July 2017 - Devel - libvirt List Archives

[libvirt] [PATCH RESEND] Interface: return appropriate status for bridge device
by Lin Ma 03 Aug '17

03 Aug '17

In function udevInterfaceIsActive, The current design relies on the sys attribute 'operstate' to determine whether the interface is active. For the device node of virtual network(say virbr0), There is already a dummy tap device to maintain a fixed mac on it, So it's available and its status should be considered as active. But if no anyelse tap device connects to it, The value of operstate of virbr0 in sysfs is 'down', So udevInterfaceIsActive returns 0, It causes 'virsh iface-list --all' to report the status of virbr0 as 'inactive'. This patch fixes the issue by counting slave devices for bridge device. Signed-off-by: Lin Ma <lma(a)suse.com> --- src/interface/interface_backend_udev.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/src/interface/interface_backend_udev.c b/src/interface/interface_backend_udev.c index 5d0fc64..9ac4674 100644 --- a/src/interface/interface_backend_udev.c +++ b/src/interface/interface_backend_udev.c @@ -1127,6 +1127,11 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) struct udev_device *dev; virInterfaceDefPtr def = NULL; int status = -1; + struct dirent **member_list = NULL; + const char *devtype; + int member_count = 0; + char *member_path; + size_t i; dev = udev_device_new_from_subsystem_sysname(udev, "net", ifinfo->name); @@ -1146,6 +1151,23 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) /* Check if it's active or not */ status = STREQ(udev_device_get_sysattr_value(dev, "operstate"), "up"); + if (!status) { + devtype = udev_device_get_devtype(dev); + if (STREQ_NULLABLE(devtype, "bridge")) { + if (virAsprintf(&member_path, "%s/%s", + udev_device_get_syspath(dev), "brif") < 0) + goto cleanup; + member_count = scandir(member_path, &member_list, + udevBridgeScanDirFilter, alphasort); + if (member_count > 0) + status = 1; + for (i = 0; i < member_count; i++) + VIR_FREE(member_list[i]); + VIR_FREE(member_list); + VIR_FREE(member_path); + } + } + udev_device_unref(dev); cleanup: -- 2.9.2

3 2

[libvirt] [PATCH 0/3] qemu: command: clean up default cpu formatting
by Cole Robinson 03 Aug '17

03 Aug '17

Just some minor patches to do with encoding a default CPU on the qemu command line. Patch 1 adds a test case Patch 2 simplifies the code Patch 3 throws an explicit error for a confusing situation Cole Robinson (3): tests: add qemu x86 kvm 32-on-64 test qemu: command: rework adding of default cpu model qemu: command: explicitly error for non-x86 default CPU src/qemu/qemu_command.c | 95 +++++++++------------- .../qemuxml2argvdata/qemuxml2argv-hyperv-off.args | 1 - .../qemuxml2argv-kvm-features-off.args | 1 - .../qemuxml2argv-x86-kvm-32-on-64.args | 21 +++++ .../qemuxml2argv-x86-kvm-32-on-64.xml | 13 +++ tests/qemuxml2argvtest.c | 1 + tests/testutilsqemu.c | 18 +++- 7 files changed, 90 insertions(+), 60 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-x86-kvm-32-on-64.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-x86-kvm-32-on-64.xml -- 2.13.3

4 10

[libvirt] [PATCH v3 0/2] Restrict usage of unsupported clock timers
by Kothapally Madhu Pavan 03 Aug '17

03 Aug '17

hpet and kvm.pit clock timers are specific to x86 architecture and are not to be used by unsuported architectures. Similarly sPAPR guests only allow RTC timer. This patchset will restrict the usage of unsupported clock timers. Kothapally Madhu Pavan (2): qemu: Restrict usage of hpet and kvm.pit timers by unsupported architectures qemu: Default hwclock source for sPAPR to RTC src/qemu/qemu_domain.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) -- 1.8.3.1

2 3

[libvirt] [PATCH 0/8] Delete a function
by Ján Tomko 03 Aug '17

03 Aug '17

*** BLURB HERE *** Ján Tomko (8): Remove superfluous usage of virDomainDeviceInfoNeedsFormat virDomainDeviceInfoFormat: delete outdated comments Use a separate buffer for <smartcard> subelements Use a separate buffer for <sound> subelements Use a separate buffer for <watchdog> subelements Use a separate buffer for <hub> subelements conf: check for buffer errors before virBufferUse Turn virDomainDeviceInfoFormat into void src/conf/capabilities.c | 5 + src/conf/cpu_conf.c | 4 + src/conf/domain_conf.c | 237 +++++++++++++++++++++--------------------------- 3 files changed, 114 insertions(+), 132 deletions(-) -- 2.13.0

3 21

[libvirt] [PATCH 00/24] qemu: refactor node name detection
by Peter Krempa 03 Aug '17

03 Aug '17

Since query-blockstats provides full hierarchy of node names without any doubtful matching algorithm, use it to make the node name detector reliable. Unfortunately some of the changes are ugly when viewed as a diff since they delete and rewrite chunks of code. This series can be fetched at: git fetch git://pipo.sk/pipo/libvirt.git node-name-detect-refactor-1 Peter Krempa (24): qemu: monitor: Extract call of 'query-blockstats' and add new API for it tests: qemumonitorjson: Consolidate basic node name detection test cases tests: qemumonitor: Prepare for more test data in testBlockNodeNameDetect tests: qemumonitorjson: Drop redundant data from testBlockNodeNameDetectFormat tests: utils: Don't calculate file size in virTestLoadFile util: buffer: Add virBufferStrcatVArgs tests: utils: Add virTestLoadFilePath helper tests: qemuhelp: convert to virTestLoadFilePath tests: utils: Add virTestLoadFileJSON helper tests: qemumonitorjson: simplify path handling in testBlockNodeNameDetect tests: qemumontitorjson: temporarily disable node name detection tests qemu: block: Rename qemuBlockFillNodeData and move it to the top util: storagefile: rename 'nodebacking' to 'nodestorage' in virStorageSource qemu: block: Refactor node name detection code tests: Extract mock library for making hash table deterministic tests: qemumonitorjson: Simplify node name detection test qemu: block: store and test driver names for detected storage nodes tests: qemumonitorjson: Add data and fix 'same-backing' node detection case tests: qemumonitorjson: Fix 'relative' node name detection test case tests: qemumonitorjson: Fix 'gluster' node name detection test case tests: qemumonitorjson: Add test case for node name detection with blockjob tests: qemumonitorjson: Old and empty test case for node name detection tests: qemumonitorjson: Test extraction of LUKS node names tests: qemumonitorjson: Test extraction of iSCSI device node names src/libvirt_private.syms | 1 + src/qemu/qemu_block.c | 344 ++- src/qemu/qemu_block.h | 12 +- src/qemu/qemu_driver.c | 10 +- src/qemu/qemu_monitor.c | 15 + src/qemu/qemu_monitor.h | 2 + src/qemu/qemu_monitor_json.c | 44 +- src/qemu/qemu_monitor_json.h | 2 + src/util/virbuffer.c | 27 +- src/util/virbuffer.h | 2 + src/util/virstoragefile.c | 6 +- src/util/virstoragefile.h | 2 +- tests/Makefile.am | 17 +- tests/qemuhelptest.c | 14 +- .../qemumonitorjson-nodename-1.result | 15 - .../qemumonitorjson-nodename-2.json | 2270 -------------------- .../qemumonitorjson-nodename-2.result | 60 - .../qemumonitorjson-nodename-basic-blockstats.json | 166 ++ ...emumonitorjson-nodename-basic-named-nodes.json} | 18 +- .../qemumonitorjson-nodename-basic.result | 16 + ...mumonitorjson-nodename-blockjob-blockstats.json | 301 +++ ...umonitorjson-nodename-blockjob-named-nodes.json | 682 ++++++ .../qemumonitorjson-nodename-blockjob.result | 26 + .../qemumonitorjson-nodename-empty-blockstats.json | 2 + ...qemumonitorjson-nodename-empty-named-nodes.json | 2 + .../qemumonitorjson-nodename-empty.result | 1 + ...emumonitorjson-nodename-gluster-blockstats.json | 111 + ...umonitorjson-nodename-gluster-named-nodes.json} | 90 +- .../qemumonitorjson-nodename-gluster.result | 19 +- .../qemumonitorjson-nodename-iscsi-blockstats.json | 113 + ...qemumonitorjson-nodename-iscsi-named-nodes.json | 114 + .../qemumonitorjson-nodename-iscsi.result | 13 + .../qemumonitorjson-nodename-luks-blockstats.json | 58 + .../qemumonitorjson-nodename-luks-named-nodes.json | 109 + .../qemumonitorjson-nodename-luks.result | 6 + .../qemumonitorjson-nodename-old-blockstats.json | 160 ++ .../qemumonitorjson-nodename-old-named-nodes.json | 2 + .../qemumonitorjson-nodename-old.result | 1 + ...mumonitorjson-nodename-relative-blockstats.json | 329 +++ ...monitorjson-nodename-relative-named-nodes.json} | 26 +- .../qemumonitorjson-nodename-relative.result | 56 +- ...nitorjson-nodename-same-backing-blockstats.json | 221 ++ ...torjson-nodename-same-backing-named-nodes.json} | 16 +- .../qemumonitorjson-nodename-same-backing.result | 32 +- tests/qemumonitorjsontest.c | 97 +- tests/testutils.c | 101 +- tests/testutils.h | 6 + ...{virmacmapmock.c => virdeterministichashmock.c} | 0 tests/virmacmaptest.c | 2 +- 49 files changed, 2961 insertions(+), 2778 deletions(-) delete mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-1.result delete mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-2.json delete mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-2.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-basic-blockstats.json rename tests/qemumonitorjsondata/{qemumonitorjson-nodename-1.json => qemumonitorjson-nodename-basic-named-nodes.json} (96%) create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-basic.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-blockjob-blockstats.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-blockjob-named-nodes.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-blockjob.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-empty-blockstats.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-empty-named-nodes.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-empty.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-gluster-blockstats.json rename tests/qemumonitorjsondata/{qemumonitorjson-nodename-gluster.json => qemumonitorjson-nodename-gluster-named-nodes.json} (89%) create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-iscsi-blockstats.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-iscsi-named-nodes.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-iscsi.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-luks-blockstats.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-luks-named-nodes.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-luks.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-old-blockstats.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-old-named-nodes.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-old.result create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-relative-blockstats.json rename tests/qemumonitorjsondata/{qemumonitorjson-nodename-relative.json => qemumonitorjson-nodename-relative-named-nodes.json} (97%) create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-nodename-same-backing-blockstats.json rename tests/qemumonitorjsondata/{qemumonitorjson-nodename-same-backing.json => qemumonitorjson-nodename-same-backing-named-nodes.json} (96%) rename tests/{virmacmapmock.c => virdeterministichashmock.c} (100%) -- 2.13.2

4 58

[libvirt] [PATCH v3 0/2] Allow saving VM state to pipe
by Chen Hanxiao 03 Aug '17

03 Aug '17

This series introduce flag VIR_DOMAIN_SAVE_DIRECT to enable command 'save' to write to PIPE. This will write QEMU_SAVE_MAGIC directly. Base upon patches from Roy Keene <rkeene(a)knightpoint.com> with some fixes. Change from original patch: 1) Check whether the specified path is a PIPE. 2) Rebase on upstream. 3) Add doc for virsh command v3: add doc/news.xml rebase on upstream v2-resend: rebase on upstream v2: rename VIR_DOMAIN_SAVE_PIPE to VIR_DOMAIN_SAVE_DIRECT remove S_ISFIFO check Chen Hanxiao (2): qemu: Allow qemuDomainSaveMemory saving VM state to a pipe virsh: introduce flage --direct for save command docs/news.xml | 9 +++++++ include/libvirt/libvirt-domain.h | 1 + src/qemu/qemu_driver.c | 54 ++++++++++++++++++++++++++-------------- tools/virsh-domain.c | 6 +++++ tools/virsh.pod | 5 +++- 5 files changed, 56 insertions(+), 19 deletions(-) -- 2.7.4

2 6

[libvirt] [dbus PATCH 0/9] code cleanup
by Pavel Hrdina 02 Aug '17

02 Aug '17

Pavel Hrdina (9): util: move bus_path_for_domain and domain_from_bus_path util: move and rename virDomainsFreep domain: split domain code into separate file events: split event code into separate file util: rename function to follow libvirt naming rules main: rename functions to follow libvirt naming rules manager: rename functions and structures to follow libvirt naming rules domain: rename functions to follow libvirt naming rules events: rename functions to follow libvirt naming rules src/Makefile.am | 4 +- src/domain.c | 549 +++++++++++++++++++++++++++++++++ src/domain.h | 10 + src/events.c | 252 ++++++++++++++++ src/events.h | 9 + src/main.c | 44 +-- src/manager.c | 920 ++++---------------------------------------------------- src/manager.h | 18 +- src/util.c | 35 ++- src/util.h | 39 ++- 10 files changed, 970 insertions(+), 910 deletions(-) create mode 100644 src/domain.c create mode 100644 src/domain.h create mode 100644 src/events.c create mode 100644 src/events.h -- 2.13.3

2 14

[libvirt] [libvirt-php PATCH 0/7] add bindings for NWFilter APIs
by Dawid Zamirski 01 Aug '17

01 Aug '17

Hello, This series adds support for libvirt's virNWFilter* APIs. Since it introduces new resource type, I took the opportunity to cleanup the driver code a little: * in patches 1-3: added macros to take care of the differences in how PHP5 and PHP7 handle resource types. * in patch 4: libvirt_doman_get_connect was segfaulting when called multiple times because it was not bumping reference count on the resource it was returning to the calling code. * in patch 5: added a macro to take care of the differences in how PHP5 and PHP7 initialize arrays. * patches 6 and 7: implement the missing binding to NWFilter APIs. Dawid Zamirski (7): move macros to header file. add wrappers for PHP resource handling. update code to use resource handling macros fix libvirt_doman_get_connect implementation. add and use VIRT_ARRAY_INIT macro add nwfilter resource type implement NWFilter API bindings. src/libvirt-php.c | 923 +++++++++++++++++++++++++++++++----------------------- src/libvirt-php.h | 150 ++++++++- 2 files changed, 672 insertions(+), 401 deletions(-) -- 2.13.0

4 16

[libvirt] [PATCH v2] tools: virsh: Adding unix socket support to 'domdisplay' command.
by Julio Faracco 01 Aug '17

01 Aug '17

This commit adds the unix socket URL support to 'domdisplay' command. Before, even if an user was using unix socket to define a spice graphics, the command 'domdisplay' showed that the settings were not supported. Now, the command shows the proper URL: spice+unix://foo/bar.sock. Settings: <graphics type='vnc' port='-1' autoport='yes' listen='127.0.0.1'> <listen type='address' address='127.0.0.1'/> </graphics> <graphics type='spice'> <listen type='socket' socket='/tmp/spice.sock'/> </graphics> Before: virsh # domdisplay --all Windows7 vnc://127.0.0.1:0 After: virsh # domdisplay --all Windows7 vnc://127.0.0.1:0 spice+unix:///tmp/spice.sock Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1336720 Signed-off-by: Julio Faracco <jcfaracco(a)gmail.com> --- tools/virsh-domain.c | 52 +++++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 41 insertions(+), 11 deletions(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 0684979..935ef8a 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -10948,6 +10948,8 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) char *xpath = NULL; char *listen_addr = NULL; int port, tls_port = 0; + char *type_conn = NULL; + char *socket = NULL; char *passwd = NULL; char *output = NULL; const char *scheme[] = { "vnc", "spice", "rdp", NULL }; @@ -11008,9 +11010,6 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) if (tmp) tls_port = 0; - if (!port && !tls_port) - continue; - /* Create our XPATH lookup for the current display's address */ if (virAsprintf(&xpath, xpath_fmt, scheme[iter], "@listen") < 0) goto cleanup; @@ -11021,6 +11020,29 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) listen_addr = virXPathString(xpath, ctxt); VIR_FREE(xpath); + /* Create our XPATH lookup for the current spice type. */ + if (virAsprintf(&xpath, xpath_fmt, scheme[iter], "listen/@type") < 0) + goto cleanup; + + /* Attempt to get the type of spice connection */ + VIR_FREE(type_conn); + type_conn = virXPathString(xpath, ctxt); + VIR_FREE(xpath); + + if (STREQ_NULLABLE(type_conn, "socket")) { + if (!socket) { + if (virAsprintf(&xpath, xpath_fmt, scheme[iter], "listen/@socket") < 0) + goto cleanup; + + socket = virXPathString(xpath, ctxt); + + VIR_FREE(xpath); + } + } + + if (!port && !tls_port && !socket) + continue; + if (!listen_addr) { /* The subelement address - <listen address='xyz'/> - * *should* have been automatically backfilled into its @@ -11035,11 +11057,9 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) listen_addr = virXPathString(xpath, ctxt); VIR_FREE(xpath); - } - - /* If listen_addr is 0.0.0.0 or [::] we should try to parse URI and set - * listen_addr based on current URI. */ - if (listen_addr) { + } else { + /* If listen_addr is 0.0.0.0 or [::] we should try to parse URI and set + * listen_addr based on current URI. */ if (virSocketAddrParse(&addr, listen_addr, AF_UNSPEC) > 0 && virSocketAddrIsWildcard(&addr)) { @@ -11078,20 +11098,28 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) VIR_FREE(xpath); /* Build up the full URI, starting with the scheme */ - virBufferAsprintf(&buf, "%s://", scheme[iter]); + if (socket) + virBufferAsprintf(&buf, "%s+unix://", scheme[iter]); + else + virBufferAsprintf(&buf, "%s://", scheme[iter]); /* There is no user, so just append password if there's any */ if (STREQ(scheme[iter], "vnc") && passwd) virBufferAsprintf(&buf, ":%s@", passwd); /* Then host name or IP */ - if (!listen_addr) + if (!listen_addr && !socket) virBufferAddLit(&buf, "localhost"); - else if (strchr(listen_addr, ':')) + else if (!socket && strchr(listen_addr, ':')) virBufferAsprintf(&buf, "[%s]", listen_addr); + else if (socket) + virBufferAsprintf(&buf, "%s", socket); else virBufferAsprintf(&buf, "%s", listen_addr); + /* Free socket to prepare the pointer for the next iteration */ + VIR_FREE(socket); + /* Add the port */ if (port) { if (STREQ(scheme[iter], "vnc")) { @@ -11148,6 +11176,8 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd) cleanup: VIR_FREE(xpath); + VIR_FREE(type_conn); + VIR_FREE(socket); VIR_FREE(passwd); VIR_FREE(listen_addr); VIR_FREE(output); -- 2.7.4

2 2

[libvirt] [PATCH 0/7] Misc improvements
by Michal Privoznik 31 Jul '17

31 Jul '17

As I started to turn more object into using RW locks, I've found couple of areas for improvement too. Michal Privoznik (7): virConnect: Update comment for @privateData Report error if virMutexInit fails virnetworkobj: Make virNetworkObjFindBy{UUID,Name}Locked() static again virNetworkObjList: Derive from virObjectRWLockable virNodeDeviceObjList: Derive from virObjectRWLockable virConnect: Derive from virObjectRWLockable storageDriver: Use RW locks src/bhyve/bhyve_driver.c | 1 + src/conf/virnetworkobj.c | 42 ++++++++++---------------------- src/conf/virnetworkobj.h | 8 ------- src/conf/virnodedeviceobj.c | 16 ++++++------- src/conf/virstorageobj.h | 2 +- src/datatypes.c | 6 +++-- src/datatypes.h | 6 ++--- src/libvirt_private.syms | 2 -- src/lxc/lxc_driver.c | 1 + src/lxc/lxc_fuse.c | 4 +++- src/network/bridge_driver.c | 1 + src/node_device/node_device_hal.c | 1 + src/nwfilter/nwfilter_dhcpsnoop.c | 12 +++++++--- src/nwfilter/nwfilter_driver.c | 5 +++- src/nwfilter/nwfilter_gentech_driver.c | 4 +++- src/secret/secret_driver.c | 2 ++ src/storage/storage_driver.c | 44 +++++++++++++++++++--------------- src/uml/uml_driver.c | 1 + src/util/virerror.c | 2 +- src/util/virnetlink.c | 1 + src/util/virthreadpool.c | 4 +++- src/vmware/vmware_driver.c | 5 +++- src/vz/vz_driver.c | 4 +++- tools/virsh-console.c | 4 +++- 24 files changed, 94 insertions(+), 84 deletions(-) -- 2.13.0

4 18