March 2010 - Devel - libvirt List Archives

[libvirt] [PATCH] Add entry point logging for cpu functions
by Jiri Denemark 25 Mar '10

25 Mar '10

--- src/cpu/cpu.c | 46 ++++++++++++++++++++++++++++++++++++++++++++-- 1 files changed, 44 insertions(+), 2 deletions(-) diff --git a/src/cpu/cpu.c b/src/cpu/cpu.c index be0f15e..183862a 100644 --- a/src/cpu/cpu.c +++ b/src/cpu/cpu.c @@ -23,6 +23,7 @@ #include <config.h> +#include "logging.h" #include "memory.h" #include "xml.h" #include "cpu.h" @@ -73,6 +74,8 @@ cpuCompareXML(virCPUDefPtr host, virCPUDefPtr cpu = NULL; virCPUCompareResult ret = VIR_CPU_COMPARE_ERROR; + VIR_DEBUG("host=%p, xml=%s", host, NULLSTR(xml)); + if (!(doc = virXMLParseString(xml, "cpu.xml"))) goto cleanup; @@ -104,6 +107,8 @@ cpuCompare(virCPUDefPtr host, { struct cpuArchDriver *driver; + VIR_DEBUG("host=%p, cpu=%p", host, cpu); + if ((driver = cpuGetSubDriver(host->arch)) == NULL) return VIR_CPU_COMPARE_ERROR; @@ -126,6 +131,13 @@ cpuDecode(virCPUDefPtr cpu, { struct cpuArchDriver *driver; + VIR_DEBUG("cpu=%p, data=%p, nmodels=%u", cpu, data, nmodels); + if (models) { + unsigned int i; + for (i = 0; i < nmodels; i++) + VIR_DEBUG("models[%u]=%s", i, NULLSTR(models[i])); + } + if (models == NULL && nmodels != 0) { virCPUReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("nonzero nmodels doesn't match with NULL models")); @@ -163,6 +175,11 @@ cpuEncode(const char *arch, { struct cpuArchDriver *driver; + VIR_DEBUG("arch=%s, cpu=%p, forced=%p, required=%p, " + "optional=%p, disabled=%p, forbidden=%p", + NULLSTR(arch), cpu, forced, required, + optional, disabled, forbidden); + if ((driver = cpuGetSubDriver(arch)) == NULL) return -1; @@ -184,6 +201,8 @@ cpuDataFree(const char *arch, { struct cpuArchDriver *driver; + VIR_DEBUG("arch=%s, data=%p", NULLSTR(arch), data); + if (data == NULL) return; @@ -206,6 +225,8 @@ cpuNodeData(const char *arch) { struct cpuArchDriver *driver; + VIR_DEBUG("arch=%s", NULLSTR(arch)); + if ((driver = cpuGetSubDriver(arch)) == NULL) return NULL; @@ -227,6 +248,8 @@ cpuGuestData(virCPUDefPtr host, { struct cpuArchDriver *driver; + VIR_DEBUG("host=%p, guest=%p, data=%p", host, guest, data); + if ((driver = cpuGetSubDriver(host->arch)) == NULL) return VIR_CPU_COMPARE_ERROR; @@ -254,6 +277,16 @@ cpuBaselineXML(const char **xmlCPUs, char *cpustr; unsigned int i; + VIR_DEBUG("ncpus=%u, nmodels=%u", ncpus, nmodels); + if (xmlCPUs) { + for (i = 0; i < ncpus; i++) + VIR_DEBUG("xmlCPUs[%u]=%s", i, NULLSTR(xmlCPUs[i])); + } + if (models) { + for (i = 0; i < nmodels; i++) + VIR_DEBUG("models[%u]=%s", i, NULLSTR(models[i])); + } + if (xmlCPUs == NULL && ncpus != 0) { virCPUReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("nonzero ncpus doesn't match with NULL xmlCPUs")); @@ -320,6 +353,17 @@ cpuBaseline(virCPUDefPtr *cpus, { struct cpuArchDriver *driver; virCPUDefPtr cpu; + unsigned int i; + + VIR_DEBUG("ncpus=%u, nmodels=%u", ncpus, nmodels); + if (cpus) { + for (i = 0; i < ncpus; i++) + VIR_DEBUG("cpus[%u]=%p", i, cpus[i]); + } + if (models) { + for (i = 0; i < nmodels; i++) + VIR_DEBUG("models[%u]=%s", i, NULLSTR(models[i])); + } if (cpus == NULL && ncpus != 0) { virCPUReportError(VIR_ERR_INTERNAL_ERROR, @@ -349,8 +393,6 @@ cpuBaseline(virCPUDefPtr *cpus, } if ((cpu = driver->baseline(cpus, ncpus, models, nmodels))) { - int i; - cpu->type = VIR_CPU_TYPE_GUEST; cpu->match = VIR_CPU_MATCH_EXACT; VIR_FREE(cpu->arch); -- 1.7.0.3

2 2

[libvirt] Sleeping 50 microseconds in qemuDomainWaitForMigrationComplete?
by Matthias Bolte 25 Mar '10

25 Mar '10

There is this nanosleep call in the QEMU driver. /* Poll every 50ms for progress & to allow cancellation */ struct timespec ts = { .tv_sec = 0, .tv_nsec = 50 * 1000ull }; [...] nanosleep(&ts, NULL); The comment says 50ms (assuming m is for milli and not for micro. because normally u is used for micro) but the actual nanosleep call is done with 50 microseconds. Is the comment right and the code wrong? I assume so, because 50 microseconds is pretty short and 50 milliseconds sounds more reasonable. Matthias

2 1

[libvirt] [PATCH 0/6] Support VNC password changes on running guests v2
by Daniel P. Berrange 25 Mar '10

25 Mar '10

This patch introduces a new virDomainUpdateDeviceFlags() API to allow the configuration of a device to be changed on the fly. It ports the existing CDROM media change code in all drivers to use this new API (keeps old support too). It then adds VNC password change support in the QEMU driver using this new API v2: Fix feedback from first posting

3 14

Re: [libvirt] [Qemu-devel] Re: Supporting hypervisor specific APIs in libvirt
by Anthony Liguori 24 Mar '10

24 Mar '10

On 03/24/2010 07:25 AM, Paul Brook wrote: >> I can't quite see what such a libqemu would buy us compared to straight >> QMP. >> >> Talking QMP should be easy, provided you got a suitable JSON library. >> > I agree. My undesranding is this was one of the large motivations behind using > JSON: It's a common protocol that already has convenient bindings in most > languages. If it's hard[1] for third parties to bind QMP to their favourite > language/framework then IMHO we've done it wrong. > You can't have convenient bindings to an RPC in C because it doesn't support dynamic dispatch. With most types of RPC, you have an IDL description and a code generator. But regardless of that, there are advantages to us providing a libqemu. The biggest one is that we can standardize transport implementations that include discovery mechanisms. If the core of libqemu provided an extensible transport interface, and a generic QMP request/completion mechanism, in a Python binding, you would never use the IDL generated wrappers but instead use dynamic dispatch to invoke arbitrary QMP requests. But the advantage is that if libvirt provided an API for a QMP transport encapsulated in their secure protocol, then provided the plumbed that API through their Python interface, you could use it for free in Python without having to reinvent the wheel. Regards, Anthony Liguori > Paul > > [1] Hard compared to any other sane RPC mechanism. Some languages make > everything hard :-) >

2 1

[libvirt] [PATCH] build: suppress distracting build output
by Jim Meyering 24 Mar '10

24 Mar '10

While building "make check", I saw this scroll by (too quickly to see what it was), but it was different enough from the rest that... CCLD libvirt.la CCLD libvirt_test.la make check-local make[3]: Entering directory `/home/j/w/co/libvirt/src' if test -x '/usr/bin/augparse'; then \ '/usr/bin/augparse' -I ./qemu \ ./qemu/test_libvirtd_qemu.aug; \ fi if test -x '/usr/bin/augparse'; then \ '/usr/bin/augparse' -I ./lxc \ ./lxc/test_libvirtd_lxc.aug; \ fi CC libvirtd-event.o CC libvirtd-libvirtd.o I scrolled back to investigate. Not a problem, but let's eliminate the distraction. Here's a patch. With it, instead of the above "if test..." lines, you'll see this single line: GEN augeas-check >From 84eb6effdd0241a8bae196d4b77d873aea7fd3db Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Wed, 24 Mar 2010 09:39:54 +0100 Subject: [PATCH] build: suppress distracting build output * src/Makefile.am (augeas-check): New target, just to give the existing rule a name. At the same time, prefix the commands with $(AM_V_GEN), to avoid unexpected build output with V=0 which is the default. --- src/Makefile.am | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 0aa3443..5f6b325 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -766,15 +766,18 @@ EXTRA_DIST += \ $(SECRET_DRIVER_SOURCES) \ $(VBOX_DRIVER_EXTRA_DIST) -check-local: +check-local: augeas-check + +.PHONY: augeas-check +augeas-check: if WITH_QEMU - if test -x '$(AUGPARSE)'; then \ + $(AM_V_GEN)if test -x '$(AUGPARSE)'; then \ '$(AUGPARSE)' -I $(srcdir)/qemu \ $(srcdir)/qemu/test_libvirtd_qemu.aug; \ fi endif if WITH_LXC - if test -x '$(AUGPARSE)'; then \ + $(AM_V_GEN)if test -x '$(AUGPARSE)'; then \ '$(AUGPARSE)' -I $(srcdir)/lxc \ $(srcdir)/lxc/test_libvirtd_lxc.aug; \ fi -- 1.7.0.3.435.g097f4

2 1

[libvirt] [PATCH] vbox: Fix segfault on empty device source
by Matthias Bolte 24 Mar '10

24 Mar '10

<source file=''/> results in def->disks[i]->src == NULL. But vboxDomainDefineXML didn't check def->disks[i]->src for NULL and expected it to be a valid string. Add checks for def->disks[i]->src != NULL to fix the segfault. --- src/vbox/vbox_tmpl.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/vbox/vbox_tmpl.c b/src/vbox/vbox_tmpl.c index 1765d63..fefed1f 100644 --- a/src/vbox/vbox_tmpl.c +++ b/src/vbox/vbox_tmpl.c @@ -3519,7 +3519,8 @@ static virDomainPtr vboxDomainDefineXML(virConnectPtr conn, const char *xml) { DEBUG("disk(%d) shared: %s", i, def->disks[i]->shared ? "True" : "False"); if (def->disks[i]->device == VIR_DOMAIN_DISK_DEVICE_CDROM) { - if (def->disks[i]->type == VIR_DOMAIN_DISK_TYPE_FILE) { + if (def->disks[i]->type == VIR_DOMAIN_DISK_TYPE_FILE && + def->disks[i]->src != NULL) { IDVDDrive *dvdDrive = NULL; /* Currently CDROM/DVD Drive is always IDE * Secondary Master so neglecting the following @@ -3801,7 +3802,8 @@ static virDomainPtr vboxDomainDefineXML(virConnectPtr conn, const char *xml) { DEBUG("disk(%d) readonly: %s", i, def->disks[i]->readonly ? "True" : "False"); DEBUG("disk(%d) shared: %s", i, def->disks[i]->shared ? "True" : "False"); - if (def->disks[i]->type == VIR_DOMAIN_DISK_TYPE_FILE) { + if (def->disks[i]->type == VIR_DOMAIN_DISK_TYPE_FILE && + def->disks[i]->src != NULL) { IMedium *medium = NULL; PRUnichar *mediumUUID = NULL; PRUnichar *mediumFileUtf16 = NULL; -- 1.6.3.3

3 4

[libvirt] error: internal error canonical hostname pointed to localhost, but this is not allowed
by Kenneth Nagin 24 Mar '10

24 Mar '10

When trying to migrate a VM the destination libvirtd virGetHostnameLocalhost throws an internal error. I issue this command from the source host. > virsh migrate rel-vm1 qemu+tls://croton/system error: internal error canonical hostname pointed to localhost, but this is not allowed The destination host throws the error: 14:14:57.914: debug : virDomainMigratePrepare2:3740 : dconn=0xed17e0, cookie=0x7f983a352d68, cookielen=0x7f983a352d74, uri_in=(null), uri_out=0xe8c360,flags=0, dname=(null), bandwidth=0, dom_xml=<domain type='kvm' id='1'> <name>rel-vm1</name> <uuid>c0c98fc9-8cd6-e246-a06e-b5ce37c4f05e</uuid> <memory>262144</memory> <currentMemory>262144</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64' machine='pc-0.11'>hvm</type> <boot dev='hd'/> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/libexec/qemu-kvm</emulator> <disk type='file' device='disk'> <source file='/share/images/nagin/rel-vm1/rel-vm1.img'/> <target dev='hda' bus='ide'/> <alias name='ide0-0-0'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <controller type='ide' index='0'> <alias name='ide0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='5900' autoport='yes'/> <video> <model type='cirrus' vram='9216' heads='1'/> <alias name='video0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> </devices> </domain> 14:14:57.914: error : virGetHostnameLocalhost:2372 : internal error canonical hostname pointed to localhost, but this is not allowed libvir: error : internal error canonical hostname pointed to localhost, but this is not allowed 14:14:57.914: debug : remoteSerializeError:141 : prog=536903814 ver=1 proc=108 type=1 serial=3, msg=internal error canonical hostname pointed to localhost, but this is not allowed The libvirt version is the tip of git: [root@croton rel-vm1]# virsh version Compiled against library: libvir 0.7.7 Using library: libvir 0.7.7 Using API: QEMU 0.7.7 Running hypervisor: QEMU 0.12.1 Changing qemudDomainMigratePrepare2 call to virGetHostnameLocalhost from: if ((hostname = virGetHostnameLocalhost(0)) == NULL) to: if ((hostname = virGetHostnameLocalhost(1)) == NULL) seems to fix the problem. But maybe this is only masking a another problem. --Kenneth Nagin

3 3

[libvirt] ANNOUNCE: virt-manager 0.8.4 and virtinst 0.500.3
by Cole Robinson 24 Mar '10

24 Mar '10

I'm happy to announce two new releases: virt-manager 0.8.4: virt-manager is a desktop application for managing KVM and Xen virtual machines via libvirt. virtinst 0.500.3: virtinst is a collection of command line tools for provisioning libvirt virtual machines, including virt-install and virt-clone. The releases can be downloaded from: http://virt-manager.org/download.html The direct download links are: http://virt-manager.org/download/sources/virt-manager/virt-manager-0.8.4.ta… http://virt-manager.org/download/sources/virtinst/virtinst-0.500.3.tar.gz The virt-manager release includes: - Import install option: create a VM using an existing OS image - Support multiple boot devices and boot order - Watchdog device support - Enable setting a human readable VM description. - Option to manually specify a bridge name, if bridge isn't detected by virt-manager The virtinst release includes: - virt-install: New --watchdog option: configure a virtual watchdog dev - virt-install: New --soundhw option: More flexible sound configuration deprecates --sound, though back compat is maintained - virt-install: New --security option: configure VM security driver - virt-install: New --description option: set a human readable desc - Better OS defaults: Use <video> VGA and <sound> AC97 if supported Thanks to everyone who has contributed to this release through testing, bug reporting, submitting patches, and otherwise sending in feedback! Thanks, Cole

1 0

[libvirt] Two issues regarding TFTP support in virtual networks...
by Darryl L. Pierce 24 Mar '10

24 Mar '10

The first is just a wiki fix: the wiki says this functionality is available as of 0.7.1 of libvirt. The code though is only in the 0.7.2 tag and later. So the wiki should say 0.7.2 instead. The second regards how I'm using it and what I'm doing wrong. I'm creating a virtual network and pointing it to a temporary directory where I've run livecd-iso-to-pxeboot to setup an ISO file for PXE booting. So the network XML looks like this: (mcpierce@mcpierce-desktop:node-image)$ sudo virsh net-dumpxml testbr541 <network> <name>testbr541</name> <uuid>f389317f-8420-7516-df9d-756b7deb3d37</uuid> <forward mode='nat'/> <bridge name='testbr541' stp='on' delay='0' /> <ip address='192.168.31.1' netmask='255.255.255.0'> <tftp root='/tmp/tmp.3B8opJfBXw/tftpboot/' /> <dhcp> <range start='192.168.31.100' end='192.168.31.199' /> </dhcp> </ip> </network> When I start up the VM, I see it get an IP address within the range specified, but it never pxeboots the ISO image. To what directory should the root attribute be pointed? -- Darryl L. Pierce, Sr. Software Engineer @ Red Hat, Inc. Delivering value year after year. Red Hat ranks #1 in value among software vendors. http://www.redhat.com/promo/vendor/

2 14

Re: [libvirt] [Qemu-devel] Re: Supporting hypervisor specific APIs in libvirt
by Markus Armbruster 24 Mar '10

24 Mar '10

Paul Brook <paul(a)codesourcery.com> writes: >> > IMO the no_user flag is a bug, and should not exist. >> >> Sorry, what's that? > > Usually an indication that a device has been incorrectly or inproperly > converted to the qdev interface. Can also be an indication that the device can't support multiple instances. For instance: commit 39a51dfda835a75c0ebbfd92705b96e4de77f795 Author: Markus Armbruster <armbru(a)redhat.com> Date: Tue Oct 27 13:52:13 2009 +0100 qdev: Tag isa-fdc, PIIX3 IDE and PIIX4 IDE as no-user These devices are created automatically, and attempting to create another one with -device fails with "qemu: hardware error: register_ioport_write: invalid opaque". Signed-off-by: Markus Armbruster <armbru(a)redhat.com> Signed-off-by: Anthony Liguori <aliguori(a)us.ibm.com> With no-user, we at least fail with a decent error message. I don't think it's fair to demand that a qdev conversion must relax restrictions that haven't otherwise bothered us to be correct and proper. We'll relax them if and when they bother us enough to make somebody send a decent patch. And yes, there are better ways to disallow multiple instances of a device than declaring it no-user. Patches welcome.

1 0