September 2014 - Devel - libvirt List Archives

[libvirt] [PATCHv6 00/11] bulk stats: QEMU implementation
by Peter Krempa 17 Sep '14

17 Sep '14

After my review on Francesco's series I've tested the patches a bit more and found a few problems. I'm re-sending the series and also I've added a few patches that tweak the documentation for this. Francesco Romani (8): qemu: bulk stats: extend internal collection API qemu: bulk stats: implement CPU stats group qemu: bulk stats: implement balloon group qemu: bulk stats: implement VCPU group qemu: bulk stats: implement interface group qemu: bulk stats: implement block group virsh: add options to query bulk stats group qemu: bulk stats: add block allocation information Peter Krempa (3): lib: De-duplicate stats group documentation for all stats functions lib: Document that virConnectGetAllDomainStats may omit some stats fields man: virsh: Add docs for supported stats groups include/libvirt/libvirt.h.in | 5 + src/libvirt.c | 76 +++++- src/qemu/qemu_driver.c | 543 +++++++++++++++++++++++++++++++++++++------ src/qemu/qemu_monitor.c | 26 +++ src/qemu/qemu_monitor.h | 21 ++ src/qemu/qemu_monitor_json.c | 227 +++++++++++++----- src/qemu/qemu_monitor_json.h | 4 + tools/virsh-domain-monitor.c | 35 +++ tools/virsh.pod | 51 +++- 9 files changed, 844 insertions(+), 144 deletions(-) -- 2.1.0

2 24

[libvirt] [PATCH 0/5] Add support for turning off offloading for virtio-net
by Ján Tomko 17 Sep '14

17 Sep '14

Ján Tomko (5): conf: split out virtio net driver formatting Add virSwitch to data types conf: remove redundant local variable conf: add options for disabling segment offloading qemu: wire up virtio-net segment offloading options docs/formatdomain.html.in | 27 ++++ docs/schemas/basictypes.rng | 6 + docs/schemas/domaincommon.rng | 25 +++ src/conf/domain_conf.c | 176 ++++++++++++++++----- src/conf/domain_conf.h | 5 + src/qemu/qemu_command.c | 20 +++ .../qemuxml2argv-net-virtio-disable-offloads.args | 8 + .../qemuxml2argv-net-virtio-disable-offloads.xml | 32 ++++ tests/qemuxml2argvtest.c | 2 + tests/qemuxml2xmltest.c | 1 + 10 files changed, 262 insertions(+), 40 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-disable-offloads.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-disable-offloads.xml -- 1.8.5.5

4 19

[libvirt] [PATCH v2 0/2] Expose UEFI binary path
by Michal Privoznik 17 Sep '14

17 Sep '14

This is practically reworked v1 from Cole. Cole Robinson (1): domaincaps: Expose UEFI binary path, if it exists Michal Privoznik (1): qemu_capabilities: Change virQEMUCapsFillDomainCaps signature docs/formatdomaincaps.html.in | 6 ++ docs/schemas/domaincaps.rng | 17 ++++-- src/conf/domain_capabilities.c | 29 ++++++++++ src/conf/domain_capabilities.h | 8 +++ src/qemu/qemu_capabilities.c | 53 +++++++++++++---- src/qemu/qemu_capabilities.h | 9 ++- src/qemu/qemu_driver.c | 7 ++- tests/domaincapsschemadata/domaincaps-full.xml | 2 + .../domaincaps-qemu_1.6.50-1.xml | 1 + tests/domaincapstest.c | 66 ++++++++++++++++++---- 10 files changed, 167 insertions(+), 31 deletions(-) -- 1.8.5.5

5 11

Re: [libvirt] [Qemu-devel] migration: qemu-coroutine-lock.c:141: qemu_co_mutex_unlock: Assertion `mutex->locked == 1' failed
by Eric Blake 17 Sep '14

17 Sep '14

[adding libvirt list] On 09/17/2014 09:04 AM, Stefan Hajnoczi wrote: > On Wed, Sep 17, 2014 at 10:25 AM, Paolo Bonzini <pbonzini(a)redhat.com> wrote: >> -----BEGIN PGP SIGNED MESSAGE----- >> Hash: SHA1 >> >> Il 17/09/2014 11:06, Stefan Hajnoczi ha scritto: >>> I think the fundamental problem here is that the mirror block job >>> on the source host does not synchronize with live migration. >>> >>> Remember the mirror block job iterates on the dirty bitmap >>> whenever it feels like. >>> >>> There is no guarantee that the mirror block job has quiesced before >>> migration handover takes place, right? >> >> Libvirt does that. Migration is started only once storage mirroring >> is out of the bulk phase, and the handover looks like: >> >> 1) migration completes >> >> 2) because the source VM is stopped, the disk has quiesced on the source > > But the mirror block job might still be writing out dirty blocks. > >> 3) libvirt sends block-job-complete > > No, it sends block-job-cancel after the source QEMU's migration has > completed. See the qemuMigrationCancelDriveMirror() call in > src/qemu/qemu_migration.c:qemuMigrationRun(). > >> 4) libvirt receives BLOCK_JOB_COMPLETED. The disk has now quiesced on >> the destination as well. > > I don't see where this happens in the libvirt source code. Libvirt > doesn't care about block job events for drive-mirror during migration. > > And that's why there could still be I/O going on (since > block-job-cancel is asynchronous). > >> 5) the VM is started on the destination >> >> 6) the NBD server is stopped on the destination and the source VM is quit. >> >> It is actually a feature that storage migration is completed >> asynchronously with respect to RAM migration. The problem is that >> qcow2_invalidate_cache happens between (3) and (5), and it doesn't >> like the concurrent I/O received by the NBD server. > > I agree that qcow2_invalidate_cache() (and any other invalidate cache > implementations) need to allow concurrent I/O requests. > > Either I'm misreading the libvirt code or libvirt is not actually > ensuring that the block job on the source has cancelled/completed > before the guest is resumed on the destination. So I think there is > still a bug, maybe Eric can verify this? You may indeed be correct that libvirt is not waiting long enough for the block job to be gone on the source before resuming on the destination. I didn't write that particular code, so I'm cc'ing the libvirt list, but I can try and take a look into it, since it's related to code I've recently touched in getting libvirt to support active layer block commit. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

1 0

[libvirt] [PATCH v2] rpc: make daemon spawning a bit more intelligent
by Martin Kletzander 17 Sep '14

17 Sep '14

This way it behaves more like the daemon itself does (acquiring a pidfile, deleting the socket before binding, etc.). Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=927369 Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1138604 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/rpc/virnetsocket.c | 65 +++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 57 insertions(+), 8 deletions(-) diff --git a/src/rpc/virnetsocket.c b/src/rpc/virnetsocket.c index 80aeddf..7be1492 100644 --- a/src/rpc/virnetsocket.c +++ b/src/rpc/virnetsocket.c @@ -51,9 +51,11 @@ #include "virlog.h" #include "virfile.h" #include "virthread.h" +#include "virpidfile.h" #include "virprobe.h" #include "virprocess.h" #include "virstring.h" +#include "dirname.h" #include "passfd.h" #if WITH_SSH2 @@ -541,7 +543,10 @@ int virNetSocketNewConnectUNIX(const char *path, const char *binary, virNetSocketPtr *retsock) { + char *binname = NULL; + char *pidpath = NULL; int fd, passfd = -1; + int pidfd = -1; virSocketAddr localAddr; virSocketAddr remoteAddr; @@ -580,16 +585,47 @@ int virNetSocketNewConnectUNIX(const char *path, goto error; } + if (!(binname = last_component(binary)) || binname[0] == '\0') { + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Cannot determine basename for binary '%s'"), + binary); + goto error; + } + + if (virPidFileConstructPath(false, NULL, binname, &pidpath) < 0) + goto error; + + pidfd = virPidFileAcquirePath(pidpath, false, getpid()); + VIR_FREE(pidpath); + if (pidfd < 0) { + /* + * This can happen in a very rare case of two clients spawning two + * daemons at the same time, and the error in the logs that gets + * reset here can be a clue to some future debugging. + */ + virResetLastError(); + spawnDaemon = false; + goto retry; + } + if ((passfd = socket(PF_UNIX, SOCK_STREAM, 0)) < 0) { virReportSystemError(errno, "%s", _("Failed to create socket")); goto error; } /* - * We have to fork() here, because umask() is set - * per-process, chmod() is racy and fchmod() has undefined - * behaviour on sockets according to POSIX, so it doesn't - * work outside Linux. + * We already even acquired the pidfile, so no one else should be using + * @path right now. So we're OK to unlink it and paying attention to + * the return value makes no real sense here. Only if it's not an + * abstract socket, of course. + */ + if (path[0] != '@') + unlink(path); + + /* + * We have to fork() here, because umask() is set per-process, chmod() + * is racy and fchmod() has undefined behaviour on sockets according to + * POSIX, so it doesn't work outside Linux. */ if ((pid = virFork()) < 0) goto error; @@ -607,12 +643,15 @@ int virNetSocketNewConnectUNIX(const char *path, if (status != EXIT_SUCCESS) { /* - * OK, so the subprocces failed to bind() the socket. This may mean - * that another daemon was starting at the same time and succeeded - * with its bind(). So we'll try connecting again, but this time - * without spawning the daemon. + * OK, so the child failed to bind() the socket. This may mean that + * another daemon was starting at the same time and succeeded with + * its bind() (even though it should not happen because we using a + * pidfile for the race). So we'll try connecting again, but this + * time without spawning the daemon. */ spawnDaemon = false; + VIR_FORCE_CLOSE(pidfd); + VIR_FORCE_CLOSE(passfd); goto retry; } @@ -629,6 +668,12 @@ int virNetSocketNewConnectUNIX(const char *path, goto error; } + /* + * Do we need to eliminate the super-rare race here any more? It would + * need incorporating the following VIR_FORCE_CLOSE() into a + * virCommandHook inside a virNetSocketForkDaemon(). + */ + VIR_FORCE_CLOSE(pidfd); if (virNetSocketForkDaemon(binary, passfd) < 0) goto error; } @@ -645,8 +690,12 @@ int virNetSocketNewConnectUNIX(const char *path, return 0; error: + if (pidfd > 0) + virPidFileDeletePath(pidpath); + VIR_FREE(pidpath); VIR_FORCE_CLOSE(fd); VIR_FORCE_CLOSE(passfd); + VIR_FORCE_CLOSE(pidfd); if (spawnDaemon) unlink(path); return -1; -- 2.1.0

2 4

[libvirt] [PATCH] domaincaps: Expose UEFI binary path, if it exists
by Cole Robinson 17 Sep '14

17 Sep '14

Check to see if the UEFI binary mentioned in qemu.conf actually exists, and if so expose it in domcapabilities like <loader ...> <value>/path/to/ovmf</value> </loader> We introduce some generic domcaps infrastructure for handling a dynamic list of string values, it may be of use for future bits. --- docs/formatdomaincaps.html.in | 6 ++++ docs/schemas/domaincaps.rng | 17 ++++++--- src/conf/domain_capabilities.c | 23 ++++++++++++ src/conf/domain_capabilities.h | 8 +++++ src/qemu/qemu_driver.c | 24 +++++++++++++ tests/domaincapsschemadata/domaincaps-full.xml | 1 + tests/domaincapstest.c | 49 +++++++++++++++++++++----- 7 files changed, 115 insertions(+), 13 deletions(-) diff --git a/docs/formatdomaincaps.html.in b/docs/formatdomaincaps.html.in index 34d746d..6959dfe 100644 --- a/docs/formatdomaincaps.html.in +++ b/docs/formatdomaincaps.html.in @@ -105,6 +105,7 @@ ... <os supported='yes'> <loader supported='yes'> + <value>/usr/share/OVMF/OVMF_CODE.fd</value> <enum name='type'> <value>rom</value> <value>pflash</value> @@ -122,6 +123,11 @@ <p>For the <code>loader</code> element, the following can occur:</p> <dl> + <dt>value</dt> + <dd>List of known loader paths. Currently this is only used + to advertise known locations of OVMF binaries for qemu. Binaries + will only be listed if they actually exist on disk.</dd> + <dt>type</dt> <dd>Whether loader is a typical BIOS (<code>rom</code>) or an UEFI binary (<code>pflash</code>). This refers to diff --git a/docs/schemas/domaincaps.rng b/docs/schemas/domaincaps.rng index ad8d966..dfdb9b9 100644 --- a/docs/schemas/domaincaps.rng +++ b/docs/schemas/domaincaps.rng @@ -46,6 +46,9 @@ <define name='loader'> <element name='loader'> + <optional> + <ref name='value'/> + </optional> <ref name='supported'/> <ref name='enum'/> </element> @@ -85,6 +88,14 @@ </element> </define> + <define name='value'> + <zeroOrMore> + <element name='value'> + <text/> + </element> + </zeroOrMore> + </define> + <define name='supported'> <attribute name='supported'> <choice> @@ -100,11 +111,7 @@ <attribute name='name'> <text/> </attribute> - <zeroOrMore> - <element name='value'> - <text/> - </element> - </zeroOrMore> + <ref name='value'/> </element> </zeroOrMore> </define> diff --git a/src/conf/domain_capabilities.c b/src/conf/domain_capabilities.c index 5a3c8e7..44e422a 100644 --- a/src/conf/domain_capabilities.c +++ b/src/conf/domain_capabilities.c @@ -46,6 +46,15 @@ static int virDomainCapsOnceInit(void) VIR_ONCE_GLOBAL_INIT(virDomainCaps) +static void +virDomainCapsValuesFree(virDomainCapsValuesPtr values) +{ + size_t i; + + for (i = 0; i < values->nvalues; i++) { + VIR_FREE(values->values[i]); + } +} static void virDomainCapsDispose(void *obj) @@ -54,6 +63,8 @@ virDomainCapsDispose(void *obj) VIR_FREE(caps->path); VIR_FREE(caps->machine); + + virDomainCapsValuesFree(&caps->os.loader.values); } @@ -156,6 +167,17 @@ virDomainCapsEnumFormat(virBufferPtr buf, return ret; } +static void +virDomainCapsValuesFormat(virBufferPtr buf, + virDomainCapsValuesPtr values) +{ + size_t i; + + for (i = 0; i < values->nvalues; i++) { + virBufferAsprintf(buf, "<value>%s</value>\n", values->values[i]); + } +} + #define FORMAT_PROLOGUE(item) \ do { \ virBufferAsprintf(buf, "<" #item " supported='%s'%s\n", \ @@ -185,6 +207,7 @@ virDomainCapsLoaderFormat(virBufferPtr buf, { FORMAT_PROLOGUE(loader); + virDomainCapsValuesFormat(buf, &loader->values); ENUM_PROCESS(loader, type, virDomainLoaderTypeToString); ENUM_PROCESS(loader, readonly, virTristateBoolTypeToString); diff --git a/src/conf/domain_capabilities.h b/src/conf/domain_capabilities.h index 768646b..3d5aaa3 100644 --- a/src/conf/domain_capabilities.h +++ b/src/conf/domain_capabilities.h @@ -37,6 +37,13 @@ struct _virDomainCapsEnum { unsigned int values; /* Bitmask of values supported in the corresponding enum */ }; +typedef struct _virDomainCapsValues virDomainCapsValues; +typedef virDomainCapsValues *virDomainCapsValuesPtr; +struct _virDomainCapsValues { + char **values; /* raw string values */ + size_t nvalues; /* number of strings */ +}; + typedef struct _virDomainCapsDevice virDomainCapsDevice; typedef virDomainCapsDevice *virDomainCapsDevicePtr; struct _virDomainCapsDevice { @@ -47,6 +54,7 @@ typedef struct _virDomainCapsLoader virDomainCapsLoader; typedef virDomainCapsLoader *virDomainCapsLoaderPtr; struct _virDomainCapsLoader { virDomainCapsDevice device; + virDomainCapsValues values; virDomainCapsEnum type; /* Info about virDomainLoader */ virDomainCapsEnum readonly; /* Info about readonly:virTristateBool */ }; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 6008aeb..4dd9d14 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -17282,6 +17282,8 @@ qemuConnectGetDomainCapabilities(virConnectPtr conn, int virttype; /* virDomainVirtType */ virDomainCapsPtr domCaps = NULL; int arch = virArchFromHost(); /* virArch */ + virQEMUDriverConfigPtr cfg = NULL; + size_t i; virCheckFlags(0, ret); @@ -17356,8 +17358,30 @@ qemuConnectGetDomainCapabilities(virConnectPtr conn, virQEMUCapsFillDomainCaps(domCaps, qemuCaps); + cfg = virQEMUDriverGetConfig(driver); + for (i = 0; i < cfg->nloader; i++) { + char *filename = cfg->loader[i]; + + if (access(filename, F_OK) == 0) { + if (VIR_REALLOC_N(domCaps->os.loader.values.values, + domCaps->os.loader.values.nvalues + 1) < 0) { + goto cleanup; + } + domCaps->os.loader.values.nvalues++; + + if (VIR_STRDUP(domCaps->os.loader.values.values[ + domCaps->os.loader.values.nvalues - 1], + filename) < 0) { + goto cleanup; + } + } else { + VIR_DEBUG("loader filename=%s doesn't exist", filename); + } + } + ret = virDomainCapsFormat(domCaps); cleanup: + virObjectUnref(cfg); virObjectUnref(domCaps); virObjectUnref(qemuCaps); return ret; diff --git a/tests/domaincapsschemadata/domaincaps-full.xml b/tests/domaincapsschemadata/domaincaps-full.xml index 9722772..e7f41d7 100644 --- a/tests/domaincapsschemadata/domaincaps-full.xml +++ b/tests/domaincapsschemadata/domaincaps-full.xml @@ -6,6 +6,7 @@ <vcpu max='255'/> <os supported='yes'> <loader supported='yes'> + <value>/foo/test</value> <enum name='type'> <value>rom</value> <value>pflash</value> diff --git a/tests/domaincapstest.c b/tests/domaincapstest.c index f240643..ddc4d02 100644 --- a/tests/domaincapstest.c +++ b/tests/domaincapstest.c @@ -28,16 +28,37 @@ #define VIR_FROM_THIS VIR_FROM_NONE -typedef void (*virDomainCapsFill)(virDomainCapsPtr domCaps, - void *opaque); +typedef int (*virDomainCapsFill)(virDomainCapsPtr domCaps, + void *opaque); #define SET_ALL_BITS(x) \ memset(&(x.values), 0xff, sizeof(x.values)) -static void +static int +fillValues(virDomainCapsValuesPtr values) +{ + int ret = -1; + + if (VIR_ALLOC_N(values->values, 1) < 0) { + goto cleanup; + } + values->nvalues = 1; + + if (VIR_STRDUP(values->values[0], "/foo/test") < 0) { + goto cleanup; + } + + ret = 0; + cleanup: + return ret; +} + +static int fillAll(virDomainCapsPtr domCaps, void *opaque ATTRIBUTE_UNUSED) { + int ret = -1; + virDomainCapsOSPtr os = &domCaps->os; virDomainCapsLoaderPtr loader = &os->loader; virDomainCapsDeviceDiskPtr disk = &domCaps->disk; @@ -49,6 +70,9 @@ fillAll(virDomainCapsPtr domCaps, loader->device.supported = true; SET_ALL_BITS(loader->type); SET_ALL_BITS(loader->readonly); + if (fillValues(&loader->values) < 0) { + goto cleanup; + } disk->device.supported = true; SET_ALL_BITS(disk->diskDevice); @@ -60,12 +84,16 @@ fillAll(virDomainCapsPtr domCaps, SET_ALL_BITS(hostdev->subsysType); SET_ALL_BITS(hostdev->capsType); SET_ALL_BITS(hostdev->pciBackend); + + ret = 0; + cleanup: + return ret; } #ifdef WITH_QEMU # include "testutilsqemu.h" -static void +static int fillQemuCaps(virDomainCapsPtr domCaps, void *opaque) { @@ -82,6 +110,8 @@ fillQemuCaps(virDomainCapsPtr domCaps, VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT, VIR_DOMAIN_HOSTDEV_PCI_BACKEND_KVM, VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO); + + return 0; } #endif /* WITH_QEMU */ @@ -94,16 +124,19 @@ buildVirDomainCaps(const char *emulatorbin, virDomainCapsFill fillFunc, void *opaque) { - virDomainCapsPtr domCaps; + virDomainCapsPtr domCaps, ret = NULL; if (!(domCaps = virDomainCapsNew(emulatorbin, machine, arch, type))) goto cleanup; - if (fillFunc) - fillFunc(domCaps, opaque); + if (fillFunc && fillFunc(domCaps, opaque) < 0) { + virObjectUnref(domCaps); + goto cleanup; + } + ret = domCaps; cleanup: - return domCaps; + return ret; } struct test_virDomainCapsFormatData { -- 2.1.0

3 2

[libvirt] [PATCH 1/1] qemu: Add support for multiple versions of 'pseries' machine type
by Pradipta Kr. Banerjee 17 Sep '14

17 Sep '14

qemu: Add support for multiple versions of 'pseries' machine type qemu for IBM Power processor architecture is adding functionality for supporting multiple 'pseries' machine type versions, each with different capabilities. This patch is for supporting the same Signed-off-by: Pradipta Kr. Banerjee <bpradip(a)in.ibm.com> --- src/qemu/qemu_command.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index e5270bd..3bc3a3f 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -706,7 +706,7 @@ qemuSetSCSIControllerModel(virDomainDefPtr def, } } else { if ((def->os.arch == VIR_ARCH_PPC64) && - STREQ(def->os.machine, "pseries")) { + STRPREFIX(def->os.machine, "pseries")) { *model = VIR_DOMAIN_CONTROLLER_MODEL_SCSI_IBMVSCSI; } else if (virQEMUCapsGet(qemuCaps, QEMU_CAPS_SCSI_LSI)) { *model = VIR_DOMAIN_CONTROLLER_MODEL_SCSI_LSILOGIC; @@ -1253,7 +1253,7 @@ int qemuDomainAssignSpaprVIOAddresses(virDomainDefPtr def, for (i = 0; i < def->nserials; i++) { if (def->serials[i]->deviceType == VIR_DOMAIN_CHR_DEVICE_TYPE_SERIAL && (def->os.arch == VIR_ARCH_PPC64) && - STREQ(def->os.machine, "pseries")) + STRPREFIX(def->os.machine, "pseries")) def->serials[i]->info.type = VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO; if (qemuAssignSpaprVIOAddress(def, &def->serials[i]->info, VIO_ADDR_SERIAL) < 0) @@ -1262,7 +1262,7 @@ int qemuDomainAssignSpaprVIOAddresses(virDomainDefPtr def, if (def->nvram) { if (def->os.arch == VIR_ARCH_PPC64 && - STREQ(def->os.machine, "pseries")) + STRPREFIX(def->os.machine, "pseries")) def->nvram->info.type = VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO; if (qemuAssignSpaprVIOAddress(def, &def->nvram->info, VIO_ADDR_NVRAM) < 0) @@ -9454,7 +9454,7 @@ qemuBuildCommandLine(virConnectPtr conn, if (def->nvram) { if (def->os.arch == VIR_ARCH_PPC64 && - STREQ(def->os.machine, "pseries")) { + STRPREFIX(def->os.machine, "pseries")) { if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_DEVICE_NVRAM)) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("nvram device is not supported by " @@ -9571,7 +9571,7 @@ qemuBuildSerialChrDeviceStr(char **deviceStr, { virBuffer cmd = VIR_BUFFER_INITIALIZER; - if ((arch == VIR_ARCH_PPC64) && STREQ(machine, "pseries")) { + if ((arch == VIR_ARCH_PPC64) && STRPREFIX(machine, "pseries")) { if (serial->deviceType == VIR_DOMAIN_CHR_DEVICE_TYPE_SERIAL && serial->info.type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO) { virBufferAsprintf(&cmd, "spapr-vty,chardev=char%s", @@ -9994,7 +9994,7 @@ qemuParseCommandLineDisk(virDomainXMLOptionPtr xmlopt, goto error; if (((dom->os.arch == VIR_ARCH_PPC64) && - dom->os.machine && STREQ(dom->os.machine, "pseries"))) + dom->os.machine && STRPREFIX(dom->os.machine, "pseries"))) def->bus = VIR_DOMAIN_DISK_BUS_SCSI; else def->bus = VIR_DOMAIN_DISK_BUS_IDE; @@ -10087,7 +10087,7 @@ qemuParseCommandLineDisk(virDomainXMLOptionPtr xmlopt, if (STREQ(values[i], "ide")) { def->bus = VIR_DOMAIN_DISK_BUS_IDE; if (((dom->os.arch == VIR_ARCH_PPC64) && - dom->os.machine && STREQ(dom->os.machine, "pseries"))) { + dom->os.machine && STRPREFIX(dom->os.machine, "pseries"))) { virReportError(VIR_ERR_INTERNAL_ERROR, _("pseries systems do not support ide devices '%s'"), val); goto error; @@ -11332,7 +11332,7 @@ qemuParseCommandLine(virCapsPtr qemuCaps, if (STREQ(arg, "-cdrom")) { disk->device = VIR_DOMAIN_DISK_DEVICE_CDROM; if (((def->os.arch == VIR_ARCH_PPC64) && - def->os.machine && STREQ(def->os.machine, "pseries"))) + def->os.machine && STRPREFIX(def->os.machine, "pseries"))) disk->bus = VIR_DOMAIN_DISK_BUS_SCSI; if (VIR_STRDUP(disk->dst, "hdc") < 0) goto error; @@ -11348,7 +11348,7 @@ qemuParseCommandLine(virCapsPtr qemuCaps, else disk->bus = VIR_DOMAIN_DISK_BUS_SCSI; if (((def->os.arch == VIR_ARCH_PPC64) && - def->os.machine && STREQ(def->os.machine, "pseries"))) + def->os.machine && STRPREFIX(def->os.machine, "pseries"))) disk->bus = VIR_DOMAIN_DISK_BUS_SCSI; } if (VIR_STRDUP(disk->dst, arg + 1) < 0) -- 1.9.3

3 3

[libvirt] [PATCH V2] libxl: support hvm direct kernel boot
by Chunyan Liu 16 Sep '14

16 Sep '14

Xen libxl can support Xen HVM direct kernel boot now. To support Xen HVM direct kernel boot in libvirt, it still needs some changes to libvirt code: accept HVM direct kernel boot related configuration in xml, parse those info into virDomainDefPtr, and fill in related parts of libxl_domain_build_info, so that libxl can handle. This patch is just to do this. Signed-off-by: Chunyan Liu <cyliu(a)suse.com> --- Changes: - fix Jim's comments src/libxl/libxl_conf.c | 18 +++++++++++++++ src/xenconfig/xen_common.c | 57 ++++++++++++++++++++++++++++++++++------------ 2 files changed, 60 insertions(+), 15 deletions(-) diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c index acba69c..a5bda64 100644 --- a/src/libxl/libxl_conf.c +++ b/src/libxl/libxl_conf.c @@ -704,6 +704,15 @@ libxlMakeDomBuildInfo(virDomainDefPtr def, if (VIR_STRDUP(b_info->u.hvm.boot, bootorder) < 0) goto error; +#ifdef LIBXL_HAVE_BUILDINFO_KERNEL + if (VIR_STRDUP(b_info->cmdline, def->os.cmdline) < 0) + goto error; + if (VIR_STRDUP(b_info->kernel, def->os.kernel) < 0) + goto error; + if (VIR_STRDUP(b_info->ramdisk, def->os.initrd) < 0) + goto error; +#endif + if (def->nserials) { if (def->nserials > 1) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, @@ -748,6 +757,14 @@ libxlMakeDomBuildInfo(virDomainDefPtr def, virStringSplit(def->os.bootloaderArgs, " \t\n", 0))) goto error; } +#ifdef LIBXL_HAVE_BUILDINFO_KERNEL + if (VIR_STRDUP(b_info->cmdline, def->os.cmdline) < 0) + goto error; + if (VIR_STRDUP(b_info->kernel, def->os.kernel) < 0) + goto error; + if (VIR_STRDUP(b_info->ramdisk, def->os.initrd) < 0) + goto error; +#else if (VIR_STRDUP(b_info->u.pv.cmdline, def->os.cmdline) < 0) goto error; if (def->os.kernel) { @@ -758,6 +775,7 @@ libxlMakeDomBuildInfo(virDomainDefPtr def, } if (VIR_STRDUP(b_info->u.pv.ramdisk, def->os.initrd) < 0) goto error; +#endif } return 0; diff --git a/src/xenconfig/xen_common.c b/src/xenconfig/xen_common.c index 32954f3..3cf7553 100644 --- a/src/xenconfig/xen_common.c +++ b/src/xenconfig/xen_common.c @@ -39,6 +39,9 @@ #include "virstring.h" #include "xen_common.h" +#if WITH_LIBXL +# include <libxl.h> +#endif /* * Convenience method to grab a long int from the config file object @@ -1053,6 +1056,27 @@ xenParseGeneralMeta(virConfPtr conf, virDomainDefPtr def, virCapsPtr caps) return 0; } +static int +xenParseCmdline(virConfPtr conf, virDomainDefPtr def) +{ + const char *extra, *root; + + if (xenConfigGetString(conf, "extra", &extra, NULL) < 0) + return -1; + + if (xenConfigGetString(conf, "root", &root, NULL) < 0) + return -1; + + if (root) { + if (virAsprintf(&def->os.cmdline, "root=%s %s", root, extra) < 0) + return -1; + } else { + if (VIR_STRDUP(def->os.cmdline, extra) < 0) + return -1; + } + + return 0; +} static int xenParseOS(virConfPtr conf, virDomainDefPtr def) @@ -1065,9 +1089,25 @@ xenParseOS(virConfPtr conf, virDomainDefPtr def) if (STREQ(def->os.type, "hvm")) { const char *boot; +#ifdef LIBXL_HAVE_BUILDINFO_KERNEL + if (xenConfigCopyStringOpt(conf, "kernel", &def->os.kernel) < 0) + return -1; + + if (def->os.kernel && strstr(def->os.kernel, "hvmloader")) { + /* 'kernel' set to 'hvmloader' will be ignored in libxl */ + VIR_FREE(def->os.kernel); + } + + if (xenConfigCopyStringOpt(conf, "ramdisk", &def->os.initrd) < 0) + return -1; + + if (xenParseCmdline(conf, def) < 0) + return -1; +#else if (VIR_ALLOC(def->os.loader) < 0 || - xenConfigCopyString(conf, "kernel", &def->os.loader->path) < 0) + xenConfigCopyStringOpt(conf, "kernel", &def->os.loader->path) < 0) return -1; +#endif if (xenConfigGetString(conf, "boot", &boot, "c") < 0) return -1; @@ -1091,8 +1131,6 @@ xenParseOS(virConfPtr conf, virDomainDefPtr def) def->os.nBootDevs++; } } else { - const char *extra, *root; - if (xenConfigCopyStringOpt(conf, "bootloader", &def->os.bootloader) < 0) return -1; if (xenConfigCopyStringOpt(conf, "bootargs", &def->os.bootloaderArgs) < 0) @@ -1104,19 +1142,8 @@ xenParseOS(virConfPtr conf, virDomainDefPtr def) if (xenConfigCopyStringOpt(conf, "ramdisk", &def->os.initrd) < 0) return -1; - if (xenConfigGetString(conf, "extra", &extra, NULL) < 0) + if (xenParseCmdline(conf, def) < 0) return -1; - - if (xenConfigGetString(conf, "root", &root, NULL) < 0) - return -1; - - if (root) { - if (virAsprintf(&def->os.cmdline, "root=%s %s", root, extra) < 0) - return -1; - } else { - if (VIR_STRDUP(def->os.cmdline, extra) < 0) - return -1; - } } return 0; -- 1.8.5.2

2 1

[libvirt] [PATCH] qemu: time: Report errors if agent command fails
by Peter Krempa 16 Sep '14

16 Sep '14

Commit b606bbb4 broke reporting of errors when setting of guest time fails via the guest agent as the return value is not checked and later overwritten by the return value qemuMonitorRTCResetReinjection(); Fix this by checking the return value before resetting the RTC reinjection. --- src/qemu/qemu_driver.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 6008aeb..15ad64d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -17134,6 +17134,9 @@ qemuDomainSetTime(virDomainPtr dom, rv = qemuAgentSetTime(priv->agent, seconds, nseconds, rtcSync); qemuDomainObjExitAgent(vm); + if (rv < 0) + goto endjob; + if (!virDomainObjIsActive(vm)) { virReportError(VIR_ERR_OPERATION_INVALID, "%s", _("domain is not running")); -- 2.1.0

2 2

[libvirt] [PATCH] util: storage: Copy driver type when initializing chain element
by Peter Krempa 16 Sep '14

16 Sep '14

virStorageSourceInitChainElement initializes a new storage chain element for use as a new disk source. If the new element doesn't contain the driver name, copy it from the old source. This fixes issue where a disk would forget the driver after a snapshot. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1140984 --- src/util/virstoragefile.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/util/virstoragefile.c b/src/util/virstoragefile.c index 13056a7..960aa23 100644 --- a/src/util/virstoragefile.c +++ b/src/util/virstoragefile.c @@ -1927,6 +1927,10 @@ virStorageSourceInitChainElement(virStorageSourcePtr newelem, virStorageSourceSeclabelsCopy(newelem, old) < 0) goto cleanup; + if (!newelem->driverName && + VIR_STRDUP(newelem->driverName, old->driverName) < 0) + goto cleanup; + newelem->shared = old->shared; newelem->readonly = old->readonly; -- 2.1.0

2 2