- Devel - libvirt List Archives

[libvirt] [PATCH] Add PCI sysfs reset access
by Alex Williamson 18 Mar '11

18 Mar '11

I'm proposing we make use of $PCIDIR/reset in qemu-kvm to reset devices on VM reset. We need to add it to libvirt's list of files that get ownership for device assignment. Signed-off-by: Alex Williamson <alex.williamson(a)redhat.com> --- src/util/pci.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/util/pci.c b/src/util/pci.c index 095ad3f..8d2dbb0 100644 --- a/src/util/pci.c +++ b/src/util/pci.c @@ -1349,11 +1349,13 @@ int pciDeviceFileIterate(pciDevice *dev, while ((ent = readdir(dir)) != NULL) { /* Device assignment requires: - * $PCIDIR/config, $PCIDIR/resource, $PCIDIR/resourceNNN, $PCIDIR/rom + * $PCIDIR/config, $PCIDIR/resource, $PCIDIR/resourceNNN, + * $PCIDIR/rom, $PCIDIR/reset */ if (STREQ(ent->d_name, "config") || STRPREFIX(ent->d_name, "resource") || - STREQ(ent->d_name, "rom")) { + STREQ(ent->d_name, "rom") || + STREQ(ent->d_name, "reset")) { if (virAsprintf(&file, "%s/%s", pcidir, ent->d_name) < 0) { virReportOOMError(); goto cleanup;

3 4

[libvirt] [PATCH 0/2] Last debug buffer patches: locking and docs
by Daniel Veillard 18 Mar '11

18 Mar '11

That's the remaining changes left for the debug buffer. The first patch is the one removing locking in the signal error handler. As discussed it's better to have a risk of slightly corrupted output in the log dump than have a risk crashing due to the unsafe locking operations in the signal handler. The second patch documents the debug buffer changes and improve the logging page to add more structure, and more examples. Daniel

2 6

[libvirt] [PATCH] do not unref obj in qemuDomainObjExitMonitorWithDriver
by Wen Congyang 18 Mar '11

18 Mar '11

Steps to reproduce this bug: # cat test.sh #! /bin/bash -x virsh start domain sleep 5 virsh qemu-monitor-command domain 'cpu_set 2 online' --hmp # while true; do ./test.sh ; done Then libvirtd will crash. The reason is that: we add a reference of obj when we open the monitor. We will reduce this reference when we free the monitor. If the reference of monitor is 0, we will free monitor automatically and the reference of obj is reduced. But in the function qemuDomainObjExitMonitorWithDriver(), we reduce this reference again when the reference of monitor is 0. It will cause the obj be freed in the function qemuDomainObjEndJob(). Then we start the domain again, and libvirtd will crash in the function virDomainObjListSearchName(), because we pass a null pointer(obj->def->name) to strcmp(). Signed-off-by: Wen Congyang <wency(a)cn.fujitsu.com> --- src/qemu/qemu_domain.c | 1 - 1 files changed, 0 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 8a2b9cc..ae28b1c 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -634,7 +634,6 @@ void qemuDomainObjExitMonitorWithDriver(struct qemud_driver *driver, virDomainObjLock(obj); if (refs == 0) { - virDomainObjUnref(obj); priv->mon = NULL; } } -- 1.7.1

3 3

[libvirt] [PATCH] check driver name while attaching disk
by Wen Congyang 18 Mar '11

18 Mar '11

This bug was reported by Shi Jin(jinzishuai(a)gmail.com): ============= # virsh attach-disk RHEL6RC /var/lib/libvirt/images/test3.img vdb --driver file --subdriver qcow2 Disk attached successfully # virsh save RHEL6RC /var/lib/libvirt/images/memory.save Domain RHEL6RC saved to /var/lib/libvirt/images/memory.save # virsh restore /var/lib/libvirt/images/memory.save error: Failed to restore domain from /var/lib/libvirt/images/memory.save error: internal error unsupported driver name 'file' for disk '/var/lib/libvirt/images/test3.img' ============= We have checked the driver name when we start or restore VM, but we do not check it while attaching a disk. Signed-off-by: Wen Congyang <wency(a)cn.fujitsu.com> --- src/qemu/qemu_driver.c | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 2892dfe..e94080d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -3994,6 +3994,14 @@ static int qemudDomainAttachDevice(virDomainPtr dom, goto endjob; if (dev->type == VIR_DOMAIN_DEVICE_DISK) { + if (dev->data.disk->driverName != NULL && + !STREQ(dev->data.disk->driverName, "qemu")) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, + _("unsupported driver name '%s' for disk '%s'"), + dev->data.disk->driverName, dev->data.disk->src); + goto endjob; + } + if (qemuCgroupControllerActive(driver, VIR_CGROUP_CONTROLLER_DEVICES)) { if (virCgroupForDomain(driver->cgroup, vm->def->name, &cgroup, 0) !=0 ) { qemuReportError(VIR_ERR_INTERNAL_ERROR, -- 1.7.1

4 6

[libvirt] xen:/// vs. xen://FQDN/ vs xen+unix:/// discrepancy
by Philipp Hahn 18 Mar '11

18 Mar '11

Hello, I regularly observe the problem, that depending on the libvirt-URL I get different information: root@xen4# virsh -c xen://xen4.domain.name/ list Id Name State ---------------------------------- root@xen4# virsh -c xen:/// list Id Name State ---------------------------------- 0 Domain-0 running 1 dos4 idle root@xen4# virsh -c xen+unix:/// list Id Name State ---------------------------------- root@xen4# xm li Name ID Mem VCPUs State Time(s) Domain-0 0 2044 2 r----- 419.7 dos4 1 4 1 -b---- 33.4 Has somebody observed the same problem and/or can give me any advise on where to look for the problem? I'm running a Debian-Lenny based OS with libvirt-0.8.2 on Xen-3.4.3 with a 2.6.32 kernel. Sincerely Philipp Hahn -- Philipp Hahn Open Source Software Engineer hahn(a)univention.de Univention GmbH Linux for Your Business fon: +49 421 22 232- 0 Mary-Somerville-Str.1 28359 Bremen fax: +49 421 22 232-99 http://www.univention.de

3 5

[libvirt] [PATCH] Add vim configuration that makes vim auto-indent code
by Hu Tao 18 Mar '11

18 Mar '11

--- .lvimrc | 10 ++++++++++ HACKING | 5 +++++ 2 files changed, 15 insertions(+), 0 deletions(-) create mode 100644 .lvimrc diff --git a/.lvimrc b/.lvimrc new file mode 100644 index 0000000..3a732ce --- /dev/null +++ b/.lvimrc @@ -0,0 +1,10 @@ +set nocompatible +filetype on +set autoindent +set smartindent +set cindent +set tabstop=4 +set shiftwidth=4 +set expandtab +set cinoptions=(0,:0,l1,t0 +filetype plugin indent on diff --git a/HACKING b/HACKING index 4a71b37..adfb7ae 100644 --- a/HACKING +++ b/HACKING @@ -87,6 +87,11 @@ If you use Emacs, add the following to one of one of your start-up files '(lambda () (if (string-match "/libvirt" (buffer-file-name)) (libvirt-c-mode)))) +If you use vim, install vim script at this page: +http://www.vim.org/scripts/script.php?script_id=1408 +it will auto load .lvimrc at the root of libvirt source only when you're +editing libvirt's code. Or you can append the contents of .lvimrc to you +.vimrc file. Code formatting (especially for new code) ========================================= -- 1.7.3.1

4 9

[libvirt] [PATCH] virsh: fix memtune's help message for swap_hard_limit
by KAMEZAWA Hiroyuki 18 Mar '11

18 Mar '11

I found this when I used virsh memtune... BTW, how to fix http://libvirt.org/formatdomain.html 's memtune description ? == >From 541ae04430f376e8168b413a20b35dce49779816 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com> Date: Thu, 3 Mar 2011 14:24:45 +0900 Subject: [PATCH 6/6] fix virsh commands' message for memtune'swap_hard_limit. cgroup's /cgroup/memory/memory.memsw.limit_in_bytes is not for limitinit 'swap' but for 'memory+swap' (then, it's memsw) (So, this number cannot be smaller than memory.limit_in_bytes) Note: If other hypervisors than Linux support this and meaning is not same as memory+swap, the name swap_hard_limit will have confusion. Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com> --- tools/virsh.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 62fca17..3a8fdaa 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -3033,7 +3033,7 @@ static const vshCmdOptDef opts_memtune[] = { {"soft-limit", VSH_OT_INT, VSH_OFLAG_NONE, N_("Memory during contention in kilobytes")}, {"swap-hard-limit", VSH_OT_INT, VSH_OFLAG_NONE, - N_("Max swap in kilobytes")}, + N_("Max memory+swap in kilobytes")}, {"min-guarantee", VSH_OT_INT, VSH_OFLAG_NONE, N_("Min guaranteed memory in kilobytes")}, {NULL, 0, 0, NULL} -- 1.7.4.1

5 14

[libvirt] [PATCH] Make virDomainObjParseNode() static
by Hu Tao 18 Mar '11

18 Mar '11

Make virDomainObjParseNode() static since it is called only in one file. --- src/conf/domain_conf.c | 38 ++++++++++++++++++++------------------ src/conf/domain_conf.h | 3 --- 2 files changed, 20 insertions(+), 21 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 16e1291..c637300 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -6081,24 +6081,10 @@ cleanup: } -virDomainObjPtr virDomainObjParseFile(virCapsPtr caps, - const char *filename) -{ - xmlDocPtr xml; - virDomainObjPtr obj = NULL; - - if ((xml = virXMLParseFile(filename))) { - obj = virDomainObjParseNode(caps, xml, xmlDocGetRootElement(xml)); - xmlFreeDoc(xml); - } - - return obj; -} - - -virDomainObjPtr virDomainObjParseNode(virCapsPtr caps, - xmlDocPtr xml, - xmlNodePtr root) +static virDomainObjPtr +virDomainObjParseNode(virCapsPtr caps, + xmlDocPtr xml, + xmlNodePtr root) { xmlXPathContextPtr ctxt = NULL; virDomainObjPtr obj = NULL; @@ -6123,6 +6109,22 @@ cleanup: return obj; } + +virDomainObjPtr virDomainObjParseFile(virCapsPtr caps, + const char *filename) +{ + xmlDocPtr xml; + virDomainObjPtr obj = NULL; + + if ((xml = virXMLParseFile(filename))) { + obj = virDomainObjParseNode(caps, xml, xmlDocGetRootElement(xml)); + xmlFreeDoc(xml); + } + + return obj; +} + + static int virDomainDefMaybeAddController(virDomainDefPtr def, int type, int idx) diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 30aeccc..cbd0f98 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -1246,9 +1246,6 @@ virDomainDefPtr virDomainDefParseNode(virCapsPtr caps, virDomainObjPtr virDomainObjParseFile(virCapsPtr caps, const char *filename); -virDomainObjPtr virDomainObjParseNode(virCapsPtr caps, - xmlDocPtr xml, - xmlNodePtr root); int virDomainDefAddImplicitControllers(virDomainDefPtr def); -- 1.7.3.1

2 1

Re: [libvirt] [PATCH] device-assignment: Reset device on system reset
by Chris Wright 17 Mar '11

17 Mar '11

* Alex Williamson (alex.williamson(a)redhat.com) wrote: > static void reset_assigned_device(DeviceState *dev) > { > - PCIDevice *d = DO_UPCAST(PCIDevice, qdev, dev); > + PCIDevice *pci_dev = DO_UPCAST(PCIDevice, qdev, dev); > + AssignedDevice *adev = DO_UPCAST(AssignedDevice, dev, pci_dev); > + char reset_file[64]; > + const char reset[] = "1"; > + int fd, ret; > + > + snprintf(reset_file, sizeof(reset_file), > + "/sys/bus/pci/devices/0000:%02x:%02x.%01x/reset", > + adev->host.bus, adev->host.dev, adev->host.func); need to consider segment: %04x:..., adev->host.seg, ... > + /* > + * Issue a device reset via pci-sysfs. Note that we use write(2) here > + * and ignore the return value because some kernels have a bug that > + * returns 0 rather than bytes written on success, sending us into an > + * infinite retry loop using other write mechanisms. > + */ > + fd = open(reset_file, O_WRONLY); > + if (fd != -1) { > + ret = write(fd, reset, strlen(reset)); > + close(fd); > + } This will probably fail when it's managed by libvirt. I expect it will need some file ownership and security label mgmt added to device assignement path I expect.

2 2

[libvirt] [PATCH] Support Xen sysctl v8, domctl v7
by Jim Fehlig 17 Mar '11

17 Mar '11

xen-unstable c/s 21118:28e5409e3fb3 bumped sysctl version to 8. xen-unstable c/s 21212:de94884a669c introduced CPU pools feature, adding another member to xen_domctl_getdomaininfo struct. Add a corresponding domctl v7 struct in xen hypervisor sub-driver and detect sysctl v8 during initialization. --- src/xen/xen_hypervisor.c | 125 ++++++++++++++++++++++++++++++++++------------ 1 files changed, 93 insertions(+), 32 deletions(-) diff --git a/src/xen/xen_hypervisor.c b/src/xen/xen_hypervisor.c index 9cb8e8c..47355ce 100644 --- a/src/xen/xen_hypervisor.c +++ b/src/xen/xen_hypervisor.c @@ -223,11 +223,28 @@ struct xen_v2d6_getdomaininfo { }; typedef struct xen_v2d6_getdomaininfo xen_v2d6_getdomaininfo; +struct xen_v2d7_getdomaininfo { + domid_t domain; /* the domain number */ + uint32_t flags; /* flags, see before */ + uint64_t tot_pages ALIGN_64; /* total number of pages used */ + uint64_t max_pages ALIGN_64; /* maximum number of pages allowed */ + uint64_t shr_pages ALIGN_64; /* number of shared pages */ + uint64_t shared_info_frame ALIGN_64; /* MFN of shared_info struct */ + uint64_t cpu_time ALIGN_64; /* CPU time used */ + uint32_t nr_online_vcpus; /* Number of VCPUs currently online. */ + uint32_t max_vcpu_id; /* Maximum VCPUID in use by this domain. */ + uint32_t ssidref; + xen_domain_handle_t handle; + uint32_t cpupool; +}; +typedef struct xen_v2d7_getdomaininfo xen_v2d7_getdomaininfo; + union xen_getdomaininfo { struct xen_v0_getdomaininfo v0; struct xen_v2_getdomaininfo v2; struct xen_v2d5_getdomaininfo v2d5; struct xen_v2d6_getdomaininfo v2d6; + struct xen_v2d7_getdomaininfo v2d7; }; typedef union xen_getdomaininfo xen_getdomaininfo; @@ -236,6 +253,7 @@ union xen_getdomaininfolist { struct xen_v2_getdomaininfo *v2; struct xen_v2d5_getdomaininfo *v2d5; struct xen_v2d6_getdomaininfo *v2d6; + struct xen_v2d7_getdomaininfo *v2d7; }; typedef union xen_getdomaininfolist xen_getdomaininfolist; @@ -273,147 +291,179 @@ typedef struct xen_v2s5_availheap xen_v2s5_availheap; #define XEN_GETDOMAININFOLIST_ALLOC(domlist, size) \ (hypervisor_version < 2 ? \ (VIR_ALLOC_N(domlist.v0, (size)) == 0) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + (VIR_ALLOC_N(domlist.v2d7, (size)) == 0) : \ + (dom_interface_version == 6 ? \ (VIR_ALLOC_N(domlist.v2d6, (size)) == 0) : \ (dom_interface_version == 5 ? \ (VIR_ALLOC_N(domlist.v2d5, (size)) == 0) : \ - (VIR_ALLOC_N(domlist.v2, (size)) == 0)))) + (VIR_ALLOC_N(domlist.v2, (size)) == 0))))) #define XEN_GETDOMAININFOLIST_FREE(domlist) \ (hypervisor_version < 2 ? \ VIR_FREE(domlist.v0) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + VIR_FREE(domlist.v2d7) : \ + (dom_interface_version == 6 ? \ VIR_FREE(domlist.v2d6) : \ (dom_interface_version == 5 ? \ VIR_FREE(domlist.v2d5) : \ - VIR_FREE(domlist.v2)))) + VIR_FREE(domlist.v2))))) #define XEN_GETDOMAININFOLIST_CLEAR(domlist, size) \ (hypervisor_version < 2 ? \ memset(domlist.v0, 0, sizeof(*domlist.v0) * size) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + memset(domlist.v2d7, 0, sizeof(*domlist.v2d7) * size) : \ + (dom_interface_version == 6 ? \ memset(domlist.v2d6, 0, sizeof(*domlist.v2d6) * size) : \ (dom_interface_version == 5 ? \ memset(domlist.v2d5, 0, sizeof(*domlist.v2d5) * size) : \ - memset(domlist.v2, 0, sizeof(*domlist.v2) * size)))) + memset(domlist.v2, 0, sizeof(*domlist.v2) * size))))) #define XEN_GETDOMAININFOLIST_DOMAIN(domlist, n) \ (hypervisor_version < 2 ? \ domlist.v0[n].domain : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + domlist.v2d7[n].domain : \ + (dom_interface_version == 6 ? \ domlist.v2d6[n].domain : \ (dom_interface_version == 5 ? \ domlist.v2d5[n].domain : \ - domlist.v2[n].domain))) + domlist.v2[n].domain)))) #define XEN_GETDOMAININFOLIST_UUID(domlist, n) \ (hypervisor_version < 2 ? \ domlist.v0[n].handle : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + domlist.v2d7[n].handle : \ + (dom_interface_version == 6 ? \ domlist.v2d6[n].handle : \ (dom_interface_version == 5 ? \ domlist.v2d5[n].handle : \ - domlist.v2[n].handle))) + domlist.v2[n].handle)))) #define XEN_GETDOMAININFOLIST_DATA(domlist) \ (hypervisor_version < 2 ? \ (void*)(domlist->v0) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + (void*)(domlist->v2d7) : \ + (dom_interface_version == 6 ? \ (void*)(domlist->v2d6) : \ (dom_interface_version == 5 ? \ (void*)(domlist->v2d5) : \ - (void*)(domlist->v2)))) + (void*)(domlist->v2))))) #define XEN_GETDOMAININFO_SIZE \ (hypervisor_version < 2 ? \ sizeof(xen_v0_getdomaininfo) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + sizeof(xen_v2d7_getdomaininfo) : \ + (dom_interface_version == 6 ? \ sizeof(xen_v2d6_getdomaininfo) : \ (dom_interface_version == 5 ? \ sizeof(xen_v2d5_getdomaininfo) : \ - sizeof(xen_v2_getdomaininfo)))) + sizeof(xen_v2_getdomaininfo))))) #define XEN_GETDOMAININFO_CLEAR(dominfo) \ (hypervisor_version < 2 ? \ memset(&(dominfo.v0), 0, sizeof(xen_v0_getdomaininfo)) : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + memset(&(dominfo.v2d7), 0, sizeof(xen_v2d7_getdomaininfo)) : \ + (dom_interface_version == 6 ? \ memset(&(dominfo.v2d6), 0, sizeof(xen_v2d6_getdomaininfo)) : \ (dom_interface_version == 5 ? \ memset(&(dominfo.v2d5), 0, sizeof(xen_v2d5_getdomaininfo)) : \ - memset(&(dominfo.v2), 0, sizeof(xen_v2_getdomaininfo))))) + memset(&(dominfo.v2), 0, sizeof(xen_v2_getdomaininfo)))))) #define XEN_GETDOMAININFO_DOMAIN(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.domain : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.domain : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.domain : \ (dom_interface_version == 5 ? \ dominfo.v2d5.domain : \ - dominfo.v2.domain))) + dominfo.v2.domain)))) #define XEN_GETDOMAININFO_CPUTIME(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.cpu_time : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.cpu_time : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.cpu_time : \ (dom_interface_version == 5 ? \ dominfo.v2d5.cpu_time : \ - dominfo.v2.cpu_time))) + dominfo.v2.cpu_time)))) #define XEN_GETDOMAININFO_CPUCOUNT(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.nr_online_vcpus : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.nr_online_vcpus : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.nr_online_vcpus : \ (dom_interface_version == 5 ? \ dominfo.v2d5.nr_online_vcpus : \ - dominfo.v2.nr_online_vcpus))) + dominfo.v2.nr_online_vcpus)))) #define XEN_GETDOMAININFO_MAXCPUID(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.max_vcpu_id : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.max_vcpu_id : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.max_vcpu_id : \ (dom_interface_version == 5 ? \ dominfo.v2d5.max_vcpu_id : \ - dominfo.v2.max_vcpu_id))) + dominfo.v2.max_vcpu_id)))) #define XEN_GETDOMAININFO_FLAGS(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.flags : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.flags : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.flags : \ (dom_interface_version == 5 ? \ dominfo.v2d5.flags : \ - dominfo.v2.flags))) + dominfo.v2.flags)))) #define XEN_GETDOMAININFO_TOT_PAGES(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.tot_pages : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.tot_pages : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.tot_pages : \ (dom_interface_version == 5 ? \ dominfo.v2d5.tot_pages : \ - dominfo.v2.tot_pages))) + dominfo.v2.tot_pages)))) #define XEN_GETDOMAININFO_MAX_PAGES(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.max_pages : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.max_pages : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.max_pages : \ (dom_interface_version == 5 ? \ dominfo.v2d5.max_pages : \ - dominfo.v2.max_pages))) + dominfo.v2.max_pages)))) #define XEN_GETDOMAININFO_UUID(dominfo) \ (hypervisor_version < 2 ? \ dominfo.v0.handle : \ - (dom_interface_version >= 6 ? \ + (dom_interface_version >= 7 ? \ + dominfo.v2d7.handle : \ + (dom_interface_version == 6 ? \ dominfo.v2d6.handle : \ (dom_interface_version == 5 ? \ dominfo.v2d5.handle : \ - dominfo.v2.handle))) + dominfo.v2.handle)))) static int @@ -2091,6 +2141,17 @@ xenHypervisorInit(void) goto done; } + /* Xen 4.1 + * sysctl version 8 -> xen-unstable c/s 21118:28e5409e3fb3 + * domctl version 7 -> xen-unstable c/s 21212:de94884a669c + */ + sys_interface_version = 8; /* XEN_SYSCTL_INTERFACE_VERSION */ + if (virXen_getdomaininfo(fd, 0, &info) == 1) { + dom_interface_version = 7; /* XEN_DOMCTL_INTERFACE_VERSION */ + VIR_DEBUG0("Using hypervisor call v2, sys ver8 dom ver7\n"); + goto done; + } + hypervisor_version = 1; sys_interface_version = -1; if (virXen_getdomaininfo(fd, 0, &info) == 1) { -- 1.7.3.1

2 2

[libvirt] [PATCH] Fix delayed event delivery when SASL is active
by Daniel P. Berrange 17 Mar '11

17 Mar '11

When SASL is active, it was possible that we read and decoded more data off the wire than we initially wanted. The loop processing this data terminated after only one message to avoid delaying the calling thread, but this could delay event delivery. As long as there is decoded SASL data in memory, we must process it, before returning to the poll() event loop. This is a counterpart to the same kind of issue solved in commit 68d2c3482fa16801f8e6ca5c42698319bb87f385 in a different area of the code * src/remote/remote_driver.c: Process all pending SASL data --- src/remote/remote_driver.c | 25 ++++++++++++++++++++----- 1 files changed, 20 insertions(+), 5 deletions(-) diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index 064fc8b..4245fee 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -10426,12 +10426,27 @@ remoteIOHandleInput(virConnectPtr conn, struct private_data *priv, ret = processCallDispatch(conn, priv, flags); priv->bufferOffset = priv->bufferLength = 0; /* - * We've completed one call, so return even - * though there might still be more data on - * the wire. We need to actually let the caller - * deal with this arrived message to keep good - * response, and also to correctly handle EOF. + * We've completed one call, but we don't want to + * spin around the loop forever if there are many + * incoming async events, or replies for other + * thread's RPC calls. We want to get out & let + * any other thread take over as soon as we've + * got our reply. When SASL is active though, we + * may have read more data off the wire than we + * initially wanted & cached it in memory. In this + * case, poll() would not detect that there is more + * ready todo. + * + * So if SASL is active *and* some SASL data is + * already cached, then we'll process that now, + * before returning. */ +#if HAVE_SASL + if (ret == 0 && + priv->saslconn && + priv->saslDecoded) + continue; +#endif return ret; } } -- 1.7.4

2 1

[libvirt] [PATCH] remote: Add missing virCondDestroy calls
by Matthias Bolte 17 Mar '11

17 Mar '11

The virCond of the remote_thread_call struct was leaked in some places. This results in leaking the underlying mutex. Which in turn leaks a handle on Windows. Reported by Aliaksandr Chabatar and Ihar Smertsin. --- See https://www.redhat.com/archives/libvir-list/2011-March/msg00789.html for the report. src/remote/remote_driver.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index 82e094b..8aa8801 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -8609,6 +8609,7 @@ remoteStreamPacket(virStreamPtr st, xdr_destroy (&xdr); ret = remoteIO(st->conn, priv, 0, thiscall); + ignore_value(virCondDestroy(&thiscall->cond)); VIR_FREE(thiscall); if (ret < 0) return -1; @@ -8617,6 +8618,7 @@ remoteStreamPacket(virStreamPtr st, error: xdr_destroy (&xdr); + ignore_value(virCondDestroy(&thiscall->cond)); VIR_FREE(thiscall); return -1; } @@ -8746,6 +8748,7 @@ remoteStreamRecv(virStreamPtr st, } ret = remoteIO(st->conn, priv, 0, thiscall); + ignore_value(virCondDestroy(&thiscall->cond)); VIR_FREE(thiscall); if (ret < 0) goto cleanup; @@ -9719,6 +9722,7 @@ prepareCall(struct private_data *priv, error: xdr_destroy (&xdr); + ignore_value(virCondDestroy(&rv->cond)); VIR_FREE(rv); return NULL; } @@ -10818,6 +10822,7 @@ call (virConnectPtr conn, struct private_data *priv, } rv = remoteIO(conn, priv, flags, thiscall); + ignore_value(virCondDestroy(&thiscall->cond)); VIR_FREE(thiscall); return rv; } -- 1.7.0.4

3 3

[libvirt] FW: Using dlls for Windows provided in http://libvirt.org/sources/win32_experimental/Libvirt-0.8.7-2.exe
by Aliaksandr Chabatar 17 Mar '11

17 Mar '11

Hello libvirt-list, We have a problem. It concerns CPU usage of libvirt library in Windows. It's not a problem in Linux. See attach. At the moment we have a workaround for item 1 - we just calculate the number of handles which are leaked and restart our service if the number exceeds 10.000 As for item 2 - we have no real workaround. In 99.99% it should not happen, but there is still 0.01% In libvirt.log you can find more info as suggested by Daniel. ================================================================ In the attached file, you will find detailed information regarding the case 100 percent CPU usage. Our test was performed on the following system: Windows XP SP3; Libvirt-0.8.8; Run the following command: virsh -c qemu+tcp: / /172.17.46.88:135/system Port 135 was one of the ports on which our service is trying to connect. ================================================================ Could you help us here? Thanks Alexander -----Original Message----- From: Aliaksandr Chabatar Sent: Tuesday, March 15, 2011 3:52 PM To: Ihar Smertsin Subject: FW: [libvirt] Using dlls for Windows provided in http://libvirt.org/sources/win32_experimental/Libvirt-0.8.7-2.exe Hi Ihar, Could you provide more information (log files, see below) so we could address this issue to libvir-list(a)redhat.com ? Mfg Alexander -----Original Message----- From: Daniel P. Berrange [mailto:berrange@redhat.com] Sent: Tuesday, March 15, 2011 3:08 PM To: Aliaksandr Chabatar Cc: Hempfer, Siegfried; Boehme, Alfred; Schnizer, Monika Subject: Re: [libvirt] Using dlls for Windows provided in http://libvirt.org/sources/win32_experimental/Libvirt-0.8.7-2.exe On Tue, Mar 15, 2011 at 04:01:31PM +0200, Aliaksandr Chabatar wrote: > Dear Daniel, > > I have another question. It concerns CPU usage of libvirt library > in Windows. It's not a problem in Linux. See attach. > > At the moment we have a workaround for item 1 - we just calculate > the number of handles which are leaked and restart our service if > the number exceeds 10.000 It sounds like we have some crazy resource leak in a piece of code. I'm not too familiar with Windows, but if you re-send this mail of yours to libvir-list(a)redhat.com, I expect one of the community members who knows Windows will be able to advise. > As for item 2 - we have no real workaround. In 99.99% it should > not happen, but there is still 0.01% Yeah, that sounds like some piece of code is missing correct error checking. It would be useful to try and obtain a couple of stack traces when it is showing 100% cpu usage. Or capture a libvirt debug log, eg by setting an environment variable in your client application LIBVIRT_LOG_FILTERS="1:libvirt 1:util 1:remote" LIBVIRT_LOG_OUTPUTS="1:file:libvirt.log" Again, sending this log + the info from your mail to the libvir-list would be best. Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|

2 2

[libvirt] [PATCH] docs: Document first release with spice and qxl
by Cole Robinson 17 Mar '11

17 Mar '11

--- docs/formatdomain.html.in | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 0c7c965..0835472 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1566,6 +1566,7 @@ qemu-kvm -net nic,model=? /dev/null to use. It is possible to set a limit on the validity of the password be giving an timestamp <code>passwdValidTo='2010-04-09T15:51:00'</code> assumed to be in UTC. NB, this may not be supported by all hypervisors. + "spice" since 0.8.6. When SPICE has both a normal and TLS secured TCP port configured, it @@ -1632,8 +1633,9 @@ qemu-kvm -net nic,model=? /dev/null <dt><code>model</code></dt> <dd> The <code>model</code> element has a mandatory <code>type</code> - attribute which takes the value "vga", "cirrus", "vmvga", "qxl", - "xen" or "vbox", depending on the hypervisor features available. + attribute which takes the value "vga", "cirrus", "vmvga", "xen", + "vbox", or "qxl" (since 0.8.6) + depending on the hypervisor features available. You can also provide the amount of video memory in kilobytes using <code>vram</code> and the number of screen with <code>heads</code>. </dd> -- 1.7.4

2 1

[libvirt] [PATCH] qemu: Fix copy&paste error messages in text monitor
by Jiri Denemark 17 Mar '11

17 Mar '11

--- src/qemu/qemu_monitor_text.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 5df7933..91ecc8b 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -2055,7 +2055,7 @@ try_command: if (qemuMonitorHMPCommand(mon, cmd, &reply) < 0) { qemuReportError(VIR_ERR_OPERATION_FAILED, - _("failed to close fd in qemu with '%s'"), cmd); + _("failed to attach drive '%s'"), drivestr); goto cleanup; } @@ -2324,7 +2324,7 @@ int qemuMonitorTextAddDrive(qemuMonitorPtr mon, if (qemuMonitorHMPCommand(mon, cmd, &reply) < 0) { qemuReportError(VIR_ERR_OPERATION_FAILED, - _("failed to close fd in qemu with '%s'"), cmd); + _("failed to add drive '%s'"), drivestr); goto cleanup; } @@ -2425,8 +2425,8 @@ int qemuMonitorTextSetDrivePassphrase(qemuMonitorPtr mon, } if (qemuMonitorHMPCommand(mon, cmd, &reply) < 0) { - qemuReportError(VIR_ERR_OPERATION_FAILED, - _("failed to close fd in qemu with '%s'"), cmd); + qemuReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("failed to set disk password")); goto cleanup; } -- 1.7.4.1

2 1

[libvirt] [PATCH] do not return -1 when human-monitor-command is not supported
by Wen Congyang 17 Mar '11

17 Mar '11

It is not a fatal error if some monitor command is not supported. So we should not return -1 when human-monitor-command is not supported. Set reply_str to "unknown command: 'human-monitor-command'", and the caller can deal this error the same as the command that we want to run by human-monitor-command. Signed-off-by: Wen Congyang <wency(a)cn.fujitsu.com> --- src/qemu/qemu_monitor_json.c | 10 ++++++++++ 1 files changed, 10 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 13d12c8..7ec4ce7 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -694,6 +694,16 @@ qemuMonitorJSONHumanCommandWithFd(qemuMonitorPtr mon, if (!cmd || qemuMonitorJSONCommandWithFd(mon, cmd, scm_fd, &reply) < 0) goto cleanup; + if (qemuMonitorJSONHasError(reply, "CommandNotFound")) { + *reply_str = strdup("unknown command: 'human-monitor-command'"); + if (!*reply_str) { + virReportOOMError(); + goto cleanup; + } + ret = 0; + goto cleanup; + } + if (qemuMonitorJSONCheckError(cmd, reply)) goto cleanup; -- 1.7.1

2 1

[libvirt] [PATCH] spec: Guess rhel/fedora macros from dist
by Jiri Denemark 17 Mar '11

17 Mar '11

--- If anyone knows how to make rpm to evaluate %(...) immediately and not at every appearance of rhel/fedora, that would be really cool. libvirt.spec.in | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/libvirt.spec.in b/libvirt.spec.in index 02be928..bfc9206 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -1,5 +1,11 @@ # -*- rpm-spec -*- +# if neither fedora nor rhel was defined, try to guess them from %{dist} +%if !0%{?rhel} && !0%{?fedora} +%define rhel %(echo "%{?dist}" | sed -ne 's/^\\.el\$[0-9]\\+\$.*/\\1/p') +%define fedora %(echo "%{?dist}" | sed -ne 's/^\\.fc\\?\$[0-9]\\+\$.*/\\1/p') +%endif + # A client only build will create a libvirt.so only containing # the generic RPC driver, and test driver and no libvirtd # Default to a full server + client build -- 1.7.4.1

3 10

[libvirt] /sys/devices/system/cpu/cpu0/topology/physical_package_id: No such file or directory (was: Re: Need help with virt_install)
by Richard W.M. Jones 17 Mar '11

17 Mar '11

On Wed, Mar 16, 2011 at 12:12:24PM -0700, JD wrote: > /virt-install \ > --connect qemu:///system \ > --name Fedora_14_i386 \ > --ram 500 \ > --disk path=/dev/sdc \ > --network network:default \ > --arch i386 \ > --os-variant fedora14 \ > --vnc \ > --location > http://mirrors.kernel.org/fedora/releases/14/Fedora/i386/os/ > > Starting install.../ / > Retrieving file .treeinfo... | 2.4 kB > 00:00 ... > Retrieving file vmlinuz... | 7.1 MB > 00:05 ... > Retrieving file initrd.img... | 60 MB > 00:45 ... > ERROR internal error Process exited while reading console log output: > libvir: error : cannot open > /sys/devices/system/cpu/cpu0/topology/physical_package_id: No such file > or directory > > Domain installation does not appear to have been > successful. If it was, you can restart your domain > by running 'virsh start Fedora_14_i386'; otherwise, please > restart your installation./ > > > is the missing /sys/devices/system/cpu/cpu0/topology/physical_package_id > caused by kernel configuration settings? > > > The relevant installed packages are: > > libguestfs-tools-1.6.2-1.fc13.4.i686 > python-virtinst-0.500.4-1.fc13.noarch > python-virtinst-0.500.4-1.fc13.noarch > virt-ctrl-1.0.1-4.fc12.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > virt-mem-0.3.1-9.fc12.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > virt-mem-0.3.1-9.fc12.i686 > python-virtinst-0.500.4-1.fc13.noarch > libguestfs-tools-1.6.2-1.fc13.4.i686 > python-virtinst-0.500.4-1.fc13.noarch > libguestfs-tools-1.6.2-1.fc13.4.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > virt-manager-0.8.5-1.fc13.noarch > virt-mem-0.3.1-9.fc12.i686 > libvirt-client-0.8.2-2.fc13.i686 > virt-mem-0.3.1-9.fc12.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > virt-top-1.0.4-3.fc13.i686 > virt-mem-0.3.1-9.fc12.i686 > virt-viewer-0.2.1-1.fc13.i686 > libguestfs-tools-1.6.2-1.fc13.4.i686 > libvirt-client-0.8.2-2.fc13.i686 (CCing libvir-list) I would say that this is a bug in libvirt. Since you are using quite an old version, can you try upgrading to the new libvirt in Fedora 14/15 and see if that fixes it? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top

1 0

Re: [libvirt] [Qemu-devel] KVM call minutes for Mar 15
by Chris Wright 17 Mar '11

17 Mar '11

* Anthony Liguori (anthony(a)codemonkey.ws) wrote: > On 03/15/2011 09:53 AM, Chris Wright wrote: > > QAPI <snip> > >- c library implementation is critical to have unit tests and test > > driven development > > - thread safe? > > - no shared state, no statics. > > - threading model requires lock for the qmp session > > - licensiing? > > - LGPL > > - forwards/backwards compat? > > - designed with that in mind see wiki: > > > > http://wiki.qemu.org/Features/QAPI > > One neat feature of libqmp is that once libvirt has a better QMP > passthrough interface, we can create a QmpSession that uses libvirt. > > It would look something like: > > QmpSession *libqmp_session_new_libvirt(virDomainPtr dom); Looks like you mean this? -> request QmpSession -> client libvirt <- return QmpSession <- client -> QmpSession -> QMP -> QEMU So bypassing libvirt completely to actually use the session? Currently, it's more like: client -> QemuMonitorCommand -> libvirt -> QMP -> QEMU > The QmpSession returned by this call can then be used with all of > the libqmp interfaces. This means we can still exercise our test > suite with a guest launched through libvirt. It also should make > the libvirt pass through interface a bit easier to consume by third > parties. This sounds like it's something libvirt folks should be involved with. At the very least, this mode is there now and considered basically unstable/experimental/developer use: "Qemu monitor command '%s' executed; libvirt results may be unpredictable!" So likely some concern about making it easier to use, esp. assuming that third parties above are mgmt apps, not just developers. thanks, -chris

3 2

[libvirt] Network Interface: How to disable virbr0 correctly?
by Jake Xu 17 Mar '11

17 Mar '11

Hi all, Recently, I have been using libvirt to create virtual machines on ESX servers. It has been very well until to the point where I couldn't find any way to disable/remove the virbr0 interface properly. We use static configuration for VMs on ESX so we do not need to use virbr0 interface and we do not want libvirt to configure iptables as a result of the newly created virbr0 interface. I have tried to remove/rename the /etc/libvirt/qemu/networks/autostart.xml file, and that prevents virbr0 being installed. However, that also seems to prevent the second interface eth1 being configured properly. After libvirt defines a VM, eth0 and eht1 are both installed, but only eth0 is configured properly. The eth1 does not appear in the /etc/sysconfig/network-script/ifcfg-xxx file. It seems to me that the 'autostart.xml' file interrupts the configuration of the eth1 interface. Can anyone please advise how I can do this properly? Thanks a lot, Jake

3 6

[libvirt] attach/detach device in json mode
by Wen Congyang 17 Mar '11

17 Mar '11

Hi, all The qemu does not support monitor_command drive_add and drive_del in json mode. These two commands are important because attaching/detaching commands need them. AFAIK, if we use libvirt provided by RHEL6(not upstream), we connect to qemu in json mode. So, it is important to make qemu to support these monitor commands in json mode. Markus Armbruster(armbru(a)redhat.com) said these commands are designed badly, and he has started to work on saner commands some time ago: http://lists.nongnu.org/archive/html/qemu-devel/2011-02/msg02042.html He suggested us to use human monitor passthrough in the meantime. Should we modify libvirt to use human monitor passthrough command to support attaching/detaching device before the saner commands are ready?

5 11

[libvirt] [PATCH] macvtap: log an error if on failure to connect to netlink socket
by Laine Stump 16 Mar '11

16 Mar '11

A bug in libnl (see https://bugzilla.redhat.com/show_bug.cgi?id=677724 and https://bugzilla.redhat.com/show_bug.cgi?id=677725) makes it very easy to create a failure to connect to the netlink socket when trying to open a macvtap network device ("type='direct'" in domain interface XML). When that error occurred, there was no log message, leading virsh (for example) to report "unknown error". This patch logs a message which will hopefully be more useful. (All other error exits from the same function were already logging a message). --- src/util/macvtap.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/src/util/macvtap.c b/src/util/macvtap.c index a71db86..66bdc73 100644 --- a/src/util/macvtap.c +++ b/src/util/macvtap.c @@ -127,6 +127,8 @@ int nlComm(struct nl_msg *nl_msg, return -1; if (nl_connect(nlhandle, NETLINK_ROUTE) < 0) { + virReportSystemError(errno, + "%s", _("cannot connect to netlink socket")); rc = -1; goto err_exit; } -- 1.7.3.4

3 5

[libvirt] [PATCH] storage: Fix a problem which will cause libvirtd crashed
by Osier Yang 16 Mar '11

16 Mar '11

--- src/conf/storage_conf.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c index 6e3fe0e..13a3622 100644 --- a/src/conf/storage_conf.c +++ b/src/conf/storage_conf.c @@ -396,7 +396,7 @@ virStoragePoolDefParseSource(xmlXPathContextPtr ctxt, char *authType = NULL; int nsource, i; virStoragePoolOptionsPtr options; - char *port; + char *port = NULL; relnode = ctxt->node; ctxt->node = node; -- 1.7.4

2 2

[libvirt] [PATCH 0/5] libvirt/qemu - allow modification of devices in inactive domain
by KAMEZAWA Hiroyuki 16 Mar '11

16 Mar '11

Now, Xen supports at(de)tach-disk, at(de)tach-interface for inactive domains with --perisitent option. (Using --persisten against live domain will cause error.) Now, virsh attach-xxx/detach-xxx cannot modify inactive domains in qemu and --persistent option just returns error. This series adds support for modification of inactive domain's disk, interfaces. Short description 1 ... clean up for Xen's driver 2 ... add my name to AUTHORS 3 ... add --persistent modification support, just a wrapper function. 4 ... add disk support 5 ... add nics support Thanks, -Kame

4 28

[libvirt] Possibly addition to python api binding virDomain class
by Russell Ballestrini 16 Mar '11

16 Mar '11

*In the python bindings of libvirt, could we please add the following code to the virDomain class? * def state(self): """return the state of the dom object""" return self.info()[0] *This helper function would make the world of difference when coding against the libvirt api. * -- *Russell J Ballestrini*

1 0