- Devel - libvirt List Archives

[libvirt] Memory Ballooning for VM Issue
by xuanmao_001 07 Aug '12

07 Aug '12

Hi, I have some problems with qemu ballooning. I saw the qemu docs. I found the qemu monitor command "balloon" that can request VM to change its memory allocation to value(in MB). it requested qemu command line with "-balloon virtio", then I start qemu into monitor mode with memory 512MB. first show balloon information. (qemu) info balloon (qemu)balloon: actual=512 the command above if like this, I think the balloon device worked fine. (qemu) balloon 400 there was no error when I executed this command. but when I executed "info balloon", the result was still 512. Is there something wrong? please give me some ideas. thanks. qemu-kvm version: 1.0.1 linux kernel version: 3.1.6 xuanmao_001

1 0

[libvirt] [PATCH] qemu_agent: support guest-info command
by MATSUDA, Daiki 06 Aug '12

06 Aug '12

Currently, libvirt qemu agent supports some QEMU Guest Agent's commands or use them. But they are not adapted to communication test to the Domain OS. So, QEMU Guest Agent provide 'guest-info' command to display its version and their commands. And I wrote the codes for supporting it. virsh # guest-agent-info RHEL62_32 Version: 1.1.0 Commands: guest-network-get-interfaces guest-suspend-hybrid guest-suspend-ram guest-suspend-disk guest-fsfreeze-thaw guest-fsfreeze-freeze guest-fsfreeze-status guest-file-flush guest-file-seek guest-file-write guest-file-read guest-file-close guest-file-open guest-shutdown guest-info guest-ping guest-sync guest-sync-delimited I am sorry that attached patch is against libvirt-0.9.13-rc1, because my network environment can not access via git and header code is not built with the problem gnulib... Regards MATSUDA Daiki

4 16

[libvirt] [Patch v3 0/3] Add QEMU network helper support
by rmarwah＠linux.vnet.ibm.com 06 Aug '12

06 Aug '12

From: Richa Marwaha <rmarwah(a)linux.vnet.ibm.com> QEMU has a new feature which allows QEMU to execute under an unprivileged user ID and still be able to add a tap device to a Linux network bridge. Below is the link to the QEMU patches for the bridge helper feature: http://lists.gnu.org/archive/html/qemu-devel/2012-01/msg03562.html The existing libvirt tap network device support for adding a tap device to a bridge (-netdev tap) works only when connected to a libvirtd instance running as the privileged system account 'root'. When connected to a libvirtd instance running as an unprivileged user (ie. using the session URI) creation of the tap device fails as follows: error: Failed to start domain F14_64 error: Unable to create tap device vnet%d: Operation not permitted With this support, creating a tap device in the above scenario will be possible. Additionally, hot attaching a tap device to a bridge while running when connected to a libvirtd instance running as an unprivileged user will be possible. Richa Marwaha (3): Add -netdev bridge capabilities Add -netdev bridge support apparmor: QEMU bridge helper policy updates AUTHORS | 1 + examples/apparmor/libvirt-qemu | 21 ++++++++++++++- src/qemu/qemu_capabilities.c | 13 ++++++--- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 57 +++++++++++++++++++++++++++++---------- src/qemu/qemu_command.h | 2 + src/qemu/qemu_hotplug.c | 31 ++++++++++++++------- tests/qemuhelptest.c | 3 +- 8 files changed, 98 insertions(+), 31 deletions(-)

4 10

[libvirt] [PATCH] Fix save/restore with USB controller in XML.
by Ján Tomko 06 Aug '12

06 Aug '12

USB controller gets put on the first place in XML, but the default one is added at the end of the controllers array. Sorting them before checking ABI compatibility solves this. The default USB controller also doesn't get a PCI address assigned, making virDomainDeviceInfoCheckABIStability fail. --- src/conf/domain_conf.c | 55 +++++++++++++++++++++++++++++++++++++++++++++-- 1 files changed, 52 insertions(+), 3 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 58603a3..c98802a 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -9765,8 +9765,10 @@ static bool virDomainControllerDefCheckABIStability(virDomainControllerDefPtr sr } } - if (!virDomainDeviceInfoCheckABIStability(&src->info, &dst->info)) - goto cleanup; + /* don't check device info for the default USB controller */ + if (!(src->type == VIR_DOMAIN_CONTROLLER_TYPE_USB && src->idx == 0 && src->model == -1)) + if (!virDomainDeviceInfoCheckABIStability(&src->info, &dst->info)) + goto cleanup; identical = true; @@ -10180,6 +10182,27 @@ cleanup: return identical; } +static int virDomainControllerCmp(const void *ctrl1, + const void *ctrl2) +{ + virDomainControllerDefPtr c1 = *(virDomainControllerDefPtr*) ctrl1; + virDomainControllerDefPtr c2 = *(virDomainControllerDefPtr*) ctrl2; + + if (c1->type < c2->type) + return -1; + if (c1->type > c2->type) + return 1; + if (c1->idx < c2->idx) + return -1; + if (c1->idx > c2->idx) + return 1; + if (c1->model < c1->model) + return -1; + if (c1->model > c2->model) + return 1; + return 0; +} + /* This compares two configurations and looks for any differences * which will affect the guest ABI. This is primarily to allow @@ -10191,6 +10214,9 @@ bool virDomainDefCheckABIStability(virDomainDefPtr src, bool identical = false; int i; + virDomainControllerDefPtr *scontrollers = NULL; + virDomainControllerDefPtr *dcontrollers = NULL; + if (src->virtType != dst->virtType) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, _("Target domain virt type %s does not match source %s"), @@ -10312,10 +10338,33 @@ bool virDomainDefCheckABIStability(virDomainDefPtr src, goto cleanup; } + /* sort the controllers before comparison */ + if (VIR_ALLOC_N(scontrollers, src->ncontrollers) < 0) { + virReportOOMError(); + goto cleanup; + } + + memcpy(scontrollers, src->controllers, src->ncontrollers*sizeof(src->controllers[0])); + qsort(scontrollers, src->ncontrollers, sizeof(virDomainControllerDefPtr), + virDomainControllerCmp); + + + if (VIR_ALLOC_N(dcontrollers, dst->ncontrollers) < 0) { + virReportOOMError(); + goto cleanup; + } + + memcpy(dcontrollers, dst->controllers, dst->ncontrollers*sizeof(dst->controllers[0])); + qsort(dcontrollers, dst->ncontrollers, sizeof(virDomainControllerDefPtr), + virDomainControllerCmp); + for (i = 0 ; i < src->ncontrollers ; i++) - if (!virDomainControllerDefCheckABIStability(src->controllers[i], dst->controllers[i])) + if (!virDomainControllerDefCheckABIStability(scontrollers[i], dcontrollers[i])) goto cleanup; + VIR_FREE(scontrollers); + VIR_FREE(dcontrollers); + if (src->nfss != dst->nfss) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, _("Target domain filesystem count %d does not match source %d"), -- 1.7.8.6

1 1

[libvirt] [PATCH] json: fix interface locale dependency
by Martin Kletzander 06 Aug '12

06 Aug '12

libvirt creates invalid commands if wrong locale is selected. For example with locale that uses comma as a decimal point, JSON commands created with decimal numbers are invalid because comma separates the entries in JSON. But even when decimal point is affected, grouping is not, because for grouping to be enabled with *printf, there has to be a apostrophe flag specified (and supported). --- Fortunately, there should be no other place where output-formatting is affected by this problem. I tried to change this in various ways with this posted one being the cleanest from my point of view, because: - setting locale is per-proccess, not per-thread (not thread-safe) - there is no number parsing that mangles the output number because of floating point precision src/util/json.c | 23 +++++++++++++++++++++-- 1 files changed, 21 insertions(+), 2 deletions(-) diff --git a/src/util/json.c b/src/util/json.c index 5132989..753a548 100644 --- a/src/util/json.c +++ b/src/util/json.c @@ -23,6 +23,8 @@ #include <config.h> +#include <locale.h> + #include "json.h" #include "memory.h" #include "virterror_internal.h" @@ -44,7 +46,6 @@ /* XXX fixme */ #define VIR_FROM_THIS VIR_FROM_NONE - typedef struct _virJSONParserState virJSONParserState; typedef virJSONParserState *virJSONParserStatePtr; struct _virJSONParserState { @@ -200,9 +201,27 @@ virJSONValuePtr virJSONValueNewNumberUlong(unsigned long long data) virJSONValuePtr virJSONValueNewNumberDouble(double data) { virJSONValuePtr val = NULL; - char *str; + char *str, *radix, *tmp; + if (virAsprintf(&str, "%lf", data) < 0) return NULL; + + /* because printing double is locale-dependent, we could end up + * with invalid JSON code, so we have to do something like this */ + radix = localeconv()->decimal_point; + tmp = strstr(str, radix); + if (tmp) { + *tmp = '.'; + if (strlen(radix) > 1) { + /* if the current locale specifies more characters as + * decimal point then cover the others with decimal + * numbers */ + memcpy(tmp + 1, + tmp + strlen(radix), + strlen(str) - (tmp - str)); + } + } + val = virJSONValueNewNumber(str); VIR_FREE(str); return val; -- 1.7.8.6

4 7

[libvirt] [PATCH] build: drop conditional use of mdns code
by Eric Blake 06 Aug '12

06 Aug '12

Commit 1f6f723 missed a step. At first I was worried that scrubbing the conditionals would lead to a runtime failure when compiled without avahi, but my testing makes it appear that the runtime error will only occur if the .conf files in /etc request mdns advertisement; and the old behavior was to silently ignore the request, so this is actually a better behavior of explicitly failing only when the config requests the impossible. * src/rpc/virnetserver.c: Drop HAVE_AVAHI conditionals; all callers already passed NULL if mdns_adv was not configured. --- In response to: https://www.redhat.com/archives/libvir-list/2012-August/msg00130.html I originally thought it would be harder than this, but my testing seems to show that this works. Caveat: my testing without avahi was done in a VM and not on bare metal, and since nested virt is slower, I may have inadvertently cut an important corner and missed a flaw in my above reasoning about why this is safe. So a close review, including checking all call sites, would be appreciated. src/rpc/virnetserver.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-) diff --git a/src/rpc/virnetserver.c b/src/rpc/virnetserver.c index 248ad9f..295e8fd 100644 --- a/src/rpc/virnetserver.c +++ b/src/rpc/virnetserver.c @@ -36,9 +36,7 @@ #include "util.h" #include "virfile.h" #include "event.h" -#if HAVE_AVAHI -# include "virnetservermdns.h" -#endif +#include "virnetservermdns.h" #ifndef SA_SIGINFO # define SA_SIGINFO 0 @@ -81,10 +79,8 @@ struct _virNetServer { int sigwatch; char *mdnsGroupName; -#if HAVE_AVAHI virNetServerMDNSPtr mdns; virNetServerMDNSGroupPtr mdnsGroup; -#endif size_t nservices; virNetServerServicePtr *services; @@ -364,7 +360,6 @@ virNetServerPtr virNetServerNew(size_t min_workers, virReportOOMError(); goto error; } -#if HAVE_AVAHI if (srv->mdnsGroupName) { if (!(srv->mdns = virNetServerMDNSNew())) goto error; @@ -372,7 +367,6 @@ virNetServerPtr virNetServerNew(size_t min_workers, srv->mdnsGroupName))) goto error; } -#endif if (virMutexInit(&srv->lock) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", @@ -592,14 +586,13 @@ error: int virNetServerAddService(virNetServerPtr srv, virNetServerServicePtr svc, - const char *mdnsEntryName ATTRIBUTE_UNUSED) + const char *mdnsEntryName) { virNetServerLock(srv); if (VIR_EXPAND_N(srv->services, srv->nservices, 1) < 0) goto no_memory; -#if HAVE_AVAHI if (mdnsEntryName) { int port = virNetServerServiceGetPort(svc); @@ -608,7 +601,6 @@ int virNetServerAddService(virNetServerPtr srv, port)) goto error; } -#endif srv->services[srv->nservices-1] = svc; virNetServerServiceRef(svc); @@ -622,9 +614,7 @@ int virNetServerAddService(virNetServerPtr srv, no_memory: virReportOOMError(); -#if HAVE_AVAHI error: -#endif virNetServerUnlock(srv); return -1; } @@ -694,11 +684,9 @@ void virNetServerRun(virNetServerPtr srv) virNetServerLock(srv); -#if HAVE_AVAHI if (srv->mdns && virNetServerMDNSStart(srv->mdns) < 0) goto cleanup; -#endif srv->quit = 0; @@ -826,9 +814,7 @@ void virNetServerFree(virNetServerPtr srv) VIR_FREE(srv->clients); VIR_FREE(srv->mdnsGroupName); -#if HAVE_AVAHI virNetServerMDNSFree(srv->mdns); -#endif virMutexDestroy(&srv->lock); VIR_FREE(srv); -- 1.7.11.2

2 2

[libvirt] [PATCH 00/13 v2] Introduce a virObject module fo reference counting
by Daniel P. Berrange 06 Aug '12

06 Aug '12

A followup to https://www.redhat.com/archives/libvir-list/2012-July/msg00476.html Notable changes - Change virAtomicIntInc to return new value, instead of void for compat with existing API - Use verify_true instead of verify in macros - Most other fixes from previous series

4 29

[libvirt] Debian packaging for the Java bindings
by Wido den Hollander 06 Aug '12

06 Aug '12

Currently the Java bindings for libvirt do not provide a Debian package. Users with Debian based systems like Ubuntu are now forced to compile from source since no binary distribution of the JAR file is available. This patch adds building Debian packages with ant $ ant deb This will build a .deb file in the target directory. It requires dpkg-dev to be installed. Wido

2 5

[libvirt] [test-API][PATCH] Fix xml parser problem when node have both attribute and value
by Wayne Sun 06 Aug '12

06 Aug '12

When xml node have both attribute and value at first level, the parser will broke. After fix, the node key will have a dictionary with both value and attr inside. For example, the xml node: <capacity unit='bytes'>536870912000</capacity> will be parsed into: {u'capacity': {'attr': {u'unit': u'bytes'}, 'value': u'536870912000'}} Also when fetch the attribute key, should use a new param (attrkey) other than exist key in outside loop. Signed-off-by: Wayne Sun <gsun(a)redhat.com> --- utils/xml_parser.py | 22 +++++++++++++++++----- 1 files changed, 17 insertions(+), 5 deletions(-) diff --git a/utils/xml_parser.py b/utils/xml_parser.py index 04e7501..01b928f 100644 --- a/utils/xml_parser.py +++ b/utils/xml_parser.py @@ -88,15 +88,21 @@ class xml_parser(object): if thenode.attributes != None: tmpattr = dict() if thenode.attributes.length > 0: - for key in thenode.attributes.keys(): + for attrkey in thenode.attributes.keys(): tmpattr.update( - {key:thenode.attributes.get(key).nodeValue}) + {attrkey:thenode.attributes.get(attrkey).nodeValue}) attrdic = { "attr":tmpattr } if key in out: if out[key] == None: - out[key] = value if attrdic != None: - out[key].update(attrdic) + if value == None: + out[key] = attrdic + else: + valdic = { "value":value } + valdic.update(attrdic) + out[key] = valdic + else: + out[key] = value elif type(out[key]) == list: if attrdic != None: newdict.update(attrdic) @@ -111,7 +117,13 @@ class xml_parser(object): else: out[key] = value if attrdic != None: - out[key].update(attrdic) + if value == None: + newdict[key] = attrdic + else: + valdic = { "value":value } + valdic.update(attrdic) + newdict = valdic + out[key] = newdict self.parseintodict(thenode, level+1, out, key) return out -- 1.7.1

2 1

[libvirt] [libvirt-php] [PATCH] libvirt_domain_update_device() not defined
by Katelyn Schiesser 04 Aug '12

04 Aug '12

This patch makes libvirt_domain_update_device() available within PHP. Previously it simply threw an error for undefinded function. Katelyn Schiesser (1): Make libvirt_domain_update_device() available. src/libvirt-php.c | 1 + 1 file changed, 1 insertion(+) -- 1.7.11.3

1 1