- Devel - libvirt List Archives

[libvirt] [PATCH] qemu_hotplug: try harder to eject media
by Pavel Hrdina 25 Jun '15

25 Jun '15

Some guests lock the tray and QEMU eject command will simply fail to eject the media. But the guest OS can handle this attempt to eject the media and can unlock the tray and open it. In this case, we should try again to actually eject the media. If the first attempt fails we will wait for tray to open and try again to eject the media, but if the second attempt fails, returns with error. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1147471 Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_hotplug.c | 33 ++++++++++++++++++++++++++++----- 1 file changed, 28 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 0628964..0cebaad 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -201,13 +201,17 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, } if (ret < 0) - goto error; + VIR_DEBUG("tray is probably locked, wait for the guest to unlock " + "the tray and open it"); virObjectRef(vm); /* we don't want to report errors from media tray_open polling */ while (retries) { - if (disk->tray_status == VIR_DOMAIN_DISK_TRAY_OPEN) + if (disk->tray_status == VIR_DOMAIN_DISK_TRAY_OPEN) { + ret = 0; + virResetLastError(); break; + } retries--; virObjectUnlock(vm); @@ -218,10 +222,29 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, virObjectUnref(vm); if (retries <= 0) { - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Unable to eject media")); - ret = -1; + /* Report a new error only if the first attempt don't fail and we don't + * receive VIR_DOMAIN_DISK_TRAY_OPEN event, otherwise report the error + * from first attempt. */ + if (ret == 0) { + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Unable to eject media")); + ret = -1; + } goto error; + } else { + /* QEMU will send eject request to guest, but if the tray is locked, + * always returns with error. However the guest can handle the eject + * request, unlock the tray and open it. In case this happens, we + * should try to eject the media once more. */ + qemuDomainObjEnterMonitor(driver, vm); + ret = qemuMonitorEjectMedia(priv->mon, driveAlias, force); + if (qemuDomainObjExitMonitor(driver, vm) < 0) { + ret = -1; + goto cleanup; + } + + if (ret < 0) + goto error; } if (!virStorageSourceIsEmpty(newsrc)) { -- 2.4.4

2 3

[libvirt] [sandbox PATCH 00/10] Patches for Libvirt-sandbox
by Eren Yagdiran 25 Jun '15

25 Jun '15

Hello, These patches provide disk support for libvirt-sandbox. Implemented '--disk' parameter will be useful when integrating Docker image support for libvirt-sandbox. --Main diffs compared to previous patches. Since many hypervisors, including kvm, will not even honour requested names for disk devices we link each device under /dev/disk/by-tag e.g /dev/disk/by-tag/foobar -> /dev/sda We populate disks.cfg with tag to device mapping when we build the sandbox. After that, in each init-process {Common,Qemu}, we basically read the configuration and populate the right symlinks under /dev/disk/by-tag The common functions for modifying directories are moved under Init-util. {Common,Qemu} inits are using them. Cédric Bosdonnat (2): Add gvir_sandbox_config_has_disks function qemu: use devtmpfs rather than tmpfs to auto-populate /dev Eren Yagdiran (8): Add an utility function for guessing filetype from file extension Add configuration object for disk support Add disk parameter to virt-sandbox Add disk support to the container builder Add disk support to machine builder Init-util : Common directory functions for init-common and init-qemu Common-init: Building symlink from disks.cfg Common-builder: /dev/disk/by-tag/thetag to /dev/vdN bin/virt-sandbox.c | 37 +++ libvirt-sandbox/Makefile.am | 7 +- .../libvirt-sandbox-builder-container.c | 37 ++- libvirt-sandbox/libvirt-sandbox-builder-machine.c | 44 ++- libvirt-sandbox/libvirt-sandbox-builder.c | 73 ++++- libvirt-sandbox/libvirt-sandbox-config-disk.c | 274 +++++++++++++++++++ libvirt-sandbox/libvirt-sandbox-config-disk.h | 82 ++++++ libvirt-sandbox/libvirt-sandbox-config.c | 300 +++++++++++++++++++++ libvirt-sandbox/libvirt-sandbox-config.h | 11 + libvirt-sandbox/libvirt-sandbox-init-common.c | 51 +++- libvirt-sandbox/libvirt-sandbox-init-qemu.c | 151 ++--------- libvirt-sandbox/libvirt-sandbox-init-util.c | 58 ++++ libvirt-sandbox/libvirt-sandbox-init-util.h | 41 +++ libvirt-sandbox/libvirt-sandbox-util.c | 72 +++++ libvirt-sandbox/libvirt-sandbox-util.h | 5 + libvirt-sandbox/libvirt-sandbox.h | 1 + libvirt-sandbox/libvirt-sandbox.sym | 5 + libvirt-sandbox/tests/test-config.c | 11 + po/POTFILES.in | 1 + 19 files changed, 1112 insertions(+), 149 deletions(-) create mode 100644 libvirt-sandbox/libvirt-sandbox-config-disk.c create mode 100644 libvirt-sandbox/libvirt-sandbox-config-disk.h create mode 100644 libvirt-sandbox/libvirt-sandbox-init-util.c create mode 100644 libvirt-sandbox/libvirt-sandbox-init-util.h create mode 100644 libvirt-sandbox/libvirt-sandbox-util.c -- 2.1.0

3 19

[libvirt] [PATCH 0/2] Handle isa-fdc removal from new q35 machine types
by Ján Tomko 25 Jun '15

25 Jun '15

https://bugzilla.redhat.com/show_bug.cgi?id=1227880 Ján Tomko (2): Separate isa-fdc options generation Explicitly format the isa-fdc controller for newer q35 machines src/qemu/qemu_command.c | 50 ++++++++++++++++------ src/qemu/qemu_domain.c | 22 ++++++++++ src/qemu/qemu_domain.h | 1 + .../qemuxml2argv-boot-floppy-q35.args | 12 ++++++ .../qemuxml2argv-boot-floppy-q35.xml | 40 +++++++++++++++++ .../qemuxml2argv-bootindex-floppy-q35.args | 11 +++++ .../qemuxml2argv-bootindex-floppy-q35.xml | 40 +++++++++++++++++ tests/qemuxml2argvtest.c | 9 ++++ 8 files changed, 172 insertions(+), 13 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-boot-floppy-q35.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-boot-floppy-q35.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-bootindex-floppy-q35.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-bootindex-floppy-q35.xml -- 2.3.6

2 3

[libvirt] [PATCH 0/5] vz: Reformat
by Michal Privoznik 25 Jun '15

25 Jun '15

There shouldn't be any functional change. Only some formatting nits fixed. Michal Privoznik (5): vz_driver: Reformat vz_network: Reformat vz_sdk: Reformat vz_storage: Reformat vz_utils: Reformat src/vz/vz_driver.c | 86 ++++++++++----------- src/vz/vz_network.c | 26 +++---- src/vz/vz_sdk.c | 209 ++++++++++++++++++++++++++-------------------------- src/vz/vz_storage.c | 58 +++++++-------- src/vz/vz_utils.h | 20 ++--- 5 files changed, 199 insertions(+), 200 deletions(-) -- 2.3.6

3 10

[libvirt] [sandbox PATCH 00/10] Patches for Libvirt-sandbox
by Eren Yagdiran 25 Jun '15

25 Jun '15

Hello, These patches provide disk support for libvirt-sandbox. Implemented '--disk' parameter will be useful when integrating Docker image support for libvirt-sandbox. --Main diffs compared to previous patches. Since many hypervisors, including kvm, will not even honour requested names for disk devices we link each device under /dev/disk/by-tag e.g /dev/disk/by-tag/foobar -> /dev/sda We populate disks.cfg with tag to device mapping when we build the sandbox. After that, in each init-process {Common,Qemu}, we basically read the configuration and populate the right symlinks under /dev/disk/by-tag The common functions for modifying directories are moved under Init-util. {Common,Qemu} inits are using them. Cédric Bosdonnat (2): Add gvir_sandbox_config_has_disks function qemu: use devtmpfs rather than tmpfs to auto-populate /dev Eren Yagdiran (8): Add an utility function for guessing filetype from file extension Add configuration object for disk support Add disk parameter to virt-sandbox Add disk support to the container builder Add disk support to machine builder Init-util : Common directory functions for init-common and init-qemu Common-init: Building symlink from disks.cfg Common-builder: /dev/disk/by-tag/thetag to /dev/vdN bin/virt-sandbox.c | 37 +++ libvirt-sandbox/Makefile.am | 7 +- .../libvirt-sandbox-builder-container.c | 37 ++- libvirt-sandbox/libvirt-sandbox-builder-machine.c | 44 ++- libvirt-sandbox/libvirt-sandbox-builder.c | 73 ++++- libvirt-sandbox/libvirt-sandbox-config-disk.c | 274 +++++++++++++++++++ libvirt-sandbox/libvirt-sandbox-config-disk.h | 82 ++++++ libvirt-sandbox/libvirt-sandbox-config.c | 300 +++++++++++++++++++++ libvirt-sandbox/libvirt-sandbox-config.h | 11 + libvirt-sandbox/libvirt-sandbox-init-common.c | 51 +++- libvirt-sandbox/libvirt-sandbox-init-qemu.c | 151 ++--------- libvirt-sandbox/libvirt-sandbox-init-util.c | 58 ++++ libvirt-sandbox/libvirt-sandbox-init-util.h | 41 +++ libvirt-sandbox/libvirt-sandbox-util.c | 72 +++++ libvirt-sandbox/libvirt-sandbox-util.h | 5 + libvirt-sandbox/libvirt-sandbox.h | 1 + libvirt-sandbox/libvirt-sandbox.sym | 5 + libvirt-sandbox/tests/test-config.c | 11 + po/POTFILES.in | 1 + 19 files changed, 1112 insertions(+), 149 deletions(-) create mode 100644 libvirt-sandbox/libvirt-sandbox-config-disk.c create mode 100644 libvirt-sandbox/libvirt-sandbox-config-disk.h create mode 100644 libvirt-sandbox/libvirt-sandbox-init-util.c create mode 100644 libvirt-sandbox/libvirt-sandbox-init-util.h create mode 100644 libvirt-sandbox/libvirt-sandbox-util.c -- 2.1.0

1 0

[libvirt] [PATCH v2 0/2] Storage fixes for libvirt.
by Prerna Saxena 25 Jun '15

25 Jun '15

From: Prerna Saxena <prerna(a)linux.vnet.ibm.com> Date: Mon, 22 Jun 2015 06:12:24 -0500 This is the second version of storage fixes. V1 : http://www.redhat.com/archives/libvir-list/2015-May/msg00050.html Summary: ------------ Prerna Saxena (2): Storage: Introduce shadow vol for refresh while the main vol builds. Storage : Fix cloning of raw, sparse volumes src/storage/storage_backend.c | 10 +++++----- src/storage/storage_driver.c | 20 +++++++++++++------- 2 files changed, 18 insertions(+), 12 deletions(-) Changelog: -------------- 1. Dropped patch 1/2 of v1 as per jtomko's suggestion; and introduced a new patch for shadow-volume based cloning. 2. Reworked patch 2/2 incorporating review comments. -- Regards, -- Prerna Saxena Linux Technology Centre, IBM Systems and Technology Lab, Bangalore, India

2 6

[libvirt] [PATCH] vz: Adapt to driver rename
by Michal Privoznik 25 Jun '15

25 Jun '15

In the e6d180f07fb06 commit the parallels driver was renamed to vz. However, there was a commit merged later, which was sent to the list before the rename. The other commit is 6de12b026b73. Fix all the missing renames. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Pushed under build breaker rule. src/vz/vz_sdk.c | 23 ++++++++++++++++------- src/vz/vz_sdk.h | 4 ++-- 2 files changed, 18 insertions(+), 9 deletions(-) diff --git a/src/vz/vz_sdk.c b/src/vz/vz_sdk.c index 1a3aa87..388ea19 100644 --- a/src/vz/vz_sdk.c +++ b/src/vz/vz_sdk.c @@ -2885,14 +2885,20 @@ static int prlsdkAddNet(PRL_HANDLE sdkdom, return ret; } -static void prlsdkDelNet(vzConnPtr privconn, virDomainNetDefPtr net) +static int +prlsdkDelNet(vzConnPtr privconn, virDomainNetDefPtr net) { + int ret = -1; PRL_RESULT pret; PRL_HANDLE vnet = PRL_INVALID_HANDLE; PRL_HANDLE job = PRL_INVALID_HANDLE; - if (net->type != VIR_DOMAIN_NET_TYPE_BRIDGE) - return; + if (net->type != VIR_DOMAIN_NET_TYPE_BRIDGE) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, + _("unplugging network device of type %s is not supported"), + virDomainNetTypeToString(net->type)); + return ret; + } pret = PrlVirtNet_Create(&vnet); prlsdkCheckRetGoto(pret, cleanup); @@ -2904,16 +2910,19 @@ static void prlsdkDelNet(vzConnPtr privconn, virDomainNetDefPtr net) if (PRL_FAILED(pret = waitJob(job))) goto cleanup; + ret = 0; + cleanup: PrlHandle_Free(vnet); + return ret; } int prlsdkAttachNet(virDomainObjPtr dom, - parallelsConnPtr privconn, + vzConnPtr privconn, virDomainNetDefPtr net) { int ret = -1; - parallelsDomObjPtr privdom = dom->privateData; + vzDomObjPtr privdom = dom->privateData; PRL_HANDLE job = PRL_INVALID_HANDLE; if (!IS_CT(dom->def)) { @@ -2998,11 +3007,11 @@ static int prlsdkDelNetAdapter(PRL_HANDLE sdkdom, int idx) } int prlsdkDetachNet(virDomainObjPtr dom, - parallelsConnPtr privconn, + vzConnPtr privconn, virDomainNetDefPtr net) { int ret = -1, idx = -1; - parallelsDomObjPtr privdom = dom->privateData; + vzDomObjPtr privdom = dom->privateData; PRL_HANDLE job = PRL_INVALID_HANDLE; if (!IS_CT(dom->def)) { diff --git a/src/vz/vz_sdk.h b/src/vz/vz_sdk.h index cde8904..80ff69a 100644 --- a/src/vz/vz_sdk.h +++ b/src/vz/vz_sdk.h @@ -67,6 +67,6 @@ prlsdkDetachVolume(virDomainObjPtr dom, virDomainDiskDefPtr disk); int prlsdkGetBlockStats(virDomainObjPtr dom, virDomainDiskDefPtr disk, virDomainBlockStatsPtr stats); int -prlsdkAttachNet(virDomainObjPtr dom, parallelsConnPtr privconn, virDomainNetDefPtr net); +prlsdkAttachNet(virDomainObjPtr dom, vzConnPtr privconn, virDomainNetDefPtr net); int -prlsdkDetachNet(virDomainObjPtr dom, parallelsConnPtr privconn, virDomainNetDefPtr net); +prlsdkDetachNet(virDomainObjPtr dom, vzConnPtr privconn, virDomainNetDefPtr net); -- 2.3.6

2 1

[libvirt] [PATCH v2 00/15] test driver refactors
by Peter Krempa 25 Jun '15

25 Jun '15

Version 2 has a few more fixes as suggested by Michal. Peter Krempa (15): test: Rename testConn to testDriver test: Drop useless forward declaration test: turn 'defaultConn' into a pointer test: Extract code to free testDriver into testDriverFree test: Extract common parts of test driver data allocation test: Drop unused attribute @path from testDriver struct test: Annotate few fields of testDriver structure test: Use atomic access to @nextDomID in struct virTestDriver test: Refactor test driver event sending test: Finalize removal of locking from driver->eventState test: Drop locked access to testDriver->domains test: Refactor test driver domain object retrieval test: Refactor testDomainSetVcpusFlags test: Refactor vcpu pinning and vcpu info retrieval test: Refactor testNodeGetCPUMap src/test/test_driver.c | 1305 +++++++++++++++--------------------------------- 1 file changed, 401 insertions(+), 904 deletions(-) -- 2.4.1

3 25

[libvirt] driver level connection close event
by Nikolay Shirokovskiy 25 Jun '15

25 Jun '15

Notify of connection close event from parallels driver (possibly) wrapped in the remote driver. Discussion. In 1 and 2 patch we forced to some decisions because we don't have a weak reference mechanics. 1 patch. ----------- virConnectCloseCallback is introduced because we can not reference the connection object itself when setting a network layer callback because of how connection close works. A connection close procedure is next: 1. client closes connection 2. a this point nobody else referencing a connection and it is disposed 3. connection dispose unreferencing network connection 4. network connection disposes Thus if we referece a connection in network close callback we never get step 2. virConnectCloseCallback broke this cycle but at cost that clients MUST unregister explicitly before closing connection. This is not good as this unregistration is not really neaded. Client is not telling that it does not want to receive events anymore but rather forced to obey some implementation-driven rules. 2 patch. ----------- We impose requirements on driver implementations which is fragile. Moreover we again need to make explicit unregistrations. Implementation of domain events illustrates this point. remoteDispatchConnectDomainEventRegister does not reference NetClient and makes unregistration before NetClient is disposed but drivers do not meet the formulated requirements. Object event system release lock before delivering event for re-entrance purposes. Shortly we have 2 undesired consequences here. 1. Mandatory unregistration. 2. Imposing multi-threading requirements. Introduction of weak pointers could free us from these artifacts. Next weak reference workflow illustrates this. 1. Take weak reference on object of interest before passing to party. This doesn't break disposing mechanics as weak eference does not prevent from disposing object. Object is disposed but memory is not freed yet if there are weak references. 2. When callback is called we are safe to check if pointer dangling as we make a weak reference before. 3. Release weak reference and this trigger memory freeing if there are no more weak references. daemon/libvirtd.h | 1 + daemon/remote.c | 87 ++++++++++++++++++++++++++++++++ src/datatypes.c | 112 +++++++++++++++++++++++++++++++---------- src/datatypes.h | 21 ++++++-- src/driver-hypervisor.h | 12 +++++ src/libvirt-host.c | 79 ++++++++++-------------------- src/remote/remote_driver.c | 108 +++++++++++++++++++++++++++++++--------- src/remote/remote_protocol.x | 24 +++++++++- src/remote_protocol-structs | 6 ++ src/vz/vz_driver.c | 26 ++++++++++ src/vz/vz_sdk.c | 32 +++++++++++- src/vz/vz_utils.h | 3 + 12 files changed, 396 insertions(+), 115 deletions(-)

2 4

[libvirt] [PATCH] qemu: fix wrong remove guest cfg if migrate fail
by Luyao Huang 25 Jun '15

25 Jun '15

If we get fail in qemuMigrationPrepareAny, we forget check if the vm is persistent then always call qemuDomainRemoveInactive to clean the inactive settings. Add a check to avoid this. This issue was introduce in commit 540c339. Signed-off-by: Luyao Huang <lhuang(a)redhat.com> --- src/qemu/qemu_migration.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 47d49cd..a57a177 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -3432,7 +3432,8 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, VIR_FREE(priv->origname); virPortAllocatorRelease(driver->migrationPorts, priv->nbdPort); priv->nbdPort = 0; - qemuDomainRemoveInactive(driver, vm); + if (!vm->persistent) + qemuDomainRemoveInactive(driver, vm); } virDomainObjEndAPI(&vm); if (event) -- 1.8.3.1

3 2