- Devel - libvirt List Archives

[libvirt] [PATCH 0/3] qemu: support update graphic device persistently
by Wang Rui 04 Dec '14

04 Dec '14

We can change vnc password by using virDomainUpdateDeviceFlags API with live flag. But it can't be changed with config flag. 1/3: improve the error number when update graphics failed 2/3: refactor the function qemuDomainFindGraphics for the future patch 3/3: support updating vnc/spice auth arguments persistently Wang Rui (3): qemu: report properer error number when change graphics failed qemu: revise qemuDomainFindGraphics to be reused in the future patch qemu: make persistent update of graphics device supported src/conf/domain_conf.c | 3 ++- src/qemu/qemu_driver.c | 38 +++++++++++++++++++++++++++++++++++++- src/qemu/qemu_hotplug.c | 23 +++++++++++------------ src/qemu/qemu_hotplug.h | 3 +++ 4 files changed, 53 insertions(+), 14 deletions(-) -- 1.7.12.4

2 5

[libvirt] [PATCHv2 0/9] Let libvirt manage a bridge's FDB
by Laine Stump 04 Dec '14

04 Dec '14

(Everyone - see the request for opinions/ideas towards the bottom) The idea behind these patches is the following: 1) most virtual machines only have a single MAC address behind each interface, and that MAC address is known by libvirt. 2) If we (i.e. libvirt) manually add an entry to the bridge's forwarding database (fdb) for the MAC address associated with a port on the bridge, we can turn off learning and unicast_flooding for that port. 3) kernels starting with 3.15 (and actually working correctly starting in kernel 3.17) will notice that all of a bridge's ports have flood and learning turned off, and in that case will turn off promiscuous mode on all ports. If all but one of the ports have flood/learning turned off, then promiscuous will be turned off on that port (and left on for all the other ports) 4) When (4) can be done, there is a measurable performance advantage. It can also *kind of* help security, as it will prevent a guest from doing anything useful if it changes its MAC address (but won't prevent the guest from *sending* packets with a spoofed MAC address). NB: These only work with a fixed MAC address, and no vlan tags set in the guest. Support for both of those will be coming. HERE IS THE REQUEST FOR OPINIONS/IDEAS: This V2 of the patchset addresses several issues brought up by jferlan on the original series, and changes the name of the attribute from: promiscLinks='yes|no' to fdb='learningWithFlood|managed' I'm somewhat more happy with this new naming than the previous but still looking for better ideas. It is closer to describing what the new code really does, but "learningWithFlood" seems a bit long and awkward, while I have been told that "fdb" is too short and unrecognizeable (I will point out that 1) "fdb" is the same name used by iproute2's "bridge" command, and 2) another bridge option, "stp" is also a three letter acronym that will only be recognized by those familiar with configuring an L2 bridge device or watching NASCAR on Saturday afternoons (or whenever it's on - not a fan myself :-)) Here is a full list of every idea that either I or someone else has come up with since I started thinking about this: promiscLinks='yes|no' After initially going with this for the v1 of the patchset, I later decided against it, because it doesn't describe what libvirt is doing, but only a *possible* side effect on *some* of the ports connected to the bridge (in practice, it only happens to the physdev port). fdb='auto|managed' I like "fdb" as the name of the attribute, because I think it really gets at what libvirt is doing - it is taking over management of the bridge's fdb (Forwarding Database), which ends up providing better performance in several ways. The problem with this proposal is that the two values are kind of ambiguous - it's not clear which one is using the bridge module's built-in management (I had figured this would be "auto"), and which is telling libvirt to manage it ("managed"). (on the other hand, the first option is "ignore the issue, let the underlying system handle it", vs. "libvirt should manage it", so maybe it *is* a reasonable choice). Also, see the comment above about the perceived terseness and obscurity of "fdb". fdb='learningWithFlood|managed' This alternate name was suggested by Michael Tsirkin as a way to unambiguously indicate what was being done in the mode where libvirt isn't involved in the fdb management. There was some criticism on IRC that the name is *too* verbose, especially when contrasted with "fdb". fdb='learning|managed' A suggested shortening of "learningWithFlood". forwardingDatabase='blah' A way to get around criticism of "fdb". I think this is too verbose, but maybe I'm biased :-) [specify each minor item that separately] In order to manage the fdb by itself, libvirt disables "learning" and "unicast_flood" for each tap device, enables "vlan_filtering" for the bridge itself, then adds an fdb entry for each tap to the bridge. There was one suggestion that, rather than trying to come up with a single option that says "do all of these things", we should instead make each of them separately configured. The problem with this is that it makes it too easy to screw up the configuration such that it causes sub-par performance, or simply doesn't work at all. Part of libvirt's job is making it difficult to screw up (or at least easier to succeed); for example, libvirt's virtual networks do a lot of things automatically - create a bridge, add iptables rules for filtering and NAT, run an instance of dnsmasq - over time we've offered the option of tweaking more and more of the details of this setup, but the initial aim was to provide something that worked with as few required (or even permitted) tweaks as possible. I guess what I'm getting at is that I think it would be a mistake to require turning on several different knobs (which individually make little/no sense) in order to get the bridge into this higher performing mode. So - does anyone have an opinion of any of the options offered above, or any ideas for alternates? In the meantime, note that while the default is currently "learningWithFlood" (meaning that that name is never actually directly used/required anywhere, but is just in the RNG and the enum definition), the intent of the people who developed this functionality in the kernel is that eventually it will work so well that libvirt management of the fdb can silently become the default with no visible change in behavior. NOTE: If you want to actually try out these patches, you'll need to apply the following patch which I haven't yet pushed: https://www.redhat.com/archives/libvir-list/2014-November/msg00948.html Also, while the description of V1 stated that patches 08 and 09 were not intended to be pushed yet, due to a problem they caused when restarting libvirtd after an update, that problem has been solved, so I now intend to push patches 08 and 09 along with the rest. Laine Stump (9): util: new functions for setting bridge and bridge port attributes util: functions to manage bridge fdb (forwarding database) conf: new network bridge device attribute fdb network: save bridge name in ActualNetDef when actualType==network too network: store network fdb setting in NetDef actual object network: setup bridge devices for fdb='managed' qemu: setup tap devices for fdb='managed' qemu: always use virDomainNetGetActualBridgeName to get interface's bridge lxc: always use virDomainNetGetActualBridgeName to get interface's bridge docs/formatnetwork.html.in | 42 ++- docs/schemas/network.rng | 9 + src/conf/domain_conf.c | 129 ++++--- src/conf/domain_conf.h | 2 + src/conf/network_conf.c | 50 ++- src/conf/network_conf.h | 11 + src/libvirt_private.syms | 11 + src/lxc/lxc_driver.c | 32 +- src/lxc/lxc_process.c | 32 +- src/network/bridge_driver.c | 74 ++++ src/qemu/qemu_command.c | 53 ++- src/qemu/qemu_hotplug.c | 60 +--- src/util/virnetdevbridge.c | 382 ++++++++++++++++++++- src/util/virnetdevbridge.h | 44 ++- tests/networkxml2xmlin/host-bridge-no-flood.xml | 6 + .../nat-network-explicit-flood.xml | 21 ++ tests/networkxml2xmlout/host-bridge-no-flood.xml | 6 + .../nat-network-explicit-flood.xml | 23 ++ tests/networkxml2xmltest.c | 2 + 19 files changed, 778 insertions(+), 211 deletions(-) create mode 100644 tests/networkxml2xmlin/host-bridge-no-flood.xml create mode 100644 tests/networkxml2xmlin/nat-network-explicit-flood.xml create mode 100644 tests/networkxml2xmlout/host-bridge-no-flood.xml create mode 100644 tests/networkxml2xmlout/nat-network-explicit-flood.xml -- 1.9.3

3 24

[libvirt] [PATCH] Fix handling of whitespae in preprocessor macros for API generator
by Daniel P. Berrange 04 Dec '14

04 Dec '14

The apibuild.py script did not handle whitespace in preprocessor macros, so it failed to detect constants declared with '# define' instead of '#define'. Since we now correctly indent our public header files, we have silently lost all constants from libvirt-api.xml. This also caused us to not detect formatting errors in constant docs --- docs/apibuild.py | 8 ++++++++ include/libvirt/libvirt-host.h | 12 ++++++++---- 2 files changed, 16 insertions(+), 4 deletions(-) diff --git a/docs/apibuild.py b/docs/apibuild.py index 7549a63..9fa9361 100755 --- a/docs/apibuild.py +++ b/docs/apibuild.py @@ -459,6 +459,14 @@ class CLexer: if line[0] == '#': self.tokens = map((lambda x: ('preproc', x)), string.split(line)) + + # We might have whitespace between the '#' and preproc + # macro name, so instead of having a single token element + # of '#define' we might end up with '#' and 'define'. This + # merges them back together + if self.tokens[0][1] == "#": + self.tokens[0] = ('preproc', self.tokens[0][1] + self.tokens[1][1]) + self.tokens = self.tokens[:1] + self.tokens[2:] break l = len(line) if line[0] == '"' or line[0] == "'": diff --git a/include/libvirt/libvirt-host.h b/include/libvirt/libvirt-host.h index 5bd9563..53b529f 100644 --- a/include/libvirt/libvirt-host.h +++ b/include/libvirt/libvirt-host.h @@ -536,28 +536,32 @@ struct _virNodeMemoryStats { */ # define VIR_NODE_MEMORY_SHARED_PAGES_SHARING "shm_pages_sharing" -/* VIR_NODE_MEMORY_SHARED_PAGES_UNSHARED: +/* + * VIR_NODE_MEMORY_SHARED_PAGES_UNSHARED: * * Macro for typed parameter that represents how many pages unique * but repeatedly checked for merging. */ # define VIR_NODE_MEMORY_SHARED_PAGES_UNSHARED "shm_pages_unshared" -/* VIR_NODE_MEMORY_SHARED_PAGES_VOLATILE: +/* + * VIR_NODE_MEMORY_SHARED_PAGES_VOLATILE: * * Macro for typed parameter that represents how many pages changing * too fast to be placed in a tree. */ # define VIR_NODE_MEMORY_SHARED_PAGES_VOLATILE "shm_pages_volatile" -/* VIR_NODE_MEMORY_SHARED_FULL_SCAN: +/* + * VIR_NODE_MEMORY_SHARED_FULL_SCANS: * * Macro for typed parameter that represents how many times all * mergeable areas have been scanned. */ # define VIR_NODE_MEMORY_SHARED_FULL_SCANS "shm_full_scans" -/* VIR_NODE_MEMORY_SHARED_MERGE_ACROSS_NODES: +/* + * VIR_NODE_MEMORY_SHARED_MERGE_ACROSS_NODES: * * Macro for typed parameter that represents whether pages from * different NUMA nodes can be merged. The parameter has type int, -- 2.1.0

2 1

[libvirt] [PATCH] docs: Create html documentation even if XHTML1 DTD is not available to validate
by Ian Campbell 04 Dec '14

04 Dec '14

On a Debian system lacking the w3c-dtd-xhtml package the build fails with: $ make -C docs/ formatcaps.html make: Entering directory '/local/scratch/ianc/devel/libvirt.git/docs' Generating formatcaps.html.tmp I/O error : Attempt to load network entity http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd formatcaps.html.in:2: warning: failed to load external entity "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" ^ I/O error : Attempt to load network entity http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd ../docs/sitemap.html.in:2: warning: failed to load external entity "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" ^ missing XHTML1 DTD rm formatcaps.html.tmp make: Leaving directory '/local/scratch/ianc/devel/libvirt.git/docs' $ ls docs/formatcaps* docs/formatcaps.html.in https://www.redhat.com/archives/libvir-list/2009-November/msg00413.html suggests that the XHTML1 DTD should not be a hard requirement and the docs should be generated but not validated if it is not available. Therefore when the DTD is not available arrange for the .html.tmp file to be propagated to the .html output. Signed-off-by: Ian Campbell <ian.campbell(a)citrix.com> Cc: Daniel Veillard <veillard(a)redhat.com> --- docs/Makefile.am | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/Makefile.am b/docs/Makefile.am index 5485ee9..c5ba688 100644 --- a/docs/Makefile.am +++ b/docs/Makefile.am @@ -232,7 +232,7 @@ internals/%.html.tmp: internals/%.html.in subsite.xsl page.xsl sitemap.html.in SGML_CATALOG_FILES='$(XML_CATALOG_FILE)' \ $(XMLLINT) --catalogs --nonet --format --valid $< > $(srcdir)/$@ \ || { rm $(srcdir)/$@ && exit 1; }; \ - else echo "missing XHTML1 DTD" ; fi ; fi + else echo "missing XHTML1 DTD"; cat $< > $(srcdir)/$@ ; fi ; fi %.php.tmp: %.php.in site.xsl page.xsl sitemap.html.in @if [ -x $(XSLTPROC) ] ; then \ @@ -258,7 +258,7 @@ html/index.html: libvirt-api.xml newapi.xsl page.xsl sitemap.html.in > /dev/null ; then \ SGML_CATALOG_FILES='$(XML_CATALOG_FILE)' \ $(XMLLINT) --catalogs --nonet --valid --noout $(srcdir)/html/*.html ; \ - else echo "missing XHTML1 DTD" ; fi ; fi + else echo "missing XHTML1 DTD"; cat $< > $(srcdir)/$@ ; fi ; fi $(addprefix $(srcdir)/,$(devhelphtml)): $(srcdir)/libvirt-api.xml $(devhelpxsl) $(AM_V_GEN)if [ -x $(XSLTPROC) ] ; then \ -- 2.1.1

3 2

[libvirt] [libvirt-perl][PATCH] Fix constants for VIR_DOMAIN_GET_ALL_STATS_*
by Zhe Peng 04 Dec '14

04 Dec '14

Constants of GET_ALL_STATS_* incorrect,this patch fix them in Domain.pm --- lib/Sys/Virt/Domain.pm | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/lib/Sys/Virt/Domain.pm b/lib/Sys/Virt/Domain.pm index 5680afa..e5ce2dd 100644 --- a/lib/Sys/Virt/Domain.pm +++ b/lib/Sys/Virt/Domain.pm @@ -3287,39 +3287,39 @@ bulk domain stats from C<Sys::Virt::get_all_domain_stats>. =over 4 -=item Sys::Virt::GET_ALL_STATS_ACTIVE +=item Sys::Virt::Domain::GET_ALL_STATS_ACTIVE Include stats for active domains -=item Sys::Virt::GET_ALL_STATS_INACTIVE +=item Sys::Virt::Domain::GET_ALL_STATS_INACTIVE Include stats for inactive domains -=item Sys::Virt::GET_ALL_STATS_OTHER +=item Sys::Virt::Domain::GET_ALL_STATS_OTHER Include stats for other domains -=item Sys::Virt::GET_ALL_STATS_PAUSED +=item Sys::Virt::Domain::GET_ALL_STATS_PAUSED Include stats for paused domains -=item Sys::Virt::GET_ALL_STATS_PERSISTENT +=item Sys::Virt::Domain::GET_ALL_STATS_PERSISTENT Include stats for persistent domains -=item Sys::Virt::GET_ALL_STATS_RUNNING +=item Sys::Virt::Domain::GET_ALL_STATS_RUNNING Include stats for running domains -=item Sys::Virt::GET_ALL_STATS_SHUTOFF +=item Sys::Virt::Domain::GET_ALL_STATS_SHUTOFF Include stats for shutoff domains -=item Sys::Virt::GET_ALL_STATS_TRANSIENT +=item Sys::Virt::Domain::GET_ALL_STATS_TRANSIENT Include stats for transient domains -=item Sys::Virt::GET_ALL_STATS_ENFORCE_STATS +=item Sys::Virt::Domain::GET_ALL_STATS_ENFORCE_STATS Require that all requested stats fields are returned -- 1.9.0

2 1

[libvirt] [PATCH] network: don't allow multiple dhcp sections
by Kyle DeFrancia 04 Dec '14

04 Dec '14

This resolves: https://bugzilla.redhat.com/show_bug.cgi?id=907779 A <dhcp> element can exist in only one IPv4 address and one IPv6 address per network. This patch enforces that in virNetworkUpdate. --- src/conf/network_conf.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) diff --git a/src/conf/network_conf.c b/src/conf/network_conf.c index 067334e..a914962 100644 --- a/src/conf/network_conf.c +++ b/src/conf/network_conf.c @@ -3472,6 +3472,31 @@ virNetworkIpDefByIndex(virNetworkDefPtr def, int parentIndex) } static int +virNetworkDefUpdateCheckMultiDHCP(virNetworkDefPtr def, + virNetworkIpDefPtr ipdef) +{ + int family = VIR_SOCKET_ADDR_FAMILY(&ipdef->address); + size_t i; + virNetworkIpDefPtr ip; + + for (i = 0; + (ip = virNetworkDefGetIpByIndex(def, family, i)); + i++) { + if (ip != ipdef) { + if (ip->nranges || ip->nhosts) { + virReportError(VIR_ERR_OPERATION_INVALID, + _("dhcp is supported only for a " + "single %s address on each network"), + (family == AF_INET) ? "IPv4" : "IPv6"); + return -1; + } + } + } + + return 0; +} + +static int virNetworkDefUpdateIPDHCPHost(virNetworkDefPtr def, unsigned int command, int parentIndex, @@ -3536,6 +3561,9 @@ virNetworkDefUpdateIPDHCPHost(virNetworkDefPtr def, } else if ((command == VIR_NETWORK_UPDATE_COMMAND_ADD_FIRST) || (command == VIR_NETWORK_UPDATE_COMMAND_ADD_LAST)) { + if (virNetworkDefUpdateCheckMultiDHCP(def, ipdef) < 0) + goto cleanup; + /* log error if an entry with same name/address/ip already exists */ for (i = 0; i < ipdef->nhosts; i++) { if ((host.mac && @@ -3643,6 +3671,9 @@ virNetworkDefUpdateIPDHCPRange(virNetworkDefPtr def, if ((command == VIR_NETWORK_UPDATE_COMMAND_ADD_FIRST) || (command == VIR_NETWORK_UPDATE_COMMAND_ADD_LAST)) { + if (virNetworkDefUpdateCheckMultiDHCP(def, ipdef) < 0) + goto cleanup; + if (i < ipdef->nranges) { char *startip = virSocketAddrFormat(&range.start); char *endip = virSocketAddrFormat(&range.end); -- 2.1.0

2 1

[libvirt] [PATCH] cpu-driver: Fix the cross driver function call
by Daniel Hansel 04 Dec '14

04 Dec '14

For Intel and PowerPC the implementation is calling a cpu driver function across driver layers (i.e. from qemu driver directly to cpu driver). The correct behavior is to use libvirt API functionality to perform such a inter-driver call. This patch introduces a new cpu driver API function getModels() to retrieve the cpu models. The currect implementation to process the cpu_map XML content is transferred to the INTEL and PowerPC cpu driver specific API functions. Additionally processing the cpu_map XML file is not safe due to the fact that the cpu map does not exist for all architectures. Therefore it is better to encapsulate the processing in the architecture specific cpu drivers. Signed-off-by: Daniel Hansel <daniel.hansel(a)linux.vnet.ibm.com> Reviewed-by: Boris Fiuczynski <fiuczy(a)linux.vnet.ibm.com> Reviewed-by: Viktor Mihajlovski <mihajlov(a)linux.vnet.ibm.com> --- src/cpu/cpu.c | 68 +++++++++------------------------------------------ src/cpu/cpu.h | 4 +++ src/cpu/cpu_powerpc.c | 37 ++++++++++++++++++++++++++++ src/cpu/cpu_x86.c | 33 +++++++++++++++++++++++++ 4 files changed, 86 insertions(+), 56 deletions(-) diff --git a/src/cpu/cpu.c b/src/cpu/cpu.c index 08bec5e..788f688 100644 --- a/src/cpu/cpu.c +++ b/src/cpu/cpu.c @@ -724,43 +724,6 @@ cpuModelIsAllowed(const char *model, return false; } -struct cpuGetModelsData -{ - char **data; - size_t len; /* It includes the last element of DATA, which is NULL. */ -}; - -static int -cpuGetArchModelsCb(cpuMapElement element, - xmlXPathContextPtr ctxt, - void *cbdata) -{ - char *name; - struct cpuGetModelsData *data = cbdata; - if (element != CPU_MAP_ELEMENT_MODEL) - return 0; - - name = virXPathString("string(@name)", ctxt); - if (name == NULL) - return -1; - - if (!data->data) { - VIR_FREE(name); - data->len++; - return 0; - } - - return VIR_INSERT_ELEMENT(data->data, data->len - 1, data->len, name); -} - - -static int -cpuGetArchModels(const char *arch, struct cpuGetModelsData *data) -{ - return cpuMapLoad(arch, cpuGetArchModelsCb, data); -} - - /** * cpuGetModels: * @@ -774,18 +737,17 @@ cpuGetArchModels(const char *arch, struct cpuGetModelsData *data) int cpuGetModels(const char *archName, char ***models) { - struct cpuGetModelsData data; - virArch arch; struct cpuArchDriver *driver; - data.data = NULL; - data.len = 1; + virArch arch; + + VIR_DEBUG("arch=%s", archName); arch = virArchFromString(archName); if (arch == VIR_ARCH_NONE) { virReportError(VIR_ERR_INVALID_ARG, _("cannot find architecture %s"), archName); - goto error; + return -1; } driver = cpuGetSubDriver(arch); @@ -793,21 +755,15 @@ cpuGetModels(const char *archName, char ***models) virReportError(VIR_ERR_INVALID_ARG, _("cannot find a driver for the architecture %s"), archName); - goto error; + return -1; } - if (models && VIR_ALLOC_N(data.data, data.len) < 0) - goto error; - - if (cpuGetArchModels(driver->name, &data) < 0) - goto error; - - if (models) - *models = data.data; - - return data.len - 1; + if (!driver->getModels) { + virReportError(VIR_ERR_NO_SUPPORT, + _("CPU driver for %s has no CPU model support"), + virArchToString(arch)); + return -1; + } - error: - virStringFreeList(data.data); - return -1; + return driver->getModels(models); } diff --git a/src/cpu/cpu.h b/src/cpu/cpu.h index 339964c..09e9538 100644 --- a/src/cpu/cpu.h +++ b/src/cpu/cpu.h @@ -100,6 +100,9 @@ typedef char * typedef virCPUDataPtr (*cpuArchDataParse) (const char *xmlStr); +typedef int +(*cpuArchGetModels) (char ***models); + struct cpuArchDriver { const char *name; const virArch *arch; @@ -115,6 +118,7 @@ struct cpuArchDriver { cpuArchHasFeature hasFeature; cpuArchDataFormat dataFormat; cpuArchDataParse dataParse; + cpuArchGetModels getModels; }; diff --git a/src/cpu/cpu_powerpc.c b/src/cpu/cpu_powerpc.c index 67cb9ff..155d93e 100644 --- a/src/cpu/cpu_powerpc.c +++ b/src/cpu/cpu_powerpc.c @@ -649,6 +649,42 @@ ppcBaseline(virCPUDefPtr *cpus, goto cleanup; } +static int +ppcGetModels(char ***models) +{ + struct ppc_map *map; + struct ppc_model *model; + char *name; + size_t nmodels = 0; + + if (!(map = ppcLoadMap())) + goto error; + + if (models && VIR_ALLOC_N(*models, 0) < 0) + goto error; + + model = map->models; + while (model != NULL) { + if (VIR_STRDUP(name, model->name) < 0) + goto error; + + if (VIR_INSERT_ELEMENT(*models, 0, nmodels, name) < 0) + goto error; + + model = model->next; + } + + cleanup: + ppcMapFree(map); + + return nmodels; + + error: + virStringFreeList(*models); + nmodels = -1; + goto cleanup; +} + struct cpuArchDriver cpuDriverPowerPC = { .name = "ppc64", .arch = archs, @@ -662,4 +698,5 @@ struct cpuArchDriver cpuDriverPowerPC = { .baseline = ppcBaseline, .update = ppcUpdate, .hasFeature = NULL, + .getModels = ppcGetModels, }; diff --git a/src/cpu/cpu_x86.c b/src/cpu/cpu_x86.c index 026b54e..f6dcba4 100644 --- a/src/cpu/cpu_x86.c +++ b/src/cpu/cpu_x86.c @@ -2160,6 +2160,38 @@ x86HasFeature(const virCPUData *data, return ret; } +static int +x86GetModels(char ***models) +{ + const struct x86_map *map; + struct x86_model *model; + char *name; + size_t nmodels = 0; + + if (!(map = virCPUx86GetMap())) + return -1; + + if (models && VIR_ALLOC_N(*models, 0) < 0) + goto error; + + model = map->models; + while (model != NULL) { + if (VIR_STRDUP(name, model->name) < 0) + goto error; + + if (VIR_INSERT_ELEMENT(*models, 0, nmodels, name) < 0) + goto error; + + model = model->next; + } + + return nmodels; + + error: + virStringFreeList(*models); + return -1; +} + struct cpuArchDriver cpuDriverX86 = { .name = "x86", @@ -2180,4 +2212,5 @@ struct cpuArchDriver cpuDriverX86 = { .hasFeature = x86HasFeature, .dataFormat = x86CPUDataFormat, .dataParse = x86CPUDataParse, + .getModels = x86GetModels, }; -- 1.8.5.5

4 5

[libvirt] [PATCH] qemu: Fix virsh freeze when blockcopy storage file is removed
by Erik Skultety 04 Dec '14

04 Dec '14

If someone removes blockcopy storage file when still in mirroring phase and then requesting blockjob abort using pivot, virsh cmd freezes. This is not an issue with older qemu versions which did not support asynchronous jobs (which we prefer by default). As we have reached the mirroring phase successfully, polling monitor for blockjob info always returns 1 and the loop never ends. This fix introduces a check for qemuDomainBlockPivot return code, possibly skipping the asynchronous waiting completely, if an error occured and asynchronous waiting was the preferred method. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1139567 --- src/qemu/qemu_driver.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 5dc62b0..2e11938 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -15606,6 +15606,8 @@ qemuDomainBlockJobImpl(virDomainObjPtr vm, if (disk->mirror && (flags & VIR_DOMAIN_BLOCK_JOB_ABORT_PIVOT)) { ret = qemuDomainBlockPivot(conn, driver, vm, device, disk); + if (ret < 0 && async) + goto endjob; goto waitjob; } if (disk->mirror) { -- 1.9.3

3 2

[libvirt] [PATCH] qemu: output error when try to hotplug unsupport console
by Luyao Huang 04 Dec '14

04 Dec '14

https://bugzilla.redhat.com/show_bug.cgi?id=1164627 When try to use attach-device to hotplug a qemu unsupport sonsole, command will return success, and add a console in vm's xml.But this doesn't work in qemu, qemu doesn't support these console.Add a error output when try to hotplug a unsupport console in qemuBuildConsoleChrDeviceStr. Signed-off-by: Luyao Huang <lhuang(a)redhat.com> --- src/qemu/qemu_command.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 1399ce4..2bf4a83 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -10069,13 +10069,17 @@ qemuBuildConsoleChrDeviceStr(char **deviceStr, break; case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_SERIAL: + break; case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_NONE: case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_XEN: case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_UML: case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_LXC: case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_OPENVZ: case VIR_DOMAIN_CHR_CONSOLE_TARGET_TYPE_LAST: - break; + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, + _("unsupported console target type %s"), + NULLSTR(virDomainChrConsoleTargetTypeToString(chr->targetType))); + return ret; } ret = 0; -- 1.8.3.1

4 4

[libvirt] [PATCH] network: Let domains be restricted to local DNS
by Josh Stone 04 Dec '14

04 Dec '14

This adds a new "localOnly" attribute on the domain element of the network xml. With this set to "yes", DNS requests under that domain will only be resolved by libvirt's dnsmasq, never forwarded upstream. This was how it worked before commit f69a6b987d616, and I found that functionality useful. For example, I have my host's NetworkManager dnsmasq configured to forward that domain to libvirt's dnsmasq, so I can easily resolve guest names from outside. But if libvirt's dnsmasq doesn't know a name and forwards it to the host, I'd get an endless forwarding loop. Now I can set localOnly="yes" to prevent the loop. Signed-off-by: Josh Stone <jistone(a)redhat.com> Cc: Laine Stump <laine(a)laine.org> --- docs/formatnetwork.html.in | 12 +++++++++++- docs/schemas/network.rng | 3 +++ src/conf/network_conf.c | 5 +++++ src/conf/network_conf.h | 1 + src/network/bridge_driver.c | 5 +++++ .../networkxml2confdata/nat-network-dns-local-domain.conf | 14 ++++++++++++++ tests/networkxml2confdata/nat-network-dns-local-domain.xml | 9 +++++++++ tests/networkxml2conftest.c | 1 + 8 files changed, 49 insertions(+), 1 deletion(-) create mode 100644 tests/networkxml2confdata/nat-network-dns-local-domain.conf create mode 100644 tests/networkxml2confdata/nat-network-dns-local-domain.xml diff --git a/docs/formatnetwork.html.in b/docs/formatnetwork.html.in index dc438aee8622..defcdba00930 100644 --- a/docs/formatnetwork.html.in +++ b/docs/formatnetwork.html.in @@ -82,7 +82,7 @@ <pre> ... <bridge name="virbr0" stp="on" delay="5"/> - <domain name="example.com"/> + <domain name="example.com" localOnly="no"/> <forward mode="nat" dev="eth0"/> ...</pre> @@ -113,6 +113,16 @@ a <code><forward></code> mode of "nat" or "route" (or an isolated network with no <code><forward></code> element). Since 0.4.5 + + + If the optional <code>localOnly</code> attribute on the + <code>domain</code> element is "yes", then DNS requests under + this domain will only be resolved by the virtual network's own + DNS server - they will not be forwarded to the host's upstream + DNS server. If <code>localOnly</code> is "no", and by + default, unresolved requests will be forwarded. + Since 1.2.11 + </dd> <dt><code>forward</code></dt> <dd>Inclusion of the <code>forward</code> element indicates that diff --git a/docs/schemas/network.rng b/docs/schemas/network.rng index 4546f8037580..a1da28092375 100644 --- a/docs/schemas/network.rng +++ b/docs/schemas/network.rng @@ -225,6 +225,9 @@ <optional> <element name="domain"> <attribute name="name"><ref name="dnsName"/></attribute> + <optional> + <attribute name="localOnly"><ref name="virYesNo"/></attribute> + </optional> </element> </optional> diff --git a/src/conf/network_conf.c b/src/conf/network_conf.c index 067334e87cb0..61451c39805f 100644 --- a/src/conf/network_conf.c +++ b/src/conf/network_conf.c @@ -2083,6 +2083,11 @@ virNetworkDefParseXML(xmlXPathContextPtr ctxt) /* Parse network domain information */ def->domain = virXPathString("string(./domain[1]/@name)", ctxt); + tmp = virXPathString("string(./domain[1]/@localOnly)", ctxt); + if (tmp) { + def->domain_local = STRCASEEQ(tmp, "yes"); + VIR_FREE(tmp); + } if ((bandwidthNode = virXPathNode("./bandwidth", ctxt)) != NULL && (def->bandwidth = virNetDevBandwidthParse(bandwidthNode, -1)) == NULL) diff --git a/src/conf/network_conf.h b/src/conf/network_conf.h index 660cd2d10cd1..6308a7dcfbf7 100644 --- a/src/conf/network_conf.h +++ b/src/conf/network_conf.h @@ -232,6 +232,7 @@ struct _virNetworkDef { char *bridge; /* Name of bridge device */ char *domain; + bool domain_local; /* Choose not to forward dns for this domain */ unsigned long delay; /* Bridge forward delay (ms) */ bool stp; /* Spanning tree protocol */ virMacAddr mac; /* mac address of bridge device */ diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 6cb421c52850..dfa375d3aa72 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -912,6 +912,11 @@ networkDnsmasqConfContents(virNetworkObjPtr network, } if (network->def->domain) { + if (network->def->domain_local) { + virBufferAsprintf(&configbuf, + "local=/%s/\n", + network->def->domain); + } virBufferAsprintf(&configbuf, "domain=%s\n" "expand-hosts\n", diff --git a/tests/networkxml2confdata/nat-network-dns-local-domain.conf b/tests/networkxml2confdata/nat-network-dns-local-domain.conf new file mode 100644 index 000000000000..5f41b9186cbc --- /dev/null +++ b/tests/networkxml2confdata/nat-network-dns-local-domain.conf @@ -0,0 +1,14 @@ +##WARNING: THIS IS AN AUTO-GENERATED FILE. CHANGES TO IT ARE LIKELY TO BE +##OVERWRITTEN AND LOST. Changes to this configuration should be made using: +## virsh net-edit default +## or other application using the libvirt API. +## +## dnsmasq conf file created by libvirt +strict-order +local=/example.com/ +domain=example.com +expand-hosts +except-interface=lo +bind-dynamic +interface=virbr0 +addn-hosts=/var/lib/libvirt/dnsmasq/default.addnhosts diff --git a/tests/networkxml2confdata/nat-network-dns-local-domain.xml b/tests/networkxml2confdata/nat-network-dns-local-domain.xml new file mode 100644 index 000000000000..a92d71f1f2f6 --- /dev/null +++ b/tests/networkxml2confdata/nat-network-dns-local-domain.xml @@ -0,0 +1,9 @@ +<network> + <name>default</name> + <uuid>81ff0d90-c91e-6742-64da-4a736edb9a9c</uuid> + <forward dev='eth0' mode='nat'/> + <bridge name='virbr0' stp='on' delay='0' /> + <domain name='example.com' localOnly='yes'/> + <ip address='192.168.122.1' netmask='255.255.255.0'> + </ip> +</network> diff --git a/tests/networkxml2conftest.c b/tests/networkxml2conftest.c index 4f1d9345ffe4..d2aa8c62cfcd 100644 --- a/tests/networkxml2conftest.c +++ b/tests/networkxml2conftest.c @@ -146,6 +146,7 @@ mymain(void) DO_TEST("nat-network-dns-hosts", full); DO_TEST("nat-network-dns-forward-plain", full); DO_TEST("nat-network-dns-forwarders", full); + DO_TEST("nat-network-dns-local-domain", full); DO_TEST("dhcp6-network", dhcpv6); DO_TEST("dhcp6-nat-network", dhcpv6); DO_TEST("dhcp6host-routed-network", dhcpv6); -- 2.1.0

4 9