- Devel - libvirt List Archives

[libvirt] [PATCH] docs: Move the "Network Filtering" page one level up in the hierarchy
by Matthias Bolte 17 Jan '11

17 Jan '11

"Network Filtering" is not directly related to "Networks". Suggested by Daniel P. Berrange. --- docs/sitemap.html.in | 10 ++++------ 1 files changed, 4 insertions(+), 6 deletions(-) diff --git a/docs/sitemap.html.in b/docs/sitemap.html.in index 490450e..7077038 100644 --- a/docs/sitemap.html.in +++ b/docs/sitemap.html.in @@ -111,12 +111,10 @@ <li> <a href="formatnetwork.html">Networks</a> The virtual network XML format - <ul> - <li> - <a href="formatnwfilter.html">Network Filtering</a> - Network filter XML format - </li> - </ul> + </li> + <li> + <a href="formatnwfilter.html">Network Filtering</a> + Network filter XML format </li> <li> <a href="formatstorage.html">Storage</a> -- 1.7.0.4

2 1

[libvirt] website: API reference for virterror contains libvirt API also
by Matthias Bolte 17 Jan '11

17 Jan '11

The API reference website [1] for virterror contains the API reference for virterror as expected. But the complete libvirt API reference, that is already found at [2], is appended too. I looked into the XSLT process that generates those two HTML files from the libvirt-api.xml to understand why this happens, but I couldn't locate the problem. [1] http://www.libvirt.org/html/libvirt-virterror.html [2] http://www.libvirt.org/html/libvirt-libvirt.html Matthias

1 1

[libvirt] [PATCH] tests: Remove obsolete secaatest
by Matthias Bolte 15 Jan '11

15 Jan '11

Before the security driver was refactored in d6623003 seclabeltest and secaatest were basically the same. seclabeltest was meant for SELinux and secaatest for AppArmor. Both tests exited early when the specific security driver backend wasn't enabled. With the new security manager trying to initialize a disabled security driver backend is an error that can't be distinguished from other errors anymore. Therefore, the updated seclabeltest just asks for the first available backend as this will always work even with SELinux and AppArmor backend being disabled due to the new Nop backend. Remove the obsolete secaatest and compile and run the seclabeltest unconditional. This fixes make check on systems that support AppArmor. --- tests/Makefile.am | 24 ------------------------ tests/secaatest.c | 45 --------------------------------------------- 2 files changed, 0 insertions(+), 69 deletions(-) delete mode 100644 tests/secaatest.c diff --git a/tests/Makefile.am b/tests/Makefile.am index 345cf46..72beb23 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -98,13 +98,7 @@ if WITH_VMX check_PROGRAMS += vmx2xmltest xml2vmxtest endif -if WITH_SECDRIVER_SELINUX check_PROGRAMS += seclabeltest -endif - -if WITH_SECDRIVER_APPARMOR -check_PROGRAMS += secaatest -endif if WITH_CIL check_PROGRAMS += object-locking @@ -191,13 +185,7 @@ if WITH_VMX TESTS += vmx2xmltest xml2vmxtest endif -if WITH_SECDRIVER_SELINUX TESTS += seclabeltest -endif - -if WITH_SECDRIVER_APPARMOR -TESTS += secaatest -endif if WITH_LIBVIRTD check_PROGRAMS += eventtest @@ -376,21 +364,9 @@ commandhelper_SOURCES = \ commandhelper_CFLAGS = -Dabs_builddir="\"`pwd`\"" commandhelper_LDADD = $(LDADDS) -if WITH_SECDRIVER_SELINUX seclabeltest_SOURCES = \ seclabeltest.c seclabeltest_LDADD = ../src/libvirt_driver_security.la $(LDADDS) -else -EXTRA_DIST += seclabeltest.c -endif - -if WITH_SECDRIVER_APPARMOR -secaatest_SOURCES = \ - secaatest.c -secaatest_LDADD = ../src/libvirt_driver_security.la $(LDADDS) -else -EXTRA_DIST += secaatest.c -endif qparamtest_SOURCES = \ qparamtest.c testutils.h testutils.c diff --git a/tests/secaatest.c b/tests/secaatest.c deleted file mode 100644 index d9d6b4a..0000000 --- a/tests/secaatest.c +++ /dev/null @@ -1,45 +0,0 @@ -#include <config.h> - -#include <unistd.h> -#include <stdlib.h> -#include <stdio.h> -#include <string.h> -#include <errno.h> -#include "security/security_driver.h" - -int -main (int argc ATTRIBUTE_UNUSED, char **argv ATTRIBUTE_UNUSED) -{ - int ret; - - const char *doi, *model; - virSecurityDriverPtr security_drv; - - ret = virSecurityDriverStartup (&security_drv, "apparmor", false); - if (ret == -1) - { - fprintf (stderr, "Failed to start security driver"); - exit (-1); - } - /* No security driver wanted to be enabled: just return */ - if (ret == -2) - return 0; - - model = virSecurityDriverGetModel (security_drv); - if (!model) - { - fprintf (stderr, "Failed to copy secModel model: %s", - strerror (errno)); - exit (-1); - } - - doi = virSecurityDriverGetDOI (security_drv); - if (!doi) - { - fprintf (stderr, "Failed to copy secModel DOI: %s", - strerror (errno)); - exit (-1); - } - - return 0; -} -- 1.7.0.4

2 2

[libvirt] [PATCH] virsh: Use WITH_SECDRIVER_APPARMOR to detect AppArmor support
by Matthias Bolte 15 Jan '11

15 Jan '11

There is no ENABLE_SECDRIVER_APPARMOR. --- tools/virsh.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 4f103ae..b86d92f 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -11847,7 +11847,7 @@ vshShowVersion(vshControl *ctl ATTRIBUTE_UNUSED) vshPrint(ctl, "\n"); vshPrint(ctl, "%s", _(" Miscellaneous:")); -#ifdef ENABLE_SECDRIVER_APPARMOR +#ifdef WITH_SECDRIVER_APPARMOR vshPrint(ctl, " AppArmor"); #endif #ifdef WITH_SECDRIVER_SELINUX -- 1.7.0.4

2 2

[libvirt] Bug in virsh memtune on RHEL6?
by Justin Clift 15 Jan '11

15 Jan '11

Hi all, Getting strange numbers from the virsh memtune command on RHEL 6. (git head compiled, rather than RHEL 6 packages) virsh # list Id Name State ---------------------------------- 4 Fedora_14_x64 running virsh # memtune Fedora_14_x64 hard_limit : 9007199254740991 soft_limit : 9007199254740991 swap_hard_limit: 9007199254740991 Those numbers are supposed to be kilobytes, but the host box itself only has 12GB ram. Bug? Regards and best wishes, Justin Clift

4 16

[libvirt] OpenVZ and Libvirt
by Nathan Stratton 14 Jan '11

14 Jan '11

I have been working to get libvirt to work with openvz and kvm. I have had no issues with KVM but libvirt is hit or miss. Sometimes I get it to work, but most of the time I get back "out of memory". I have tried all recent versions of libvirt and vzctl from openvz. Any pointers? 17:52:31.974: debug : virConnectOpen:1425 : name=openvz:///system 17:52:31.974: debug : virEventRunOnce:593 : Poll on 12 handles 0x7f6394001420 timeout -1 17:52:31.974: debug : do_open:1204 : name "openvz:///system" to URI components: scheme openvz opaque (null) authority (null) server (null) user (null) port 0 path /system 17:52:31.974: debug : do_open:1242 : trying driver 0 (Test) ... 17:52:31.974: debug : do_open:1248 : driver 0 Test returned DECLINED 17:52:31.974: debug : do_open:1242 : trying driver 1 (Xen) ... 17:52:31.974: debug : do_open:1248 : driver 1 Xen returned DECLINED 17:52:31.974: debug : do_open:1242 : trying driver 2 (OPENVZ) ... 17:52:32.003: debug : virEventRunOnce:595 : Poll got 1 event 17:52:32.003: debug : virEventDispatchTimeouts:405 : Dispatch 3 17:52:32.003: debug : virEventDispatchHandles:450 : Dispatch 12 17:52:32.003: debug : virEventDispatchHandles:464 : i=0 w=1 17:52:32.003: debug : virEventDispatchHandles:464 : i=1 w=2 17:52:32.003: debug : virEventDispatchHandles:477 : Dispatch n=1 f=5 w=2 e=1 0x1cfa820 17:52:32.003: info : qemudDispatchSignalEvent:401 : Received unexpected signal 17 17:52:32.003: debug : virEventDispatchHandles:464 : i=2 w=3 17:52:32.003: debug : virEventDispatchHandles:464 : i=3 w=4 17:52:32.003: debug : virEventDispatchHandles:464 : i=4 w=5 17:52:32.003: debug : virEventDispatchHandles:464 : i=5 w=6 17:52:32.003: debug : virEventDispatchHandles:464 : i=6 w=7 17:52:32.003: debug : virEventDispatchHandles:464 : i=7 w=8 17:52:32.003: debug : virEventDispatchHandles:464 : i=8 w=9 17:52:32.003: debug : virEventDispatchHandles:464 : i=9 w=10 17:52:32.003: debug : virEventDispatchHandles:464 : i=10 w=11 17:52:32.003: debug : virEventDispatchHandles:464 : i=11 w=12 17:52:32.003: debug : virEventCleanupTimeouts:495 : Cleanup 3 17:52:32.003: debug : virEventCleanupHandles:536 : Cleanupo 12 17:52:32.006: error : openvzReadNetworkConf:318 : out of memory 17:52:32.009: error : openvzReadNetworkConf:318 : out of memory 17:52:32.013: error : openvzReadNetworkConf:318 : out of memory 17:52:32.016: error : openvzReadNetworkConf:318 : out of memory 17:52:32.020: error : openvzReadNetworkConf:318 : out of memory 17:52:32.023: error : openvzReadNetworkConf:318 : out of memory 17:52:32.027: error : openvzReadNetworkConf:318 : out of memory 17:52:32.030: error : openvzReadNetworkConf:318 : out of memory 17:52:32.034: error : openvzReadNetworkConf:318 : out of memory 17:52:32.037: error : openvzReadNetworkConf:318 : out of memory 17:52:32.041: error : openvzReadNetworkConf:318 : out of memory 17:52:32.044: error : openvzReadNetworkConf:318 : out of memory 17:52:32.048: error : openvzReadNetworkConf:318 : out of memory 17:52:32.051: error : openvzReadNetworkConf:318 : out of memory 17:52:32.055: error : openvzReadNetworkConf:318 : out of memory 17:52:32.058: error : openvzReadNetworkConf:318 : out of memory 17:52:32.062: error : openvzReadNetworkConf:318 : out of memory 17:52:32.062: error : openvzLoadDomains:534 : out of memory 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384009340 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384009340 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384007780 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384007780 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384005be0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384005be0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f63840080d0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f63840080d0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f63840089f0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f63840089f0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384006500 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384006500 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f63840049b0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f63840049b0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384004ef0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384004ef0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384008e80 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384008e80 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384006070 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384006070 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384007c10 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384007c10 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384006990 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384006990 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384006e20 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384006e20 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384005330 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384005330 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f63840072e0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f63840072e0 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f6384005750 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f6384005750 17:52:32.062: debug : virDomainObjUnref:777 : obj=0x7f63840085f0 refs=0 17:52:32.062: debug : virDomainObjFree:753 : obj=0x7f63840085f0 17:52:32.062: debug : do_open:1248 : driver 2 OPENVZ returned ERROR 17:52:32.062: debug : virUnrefConnect:294 : unref connection 0x7f6384000bd0 1 17:52:32.062: debug : virReleaseConnect:249 : release connection 0x7f6384000bd0 libvir: OpenVZ error : out of memory ><> Nathan Stratton CTO, BlinkMind, Inc. nathan at robotics.net nathan at blinkmind.com http://www.robotics.net http://www.blinkmind.com

2 1

[libvirt] [PATCHv2] Enable tuning of qemu network tap device "sndbuf" size
by Laine Stump 14 Jan '11

14 Jan '11

This is in response to a request in: https://bugzilla.redhat.com/show_bug.cgi?id=665293 In short, under heavy load, it's possible for qemu's networking to lock up due to the tap device's default 1MB sndbuf being inadequate. adding "sndbuf=0" to the qemu commandline -netdevice option will alleviate this problem (sndbuf=0 actually sets it to 0xffffffff). Because we must be able to explicitly specify "0" as a value, the standard practice of "0 means not specified" won't work here. Instead, virDomainNetDef also has a sndbuf_specified, which defaults to 0, but is set to 1 if some value was given. The sndbuf value is put inside a <tune> element of each <interface> in the domain. The intent is that further tunable settings will also be placed inside this element. <interface type='network'> ... <tune> <sndbuf>0</sndbuf> ... </tune> </interface> --- Changes from V1: sndbuf_specified is now a bool rather than an int bitfield. sndbuf is now unsigned long. Made possible by eblake's patch adding new virXPath* and virStrToLong_* functions. docs/schemas/domain.rng | 10 ++++++++++ src/conf/domain_conf.c | 29 +++++++++++++++++++++++++++-- src/conf/domain_conf.h | 4 ++++ src/qemu/qemu_command.c | 19 +++++++++++++++++-- 4 files changed, 58 insertions(+), 4 deletions(-) diff --git a/docs/schemas/domain.rng b/docs/schemas/domain.rng index 5e140fb..a79ca6a 100644 --- a/docs/schemas/domain.rng +++ b/docs/schemas/domain.rng @@ -1025,6 +1025,16 @@ <ref name="filterref-node-attributes"/> </element> </optional> + <optional> + <element name="tune"> + <optional> +  + <element name="sndbuf"> + <ref name="unsignedInt"/> + </element> + </optional> + </element> + </optional> </interleave> </define> <define name="virtualPortProfile"> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 1cef112..396574e 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -2282,6 +2282,7 @@ err_exit: static virDomainNetDefPtr virDomainNetDefParseXML(virCapsPtr caps, xmlNodePtr node, + xmlXPathContextPtr ctxt, int flags ATTRIBUTE_UNUSED) { virDomainNetDefPtr def; xmlNodePtr cur; @@ -2303,12 +2304,15 @@ virDomainNetDefParseXML(virCapsPtr caps, virNWFilterHashTablePtr filterparams = NULL; virVirtualPortProfileParams virtPort; bool virtPortParsed = false; + xmlNodePtr oldnode = ctxt->node; if (VIR_ALLOC(def) < 0) { virReportOOMError(); return NULL; } + ctxt->node = node; + type = virXMLPropString(node, "type"); if (type != NULL) { if ((int)(def->type = virDomainNetTypeFromString(type)) < 0) { @@ -2596,7 +2600,12 @@ virDomainNetDefParseXML(virCapsPtr caps, } } + if (virXPathULong("string(./tune/sndbuf)", ctxt, &def->tune.sndbuf) >= 0) { + def->tune.sndbuf_specified = true; + } + cleanup: + ctxt->node = oldnode; VIR_FREE(macaddr); VIR_FREE(network); VIR_FREE(address); @@ -4306,6 +4315,7 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, { xmlDocPtr xml; xmlNodePtr node; + xmlXPathContextPtr ctxt = NULL; virDomainDeviceDefPtr dev = NULL; if (!(xml = xmlReadDoc(BAD_CAST xmlStr, "device.xml", NULL, @@ -4322,6 +4332,13 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, goto error; } + ctxt = xmlXPathNewContext(xml); + if (ctxt == NULL) { + virReportOOMError(); + goto error; + } + ctxt->node = node; + if (VIR_ALLOC(dev) < 0) { virReportOOMError(); goto error; @@ -4337,7 +4354,7 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, goto error; } else if (xmlStrEqual(node->name, BAD_CAST "interface")) { dev->type = VIR_DOMAIN_DEVICE_NET; - if (!(dev->data.net = virDomainNetDefParseXML(caps, node, flags))) + if (!(dev->data.net = virDomainNetDefParseXML(caps, node, ctxt, flags))) goto error; } else if (xmlStrEqual(node->name, BAD_CAST "input")) { dev->type = VIR_DOMAIN_DEVICE_INPUT; @@ -4375,11 +4392,12 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, } xmlFreeDoc(xml); - + xmlXPathFreeContext(ctxt); return dev; error: xmlFreeDoc(xml); + xmlXPathFreeContext(ctxt); VIR_FREE(dev); return NULL; } @@ -5050,6 +5068,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, for (i = 0 ; i < n ; i++) { virDomainNetDefPtr net = virDomainNetDefParseXML(caps, nodes[i], + ctxt, flags); if (!net) goto error; @@ -6318,6 +6337,12 @@ virDomainNetDefFormat(virBufferPtr buf, VIR_FREE(attrs); } + if (def->tune.sndbuf_specified) { + virBufferAddLit(buf, " <tune>\n"); + virBufferVSprintf(buf, " <sndbuf>%lu</sndbuf>\n", def->tune.sndbuf); + virBufferAddLit(buf, " </tune>\n"); + } + if (virDomainDeviceInfoFormat(buf, &def->info, flags) < 0) return -1; diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index a73fd14..1269ed6 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -346,6 +346,10 @@ struct _virDomainNetDef { virVirtualPortProfileParams virtPortProfile; } direct; } data; + struct { + bool sndbuf_specified; + unsigned long sndbuf; + } tune; char *ifname; virDomainDeviceInfo info; char *filter; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index a3b5ff3..c37fdb2 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1584,6 +1584,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, const char *tapfd, const char *vhostfd) { + bool is_tap = false; virBuffer buf = VIR_BUFFER_INITIALIZER; switch (net->type) { @@ -1593,6 +1594,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, virBufferAddLit(&buf, "tap"); virBufferVSprintf(&buf, "%cfd=%s", type_sep, tapfd); type_sep = ','; + is_tap = true; break; case VIR_DOMAIN_NET_TYPE_ETHERNET: @@ -1606,6 +1608,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, net->data.ethernet.script); type_sep = ','; } + is_tap = true; break; case VIR_DOMAIN_NET_TYPE_CLIENT: @@ -1659,8 +1662,11 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, type_sep, net->info.alias); } - if (vhostfd && *vhostfd) { - virBufferVSprintf(&buf, ",vhost=on,vhostfd=%s", vhostfd); + if (is_tap) { + if (vhostfd && *vhostfd) + virBufferVSprintf(&buf, ",vhost=on,vhostfd=%s", vhostfd); + if (net->tune.sndbuf_specified) + virBufferVSprintf(&buf, ",sndbuf=%lu", net->tune.sndbuf); } if (virBufferError(&buf)) { @@ -4661,6 +4667,15 @@ qemuParseCommandLineNet(virCapsPtr caps, } else if (STREQ(keywords[i], "off")) { def->backend = VIR_DOMAIN_NET_BACKEND_TYPE_QEMU; } + } else if (STREQ(keywords[i], "sndbuf") && values[i]) { + if (virStrToLong_ul(values[i], NULL, 10, &def->tune.sndbuf) < 0) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, + _("cannot parse sndbuf size in '%s'"), val); + virDomainNetDefFree(def); + def = NULL; + goto cleanup; + } + def->tune.sndbuf_specified = true; } } -- 1.7.3.4

2 2

[libvirt] [PATCH 0/2] Update qemu password support (v2)
by Marc-André Lureau 14 Jan '11

14 Jan '11

Recently, QEMU learned set_password and expire_password monitor commands. http://qemu.com/qemu.git/commit/?id=7572150c189c6553c2448334116ab717680de66d The following patches make use of these new commands if available, and fallback on VNC-only password support. Change since v1: - moved qemuInitGraphicsPasswords to qemu_hotplug, renamed to qemuDomainChangeGraphicsPasswords. - updated what looks like a typo (that appears to work anyway) in initial patch from Daniel: - ret = qemuInitGraphicsPasswords(driver, vm, - VIR_DOMAIN_GRAPHICS_TYPE_SPICE, - &vm->def->graphics[0]->data.vnc.auth, - driver->vncPassword); + ret = qemuInitGraphicsPasswords(driver, vm, + VIR_DOMAIN_GRAPHICS_TYPE_SPICE, + &vm->def->graphics[0]->data.spice.auth, + driver->spicePassword); Marc-André Lureau (2): qemu: add set_password and expire_password monitor commands Use the new set_password monitor command to set password. src/qemu/qemu_driver.c | 22 ++++---- src/qemu/qemu_hotplug.c | 114 +++++++++++++++++++++++++++++++++++++++--- src/qemu/qemu_hotplug.h | 5 ++ src/qemu/qemu_monitor.c | 77 ++++++++++++++++++++++++++++ src/qemu/qemu_monitor.h | 7 +++ src/qemu/qemu_monitor_json.c | 56 ++++++++++++++++++++ src/qemu/qemu_monitor_json.h | 7 +++ src/qemu/qemu_monitor_text.c | 69 +++++++++++++++++++++++++ src/qemu/qemu_monitor_text.h | 7 +++ 9 files changed, 346 insertions(+), 18 deletions(-) -- 1.7.3.4

3 5

[libvirt] [PATCHv2] Add XML config switch to enable/disable vhost-net support
by Laine Stump 14 Jan '11

14 Jan '11

This patch is in response to https://bugzilla.redhat.com/show_bug.cgi?id=643050 The existing libvirt support for the vhost-net backend to the virtio network driver happens automatically - if the vhost-net device is available, it is always enabled, otherwise the standard userland virtio backend is used. This patch makes it possible to force whether or not vhost-net is used with a bit of XML. Adding a <driver> element to the interface XML, eg: <interface type="network"> <model type="virtio"/> <driver name="vhost"/> will force use of vhost-net (if it's not available, the domain will fail to start). if driver name="qemu", vhost-net will not be used even if it is available. If there is no <driver name='xxx'/> in the config, libvirt will revert to the pre-existing automatic behavior - use vhost-net if it's available, and userland backend if vhost-net isn't available. --- Changes from V1: enum now starts at 0 instead of -1. I just disallow "default" when parsing. This eliminates the need for vhost_specified. Removed superfluous 3rd arg to open(). docs/schemas/domain.rng | 13 +++++++++ src/conf/domain_conf.c | 29 +++++++++++++++++++- src/conf/domain_conf.h | 10 +++++++ src/qemu/qemu_command.c | 67 ++++++++++++++++++++++++++++++++++++++-------- src/qemu/qemu_command.h | 3 -- 5 files changed, 106 insertions(+), 16 deletions(-) diff --git a/docs/schemas/domain.rng b/docs/schemas/domain.rng index 6de85fd..5e140fb 100644 --- a/docs/schemas/domain.rng +++ b/docs/schemas/domain.rng @@ -1005,6 +1005,19 @@ </element> </optional> <optional> + <element name="driver"> + <optional> + <attribute name="name"> + <choice> + <value>qemu</value> + <value>vhost</value> + </choice> + </attribute> + </optional> + <empty/> + </element> + </optional> + <optional> <ref name="address"/> </optional> <optional> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 2c54683..1cef112 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -185,6 +185,11 @@ VIR_ENUM_IMPL(virDomainNet, VIR_DOMAIN_NET_TYPE_LAST, "internal", "direct") +VIR_ENUM_IMPL(virDomainNetBackend, VIR_DOMAIN_NET_BACKEND_TYPE_LAST, + "default", + "qemu", + "vhost") + VIR_ENUM_IMPL(virDomainChrChannelTarget, VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_LAST, "guestfwd", @@ -2290,6 +2295,7 @@ virDomainNetDefParseXML(virCapsPtr caps, char *address = NULL; char *port = NULL; char *model = NULL; + char *backend = NULL; char *filter = NULL; char *internal = NULL; char *devaddr = NULL; @@ -2372,6 +2378,8 @@ virDomainNetDefParseXML(virCapsPtr caps, script = virXMLPropString(cur, "path"); } else if (xmlStrEqual (cur->name, BAD_CAST "model")) { model = virXMLPropString(cur, "type"); + } else if (xmlStrEqual (cur->name, BAD_CAST "driver")) { + backend = virXMLPropString(cur, "name"); } else if (xmlStrEqual (cur->name, BAD_CAST "filterref")) { filter = virXMLPropString(cur, "filter"); VIR_FREE(filterparams); @@ -2559,6 +2567,19 @@ virDomainNetDefParseXML(virCapsPtr caps, model = NULL; } + if ((backend != NULL) && + (def->model && STREQ(def->model, "virtio"))) { + int b; + if (((b = virDomainNetBackendTypeFromString(backend)) < 0) || + (b == 0)) { + virDomainReportError(VIR_ERR_INTERNAL_ERROR, + _("Unknown interface <driver name='%s'> " + "has been specified"), + backend); + goto error; + } + def->backend = b; + } if (filter != NULL) { switch (def->type) { case VIR_DOMAIN_NET_TYPE_ETHERNET: @@ -2585,6 +2606,7 @@ cleanup: VIR_FREE(script); VIR_FREE(bridge); VIR_FREE(model); + VIR_FREE(backend); VIR_FREE(filter); VIR_FREE(type); VIR_FREE(internal); @@ -6276,9 +6298,14 @@ virDomainNetDefFormat(virBufferPtr buf, if (def->ifname) virBufferEscapeString(buf, " <target dev='%s'/>\n", def->ifname); - if (def->model) + if (def->model) { virBufferEscapeString(buf, " <model type='%s'/>\n", def->model); + if (STREQ(def->model, "virtio") && def->backend) { + virBufferVSprintf(buf, " <driver name='%s'/>\n", + virDomainNetBackendTypeToString(def->backend)); + } + } if (def->filter) { virBufferEscapeString(buf, " <filterref filter='%s'", def->filter); diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 6a8ec64..a73fd14 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -292,6 +292,14 @@ enum virDomainNetType { VIR_DOMAIN_NET_TYPE_LAST, }; +/* the backend driver used for virtio interfaces */ +enum virDomainNetBackendType { + VIR_DOMAIN_NET_BACKEND_TYPE_DEFAULT, /* prefer kernel, fall back to user */ + VIR_DOMAIN_NET_BACKEND_TYPE_QEMU, /* userland */ + VIR_DOMAIN_NET_BACKEND_TYPE_VHOST, /* kernel */ + + VIR_DOMAIN_NET_BACKEND_TYPE_LAST, +}; /* the mode type for macvtap devices */ enum virDomainNetdevMacvtapType { @@ -310,6 +318,7 @@ struct _virDomainNetDef { enum virDomainNetType type; unsigned char mac[VIR_MAC_BUFLEN]; char *model; + enum virDomainNetBackendType backend; union { struct { char *dev; @@ -1265,6 +1274,7 @@ VIR_ENUM_DECL(virDomainControllerModel) VIR_ENUM_DECL(virDomainFS) VIR_ENUM_DECL(virDomainFSAccessMode) VIR_ENUM_DECL(virDomainNet) +VIR_ENUM_DECL(virDomainNetBackend) VIR_ENUM_DECL(virDomainChrDevice) VIR_ENUM_DECL(virDomainChrChannelTarget) VIR_ENUM_DECL(virDomainChrConsoleTarget) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 86c5bb5..a3b5ff3 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -302,24 +302,58 @@ cleanup: } -int +static int qemuOpenVhostNet(virDomainNetDefPtr net, - unsigned long long qemuCmdFlags) + unsigned long long qemuCmdFlags, + int *vhostfd) { - /* If qemu supports vhost-net mode (including the -netdev command - * option), the nic model is virtio, and we can open - * /dev/vhost_net, assume that vhost-net mode is available and - * return the fd to /dev/vhost_net. Otherwise, return -1. - */ + *vhostfd = -1; /* assume we won't use vhost */ + /* If the config says explicitly to not use vhost, return now */ + if (net->backend == VIR_DOMAIN_NET_BACKEND_TYPE_QEMU) { + return 0; + } + + /* If qemu doesn't support vhost-net mode (including the -netdev command + * option), don't try to open the device. + */ if (!(qemuCmdFlags & QEMUD_CMD_FLAG_VNET_HOST && qemuCmdFlags & QEMUD_CMD_FLAG_NETDEV && - qemuCmdFlags & QEMUD_CMD_FLAG_DEVICE && - net->model && STREQ(net->model, "virtio"))) - return -1; + qemuCmdFlags & QEMUD_CMD_FLAG_DEVICE)) { + if (net->backend == VIR_DOMAIN_NET_BACKEND_TYPE_VHOST) { + qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, + "%s", _("vhost-net is not supported with " + "this QEMU binary")); + return -1; + } + return 0; + } - return open("/dev/vhost-net", O_RDWR, 0); + /* If the nic model isn't virtio, don't try to open. */ + if (!(net->model && STREQ(net->model, "virtio"))) { + if (net->backend == VIR_DOMAIN_NET_BACKEND_TYPE_VHOST) { + qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, + "%s", _("vhost-net is only supported for " + "virtio network interfaces")); + return -1; + } + return 0; + } + + *vhostfd = open("/dev/vhost-net", O_RDWR); + + /* If the config says explicitly to use vhost and we couldn't open it, + * report an error. + */ + if ((*vhostfd < 0) && + (net->backend == VIR_DOMAIN_NET_BACKEND_TYPE_VHOST)) { + qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, + "%s", _("vhost-net was requested for an interface, " + "but is unavailable")); + return -1; + } + return 0; } @@ -3278,7 +3312,10 @@ qemuBuildCommandLine(virConnectPtr conn, net->type == VIR_DOMAIN_NET_TYPE_DIRECT) { /* Attempt to use vhost-net mode for these types of network device */ - int vhostfd = qemuOpenVhostNet(net, qemuCmdFlags); + int vhostfd; + + if (qemuOpenVhostNet(net, qemuCmdFlags, &vhostfd) < 0) + goto error; if (vhostfd >= 0) { virCommandTransferFD(cmd, vhostfd); @@ -4618,6 +4655,12 @@ qemuParseCommandLineNet(virCapsPtr caps, } else if (STREQ(keywords[i], "model")) { def->model = values[i]; values[i] = NULL; + } else if (STREQ(keywords[i], "vhost")) { + if ((values[i] == NULL) || STREQ(values[i], "on")) { + def->backend = VIR_DOMAIN_NET_BACKEND_TYPE_VHOST; + } else if (STREQ(keywords[i], "off")) { + def->backend = VIR_DOMAIN_NET_BACKEND_TYPE_QEMU; + } } } diff --git a/src/qemu/qemu_command.h b/src/qemu/qemu_command.h index 4c42a10..5439184 100644 --- a/src/qemu/qemu_command.h +++ b/src/qemu/qemu_command.h @@ -116,9 +116,6 @@ int qemuNetworkIfaceConnect(virConnectPtr conn, unsigned long long qemCmdFlags) ATTRIBUTE_NONNULL(1); -int qemuOpenVhostNet(virDomainNetDefPtr net, - unsigned long long qemuCmdFlags); - int qemuPhysIfaceConnect(virConnectPtr conn, struct qemud_driver *driver, virDomainNetDefPtr net, -- 1.7.3.4

2 2

[libvirt] [PATCH] Fix 'make check' after commit 04197350
by Jim Fehlig 14 Jan '11

14 Jan '11

I broke 'make check' with commit 04197350 by unconditionally emitting 'hap=' in xen xm driver. Only emit 'hap=' if xendConfigVersion >= 3. I've tested sending 'hap=' to a Xen 3.2 machine without support for hap setting and verified that xend silently drops the unrecognized setting. --- src/xen/xm_internal.c | 9 +++++---- 1 files changed, 5 insertions(+), 4 deletions(-) diff --git a/src/xen/xm_internal.c b/src/xen/xm_internal.c index ec618aa..bfb6698 100644 --- a/src/xen/xm_internal.c +++ b/src/xen/xm_internal.c @@ -2413,10 +2413,11 @@ virConfPtr xenXMDomainConfigFormat(virConnectPtr conn, (1 << VIR_DOMAIN_FEATURE_APIC)) ? 1 : 0) < 0) goto no_memory; - if (xenXMConfigSetInt(conf, "hap", - (def->features & - (1 << VIR_DOMAIN_FEATURE_HAP)) ? 1 : 0) < 0) - goto no_memory; + if (priv->xendConfigVersion >= 3) + if (xenXMConfigSetInt(conf, "hap", + (def->features & + (1 << VIR_DOMAIN_FEATURE_HAP)) ? 1 : 0) < 0) + goto no_memory; if (def->clock.offset == VIR_DOMAIN_CLOCK_OFFSET_LOCALTIME) { if (def->clock.data.timezone) { -- 1.7.3.1

2 2

[libvirt] [PATCH] Enable tuning of qemu network tap device "sndbuf" size
by Laine Stump 14 Jan '11

14 Jan '11

This is in response to a request in: https://bugzilla.redhat.com/show_bug.cgi?id=665293 In short, under heavy load, it's possible for qemu's networking to lock up due to the tap device's default 1MB sndbuf being inadequate. adding "sndbuf=0" to the qemu commandline -netdevice option will alleviate this problem (sndbuf=0 actually sets it to 0xffffffff). Because we must be able to explicitly specify "0" as a value, the standard practice of "0 means not specified" won't work here. Instead, virDomainNetDef also has a sndbuf_specified, which defaults to 0, but is set to 1 if some value was given. The sndbuf value is put inside a <tune> element of each <interface> in the domain. The intent is that further tunable settings will also be placed inside this elemnent. <interface type='network'> ... <tune> <sndbuf>0</sndbuf> ... </tune> </interface> --- Note that in qemuBuildHostNetStr() I have moved if (vhostfd && *vhostfd) { virBufferVSprintf(&buf, ",vhost=on,vhostfd=%s", vhostfd); into a newly created "if (is_tap) { ... }" block. This always should have been inside such a conditional, but none existed until now. (I can make this a separate patch if anyone wants, but it seemed so simple and obvious that I took the slothenly way out :-) Also, as with the vhost patch, I didn't get the html docs updated for this addition either. I will add both in a single followup patch next week. docs/schemas/domain.rng | 10 ++++++++++ src/conf/domain_conf.c | 31 +++++++++++++++++++++++++++++-- src/conf/domain_conf.h | 4 ++++ src/qemu/qemu_command.c | 19 +++++++++++++++++-- 4 files changed, 60 insertions(+), 4 deletions(-) diff --git a/docs/schemas/domain.rng b/docs/schemas/domain.rng index 6d0654d..e2883aa 100644 --- a/docs/schemas/domain.rng +++ b/docs/schemas/domain.rng @@ -1025,6 +1025,16 @@ <ref name="filterref-node-attributes"/> </element> </optional> + <optional> + <element name="tune"> + <optional> +  + <element name="sndbuf"> + <ref name="unsignedInt"/> + </element> + </optional> + </element> + </optional> </interleave> </define> <define name="virtualPortProfile"> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 04ed502..5d1b8cf 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -2280,6 +2280,7 @@ err_exit: static virDomainNetDefPtr virDomainNetDefParseXML(virCapsPtr caps, xmlNodePtr node, + xmlXPathContextPtr ctxt, int flags ATTRIBUTE_UNUSED) { virDomainNetDefPtr def; xmlNodePtr cur; @@ -2298,15 +2299,19 @@ virDomainNetDefParseXML(virCapsPtr caps, char *internal = NULL; char *devaddr = NULL; char *mode = NULL; + unsigned long sndbuf; virNWFilterHashTablePtr filterparams = NULL; virVirtualPortProfileParams virtPort; bool virtPortParsed = false; + xmlNodePtr oldnode = ctxt->node; if (VIR_ALLOC(def) < 0) { virReportOOMError(); return NULL; } + ctxt->node = node; + type = virXMLPropString(node, "type"); if (type != NULL) { if ((int)(def->type = virDomainNetTypeFromString(type)) < 0) { @@ -2593,7 +2598,13 @@ virDomainNetDefParseXML(virCapsPtr caps, } } + if (virXPathULong("string(./tune/sndbuf)", ctxt, &sndbuf) >= 0) { + def->tune.sndbuf = sndbuf; + def->tune.sndbuf_specified = 1; + } + cleanup: + ctxt->node = oldnode; VIR_FREE(macaddr); VIR_FREE(network); VIR_FREE(address); @@ -4303,6 +4314,7 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, { xmlDocPtr xml; xmlNodePtr node; + xmlXPathContextPtr ctxt = NULL; virDomainDeviceDefPtr dev = NULL; if (!(xml = xmlReadDoc(BAD_CAST xmlStr, "device.xml", NULL, @@ -4319,6 +4331,13 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, goto error; } + ctxt = xmlXPathNewContext(xml); + if (ctxt == NULL) { + virReportOOMError(); + goto error; + } + ctxt->node = node; + if (VIR_ALLOC(dev) < 0) { virReportOOMError(); goto error; @@ -4334,7 +4353,7 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, goto error; } else if (xmlStrEqual(node->name, BAD_CAST "interface")) { dev->type = VIR_DOMAIN_DEVICE_NET; - if (!(dev->data.net = virDomainNetDefParseXML(caps, node, flags))) + if (!(dev->data.net = virDomainNetDefParseXML(caps, node, ctxt, flags))) goto error; } else if (xmlStrEqual(node->name, BAD_CAST "input")) { dev->type = VIR_DOMAIN_DEVICE_INPUT; @@ -4372,11 +4391,12 @@ virDomainDeviceDefPtr virDomainDeviceDefParse(virCapsPtr caps, } xmlFreeDoc(xml); - + xmlXPathFreeContext(ctxt); return dev; error: xmlFreeDoc(xml); + xmlXPathFreeContext(ctxt); VIR_FREE(dev); return NULL; } @@ -5047,6 +5067,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, for (i = 0 ; i < n ; i++) { virDomainNetDefPtr net = virDomainNetDefParseXML(caps, nodes[i], + ctxt, flags); if (!net) goto error; @@ -6315,6 +6336,12 @@ virDomainNetDefFormat(virBufferPtr buf, VIR_FREE(attrs); } + if (def->tune.sndbuf_specified) { + virBufferAddLit(buf, " <tune>\n"); + virBufferVSprintf(buf, " <sndbuf>%d</sndbuf>\n", def->tune.sndbuf); + virBufferAddLit(buf, " </tune>\n"); + } + if (virDomainDeviceInfoFormat(buf, &def->info, flags) < 0) return -1; diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 451ccad..2d35d68 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -346,6 +346,10 @@ struct _virDomainNetDef { virVirtualPortProfileParams virtPortProfile; } direct; } data; + struct { + int sndbuf_specified : 1; + int sndbuf; + } tune; char *ifname; virDomainDeviceInfo info; char *filter; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 9eb54a1..add66cb 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1587,6 +1587,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, const char *tapfd, const char *vhostfd) { + bool is_tap = false; virBuffer buf = VIR_BUFFER_INITIALIZER; switch (net->type) { @@ -1596,6 +1597,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, virBufferAddLit(&buf, "tap"); virBufferVSprintf(&buf, "%cfd=%s", type_sep, tapfd); type_sep = ','; + is_tap = true; break; case VIR_DOMAIN_NET_TYPE_ETHERNET: @@ -1609,6 +1611,7 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, net->data.ethernet.script); type_sep = ','; } + is_tap = true; break; case VIR_DOMAIN_NET_TYPE_CLIENT: @@ -1662,8 +1665,11 @@ qemuBuildHostNetStr(virDomainNetDefPtr net, type_sep, net->info.alias); } - if (vhostfd && *vhostfd) { - virBufferVSprintf(&buf, ",vhost=on,vhostfd=%s", vhostfd); + if (is_tap) { + if (vhostfd && *vhostfd) + virBufferVSprintf(&buf, ",vhost=on,vhostfd=%s", vhostfd); + if (net->tune.sndbuf_specified) + virBufferVSprintf(&buf, ",sndbuf=%d", net->tune.sndbuf); } if (virBufferError(&buf)) { @@ -4665,6 +4671,15 @@ qemuParseCommandLineNet(virCapsPtr caps, def->backend = VIR_DOMAIN_NET_BACKEND_TYPE_QEMU; } def->backend_specified = 1; + } else if (STREQ(keywords[i], "sndbuf") && values[i]) { + if (virStrToLong_i(values[i], NULL, 10, &def->tune.sndbuf) < 0) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, + _("cannot parse sndbuf size in '%s'"), val); + virDomainNetDefFree(def); + def = NULL; + goto cleanup; + } + def->tune.sndbuf_specified = 1; } } -- 1.7.3.4

3 4

[libvirt] [PATCH] cpu: Add support for Westmere CPU model
by Jiri Denemark 14 Jan '11

14 Jan '11

--- src/cpu/cpu_map.xml | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-) diff --git a/src/cpu/cpu_map.xml b/src/cpu/cpu_map.xml index 75c6522..35dca20 100644 --- a/src/cpu/cpu_map.xml +++ b/src/cpu/cpu_map.xml @@ -142,6 +142,9 @@ <feature name='popcnt'>  <cpuid function='0x00000001' ecx='0x00800000'/> </feature> + <feature name='aes'> + <cpuid function='0x00000001' ecx='0x02000000'/> + </feature> <feature name='hypervisor'>  <cpuid function='0x00000001' ecx='0x80000000'/> </feature> @@ -453,6 +456,12 @@ <feature name='lahf_lm'/> </model> + <model name='Westmere'> + <vendor name='Intel'/> + <model name='Nehalem'/> + <feature name='aes'/> + </model> + <model name='Opteron_G1'> <vendor name='AMD'/> <feature name='sse2'/> -- 1.7.4.rc1

2 2

[libvirt] build failure on ubuntu
by Eric Blake 14 Jan '11

14 Jan '11

I haven't looked into this closely yet, but Daniel's security refactoring broke the Ubuntu 'make check' build when apparmor is enabled: CC secaatest.o secaatest.c: In function 'main': secaatest.c:18: error: implicit declaration of function 'virSecurityDriverStartup' [-Wimplicit-function-declaration] -- Eric Blake eblake(a)redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org

2 2

[libvirt] [PATCH 0/4] Add HAP to domain features
by Jim Fehlig 14 Jan '11

14 Jan '11

Hardware features such as Extended Page Table and Nested Page Table augment hypervisor software techniques such as shadow page table. Adding HAP (hardware assisted paging) to the virDomainFeature enumeration allows users to select between hardware and software memory management mechanisms for their guests. I've seen reports [1] of users asking for this capability in the past. Patch 3 is an attempt to advertise the feature in hvm guest capabilities for the Xen hypervisor. I have not found a way to discover the presence of HAP support through the xen tools. Instead, the feature is only advertised on Xen >= 3.3, which sadly won't accommodate distro backports to older Xen versions. Thanks for your comments, Jim [1] http://www.mail-archive.com/fedora-xen@redhat.com/msg02680.html Jim Fehlig (4): Add HAP to virDomainFeature enum Add support for HAP feature to xen drivers Add HAP to xen hypervisor capabilities Document HAP domain feature docs/formatdomain.html.in | 5 +++++ docs/schemas/domain.rng | 7 ++++++- src/conf/domain_conf.c | 3 ++- src/conf/domain_conf.h | 1 + src/xen/xen_hypervisor.c | 10 ++++++++++ src/xen/xend_internal.c | 4 ++++ src/xen/xm_internal.c | 8 ++++++++ src/xenapi/xenapi_driver.c | 2 ++ src/xenapi/xenapi_utils.c | 2 ++ 9 files changed, 40 insertions(+), 2 deletions(-) -- 1.7.3.1

3 13

[libvirt] virsh setmem and setmaxmem: which XML do they match?
by Justin Clift 14 Jan '11

14 Jan '11

Hi all, Updating the man pages for the virsh setmem and setmaxmem commands at the moment. These two commands only run on active guest domains. So, as a helpful pointer on how to make the change with an inactive domain, I'm going to mention what the equivalent XML tag is for use with virsh edit. Like this: Note, this command only works on active guest domains. To change the memory allocation for an inactive guest domain, use the virsh B<edit> command to update the XML <memory> element. I'm just not sure which XML element the setmem and setmaxmem commands map to. Does "setmaxmem" map to <memory>, and "setmem" map to <currentMemory>? The description in our XML format domain doc is less than clear. :/ Regards and best wishes, Justin Clift

1 0

[libvirt] [PATCH] util: add missing string->integer conversion functions
by Eric Blake 14 Jan '11

14 Jan '11

It was awkward having only int conversion in the virStrToLong family, but only long conversion in the virXPath family. Make both families support both types. * src/util/util.h (virStrToLong_l, virStrToLong_ul): New prototypes. * src/util/xml.h (virXPathInt, virXPathUInt): Likewise. * src/util/util.c (virStrToLong_l, virStrToLong_ul): New functions. * src/util/xml.c (virXPathInt, virXPathUInt): Likewise. * src/libvirt_private.syms (util.h, xml.h): Export them. --- src/libvirt_private.syms | 4 +++ src/util/util.c | 40 +++++++++++++++++++++++++++++- src/util/util.h | 8 ++++++ src/util/xml.c | 60 +++++++++++++++++++++++++++++++++++++++++++++- src/util/xml.h | 12 +++++++-- 5 files changed, 119 insertions(+), 5 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index a166bd9..d95ef31 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -886,8 +886,10 @@ virSetUIDGID; virSkipSpaces; virStrToDouble; virStrToLong_i; +virStrToLong_l; virStrToLong_ll; virStrToLong_ui; +virStrToLong_ul; virStrToLong_ull; virStrcpy; virStrncpy; @@ -926,6 +928,7 @@ virStrerror; # xml.h virXMLPropString; virXPathBoolean; +virXPathInt; virXPathLong; virXPathLongHex; virXPathLongLong; @@ -934,6 +937,7 @@ virXPathNodeSet; virXPathNumber; virXPathString; virXPathStringLimit; +virXPathUInt; virXPathULong; virXPathULongHex; virXPathULongLong; diff --git a/src/util/util.c b/src/util/util.c index 18b38f4..f412a83 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -2081,7 +2081,45 @@ virStrToLong_ui(char const *s, char **end_ptr, int base, unsigned int *result) return 0; } -/* Just like virStrToLong_i, above, but produce an "long long" value. */ +/* Just like virStrToLong_i, above, but produce a "long" value. */ +int +virStrToLong_l(char const *s, char **end_ptr, int base, long *result) +{ + long int val; + char *p; + int err; + + errno = 0; + val = strtol(s, &p, base); + err = (errno || (!end_ptr && *p) || p == s); + if (end_ptr) + *end_ptr = p; + if (err) + return -1; + *result = val; + return 0; +} + +/* Just like virStrToLong_i, above, but produce an "unsigned long" value. */ +int +virStrToLong_ul(char const *s, char **end_ptr, int base, unsigned long *result) +{ + unsigned long int val; + char *p; + int err; + + errno = 0; + val = strtoul(s, &p, base); + err = (errno || (!end_ptr && *p) || p == s); + if (end_ptr) + *end_ptr = p; + if (err) + return -1; + *result = val; + return 0; +} + +/* Just like virStrToLong_i, above, but produce a "long long" value. */ int virStrToLong_ll(char const *s, char **end_ptr, int base, long long *result) { diff --git a/src/util/util.h b/src/util/util.h index 932db03..8373038 100644 --- a/src/util/util.h +++ b/src/util/util.h @@ -188,6 +188,14 @@ int virStrToLong_ui(char const *s, char **end_ptr, int base, unsigned int *result); +int virStrToLong_l(char const *s, + char **end_ptr, + int base, + long *result); +int virStrToLong_ul(char const *s, + char **end_ptr, + int base, + unsigned long *result); int virStrToLong_ll(char const *s, char **end_ptr, int base, diff --git a/src/util/xml.c b/src/util/xml.c index e2c2c6c..de5e9de 100644 --- a/src/util/xml.c +++ b/src/util/xml.c @@ -1,7 +1,7 @@ /* * xml.c: XML based interfaces for the libvir library * - * Copyright (C) 2005, 2007-2010 Red Hat, Inc. + * Copyright (C) 2005, 2007-2011 Red Hat, Inc. * * See COPYING.LIB for the License of this software * @@ -195,6 +195,35 @@ virXPathLongBase(const char *xpath, } /** + * virXPathInt: + * @xpath: the XPath string to evaluate + * @ctxt: an XPath context + * @value: the returned int value + * + * Convenience function to evaluate an XPath number + * + * Returns 0 in case of success in which case @value is set, + * or -1 if the XPath evaluation failed or -2 if the + * value doesn't have an int format. + */ +int +virXPathInt(const char *xpath, + xmlXPathContextPtr ctxt, + int *value) +{ + long tmp; + int ret; + + ret = virXPathLongBase(xpath, ctxt, 10, &tmp); + if (ret < 0) + return ret; + if ((int) tmp != tmp) + return -2; + *value = tmp; + return 0; +} + +/** * virXPathLong: * @xpath: the XPath string to evaluate * @ctxt: an XPath context @@ -279,6 +308,35 @@ virXPathULongBase(const char *xpath, } /** + * virXPathUInt: + * @xpath: the XPath string to evaluate + * @ctxt: an XPath context + * @value: the returned int value + * + * Convenience function to evaluate an XPath number + * + * Returns 0 in case of success in which case @value is set, + * or -1 if the XPath evaluation failed or -2 if the + * value doesn't have an int format. + */ +int +virXPathUInt(const char *xpath, + xmlXPathContextPtr ctxt, + unsigned int *value) +{ + unsigned long tmp; + int ret; + + ret = virXPathULongBase(xpath, ctxt, 10, &tmp); + if (ret < 0) + return ret; + if ((unsigned int) tmp != tmp) + return -2; + *value = tmp; + return 0; +} + +/** * virXPathULong: * @xpath: the XPath string to evaluate * @ctxt: an XPath context diff --git a/src/util/xml.h b/src/util/xml.h index b1da741..b342e83 100644 --- a/src/util/xml.h +++ b/src/util/xml.h @@ -21,19 +21,25 @@ char * virXPathStringLimit(const char *xpath, int virXPathNumber(const char *xpath, xmlXPathContextPtr ctxt, double *value); +int virXPathInt(const char *xpath, + xmlXPathContextPtr ctxt, + int *value); +int virXPathUInt(const char *xpath, + xmlXPathContextPtr ctxt, + unsigned int *value); int virXPathLong(const char *xpath, xmlXPathContextPtr ctxt, long *value); -int virXPathULong(const char *xpath, +int virXPathULong(const char *xpath, xmlXPathContextPtr ctxt, unsigned long *value); int virXPathULongLong(const char *xpath, xmlXPathContextPtr ctxt, unsigned long long *value); -int virXPathLongLong(const char *xpath, +int virXPathLongLong(const char *xpath, xmlXPathContextPtr ctxt, long long *value); -int virXPathLongHex (const char *xpath, +int virXPathLongHex(const char *xpath, xmlXPathContextPtr ctxt, long *value); int virXPathULongHex(const char *xpath, -- 1.7.3.4

2 2

[libvirt] virsh setmaxmem granularity for vSphere/ESX?
by Justin Clift 14 Jan '11

14 Jan '11

Hi Matthias, Do you know if the virsh "setmaxmem" has the same MB granularity as "setmem" (on vSphere/ESX)? Regards and best wishes, Justin Clift

1 1

[libvirt] [PATCH] docs: clarify that virsh setvcpus and setmem only work on active domains
by Justin Clift 13 Jan '11

13 Jan '11

Addresses BZ # 622534: https://bugzilla.redhat.com/show_bug.cgi?id=622534 --- tools/virsh.pod | 28 ++++++++++++++++++---------- 1 files changed, 18 insertions(+), 10 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index 0e03d68..b4c6ff7 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -568,13 +568,16 @@ XEN_CREDIT scheduler and are now I<DEPRECATED>. =item B<setmem> I<domain-id> B<kilobytes> -Change the current memory allocation in the guest domain. This should take -effect immediately. The memory limit is specified in -kilobytes. +Change the current memory allocation in an B<active> guest domain. This +should take effect immediately. The memory limit is specified in kilobytes. For Xen, you can only adjust the memory of a running domain if the domain is paravirtualized or running the PV balloon driver. +B<Note>, this command only works on active guest domains. To change the +number of virtual CPUs in a inactive guest domain, use the virsh B<edit> +command to update the XML <memory> element. + =item B<setmaxmem> I<domain-id> B<kilobytes> Change the maximum memory allocation limit in the guest domain. This should @@ -593,24 +596,29 @@ QEMU/KVM supports I<--hard-limit>, I<--soft-limit>, and I<--swap-hard-limit>. =item B<setvcpus> I<domain-id> I<count> optional I<--maximum> I<--config> I<--live> -Change the number of virtual CPUs active in the guest domain. Note that -I<count> may be limited by host, hypervisor or limit coming from the -original description of domain. +Change the number of virtual CPUs active in an B<active> guest domain. + +The I<count> value may be limited by host, hypervisor or limit coming from +the original description of the guest domain. For Xen, you can only adjust the virtual CPUs of a running domain if the domain is paravirtualized. If I<--config> is specified, the change will only affect the next boot of a domain. If I<--live> is specified, the domain must be -running, and the change takes place immediately. Both flags may be -specified, if supported by the hypervisor. If neither flag is given, -then I<--live> is implied and it is up to the hypervisor whether -I<--config> is also implied. +running, and the change takes place immediately. Both I<--config> and +I<--live> flags may be specified, if supported by the hypervisor. If +neither flag is given, then I<--live> is implied and it is up to the +hypervisor whether I<--config> is also implied. If I<--maximum> is specified, then you must use I<--config> and avoid I<--live>; this flag controls the maximum limit of vcpus that can be hot-plugged the next time the domain is booted. +B<Note>, this command only works on active guest domains. To change the +number of virtual CPUs in a inactive guest domain, use the virsh B<edit> +command to update the XML <vcpu> element. + =item B<shutdown> I<domain-id> Gracefully shuts down a domain. This coordinates with the domain OS -- 1.7.3.2

3 14

[libvirt] esx driver: XML format for guest OS type/variant
by Jake Xu 13 Jan '11

13 Jan '11

Hi, I am trying to create a VM using the Python bindings of Libvirt. I can successfully create VM from a XML template, but I can't find any way to define the guest OS type/variant like CentOS 5.5 64bit for my VM. The native format converted from XML is always guestOS="other-64" - which doesn't tell us much about the guest operating system. I have looked at the C libvirt source code a bit, and it seems like libvirt does not support defining guest os type using XML description yet. Is there any way I can set the guest OS type for my VM? Thanks, Jake

6 23

[libvirt] [PATCH 0/7] security: Allow disabling security per VM
by Cole Robinson 13 Jan '11

13 Jan '11

Enabling a security driver in qemu.conf is currently all or nothing. The option to disable security on a per VM basis can be a useful debugging tool or work around for frustrated users. Patches 1-3 and 5-6 are prep and cleanup work. Patch 4 fixes an easily triggerable segfault when defining a domain in qemu. Patch 7 is the actual feature. Cole Robinson (7): tests: Add qemuxml2xml tests for <seclabel> handling security: Use virDomainSeclabelDefClear security: Add virSecurityIsSpecifiedDriver qemu: Fix segfault if defining a domain without <seclabel> domain: Handle seclabel model with an enum domain: Always validate seclabel model security: Allow disabling security on a per VM basis cfg.mk | 1 + docs/schemas/domain.rng | 13 ++- src/conf/domain_conf.c | 69 ++++++++++----- src/conf/domain_conf.h | 15 +++- src/libvirt_private.syms | 2 +- src/qemu/qemu_driver.c | 4 +- src/security/security_apparmor.c | 31 ++----- src/security/security_driver.c | 25 ++++++ src/security/security_driver.h | 3 + src/security/security_manager.c | 90 +++++++++++++------- src/security/security_selinux.c | 50 +++--------- tests/domainschematest | 2 +- .../qemuxml2xml-balloon-device-auto-out.xml | 25 ++++++ .../qemuxml2xml-channel-virtio-auto-out.xml | 54 ++++++++++++ .../qemuxml2xml-console-compat-auto-out.xml | 31 +++++++ .../qemuxml2xml-console-virtio-out.xml | 29 ++++++ .../qemuxml2xml-disk-scsi-device-auto-out.xml | 31 +++++++ .../qemuxml2xml-seclabel-dynamic-in.xml | 24 +++++ .../qemuxml2xml-seclabel-dynamic-out.xml | 21 +++++ .../qemuxml2xml-seclabel-model-none-in.xml | 21 +++++ .../qemuxml2xml-seclabel-model-none-out.xml | 21 +++++ .../qemuxml2xml-seclabel-static-in.xml | 24 +++++ .../qemuxml2xml-seclabel-static-out.xml | 23 +++++ .../qemuxml2xmlout-balloon-device-auto.xml | 25 ------ .../qemuxml2xmlout-channel-virtio-auto.xml | 54 ------------ .../qemuxml2xmlout-console-compat-auto.xml | 31 ------- .../qemuxml2xmlout-console-virtio.xml | 29 ------ .../qemuxml2xmlout-disk-scsi-device-auto.xml | 31 ------- tests/qemuxml2xmltest.c | 26 ++++-- 29 files changed, 501 insertions(+), 304 deletions(-) create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-balloon-device-auto-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-channel-virtio-auto-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-console-compat-auto-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-console-virtio-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-disk-scsi-device-auto-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-dynamic-in.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-dynamic-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-model-none-in.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-model-none-out.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-static-in.xml create mode 100644 tests/qemuxml2xmldata/qemuxml2xml-seclabel-static-out.xml delete mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-balloon-device-auto.xml delete mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-channel-virtio-auto.xml delete mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-console-compat-auto.xml delete mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-console-virtio.xml delete mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-scsi-device-auto.xml -- 1.7.3.2

3 23

[libvirt] [PATCH] Fix crash in SELinuxSecurityVerify
by Laine Stump 13 Jan '11

13 Jan '11

When attempting to edit a domain, libvirtd segfaulted in SELinuxSecurityVerify() on this line: if (!STREQ(virSecurityManagerGetModel(mgr), secdef->model)) { because secdef->model was NULL. Although I'm too tired to investigate in depth, I noticed that all the other functions in that file that do the same STREQ() will first check that def->seclabel.label is non-NULL, but this function doesn't. I also noticed that label *is* NULL in my case, so I tried adding that check to SELinuxSecurityVerify(), and the crash goes away. I have no idea if this is the correct fix, but it allowed me to continue my testing of a new (unrelated) feature. --- src/security/security_selinux.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/src/security/security_selinux.c b/src/security/security_selinux.c index d06afde..b97ca4c 100644 --- a/src/security/security_selinux.c +++ b/src/security/security_selinux.c @@ -871,6 +871,10 @@ SELinuxSecurityVerify(virSecurityManagerPtr mgr ATTRIBUTE_UNUSED, virDomainDefPtr def) { const virSecurityLabelDefPtr secdef = &def->seclabel; + + if (def->seclabel.label == NULL) + return 0; + if (!STREQ(virSecurityManagerGetModel(mgr), secdef->model)) { virSecurityReportError(VIR_ERR_INTERNAL_ERROR, _("security label driver mismatch: " -- 1.7.3.4

4 6

[libvirt] [PATCH 0/6 v2] events: Add helpers for driver dispatching
by Cole Robinson 13 Jan '11

13 Jan '11

I noticed that there is quite a bit of code duplication among the drivers that support domain events. This patch series is an attempt to consolidate the shared logic. v2: 2 patches were applied Addressed Eric's comments: NONNULL tagging Use bool for isDispatching Move libvirt_private.syms earlier Add NULL check in StateFree Cole Robinson (6): domain_event: Add virDomainEventState structure domain_event: Add common domain event queue/flush helpers qemu: Use virDomainEventState helpers lxc: Use virDomainEventState helpers test: Use virDomainEventState helpers remote: Use virDomainEventState helpers cfg.mk | 1 + src/conf/domain_event.c | 166 ++++++++++++++++++++++++++++++++++++++----- src/conf/domain_event.h | 69 +++++++++++++++---- src/libvirt_private.syms | 6 ++ src/lxc/lxc_conf.h | 6 +-- src/lxc/lxc_driver.c | 76 ++++++-------------- src/qemu/qemu_conf.h | 6 +-- src/qemu/qemu_driver.c | 80 +++++++-------------- src/remote/remote_driver.c | 162 ++++++++++++++++-------------------------- src/test/test_driver.c | 104 ++++++++------------------- 10 files changed, 353 insertions(+), 323 deletions(-) -- 1.7.3.3

3 11

[libvirt] [PATCH] virFindFileInPath: only find executable non-directory
by Eric Blake 13 Jan '11

13 Jan '11

Without this patch, at least tests/daemon-conf (which sticks $builddir/src in the PATH) tries to execute the directory $builddir/src/qemu rather than the real /usr/bin/qemu binary. That was fine when qemu_capabilities silently ignored execution failure, but not so good once it is converted to use virCommand. * src/util/util.h (virFileExists): Adjust prototype. (virFileIsExecutable): New prototype. * src/util/util.c (virFindFileInPath): Reject non-executables and directories. Avoid huge stack allocation. (virFileExists): Use lighter-weight syscall. (virFileIsExecutable): New function. * src/libvirt_private.syms (util.h): Export new function. --- Questions: Should I be requiring S_ISREG, rather than !S_ISDIR (that is, should we be rejecting devices and sockets as non-exectuable)? Should I import the gnulib module euidaccess (and/or faccessat) for the access check? Using access(F_OK) is okay regardless of uid/gid, but access(X_OK) may have different answers than euidaccess(X_OK) when the effective uid/gid do not match the current uid/gid. However, dragging in the gnulib module will require adding an extra link library for some platforms (for example, Solaris needs -lgen), which means the change is more invasive as it will also affect Makefiles. src/libvirt_private.syms | 1 + src/util/util.c | 53 ++++++++++++++++++++++++++++++--------------- src/util/util.h | 7 +++-- 3 files changed, 40 insertions(+), 21 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 65911df..948bbe1 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -850,6 +850,7 @@ virFileDeletePid; virFileExists; virFileFindMountPoint; virFileHasSuffix; +virFileIsExecutable; virFileLinkPointsTo; virFileMakePath; virFileMatchesNameSuffix; diff --git a/src/util/util.c b/src/util/util.c index 60feb79..25e6185 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -1255,7 +1255,7 @@ int virFileResolveLink(const char *linkpath, } /* - * Finds a requested file in the PATH env. e.g.: + * Finds a requested executable file in the PATH env. e.g.: * "kvm-img" will return "/usr/bin/kvm-img" * * You must free the result @@ -1263,19 +1263,18 @@ int virFileResolveLink(const char *linkpath, char *virFindFileInPath(const char *file) { char *path; - char pathenv[PATH_MAX]; - char *penv = pathenv; + char *pathiter; char *pathseg; - char fullpath[PATH_MAX]; + char *fullpath = NULL; if (file == NULL) return NULL; /* if we are passed an absolute path (starting with /), return a - * copy of that path + * copy of that path, after validating that it is executable */ - if (file[0] == '/') { - if (virFileExists(file)) + if (IS_ABSOLUTE_FILE_NAME(file)) { + if (virFileIsExecutable(file)) return strdup(file); else return NULL; @@ -1284,27 +1283,45 @@ char *virFindFileInPath(const char *file) /* copy PATH env so we can tweak it */ path = getenv("PATH"); - if (path == NULL || virStrcpyStatic(pathenv, path) == NULL) + if (path == NULL || (path = strdup(path)) == NULL) return NULL; /* for each path segment, append the file to search for and test for * it. return it if found. */ - while ((pathseg = strsep(&penv, ":")) != NULL) { - snprintf(fullpath, PATH_MAX, "%s/%s", pathseg, file); - if (virFileExists(fullpath)) - return strdup(fullpath); + pathiter = path; + while ((pathseg = strsep(&pathiter, ":")) != NULL) { + if (virAsprintf(&fullpath, "%s/%s", pathseg, file) < 0 || + virFileIsExecutable(fullpath)) + break; + VIR_FREE(fullpath); } - return NULL; + VIR_FREE(path); + return fullpath; } -int virFileExists(const char *path) + +bool virFileExists(const char *path) { - struct stat st; + return access(path, F_OK) == 0; +} - if (stat(path, &st) >= 0) - return(1); - return(0); +/* Check that a file can be executed by the current user, and is not + * a directory. */ +bool +virFileIsExecutable(const char *file) +{ + struct stat sb; + + /* The existence of ACLs means that checking (sb.st_mode&0111) for + * executable bits may give false results; plus, access is + * lighter-weight than stat for a first pass filter. + * + * XXX: should this use gnulib's euidaccess module? + */ + return (access(file, X_OK) == 0 && + stat(file, &sb) == 0 && + !S_ISDIR(sb.st_mode)); } #ifndef WIN32 diff --git a/src/util/util.h b/src/util/util.h index 989962f..54c3058 100644 --- a/src/util/util.h +++ b/src/util/util.h @@ -1,8 +1,7 @@ - /* * utils.h: common, generic utility functions * - * Copyright (C) 2010 Red Hat, Inc. + * Copyright (C) 2010-2011 Red Hat, Inc. * Copyright (C) 2006, 2007 Binary Karma * Copyright (C) 2006 Shuveb Hussain * @@ -32,6 +31,7 @@ # include <sys/select.h> # include <sys/types.h> # include <stdarg.h> +# include <stdbool.h> # ifndef MIN # define MIN(a, b) ((a) < (b) ? (a) : (b)) @@ -120,7 +120,8 @@ int virFileResolveLink(const char *linkpath, char *virFindFileInPath(const char *file); -int virFileExists(const char *path); +bool virFileExists(const char *file); +bool virFileIsExecutable(const char *file); char *virFileSanitizePath(const char *path); -- 1.7.3.4

2 7

[libvirt] [PATCH 0/5] Explicit boot device ordering
by Jiri Denemark 13 Jan '11

13 Jan '11

Currently, boot order can be specified per device class but there is no way to specify exact disk/NIC device to boot from. This patchset fixes that. There were two options suggested for how this should be modeled in domain XML: 1) <os> <boot target='net1'/> <boot target='net0'/> <boot target='hdc'/> <boot target='hdb'/> <os> Where target attributes would match /domain/devices/*/target@dev 2) Adding <boot order='n'/> elements into appropriate device elements. In addition to the two options Rich suggested a more compact variant of option 1. Per former discussion this patchset implements the second option. Jiri Denemark (5): conf: Move boot parsing into a separate function Introduce per-device boot element qemu: Refactor qemuCapsParsePCIDeviceStrs using virCommand qemu: Support per-device boot ordering tests: Add tests for per-device boot elements docs/formatcaps.html.in | 1 + docs/formatdomain.html.in | 41 +++++- docs/schemas/domain.rng | 20 +++- src/conf/capabilities.c | 3 +- src/conf/domain_conf.c | 148 ++++++++++++++------ src/conf/domain_conf.h | 2 + src/qemu/qemu_capabilities.c | 82 +++++++---- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 15 ++- src/qemu/qemu_command.h | 6 +- src/qemu/qemu_driver.c | 2 + src/qemu/qemu_hotplug.c | 8 +- .../qemuxml2argvdata/qemuxml2argv-boot-order.args | 1 + tests/qemuxml2argvdata/qemuxml2argv-boot-order.xml | 44 ++++++ tests/qemuxml2argvtest.c | 2 + tests/qemuxml2xmltest.c | 1 + 16 files changed, 291 insertions(+), 86 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-boot-order.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-boot-order.xml -- 1.7.4.rc1

2 7

[libvirt] [Patch v5 0/4] Add new feature into live migration
by Wen Congyang 13 Jan '11

13 Jan '11

If the memory of guest OS is changed constantly, the live migration can not be ended ever for ever. We can use the command 'virsh migrate-setmaxdowntime' to control the live migration. But the value of maxdowntime is diffcult to calculate because it depends on the transfer speed of network and constantly changing memroy size. We need a easy way to control the live migration. This patch set add the support of forcing guest to suspend at timeout. With this patch set, when we migrate the guest OS, we can specify a timeout. If the live migration timeouts, auto-suspend the guest OS, where the migration will complete offline. Test of this patchset on Linux: Env: a. The size of guest OS's memory: 1GB b. The transfer speed of network: about 100Mb/s c. The size of constantly changing memory: more than 900MB 1. migrate without timeout # virsh migrate --live RHEL6RC qemu+ssh://<dest IP>/system tcp://<dest IP>:49152 The migration does not end after 12 hours. 2. migrate with timeout(30 minutes): # virsh -t migrate --live --timeout 1800 RHEL6RC qemu+ssh://<dest IP>/system tcp:<dest IP>:49152 (Time: 1816768.066 ms) v5: - merge daemon/event.* and src/util/event.* - test event when builing without libvirtd v4: - move tools/timer.* into src/util/ directory v3: - use the existing virEventXXXTimeout() APT to implement timer - merge Add, Update, Delete into one single interface virSetTimeout() - start timer thread when initializing timer v2: - implement timer for Windows - implement dynamic timers Wen Congyang (4): move daemon/event.* into src/util/ directory test event when builing without libvirtd timer impl force guest to suspend at timeout cfg.mk | 1 + daemon/Makefile.am | 1 - daemon/event.c | 700 ---------------------------------------------- daemon/event.h | 134 --------- src/Makefile.am | 3 +- src/libvirt.c | 2 + src/libvirt_private.syms | 17 ++ src/util/event.c | 676 ++++++++++++++++++++++++++++++++++++++++++++- src/util/event.h | 106 +++++++ src/util/timer.c | 158 +++++++++++ src/util/timer.h | 34 +++ tests/Makefile.am | 6 +- tests/eventtest.c | 2 +- tools/Makefile.am | 1 - tools/console.c | 2 +- tools/virsh.c | 69 +++++- tools/virsh.pod | 4 + 17 files changed, 1069 insertions(+), 847 deletions(-) delete mode 100644 daemon/event.c delete mode 100644 daemon/event.h create mode 100644 src/util/timer.c create mode 100644 src/util/timer.h

2 6