - os_version == 'Rawhide' - name: Update installed packages apt: -- 2.14.3

2 1

[libvirt] [PATCH] tests: remove double space
by Ján Tomko 17 Apr '18

17 Apr '18

Commit 70255fa5 introduced a double space before the newline-escaping backslash and broke syntax check. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- Pushed as a trivial build-breaker fix. tests/qemuxml2argvtest.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 43f07efc1..509fe776e 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -668,7 +668,7 @@ mymain(void) testCompareXMLToArgv, &info) < 0) \ ret = -1; \ if (virTestRun("QEMU XML-2-startup-XML " name, \ - testCompareXMLToStartupXML, &info) < 0) \ + testCompareXMLToStartupXML, &info) < 0) \ ret = -1; \ virObjectUnref(info.qemuCaps); \ virObjectUnref(info.vm); \ -- 2.16.1

3 2

[libvirt] [libvirt PATCH v2 00/4] Require QEMU 1.5.0
by Ján Tomko 17 Apr '18

17 Apr '18

v2: * Change the version from 1.3.0 to 1.5.0 * Mention this breaks RHEL 6 QEMU and list the versions in relevant long-term releases Some of the patches are very big, available in my repo: https://repo.or.cz/libvirt/jtomko.git/shortlog/refs/heads/caps_cleanup_v2 git fetch git://repo.or.cz/libvirt/jtomko.git caps_cleanup_v2: Ján Tomko (44): qemuxml2argvtest: Fix q35-virt-manager capabilities Delete bogus cpu-host-passthrough test Force QMP capability probing Require QEMU 1.5.0 The following patches are nearly identical to v1: * rebased on the 1.3.1 and 1.4.2 caps deletion * luks-disks-source-qcow2.args is fixed Deprecate QEMU_CAPS_PCI_MULTIFUNCTION Deprecate QEMU_CAPS_NETDEV Clean up qemuDomainAttachNetDevice Remove unused qemuDomainNetVLAN Remove qemuDomainSupportsNetdev Deprecate QEMU_CAPS_MONITOR_JSON Deprecate QEMU_CAPS_NO_USER_CONFIG Deprecate QEMU_CAPS_MEM_PATH Deprecate QEMU_CAPS_DRIVE_SERIAL Deprecate QEMU_CAPS_SDL Deprecate QEMU_CAPS_RTC Deprecate QEMU_CAPS_RTC_TD_HACK Deprecate QEMU_CAPS_VHOST_NET Deprecate QEMU_CAPS_NODEFCONFIG Deprecate QEMU_CAPS_BOOT_MENU Deprecate QEMU_CAPS_FSDEV Deprecate QEMU_CAPS_NAME_PROCESS Deprecate QEMU_CAPS_SMBIOS_TYPE Deprecate QEMU_CAPS_VGA_NONE Deprecate QEMU_CAPS_DRIVE_AIO Deprecate QEMU_CAPS_CHARDEV_SPICEVMC Deprecate QEMU_CAPS_DEVICE_SPICEVMC Deprecate QEMU_CAPS_DRIVE_CACHE_DIRECTSYNC Deprecate QEMU_CAPS_NO_SHUTDOWN Deprecate QEMU_CAPS_DRIVE_CACHE_UNSAFE Deprecate QEMU_CAPS_FSDEV_READONLY Deprecate QEMU_CAPS_DRIVE_COPY_ON_READ Deprecate QEMU_CAPS_FSDEV_WRITEOUT Deprecate QEMU_CAPS_DRIVE_IOTUNE Deprecate QEMU_CAPS_WAKEUP Deprecate QEMU_CAPS_NETDEV_BRIDGE Deprecate QEMU_CAPS_SECCOMP_SANDBOX Deprecate QEMU_CAPS_DTB Deprecate QEMU_CAPS_IPV6_MIGRATION Deprecate QEMU_CAPS_MACHINE_OPT Deprecate QEMU_CAPS_DUMP_GUEST_CORE Deprecate QEMU_CAPS_VNC_SHARE_POLICY Deprecate QEMU_CAPS_HOST_PCI_MULTIDOMAIN Deprecate QEMU_CAPS_DISPLAY Deprecate QEMU_CAPS_MACHINE_USB_OPT cfg.mk | 2 +- src/qemu/qemu_capabilities.c | 936 +-------- src/qemu/qemu_capabilities.h | 74 +- src/qemu/qemu_capspriv.h | 17 +- src/qemu/qemu_command.c | 690 ++---- src/qemu/qemu_domain.c | 16 - src/qemu/qemu_domain.h | 6 - src/qemu/qemu_driver.c | 48 +- src/qemu/qemu_hotplug.c | 147 +- src/qemu/qemu_interface.c | 4 +- src/qemu/qemu_interface.h | 1 - src/qemu/qemu_migration.c | 10 +- src/qemu/qemu_process.c | 23 +- tests/Makefile.am | 8 +- .../qemucapabilitiesdata/caps_1.2.2.x86_64.replies | 1886 ----------------- tests/qemucapabilitiesdata/caps_1.2.2.x86_64.xml | 172 -- .../qemucapabilitiesdata/caps_1.3.1.x86_64.replies | 2141 ------------------- tests/qemucapabilitiesdata/caps_1.3.1.x86_64.xml | 196 -- .../qemucapabilitiesdata/caps_1.4.2.x86_64.replies | 2191 -------------------- tests/qemucapabilitiesdata/caps_1.4.2.x86_64.xml | 198 -- tests/qemucapabilitiesdata/caps_1.5.3.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_1.6.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_1.7.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.1.1.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.10.0.aarch64.xml | 35 - tests/qemucapabilitiesdata/caps_2.10.0.ppc64.xml | 35 - tests/qemucapabilitiesdata/caps_2.10.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.10.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.11.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 35 - tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 35 - tests/qemucapabilitiesdata/caps_2.12.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.4.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.5.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.6.0.aarch64.xml | 35 - tests/qemucapabilitiesdata/caps_2.6.0.ppc64.xml | 35 - tests/qemucapabilitiesdata/caps_2.6.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.7.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.7.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.8.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.8.0.x86_64.xml | 35 - tests/qemucapabilitiesdata/caps_2.9.0.ppc64.xml | 35 - tests/qemucapabilitiesdata/caps_2.9.0.s390x.xml | 35 - tests/qemucapabilitiesdata/caps_2.9.0.x86_64.xml | 35 - tests/qemucapabilitiestest.c | 3 - tests/qemucapsprobe.c | 2 +- tests/qemuhelpdata/qemu-0.12.1 | 198 -- tests/qemuhelpdata/qemu-0.12.1-device | 62 - tests/qemuhelpdata/qemu-1.0 | 253 --- tests/qemuhelpdata/qemu-1.0-device | 148 -- tests/qemuhelpdata/qemu-1.1.0 | 269 --- tests/qemuhelpdata/qemu-1.1.0-device | 170 -- tests/qemuhelpdata/qemu-1.2.0 | 272 --- tests/qemuhelpdata/qemu-kvm-0.12.3 | 215 -- tests/qemuhelpdata/qemu-kvm-0.12.3-device | 63 - tests/qemuhelpdata/qemu-kvm-0.13.0 | 242 --- tests/qemuhelpdata/qemu-kvm-0.13.0-device | 90 - tests/qemuhelpdata/qemu-kvm-1.2.0 | 277 --- tests/qemuhelptest.c | 437 ---- .../aarch64-aavmf-virtio-mmio.args | 14 +- tests/qemuxml2argvdata/aarch64-acpi-uefi.args | 9 +- [... lots of .args files ...] tests/qemuxml2argvdata/watchdog.args | 9 +- tests/qemuxml2argvdata/x86-kvm-32-on-64.args | 9 +- tests/qemuxml2argvtest.c | 744 +++---- tests/qemuxml2xmltest.c | 119 +- 670 files changed, 4582 insertions(+), 14981 deletions(-) delete mode 100644 tests/qemucapabilitiesdata/caps_1.2.2.x86_64.replies delete mode 100644 tests/qemucapabilitiesdata/caps_1.2.2.x86_64.xml delete mode 100644 tests/qemucapabilitiesdata/caps_1.3.1.x86_64.replies delete mode 100644 tests/qemucapabilitiesdata/caps_1.3.1.x86_64.xml delete mode 100644 tests/qemucapabilitiesdata/caps_1.4.2.x86_64.replies delete mode 100644 tests/qemucapabilitiesdata/caps_1.4.2.x86_64.xml delete mode 100644 tests/qemuhelpdata/qemu-0.12.1 delete mode 100644 tests/qemuhelpdata/qemu-0.12.1-device delete mode 100644 tests/qemuhelpdata/qemu-1.0 delete mode 100644 tests/qemuhelpdata/qemu-1.0-device delete mode 100644 tests/qemuhelpdata/qemu-1.1.0 delete mode 100644 tests/qemuhelpdata/qemu-1.1.0-device delete mode 100644 tests/qemuhelpdata/qemu-1.2.0 delete mode 100644 tests/qemuhelpdata/qemu-kvm-0.12.3 delete mode 100644 tests/qemuhelpdata/qemu-kvm-0.12.3-device delete mode 100644 tests/qemuhelpdata/qemu-kvm-0.13.0 delete mode 100644 tests/qemuhelpdata/qemu-kvm-0.13.0-device delete mode 100644 tests/qemuhelpdata/qemu-kvm-1.2.0 delete mode 100644 tests/qemuhelptest.c delete mode 100644 tests/qemuxml2argvdata/channel-spicevmc-old.args delete mode 100644 tests/qemuxml2argvdata/channel-spicevmc-old.xml -- 2.16.1

4 102

[libvirt] [PATCH] tests: mock qemuInterfaceOpenVhostNet
by Ján Tomko 17 Apr '18

17 Apr '18

This functions contains logic that tries to use vhost for virtio interfaces, even if <driver name='vhost'/> was not supplied. In this case, a failure is non-fatal. On my system, /dev/vhost-net was not accessible to the user running 'make check', but we should not depend on that. Mock it to prevent accessing /dev/vhost-net and return some predictable file descriptor numbers instead. Introduced by commit c1f684e - deprecate QEMU_CAPS_VHOST_NET. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> Reported-by: Jiří Denemark <jdenemar(a)redhat.com> --- tests/qemuxml2argvdata/user-aliases.args | 2 +- tests/qemuxml2argvmock.c | 19 +++++++++++++++++++ 2 files changed, 20 insertions(+), 1 deletion(-) diff --git a/tests/qemuxml2argvdata/user-aliases.args b/tests/qemuxml2argvdata/user-aliases.args index d887d53fa..774c599b5 100644 --- a/tests/qemuxml2argvdata/user-aliases.args +++ b/tests/qemuxml2argvdata/user-aliases.args @@ -51,7 +51,7 @@ id=ua-myEncryptedDisk1 \ if=none,id=drive-ua-WhatAnAwesomeCDROM,media=cdrom,readonly=on,cache=none \ -device ide-drive,bus=ide.1,unit=0,drive=drive-ua-WhatAnAwesomeCDROM,\ id=ua-WhatAnAwesomeCDROM \ --netdev tap,fd=3,id=hostua-CheckoutThisNIC \ +-netdev tap,fd=3,id=hostua-CheckoutThisNIC,vhost=on,vhostfd=44 \ -device virtio-net-pci,netdev=hostua-CheckoutThisNIC,id=ua-CheckoutThisNIC,\ mac=52:54:00:d6:c0:0b,bus=pci.0,addr=0x3 \ -netdev socket,listen=127.0.0.1:1234,id=hostua-WeCanAlsoDoServerMode \ diff --git a/tests/qemuxml2argvmock.c b/tests/qemuxml2argvmock.c index 177b24e0a..adab5c911 100644 --- a/tests/qemuxml2argvmock.c +++ b/tests/qemuxml2argvmock.c @@ -36,6 +36,7 @@ #include "virstring.h" #include "virtpm.h" #include "virutil.h" +#include "qemu/qemu_interface.h" #include <time.h> #include <unistd.h> @@ -188,3 +189,21 @@ virNetDevOpenvswitchGetVhostuserIfname(const char *path ATTRIBUTE_UNUSED, { return VIR_STRDUP(*ifname, "vhost-user0"); } + +int +qemuInterfaceOpenVhostNet(virDomainDefPtr def ATTRIBUTE_UNUSED, + virDomainNetDefPtr net, + int *vhostfd, + size_t *vhostfdSize) +{ + size_t i; + + if (!(net->model && STREQ(net->model, "virtio"))) { + *vhostfdSize = 0; + return 0; + } + + for (i = 0; i < *vhostfdSize; i++) + vhostfd[i] = STDERR_FILENO + 42 + i; + return 0; +} -- 2.16.1

2 1

[libvirt] [PATCH] tests: Don't touch /dev/vhost-net in qemuxml2argvtest
by Martin Kletzander 17 Apr '18

17 Apr '18

The code is trying to open /dev/vhost-net from the host machine and if the current user has access to it, the command line will end up having two more options for an interface (,vhost=on,vhostfd=X where X is next free file descriptor). Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- tests/qemuxml2argvmock.c | 36 ++++++++++++++++++++++++++++++++++++ 1 file changed, 36 insertions(+) diff --git a/tests/qemuxml2argvmock.c b/tests/qemuxml2argvmock.c index 177b24e0a953..11626e2794c1 100644 --- a/tests/qemuxml2argvmock.c +++ b/tests/qemuxml2argvmock.c @@ -20,6 +20,10 @@ #include <config.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> + #include "internal.h" #include "viralloc.h" #include "vircommand.h" @@ -188,3 +192,35 @@ virNetDevOpenvswitchGetVhostuserIfname(const char *path ATTRIBUTE_UNUSED, { return VIR_STRDUP(*ifname, "vhost-user0"); } + +static int (*real_open)(const char *path, int flags, ...); + +static void init_syms(void) +{ + if (real_open) + return; + + VIR_MOCK_REAL_INIT(open); +} + +int open(const char *path, int flags, ...) +{ + va_list ap; + mode_t mode = 0; + + init_syms(); + + if (STREQ(path, "/dev/vhost-net")) + return -1; + + /* The mode argument is mandatory when O_CREAT is set in flags, + * otherwise the argument is ignored. + */ + if (flags & O_CREAT) { + va_start(ap, flags); + mode = (mode_t) va_arg(ap, int); + va_end(ap); + } + + return real_open(path, flags, mode); +} -- 2.17.0

3 4