- Devel - libvirt List Archives

[libvirt] [PATCH] virsh: avoid bogus description
by Eric Blake 21 Jun '11

21 Jun '11

https://bugzilla.redhat.com/show_bug.cgi?id=682121 Gettext reserves the empty string for internal use, and it must not be passed through _(). We were violating this for commands that (for whatever reason) used "" for their description. * tools/virsh.c (vshCmddefHelp): Don't translate empty string. Reported by Tatsuo Kawasaki. --- tools/virsh.c | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index a315f05..bdd5005 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -12113,7 +12113,8 @@ vshCmddefHelp(vshControl *ctl, const char *cmdname) vshError(ctl, _("command '%s' doesn't exist"), cmdname); return false; } else { - const char *desc = _(vshCmddefGetInfo(def, "desc")); + /* Don't translate desc until after we know it isn't "". */ + const char *desc = vshCmddefGetInfo(def, "desc"); const char *help = _(vshCmddefGetInfo(def, "help")); char buf[256]; uint32_t opts_need_arg; @@ -12167,7 +12168,7 @@ vshCmddefHelp(vshControl *ctl, const char *cmdname) if (desc[0]) { /* Print the description only if it's not empty. */ fputs(_("\n DESCRIPTION\n"), stdout); - fprintf(stdout, " %s\n", desc); + fprintf(stdout, " %s\n", _(desc)); } if (def->opts) { -- 1.7.4.4

3 3

[libvirt] [libvirt-php] Announcement: Release 0.4.2 available
by Michal Novotny 21 Jun '11

21 Jun '11

Hello everybody. it's been a rather long time since the first release of libvirt-php bindings however there are many new features so I've decided to release a new version of libvirt-php, release 0.4.2. New features: - Storage pool / volume related functions (by Lyre) - VNC client support to get screenshot using gvnccapture (if present) - experimental - sending keys and client pointer events over VNC protocol - API to change domain VCPU count, memory amount and guest boot order - libvirt resources tracking - can be obtained by libvirt_print_binding_resources() API function - debug logging using libvirt_logfile_set() API function (default max log file size can be obtained in PHPInfo output) - functions to add/remove domain disks/NICs in an easy way (using a simple PHP API functions) For this release thanks goes to those guys: - Lyre <liyong(a)skybility.com> - storage-related functions as well as many bug fixes - Yukihiro Kawada <warp.kawada(a)gmail.com> - fix for the memory leak in libvirt_list_domains() - Daniel P. Berrange <berrange(a)redhat.com> - bug fixes/improvements for the build system - Radek Hladik <r.hladik(a)cybersales.cz> - patch to add API to update device - Tiziano Mueller <dev-zero(a)gentoo.org> - several bug fixes for threaded php and libvirt_connect/version failures without Xen I hope I didn't forget to mention anybody ;-) The new release can be found at: http://libvirt.org/sources/php/libvirt-php-0.4.2.tar.gz Michal -- Michal Novotny <minovotn(a)redhat.com>, RHCE, Red Hat Virtualization Team (xen userspace) | libvirt-php bindings

1 0

[libvirt] [PATCH 00/10] python: Implement missing stream/event ops
by Cole Robinson 21 Jun '11

21 Jun '11

Currently it isn't possible to implement the equivalent of 'virsh console' with the python bindings. This series implements the missing functionality required to do so. Cole Robinson (10): python: libvirt-override: use simpler debug python: generator: Don't print warning if nothing to warn about python: Implement bindings for virStreamEventAddCallback python: Implement virStreamSend/Recv python: Implement virStreamSend/RecvAll helpers Promote virEvent*Handle/Timeout to public API events: Correct virEventAddTimeout docs python: Add bindings for virEvent*Handle/Timeout python: events: Fix C->Python handle callback prototype python: Mark event callback wrappers as private daemon/libvirtd.c | 1 - daemon/mdns.c | 1 - examples/domain-events/events-python/event-test.py | 6 +- python/generator.py | 27 +- python/libvirt-override-virConnect.py | 102 ++--- python/libvirt-override-virStream.py | 123 +++++- python/libvirt-override.c | 492 ++++++++++++++----- python/libvirt-override.py | 80 +++- python/typewrappers.c | 14 + python/typewrappers.h | 2 +- src/conf/domain_event.c | 1 - src/conf/domain_event.h | 1 - src/fdstream.c | 1 - src/libvirt_private.syms | 9 - src/libvirt_public.syms | 6 + src/libxl/libxl_driver.c | 1 - src/lxc/lxc_driver.c | 1 - src/network/bridge_driver.c | 1 - src/node_device/node_device_hal.c | 1 - src/node_device/node_device_udev.c | 1 - src/openvz/openvz_driver.c | 1 - src/qemu/qemu_domain.c | 1 - src/qemu/qemu_driver.c | 1 - src/qemu/qemu_monitor.c | 1 - src/remote/remote_driver.c | 1 - src/test/test_driver.c | 1 - src/uml/uml_driver.c | 1 - src/util/event.c | 56 +++ src/util/event.h | 73 --- src/util/util.c | 1 - src/vbox/vbox_tmpl.c | 1 - src/xen/xen_inotify.c | 1 - src/xen/xs_internal.c | 1 - tools/console.c | 1 - 34 files changed, 692 insertions(+), 320 deletions(-) -- 1.7.4.4

3 31

Re: [libvirt] Network device abstraction aka virtual switch - V3
by Neil Wilson 21 Jun '11

21 Jun '11

> The current modes are: > > <forward layer='network' mode='route|nat'/> > > (in addition to not listing any mode, which equates to "isolated") > > Here are suggested new modes: Has anybody considered the migration requirements of networks in this new layout? If you move a machine attached to a 'route|nat' network to another Host you need to extend the network to the new Host and eventually you need to move the supporting processes that hand out the IP addresses/DNS addresses etc (if you're decommissioning a Host server for example). If you try to start a copy of the network on the new host, then it generally clashes with the pre-existing routing It would be useful to be able to extend a network onto another Host somehow and be able to migrate the supporting processes (DNSMasq, IP addressing, etc) between Hosts. (And obviously remove extensions as required as well). Perhaps this suggests that the address and route management systems need separating from the definition of the bridge network. (There's no real reason why the address management systems couldn't be attached to macvtap and vepa networks as well as basic bridges). Rgs Neil Wilson

2 1

[libvirt] [PATCH] Support reboots with the QEMU driver
by Daniel P. Berrange 21 Jun '11

21 Jun '11

For controlled shutdown we issue a 'system_powerdown' command to the QEMU monitor. This triggers an ACPI event which (most) guest OS wire up to a controlled shutdown. There is no equiv ACPI event to trigger a controlled reboot. This patch attempts to fake a reboot. - In qemuDomainObjPrivatePtr we have a bool fakeReboot flag. - The virDomainReboot method sets this flag and then triggers a normal 'system_powerdown'. - The QEMU process is started with '-no-shutdown' so that the guest CPUs pause when it powers off the guest - When we receive the 'POWEROFF' event from QEMU JSON monitor if fakeReboot is not set we invoke the qemuProcessKill command and shutdown continues normally - If fakeReboot was set, we spawn a background thread which issues 'system_reset' to perform a warm reboot of the guest hardware. Then it issues 'cont' to start the CPUs again * src/qemu/qemu_command.c: Add -no-shutdown flag if we have JSON support * src/qemu/qemu_domain.h: Add 'fakeReboot' flag to qemuDomainObjPrivate struct * src/qemu/qemu_driver.c: Fake reboot using the system_powerdown command if JSON support is available * src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h, src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_json.h, src/qemu/qemu_monitor_text.c, src/qemu/qemu_monitor_text.h: Add binding for system_reset command * src/qemu/qemu_process.c: Reset the guest & start CPUs if fakeReboot is set --- src/qemu/qemu_command.c | 7 +++ src/qemu/qemu_domain.h | 2 + src/qemu/qemu_driver.c | 65 ++++++++++++++++++++++++++++- src/qemu/qemu_monitor.c | 19 ++++++++ src/qemu/qemu_monitor.h | 1 + src/qemu/qemu_monitor_json.c | 19 ++++++++ src/qemu/qemu_monitor_json.h | 1 + src/qemu/qemu_monitor_text.c | 13 ++++++ src/qemu/qemu_monitor_text.h | 1 + src/qemu/qemu_process.c | 95 +++++++++++++++++++++++++++++++++++++++++- 10 files changed, 220 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 6346243..9ae1409 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3207,6 +3207,13 @@ qemuBuildCommandLine(virConnectPtr conn, def->onReboot != VIR_DOMAIN_LIFECYCLE_RESTART) virCommandAddArg(cmd, "-no-reboot"); + /* If JSON monitor is enabled, we can receive an event + * when QEMU stops. If we use no-shutdown, then we can + * watch for this event and do a soft/warm reboot. + */ + if (monitor_json) + virCommandAddArg(cmd, "-no-shutdown"); + if (!(def->features & (1 << VIR_DOMAIN_FEATURE_ACPI))) virCommandAddArg(cmd, "-no-acpi"); diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index bacf5b5..b8469b6 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -89,6 +89,8 @@ struct _qemuDomainObjPrivate { virBitmapPtr qemuCaps; char *lockState; + + bool fakeReboot; }; struct qemuDomainWatchdogEvent diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 853c84c..4c827cf 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1429,7 +1429,7 @@ cleanup: } -static int qemudDomainShutdown(virDomainPtr dom) { +static int qemuDomainShutdown(virDomainPtr dom) { struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; int ret = -1; @@ -1461,6 +1461,8 @@ static int qemudDomainShutdown(virDomainPtr dom) { ret = qemuMonitorSystemPowerdown(priv->mon); qemuDomainObjExitMonitor(vm); + priv->fakeReboot = false; + endjob: if (qemuDomainObjEndJob(vm) == 0) vm = NULL; @@ -1472,11 +1474,66 @@ cleanup: } +static int qemuDomainReboot(virDomainPtr dom, unsigned int flags ATTRIBUTE_UNUSED) { + struct qemud_driver *driver = dom->conn->privateData; + virDomainObjPtr vm; + int ret = -1; + qemuDomainObjPrivatePtr priv; + + qemuDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + qemuDriverUnlock(driver); + + if (!vm) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + virUUIDFormat(dom->uuid, uuidstr); + qemuReportError(VIR_ERR_NO_DOMAIN, + _("no domain with matching uuid '%s'"), uuidstr); + goto cleanup; + } + priv = vm->privateData; + +#if HAVE_YAJL + if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MONITOR_JSON)) { + if (qemuDomainObjBeginJob(vm) < 0) + goto cleanup; + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is not running")); + goto endjob; + } + + qemuDomainObjEnterMonitor(vm); + ret = qemuMonitorSystemPowerdown(priv->mon); + qemuDomainObjExitMonitor(vm); + + priv->fakeReboot = true; + + endjob: + if (qemuDomainObjEndJob(vm) == 0) + vm = NULL; + } else { +#endif + qemuReportError(VIR_ERR_NO_SUPPORT, "%s", + _("Reboot is not supported without the JSON monitor")); +#if HAVE_YAJL + } +#endif + +cleanup: + if (vm) + virDomainObjUnlock(vm); + return ret; +} + + static int qemudDomainDestroy(virDomainPtr dom) { struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; int ret = -1; virDomainEventPtr event = NULL; + qemuDomainObjPrivatePtr priv; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -1488,6 +1545,9 @@ static int qemudDomainDestroy(virDomainPtr dom) { goto cleanup; } + priv = vm->privateData; + priv->fakeReboot = false; + /* Although qemuProcessStop does this already, there may * be an outstanding job active. We want to make sure we * can kill the process even if a job is active. Killing @@ -8146,7 +8206,8 @@ static virDriver qemuDriver = { .domainLookupByName = qemudDomainLookupByName, /* 0.2.0 */ .domainSuspend = qemudDomainSuspend, /* 0.2.0 */ .domainResume = qemudDomainResume, /* 0.2.0 */ - .domainShutdown = qemudDomainShutdown, /* 0.2.0 */ + .domainShutdown = qemuDomainShutdown, /* 0.2.0 */ + .domainReboot = qemuDomainReboot, /* 0.9.3 */ .domainDestroy = qemudDomainDestroy, /* 0.2.0 */ .domainGetOSType = qemudDomainGetOSType, /* 0.2.2 */ .domainGetMaxMemory = qemudDomainGetMaxMemory, /* 0.4.2 */ diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 1428921..1dafc4b 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -1096,6 +1096,25 @@ int qemuMonitorSystemPowerdown(qemuMonitorPtr mon) } +int qemuMonitorSystemReset(qemuMonitorPtr mon) +{ + int ret; + VIR_DEBUG("mon=%p", mon); + + if (!mon) { + qemuReportError(VIR_ERR_INVALID_ARG, "%s", + _("monitor must not be NULL")); + return -1; + } + + if (mon->json) + ret = qemuMonitorJSONSystemReset(mon); + else + ret = qemuMonitorTextSystemReset(mon); + return ret; +} + + int qemuMonitorGetCPUInfo(qemuMonitorPtr mon, int **pids) { diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index 3bb0269..ae74a3c 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -194,6 +194,7 @@ int qemuMonitorStartCPUs(qemuMonitorPtr mon, int qemuMonitorStopCPUs(qemuMonitorPtr mon); int qemuMonitorGetStatus(qemuMonitorPtr mon, bool *running); +int qemuMonitorSystemReset(qemuMonitorPtr mon); int qemuMonitorSystemPowerdown(qemuMonitorPtr mon); int qemuMonitorGetCPUInfo(qemuMonitorPtr mon, diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 56ec65b..6cc6ea7 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -947,6 +947,25 @@ int qemuMonitorJSONSystemPowerdown(qemuMonitorPtr mon) } +int qemuMonitorJSONSystemReset(qemuMonitorPtr mon) +{ + int ret; + virJSONValuePtr cmd = qemuMonitorJSONMakeCommand("system_reset", NULL); + virJSONValuePtr reply = NULL; + if (!cmd) + return -1; + + ret = qemuMonitorJSONCommand(mon, cmd, &reply); + + if (ret == 0) + ret = qemuMonitorJSONCheckError(cmd, reply); + + virJSONValueFree(cmd); + virJSONValueFree(reply); + return ret; +} + + /* * [ { "CPU": 0, "current": true, "halted": false, "pc": 3227107138 }, * { "CPU": 1, "current": false, "halted": true, "pc": 7108165 } ] diff --git a/src/qemu/qemu_monitor_json.h b/src/qemu/qemu_monitor_json.h index 393d8fc..a72bf7c 100644 --- a/src/qemu/qemu_monitor_json.h +++ b/src/qemu/qemu_monitor_json.h @@ -49,6 +49,7 @@ int qemuMonitorJSONStopCPUs(qemuMonitorPtr mon); int qemuMonitorJSONGetStatus(qemuMonitorPtr mon, bool *running); int qemuMonitorJSONSystemPowerdown(qemuMonitorPtr mon); +int qemuMonitorJSONSystemReset(qemuMonitorPtr mon); int qemuMonitorJSONGetCPUInfo(qemuMonitorPtr mon, int **pids); diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index a16ea91..67333aa 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -417,6 +417,19 @@ int qemuMonitorTextSystemPowerdown(qemuMonitorPtr mon) { } +int qemuMonitorTextSystemReset(qemuMonitorPtr mon) { + char *info; + + if (qemuMonitorHMPCommand(mon, "system_reset", &info) < 0) { + qemuReportError(VIR_ERR_OPERATION_FAILED, + "%s", _("system reset operation failed")); + return -1; + } + VIR_FREE(info); + return 0; +} + + int qemuMonitorTextGetCPUInfo(qemuMonitorPtr mon, int **pids) { diff --git a/src/qemu/qemu_monitor_text.h b/src/qemu/qemu_monitor_text.h index 4fa5064..7536557 100644 --- a/src/qemu/qemu_monitor_text.h +++ b/src/qemu/qemu_monitor_text.h @@ -46,6 +46,7 @@ int qemuMonitorTextStopCPUs(qemuMonitorPtr mon); int qemuMonitorTextGetStatus(qemuMonitorPtr mon, bool *running); int qemuMonitorTextSystemPowerdown(qemuMonitorPtr mon); +int qemuMonitorTextSystemReset(qemuMonitorPtr mon); int qemuMonitorTextGetCPUInfo(qemuMonitorPtr mon, int **pids); diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 6838417..694d8a9 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -348,12 +348,99 @@ qemuProcessHandleReset(qemuMonitorPtr mon ATTRIBUTE_UNUSED, } +/* + * Since we have the '-no-shutdown' flag set, the + * QEMU process will currently have guest OS shutdown + * and the CPUS stopped. To fake the reboot, we thus + * want todo a reset of the virtual hardware, followed + * by restart of the CPUs. This should result in the + * guest OS booting up again + */ +static void +qemuProcessFakeReboot(void *opaque) +{ + struct qemud_driver *driver = qemu_driver; + virDomainObjPtr vm = opaque; + qemuDomainObjPrivatePtr priv = vm->privateData; + virDomainEventPtr event = NULL; + int ret = -1; + VIR_DEBUG("vm=%p", vm); + qemuDriverLock(driver); + virDomainObjLock(vm); + if (qemuDomainObjBeginJob(vm) < 0) + goto cleanup; + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("guest unexpectedly quit")); + goto endjob; + } + + qemuDomainObjEnterMonitorWithDriver(driver, vm); + if (qemuMonitorSystemReset(priv->mon) < 0) { + qemuDomainObjExitMonitorWithDriver(driver, vm); + goto endjob; + } + qemuDomainObjExitMonitorWithDriver(driver, vm); + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("guest unexpectedly quit")); + goto endjob; + } + + if (qemuProcessStartCPUs(driver, vm, NULL, + VIR_DOMAIN_RUNNING_BOOTED) < 0) { + if (virGetLastError() == NULL) + qemuReportError(VIR_ERR_INTERNAL_ERROR, + "%s", _("resume operation failed")); + goto endjob; + } + event = virDomainEventNewFromObj(vm, + VIR_DOMAIN_EVENT_RESUMED, + VIR_DOMAIN_EVENT_RESUMED_UNPAUSED); + + ret = 0; + +endjob: + if (qemuDomainObjEndJob(vm) == 0) + vm = NULL; + +cleanup: + if (vm) { + if (ret == -1) + qemuProcessKill(vm); + if (virDomainObjUnref(vm) > 0) + virDomainObjUnlock(vm); + } + if (event) + qemuDomainEventQueue(driver, event); + qemuDriverUnlock(driver); +} + + static int qemuProcessHandleShutdown(qemuMonitorPtr mon ATTRIBUTE_UNUSED, virDomainObjPtr vm) { + qemuDomainObjPrivatePtr priv = vm->privateData; + VIR_DEBUG("vm=%p", vm); + virDomainObjLock(vm); - ((qemuDomainObjPrivatePtr) vm->privateData)->gotShutdown = true; + priv->gotShutdown = true; + if (priv->fakeReboot) { + virDomainObjRef(vm); + virThread th; + if (virThreadCreate(&th, + false, + qemuProcessFakeReboot, + vm) < 0) { + VIR_ERROR("Failed to create reboot thread, killing domain"); + qemuProcessKill(vm); + } + } else { + qemuProcessKill(vm); + } virDomainObjUnlock(vm); return 0; @@ -1920,6 +2007,11 @@ qemuProcessPrepareMonitorChr(struct qemud_driver *driver, } +/* + * Precondition: Both driver and vm must be locked, + * and a job must be active. This method will call + * {Enter,Exit}MonitorWithDriver + */ int qemuProcessStartCPUs(struct qemud_driver *driver, virDomainObjPtr vm, virConnectPtr conn, virDomainRunningReason reason) @@ -2182,6 +2274,7 @@ int qemuProcessStart(virConnectPtr conn, goto cleanup; vm->def->id = driver->nextvmid++; + priv->fakeReboot = false; /* Run an early hook to set-up missing devices */ if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) { -- 1.7.4.4

2 4

[libvirt] FreeBSD portability question, --network question
by Jason Helfman 21 Jun '11

21 Jun '11

Hey Everyone, I've ported libvirt to FreeBSD, however it has the networking disabled for the build, however I was curious what would need to be changed for this to work. Currently, we do have virtualbox-ose in the FreeBSD portstree, and I am curious if the networking would need to be reworked to support libvirt with networking in FreeBSD in the same manner in which virtualbox creates interfaces and uses the networking stack? Basically, I am seeking to learn what is missing, and how to either find resources to get this implemented, or implement it myself. The latter of implementing it myself is unlikely, as I really don't have very much source control experience at a networking stack layer. Thanks very much, Jason -- Jason Helfman System Administrator experts-exchange.com http://www.experts-exchange.com/M_4830110.html E4AD 7CF1 1396 27F6 79DD 4342 5E92 AD66 8C8C FBA5

2 1

[libvirt] [PATCH v3]: set and restore MAC address of a NIC when using PASSTHROUGH mode
by Gerhard Stenzel 21 Jun '11

21 Jun '11

This is another rework of the patch from Dirk addressing all comments received so far. The following patch addresses the problem that when a PASSTHROUGH mode DIRECT NIC connection is made the MAC address of the NIC is not automatically set and reset to the configured VM MAC and back again. The attached patch fixes this problem by setting and resetting the MAC while remembering the previous setting while the VM is running. This also works if libvirtd is restarted while the VM is running. the patch passes make syntax-check Signed-off-by: Dirk Herrendoerfer <d.herrendoerfer at herrendoerfer.name> Signed-off-by: Gerhard Stenzel <gerhard.stenzel(a)de.ibm.com> --- Index: libvirt/src/qemu/qemu_command.c =================================================================== --- libvirt.orig/src/qemu/qemu_command.c +++ libvirt/src/qemu/qemu_command.c @@ -128,7 +128,7 @@ qemuPhysIfaceConnect(virDomainDefPtr def rc = openMacvtapTap(net->ifname, net->mac, net->data.direct.linkdev, net->data.direct.mode, vnet_hdr, def->uuid, &net->data.direct.virtPortProfile, &res_ifname, - vmop); + vmop, driver->stateDir); if (rc >= 0) { qemuAuditNetDevice(def, net, res_ifname, true); VIR_FREE(net->ifname); @@ -149,7 +149,9 @@ qemuPhysIfaceConnect(virDomainDefPtr def if (err) { VIR_FORCE_CLOSE(rc); delMacvtap(net->ifname, net->mac, net->data.direct.linkdev, - &net->data.direct.virtPortProfile); + net->data.direct.mode, + &net->data.direct.virtPortProfile, + driver->stateDir); VIR_FREE(net->ifname); } } Index: libvirt/src/qemu/qemu_process.c =================================================================== --- libvirt.orig/src/qemu/qemu_process.c +++ libvirt/src/qemu/qemu_process.c @@ -2876,7 +2876,8 @@ void qemuProcessStop(struct qemud_driver virDomainNetDefPtr net = def->nets[i]; if (net->type == VIR_DOMAIN_NET_TYPE_DIRECT) { delMacvtap(net->ifname, net->mac, net->data.direct.linkdev, - &net->data.direct.virtPortProfile); + net->data.direct.mode, + &net->data.direct.virtPortProfile, driver->stateDir); VIR_FREE(net->ifname); } } Index: libvirt/src/util/macvtap.c =================================================================== --- libvirt.orig/src/util/macvtap.c +++ libvirt/src/util/macvtap.c @@ -545,6 +545,106 @@ configMacvtapTap(int tapfd, int vnet_hdr return 0; } +/** + * replaceMacAdress: + * @macaddress: new MAC address for interface + * @linkdev: name of interface + * @stateDir: directory to store old MAC address + * + * Returns 0 on success, -1 in case of fatal error, error code otherwise. + * + */ +static int +replaceMacAdress(const unsigned char *macaddress, + const char *linkdev, + char *stateDir) +{ + unsigned char oldmac[6]; + int rc; + + rc = ifaceGetMacaddr(linkdev, oldmac); + + if (rc) { + virReportSystemError(rc, + _("Getting MAC address from '%s' " + "to '%02x:%02x:%02x:%02x:%02x:%02x' failed."), + linkdev, + oldmac[0], oldmac[1], oldmac[2], + oldmac[3], oldmac[4], oldmac[5]); + } else { + char *path = NULL; + + char macstr[VIR_MAC_STRING_BUFLEN]; + if (virAsprintf(&path, "%s/%s", + stateDir, + linkdev) < 0) { + virReportOOMError(); + return errno; + } + virFormatMacAddr(oldmac, macstr); + if (virFileWriteStr(path, macstr, O_CREAT|O_TRUNC|O_WRONLY) < 0) { + virReportSystemError(errno, _("Unable to preserve mac for %s"), + linkdev); + return errno; + } + } + + rc = ifaceSetMacaddr(linkdev, macaddress); + if (rc) { + virReportSystemError(errno, + _("Setting MAC address on '%s' to " + "'%02x:%02x:%02x:%02x:%02x:%02x' failed."), + linkdev, + macaddress[0], macaddress[1], macaddress[2], + macaddress[3], macaddress[4], macaddress[5]); + } + return rc; +} + +/** + * restoreMacAddress: + * @linkdev: name of interface + * @stateDir: directory containing old MAC address + * + * Returns 0 on success, -1 in case of fatal error, error code otherwise. + * + */ +static int +restoreMacAddress(const char *linkdev, + char *stateDir) +{ + int ret; + char *oldmacname = NULL; + char *macstr = NULL; + char *path = NULL; + unsigned char oldmac[6]; + + if (virAsprintf(&path, "%s/%s", + stateDir, + linkdev) < 0) { + virReportOOMError(); + return -1; + } + + if (virFileReadAll(path, VIR_MAC_STRING_BUFLEN, &macstr) < 0) { + virReportSystemError(errno, _("Unable to preserve mac for %s"), + linkdev); + return errno; + } + + if (virParseMacAddr(macstr, &oldmac[0]) != 0) { + virReportSystemError(VIR_ERR_INTERNAL_ERROR, + _("Cannot parse MAC address from '%s'"), + oldmacname); + return -1; + } + + /*reset mac and remove file-ignore results*/ + ret = ifaceSetMacaddr(linkdev, oldmac); + ret = unlink(path); + VIR_FREE(macstr); + return 0; +} /** * openMacvtapTap: @@ -575,7 +675,8 @@ openMacvtapTap(const char *tgifname, const unsigned char *vmuuid, virVirtualPortProfileParamsPtr virtPortProfile, char **res_ifname, - enum virVMOperationType vmOp) + enum virVMOperationType vmOp, + char *stateDir) { const char *type = "macvtap"; int c, rc; @@ -589,6 +690,21 @@ openMacvtapTap(const char *tgifname, VIR_DEBUG("%s: VM OPERATION: %s", __FUNCTION__, virVMOperationTypeToString(vmOp)); + /** Note: When using PASSTHROUGH mode with MACVTAP devices the link + * device's MAC address must be set to the VMs MAC address. In + * order to not confuse the first switch or bridge in line this MAC + * address must be reset when the VM is shut down. + * This is especially important when using SRIOV capable cards that + * emulate their switch in firmware. + */ + if (mode == VIR_DOMAIN_NETDEV_MACVTAP_MODE_PASSTHRU) { + if (replaceMacAdress(macaddress, linkdev, stateDir) != 0) { + virReportSystemError(errno, + _("cannot replace MAC address on $s"), + linkdev); + } + } + if (tgifname) { if(ifaceGetIndex(false, tgifname, &ifindex) == 0) { if (STRPREFIX(tgifname, @@ -684,8 +800,18 @@ void delMacvtap(const char *ifname, const unsigned char *macaddr, const char *linkdev, - virVirtualPortProfileParamsPtr virtPortProfile) + int mode, + virVirtualPortProfileParamsPtr virtPortProfile, + char *stateDir) { + if (mode == VIR_DOMAIN_NETDEV_MACVTAP_MODE_PASSTHRU) { + if (restoreMacAddress(linkdev, stateDir) != 0) { + virReportSystemError(errno, + _("cannot restore MAC address on $s"), + linkdev); + } + } + if (ifname) { vpDisassociatePortProfileId(ifname, macaddr, linkdev, Index: libvirt/src/util/macvtap.h =================================================================== --- libvirt.orig/src/util/macvtap.h +++ libvirt/src/util/macvtap.h @@ -83,12 +83,15 @@ int openMacvtapTap(const char *ifname, const unsigned char *vmuuid, virVirtualPortProfileParamsPtr virtPortProfile, char **res_ifname, - enum virVMOperationType vmop); + enum virVMOperationType vmop, + char *stateDir); void delMacvtap(const char *ifname, const unsigned char *macaddress, const char *linkdev, - virVirtualPortProfileParamsPtr virtPortProfile); + int mode, + virVirtualPortProfileParamsPtr virtPortProfile, + char *stateDir); int vpAssociatePortProfileId(const char *macvtap_ifname, const unsigned char *macvtap_macaddr, Index: libvirt/src/qemu/qemu_hotplug.c =================================================================== --- libvirt.orig/src/qemu/qemu_hotplug.c +++ libvirt/src/qemu/qemu_hotplug.c @@ -1610,7 +1610,9 @@ int qemuDomainDetachNetDevice(struct qem #if WITH_MACVTAP if (detach->type == VIR_DOMAIN_NET_TYPE_DIRECT) { delMacvtap(detach->ifname, detach->mac, detach->data.direct.linkdev, - &detach->data.direct.virtPortProfile); + detach->data.direct.mode, + &detach->data.direct.virtPortProfile, + driver->stateDir); VIR_FREE(detach->ifname); } #endif Index: libvirt/src/util/interface.c =================================================================== --- libvirt.orig/src/util/interface.c +++ libvirt/src/util/interface.c @@ -390,3 +390,102 @@ ifaceGetVlanID(const char *vlanifname AT return ENOSYS; } #endif /* __linux__ */ + +/** + * ifaceGetMacaddr: + * @ifname: interface name to set MTU for + * @macaddr: MAC address (VIR_MAC_BUFLEN in size) + * + * This function gets the @macaddr for a given interface @ifname. + * + * Returns 0 in case of success or an errno code in case of failure. + */ +#ifdef __linux__ +int +ifaceGetMacaddr(const char *ifname, + unsigned char *macaddr) +{ + struct ifreq ifr; + int fd; + + if (!ifname) + return EINVAL; + + fd = socket(AF_INET, SOCK_STREAM, 0); + if(fd < 0){ + return errno; + } + + memset(&ifr, 0, sizeof(struct ifreq)); + if (virStrcpyStatic(ifr.ifr_name, ifname) == NULL) + return EINVAL; + + if(ioctl(fd, SIOCGIFHWADDR, (char *)&ifr) != 0){ + return errno; + } + + memcpy(macaddr, ifr.ifr_ifru.ifru_hwaddr.sa_data, VIR_MAC_BUFLEN); + + return 0; +} + +#else + +int +ifaceGetMacaddr(const char *ifname, + unsigned char *macaddr) +{ + return ENOSYS; +} + +#endif /* __linux__ */ + +/** + * ifaceSetMacaddr: + * @ifname: interface name to set MTU for + * @macaddr: MAC address (VIR_MAC_BUFLEN in size) + * + * This function sets the @macaddr for a given interface @ifname. This + * gets rid of the kernel's automatically assigned random MAC. + * + * Returns 0 in case of success or an errno code in case of failure. + */ +#ifdef __linux__ +int +ifaceSetMacaddr(const char *ifname, + const unsigned char *macaddr) +{ + struct ifreq ifr; + int fd; + + if (!ifname) + return EINVAL; + + fd = socket(AF_INET, SOCK_STREAM, 0); + if(fd < 0){ + return errno; + } + + memset(&ifr, 0, sizeof(struct ifreq)); + if (virStrcpyStatic(ifr.ifr_name, ifname) == NULL) + return EINVAL; + + /* To fill ifr.ifr_hdaddr.sa_family field */ + if (ioctl(fd, SIOCGIFHWADDR, &ifr) != 0) + return errno; + + memcpy(ifr.ifr_hwaddr.sa_data, macaddr, VIR_MAC_BUFLEN); + + return ioctl(fd, SIOCSIFHWADDR, &ifr) == 0 ? 0 : errno; +} + +#else + +int +ifaceSetMacaddr(const char *ifname, + const unsigned char *macaddr) +{ + return ENOSYS; +} + +#endif /* __linux__ */ Index: libvirt/src/util/interface.h =================================================================== --- libvirt.orig/src/util/interface.h +++ libvirt/src/util/interface.h @@ -32,4 +32,8 @@ int ifaceGetIndex(bool reportError, cons int ifaceGetVlanID(const char *vlanifname, int *vlanid); +int ifaceSetMacaddr(const char *ifname, const unsigned char *macaddr); + +int ifaceGetMacaddr(const char *ifname, unsigned char *macaddr); + #endif /* __VIR_INTERFACE_H__ */ Index: libvirt/src/libvirt_private.syms =================================================================== --- libvirt.orig/src/libvirt_private.syms +++ libvirt/src/libvirt_private.syms @@ -507,7 +507,9 @@ ifaceCheck; ifaceCtrl; ifaceGetFlags; ifaceGetIndex; +ifaceGetMacaddr; ifaceGetVlanID; +ifaceSetMacaddr; ifaceIsUp; =================================================================== Best regards, Gerhard Stenzel ------------------------------------------------------------------------------------- IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschaeftsfuehrung: Dirk Wittkopp Sitz der Gesellschaft: Boeblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

3 2

[libvirt] [PATCH 0/4] RFC: vcpupin: some improvements of virsh vcpupin command
by Taku Izumi 21 Jun '11

21 Jun '11

Hi all, I'll resend this patchset because what sent yesterday was broken. Sorry to be a nuisance. --- This patchset improves virsh vcpupin command like the following: (i) introduce special markup "-" and "^" which are similar to XML schema of "cpuset" attribute # virsh vcpupin VM 0 0-8,^4 --config # virsh dumpxml VM ... <vcpu>4</vcpu> <cputune> <vcpupin vcpu='0' cpuset='0-3,5-8'/> </cputune> ... (ii) allow to receive the special keyword 'r' as a cpulist parameter when resetting vcpupin setting. # virsh vcpupin VM 0 r --config (iii) when resetting, that is, pinning the specified virtual cpu to all host physical cpus, vcpupin setting will be deleted # virsh dumpxml VM ... <vcpu>4</vcpu> <cputune> </cputune> ... The following my patchset is the prerequisite of this: http://www.redhat.com/archives/libvir-list/2011-June/msg00441.html *[PATCH 1/4] vcpupin: improve vcpupin definition of virsh vcpupin *[PATCH 2/4] vcpupin: add reset option to virsh vcpupin command *[PATCH 3/4] vcpupin: add virDomainVcpupinDel function *[PATCH 4/4] vcpupin: add vcpupin resetting feature to qemu driver Best regards, Taku Izumi

4 12

[libvirt] [PATCH] conf: Make full use of goto lable
by Osier Yang 21 Jun '11

21 Jun '11

* virDomainDefParse: There is a goto label "no_memory", which reports OOM error, and it then fallthrough lable "error". This patch changes things like following: virReportOOMError(); goto error; into: goto no_memory; --- src/conf/domain_conf.c | 21 +++++++-------------- 1 files changed, 7 insertions(+), 14 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 5360863..f80ab43 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -5832,8 +5832,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, char *set = tmp; def->cpumasklen = VIR_DOMAIN_CPUMASK_LEN; if (VIR_ALLOC_N(def->cpumask, def->cpumasklen) < 0) { - virReportOOMError(); - goto error; + goto no_memory; } if (virDomainCpuSetParse((const char **)&set, 0, def->cpumask, @@ -5894,8 +5893,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, int nodemasklen = VIR_DOMAIN_CPUMASK_LEN; if (VIR_ALLOC_N(def->numatune.memory.nodemask, nodemasklen) < 0) { - virReportOOMError(); - goto error; + goto no_memory; } /* "nodeset" leads same syntax with "cpuset". */ @@ -6010,8 +6008,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, if (def->os.bootloader) { def->os.type = strdup("xen"); if (!def->os.type) { - virReportOOMError(); - goto error; + goto no_memory; } } else { virDomainReportError(VIR_ERR_OS_TYPE, @@ -6028,8 +6025,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, def->virtType == VIR_DOMAIN_VIRT_XEN) { VIR_FREE(def->os.type); if (!(def->os.type = strdup("xen"))) { - virReportOOMError(); - goto error; + goto no_memory; } } @@ -6056,8 +6052,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, goto error; } if (!(def->os.arch = strdup(defaultArch))) { - virReportOOMError(); - goto error; + goto no_memory; } } @@ -6069,8 +6064,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, virDomainVirtTypeToString(def->virtType)); if (defaultMachine != NULL) { if (!(def->os.machine = strdup(defaultMachine))) { - virReportOOMError(); - goto error; + goto no_memory; } } } @@ -6403,8 +6397,7 @@ static virDomainDefPtr virDomainDefParseXML(virCapsPtr caps, virDomainInputDefPtr input; if (VIR_ALLOC(input) < 0) { - virReportOOMError(); - goto error; + goto no_memory; } if (STREQ(def->os.type, "hvm")) { input->type = VIR_DOMAIN_INPUT_TYPE_MOUSE; -- 1.7.4

2 2

[libvirt] [PATCH v2] Add support for fd: protocol
by Corey Bryant 21 Jun '11

21 Jun '11

sVirt provides SELinux MAC isolation for Qemu guest processes and their corresponding resources (image files). sVirt provides this support by labeling guests and resources with security labels that are stored in file system extended attributes. Some file systems, such as NFS, do not support the extended attribute security namespace, which is needed for image file isolation when using the sVirt SELinux security driver in libvirt. The proposed solution entails a combination of Qemu, libvirt, and SELinux patches that work together to isolate multiple guests' images when they're stored in the same NFS mount. This results in an environment where sVirt isolation and NFS image file isolation can both be provided. This patch contains the Qemu code to support this solution. I would like to solicit input from the libvirt community prior to starting the libvirt patch. Currently, Qemu opens an image file in addition to performing the necessary read and write operations. The proposed solution will move the open out of Qemu and into libvirt. Once libvirt opens an image file for the guest, it will pass the file descriptor to Qemu via a new fd: protocol. If the image file resides in an NFS mount, the following SELinux policy changes will provide image isolation: - A new SELinux boolean is created (e.g. virt_read_write_nfs) to allow Qemu (svirt_t) to only have SELinux read and write permissions on nfs_t files - Qemu (svirt_t) also gets SELinux use permissions on libvirt (virtd_t) file descriptors Following is a sample invocation of Qemu using the fd: protocol on the command line: qemu -drive file=fd:4,format=qcow2 The fd: protocol is also supported by the drive_add monitor command. This requires that the specified file descriptor is passed to the monitor alongside a prior getfd monitor command. There are some additional features provided by certain image types where Qemu reopens the image file. All of these scenarios will be unsupported for the fd: protocol, at least for this patch: - The -snapshot command line option - The savevm monitor command - The snapshot_blkdev monitor command - Starting Qemu with a backing file The thought is that this support can be added in the future, but is not required for the initial fd: support. This patch was tested with the following formats: raw, cow, qcow, qcow2, qed, and vmdk, using the fd: protocol from the command line and the monitor. Tests were also run to verify existing file name support and qemu-img were not regressed. Non-valid file descriptors, fd: without format, snapshot and backing files were also tested. Signed-off-by: Corey Bryant <coreyb(a)linux.vnet.ibm.com> --- block.c | 16 ++++++++++ block.h | 1 + block/cow.c | 5 +++ block/qcow.c | 5 +++ block/qcow2.c | 5 +++ block/qed.c | 4 ++ block/raw-posix.c | 81 +++++++++++++++++++++++++++++++++++++++++++++++------ block/vmdk.c | 5 +++ block_int.h | 1 + blockdev.c | 10 ++++++ monitor.c | 5 +++ monitor.h | 1 + qemu-options.hx | 8 +++-- qemu-tool.c | 5 +++ 14 files changed, 140 insertions(+), 12 deletions(-) diff --git a/block.c b/block.c index 24a25d5..500db84 100644 --- a/block.c +++ b/block.c @@ -536,6 +536,10 @@ int bdrv_open(BlockDriverState *bs, const char *filename, int flags, char tmp_filename[PATH_MAX]; char backing_filename[PATH_MAX]; + if (bdrv_is_fd_protocol(bs)) { + return -ENOTSUP; + } + /* if snapshot, we create a temporary backing file and open it instead of opening 'filename' directly */ @@ -585,6 +589,10 @@ int bdrv_open(BlockDriverState *bs, const char *filename, int flags, /* Find the right image format driver */ if (!drv) { + /* format must be specified for fd: protocol */ + if (bdrv_is_fd_protocol(bs)) { + return -ENOTSUP; + } ret = find_image_format(filename, &drv); } @@ -1460,6 +1468,11 @@ int bdrv_enable_write_cache(BlockDriverState *bs) return bs->enable_write_cache; } +int bdrv_is_fd_protocol(BlockDriverState *bs) +{ + return bs->fd_protocol; +} + /* XXX: no longer used */ void bdrv_set_change_cb(BlockDriverState *bs, void (*change_cb)(void *opaque, int reason), @@ -1964,6 +1977,9 @@ int bdrv_snapshot_create(BlockDriverState *bs, BlockDriver *drv = bs->drv; if (!drv) return -ENOMEDIUM; + if (bdrv_is_fd_protocol(bs)) { + return -ENOTSUP; + } if (drv->bdrv_snapshot_create) return drv->bdrv_snapshot_create(bs, sn_info); if (bs->file) diff --git a/block.h b/block.h index da7d39c..dd46d52 100644 --- a/block.h +++ b/block.h @@ -182,6 +182,7 @@ int bdrv_is_removable(BlockDriverState *bs); int bdrv_is_read_only(BlockDriverState *bs); int bdrv_is_sg(BlockDriverState *bs); int bdrv_enable_write_cache(BlockDriverState *bs); +int bdrv_is_fd_protocol(BlockDriverState *bs); int bdrv_is_inserted(BlockDriverState *bs); int bdrv_media_changed(BlockDriverState *bs); int bdrv_is_locked(BlockDriverState *bs); diff --git a/block/cow.c b/block/cow.c index 4cf543c..e17f8e7 100644 --- a/block/cow.c +++ b/block/cow.c @@ -82,6 +82,11 @@ static int cow_open(BlockDriverState *bs, int flags) pstrcpy(bs->backing_file, sizeof(bs->backing_file), cow_header.backing_file); + if (bs->backing_file[0] != '\0' && bdrv_is_fd_protocol(bs)) { + /* backing file currently not supported by fd: protocol */ + goto fail; + } + bitmap_size = ((bs->total_sectors + 7) >> 3) + sizeof(cow_header); s->cow_sectors_offset = (bitmap_size + 511) & ~511; return 0; diff --git a/block/qcow.c b/block/qcow.c index a26c886..1e46bdb 100644 --- a/block/qcow.c +++ b/block/qcow.c @@ -157,6 +157,11 @@ static int qcow_open(BlockDriverState *bs, int flags) if (bdrv_pread(bs->file, header.backing_file_offset, bs->backing_file, len) != len) goto fail; bs->backing_file[len] = '\0'; + + if (bs->backing_file[0] != '\0' && bdrv_is_fd_protocol(bs)) { + /* backing file currently not supported by fd: protocol */ + goto fail; + } } return 0; diff --git a/block/qcow2.c b/block/qcow2.c index 8451ded..8b9c160 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -270,6 +270,11 @@ static int qcow2_open(BlockDriverState *bs, int flags) goto fail; } bs->backing_file[len] = '\0'; + + if (bs->backing_file[0] != '\0' && bdrv_is_fd_protocol(bs)) { + ret = -ENOTSUP; + goto fail; + } } if (qcow2_read_snapshots(bs) < 0) { ret = -EINVAL; diff --git a/block/qed.c b/block/qed.c index 3970379..5028897 100644 --- a/block/qed.c +++ b/block/qed.c @@ -446,6 +446,10 @@ static int bdrv_qed_open(BlockDriverState *bs, int flags) return ret; } + if (bs->backing_file[0] != '\0' && bdrv_is_fd_protocol(bs)) { + return -ENOTSUP; + } + if (s->header.features & QED_F_BACKING_FORMAT_NO_PROBE) { pstrcpy(bs->backing_format, sizeof(bs->backing_format), "raw"); } diff --git a/block/raw-posix.c b/block/raw-posix.c index 4cd7d7a..c72de3d 100644 --- a/block/raw-posix.c +++ b/block/raw-posix.c @@ -28,6 +28,7 @@ #include "block_int.h" #include "module.h" #include "block/raw-posix-aio.h" +#include "monitor.h" #ifdef CONFIG_COCOA #include <paths.h> @@ -183,7 +184,8 @@ static int raw_open_common(BlockDriverState *bs, const char *filename, int bdrv_flags, int open_flags) { BDRVRawState *s = bs->opaque; - int fd, ret; + int fd = -1; + int ret; ret = raw_normalize_devicepath(&filename); if (ret != 0) { @@ -205,15 +207,17 @@ static int raw_open_common(BlockDriverState *bs, const char *filename, if (!(bdrv_flags & BDRV_O_CACHE_WB)) s->open_flags |= O_DSYNC; - s->fd = -1; - fd = qemu_open(filename, s->open_flags, 0644); - if (fd < 0) { - ret = -errno; - if (ret == -EROFS) - ret = -EACCES; - return ret; + if (s->fd == -1) { + fd = qemu_open(filename, s->open_flags, 0644); + if (fd < 0) { + ret = -errno; + if (ret == -EROFS) { + ret = -EACCES; + } + return ret; + } + s->fd = fd; } - s->fd = fd; s->aligned_buf = NULL; if ((bdrv_flags & BDRV_O_NOCACHE)) { @@ -270,6 +274,7 @@ static int raw_open(BlockDriverState *bs, const char *filename, int flags) { BDRVRawState *s = bs->opaque; + s->fd = -1; s->type = FTYPE_FILE; return raw_open_common(bs, filename, flags, 0); } @@ -890,6 +895,60 @@ static BlockDriver bdrv_file = { .create_options = raw_create_options, }; +static int raw_open_fd(BlockDriverState *bs, const char *filename, int flags) +{ + BDRVRawState *s = bs->opaque; + const char *fd_str; + int fd; + + /* extract the file descriptor - fail if it's not fd: */ + if (!strstart(filename, "fd:", &fd_str)) { + return -EINVAL; + } + + if (!qemu_isdigit(fd_str[0])) { + /* get fd from monitor */ + fd = qemu_get_fd(fd_str); + if (fd == -1) { + return -EBADF; + } + } else { + char *endptr = NULL; + + fd = strtol(fd_str, &endptr, 10); + if (*endptr || (fd == 0 && fd_str == endptr)) { + return -EBADF; + } + } + + s->fd = fd; + s->type = FTYPE_FILE; + + return raw_open_common(bs, filename, flags, 0); +} + +static BlockDriver bdrv_file_fd = { + .format_name = "file", + .protocol_name = "fd", + .instance_size = sizeof(BDRVRawState), + .bdrv_probe = NULL, /* no probe for protocols */ + .bdrv_file_open = raw_open_fd, + .bdrv_read = raw_read, + .bdrv_write = raw_write, + .bdrv_close = raw_close, + .bdrv_flush = raw_flush, + .bdrv_discard = raw_discard, + + .bdrv_aio_readv = raw_aio_readv, + .bdrv_aio_writev = raw_aio_writev, + .bdrv_aio_flush = raw_aio_flush, + + .bdrv_truncate = raw_truncate, + .bdrv_getlength = raw_getlength, + + .create_options = raw_create_options, +}; + /***********************************************/ /* host device */ @@ -998,6 +1057,7 @@ static int hdev_open(BlockDriverState *bs, const char *filename, int flags) } #endif + s->fd = -1; s->type = FTYPE_FILE; #if defined(__linux__) { @@ -1168,6 +1228,7 @@ static int floppy_open(BlockDriverState *bs, const char *filename, int flags) BDRVRawState *s = bs->opaque; int ret; + s->fd = -1; s->type = FTYPE_FD; /* open will not fail even if no floppy is inserted, so add O_NONBLOCK */ @@ -1280,6 +1341,7 @@ static int cdrom_open(BlockDriverState *bs, const char *filename, int flags) { BDRVRawState *s = bs->opaque; + s->fd = -1; s->type = FTYPE_CD; /* open will not fail even if no CD is inserted, so add O_NONBLOCK */ @@ -1503,6 +1565,7 @@ static void bdrv_file_init(void) * Register all the drivers. Note that order is important, the driver * registered last will get probed first. */ + bdrv_register(&bdrv_file_fd); bdrv_register(&bdrv_file); bdrv_register(&bdrv_host_device); #ifdef __linux__ diff --git a/block/vmdk.c b/block/vmdk.c index 922b23d..2ea808e 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -353,6 +353,11 @@ static int vmdk_parent_open(BlockDriverState *bs) return -1; pstrcpy(bs->backing_file, end_name - p_name + 1, p_name); + + if (bs->backing_file[0] != '\0' && bdrv_is_fd_protocol(bs)) { + /* backing file currently not supported by fd: protocol */ + return -1; + } } return 0; diff --git a/block_int.h b/block_int.h index fa91337..a305ee2 100644 --- a/block_int.h +++ b/block_int.h @@ -152,6 +152,7 @@ struct BlockDriverState { int encrypted; /* if true, the media is encrypted */ int valid_key; /* if true, a valid encryption key has been set */ int sg; /* if true, the device is a /dev/sg* */ + int fd_protocol; /* if true, the fd: protocol was specified */ /* event callback when inserting/removing */ void (*change_cb)(void *opaque, int reason); void *change_opaque; diff --git a/blockdev.c b/blockdev.c index e81e0ab..a536c20 100644 --- a/blockdev.c +++ b/blockdev.c @@ -546,6 +546,10 @@ DriveInfo *drive_init(QemuOpts *opts, int default_to_scsi) bdrv_flags |= ro ? 0 : BDRV_O_RDWR; + if (strncmp(file, "fd:", 3) == 0) { + dinfo->bdrv->fd_protocol = 1; + } + ret = bdrv_open(dinfo->bdrv, file, bdrv_flags, drv); if (ret < 0) { error_report("could not open disk image %s: %s", @@ -606,6 +610,12 @@ int do_snapshot_blkdev(Monitor *mon, const QDict *qdict, QObject **ret_data) goto out; } + if (bdrv_is_fd_protocol(bs)) { + qerror_report(QERR_UNSUPPORTED); + ret = -1; + goto out; + } + pstrcpy(old_filename, sizeof(old_filename), bs->filename); old_drv = bs->drv; diff --git a/monitor.c b/monitor.c index 59a3e76..ea60be2 100644 --- a/monitor.c +++ b/monitor.c @@ -2832,6 +2832,11 @@ int monitor_get_fd(Monitor *mon, const char *fdname) return -1; } +int qemu_get_fd(const char *fdname) +{ + return cur_mon ? monitor_get_fd(cur_mon, fdname) : -1; +} + static const mon_cmd_t mon_cmds[] = { #include "hmp-commands.h" { NULL, NULL, }, diff --git a/monitor.h b/monitor.h index 4f2d328..de5b987 100644 --- a/monitor.h +++ b/monitor.h @@ -51,6 +51,7 @@ int monitor_read_bdrv_key_start(Monitor *mon, BlockDriverState *bs, void *opaque); int monitor_get_fd(Monitor *mon, const char *fdname); +int qemu_get_fd(const char *fdname); void monitor_vprintf(Monitor *mon, const char *fmt, va_list ap) GCC_FMT_ATTR(2, 0); diff --git a/qemu-options.hx b/qemu-options.hx index 1d5ad8b..f9b66f4 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -116,7 +116,7 @@ using @file{/dev/cdrom} as filename (@pxref{host_drives}). ETEXI DEF("drive", HAS_ARG, QEMU_OPTION_drive, - "-drive [file=file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n" + "-drive [file=[fd:]file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n" " [,cyls=c,heads=h,secs=s[,trans=t]][,snapshot=on|off]\n" " [,cache=writethrough|writeback|none|unsafe][,format=f]\n" " [,serial=s][,addr=A][,id=name][,aio=threads|native]\n" @@ -129,10 +129,12 @@ STEXI Define a new drive. Valid options are: @table @option -@item file=@var{file} +@item file=[fd:]@var{file} This option defines which disk image (@pxref{disk_images}) to use with this drive. If the filename contains comma, you must double it -(for instance, "file=my,,file" to use file "my,file"). +(for instance, "file=my,,file" to use file "my,file"). @option{fd:}@var{file} +specifies the file descriptor of an already open disk image. +@option{format=}@var{format} is required by @option{fd:}@var{file}. @item if=@var{interface} This option defines on which type on interface the drive is connected. Available types are: ide, scsi, sd, mtd, floppy, pflash, virtio. diff --git a/qemu-tool.c b/qemu-tool.c index 41e5c41..8fe6b8c 100644 --- a/qemu-tool.c +++ b/qemu-tool.c @@ -96,3 +96,8 @@ int64_t qemu_get_clock_ns(QEMUClock *clock) { return 0; } + +int qemu_get_fd(const char *fdname) +{ + return -1; +} -- 1.7.1

6 14

[libvirt] [PATCH] conf: Fix one memory leak
by Osier Yang 21 Jun '11

21 Jun '11

Free def->numatune.memory.nodemask in virDomainDefFree. --- src/conf/domain_conf.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 5360863..38a8587 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -1092,6 +1092,8 @@ void virDomainDefFree(virDomainDefPtr def) virDomainVcpupinDefFree(def->cputune.vcpupin, def->cputune.nvcpupin); + VIR_FREE(def->numatune.memory.nodemask); + virSysinfoDefFree(def->sysinfo); if (def->namespaceData && def->ns.free) -- 1.7.4

2 2

[libvirt] [PATCH 0/7 v2] python: Implement missing stream/event ops
by Cole Robinson 20 Jun '11

20 Jun '11

Currently it isn't possible to implement the equivalent of 'virsh console' with the python bindings. This series implements the missing functionality required to do so. v2: Pushed first 3 patches Actually export virEvent APIs in libvirt.h DV had comments on patch 1 and patch 6 that I didn't address, if I haven't managed to convince him (or others) with my latest mails I can do another rev of the series. Cole Robinson (7): python: Implement virStreamSend/Recv python: Implement virStreamSend/RecvAll helpers Promote virEvent*Handle/Timeout to public API events: Correct virEventAddTimeout docs python: Add bindings for virEvent*Handle/Timeout python: events: Fix C->Python handle callback prototype python: Mark event callback wrappers as private daemon/libvirtd.c | 1 - daemon/mdns.c | 1 - examples/domain-events/events-python/event-test.py | 6 +- include/libvirt/libvirt.h.in | 14 + python/generator.py | 12 +- python/libvirt-override-virConnect.py | 102 +++---- python/libvirt-override-virStream.py | 101 +++++++- python/libvirt-override.c | 274 +++++++++++++++++--- python/libvirt-override.py | 80 ++++++- python/typewrappers.c | 14 + python/typewrappers.h | 1 + src/conf/domain_event.c | 1 - src/conf/domain_event.h | 1 - src/fdstream.c | 1 - src/libvirt_private.syms | 9 - src/libvirt_public.syms | 6 + src/libxl/libxl_driver.c | 1 - src/lxc/lxc_driver.c | 1 - src/network/bridge_driver.c | 1 - src/node_device/node_device_hal.c | 1 - src/node_device/node_device_udev.c | 1 - src/openvz/openvz_driver.c | 1 - src/qemu/qemu_domain.c | 1 - src/qemu/qemu_driver.c | 1 - src/qemu/qemu_monitor.c | 1 - src/remote/remote_driver.c | 1 - src/test/test_driver.c | 1 - src/uml/uml_driver.c | 1 - src/util/event.c | 56 ++++ src/util/event.h | 73 ------ src/util/util.c | 1 - src/vbox/vbox_tmpl.c | 1 - src/xen/xen_inotify.c | 1 - src/xen/xs_internal.c | 1 - tools/console.c | 1 - 35 files changed, 559 insertions(+), 211 deletions(-) -- 1.7.4.4

2 9

[libvirt] virt-dmesg and virt-uname
by Richard W.M. Jones 20 Jun '11

20 Jun '11

I just pushed the virt-dmesg package to Fedora Rawhide[1]. http://koji.fedoraproject.org/koji/buildinfo?buildID=248890 This tool uses libvirt to let you list the kernel messages from a Linux KVM guest without needing to log in or run any special agent in the guest. It's useful when a guest crashes or gets stuck. Usage is pretty simple: # virt-dmesg F14x64 | head <6>[ 0.000000] Initializing cgroup subsys cpuset <6>[ 0.000000] Initializing cgroup subsys cpu <5>[ 0.000000] Linux version 2.6.35.13-91.fc14.x86_64 (mockbuild(a)x86-15.phx2.fedoraproject.org) (gcc version 4.5.1 20100924 (Red Hat 4.5.1-4) (GCC) ) #1 SMP Tue May 3 13:23:06 UTC 2011 <6>[ 0.000000] Command line: ro root=/dev/mapper/vg_f13x64-lv_root rd_LVM_LV=vg_f13x64/lv_root rd_LVM_LV=vg_f13x64/lv_swap rd_NO_LUKS rd_NO_MD rd_NO_DM LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYBOARDTYPE=pc KEYTABLE=uk rhgb quiet <6>[ 0.000000] BIOS-provided physical RAM map: <6>[ 0.000000] BIOS-e820: 0000000000000000 - 000000000009f400 (usable) <6>[ 0.000000] BIOS-e820: 000000000009f400 - 00000000000a0000 (reserved) <6>[ 0.000000] BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved) <6>[ 0.000000] BIOS-e820: 0000000000100000 - 000000003fffd000 (usable) <6>[ 0.000000] BIOS-e820: 000000003fffd000 - 0000000040000000 (reserved) In some cases virt-dmesg won't be able to find the symbol table, in which case you can fall back to dumping out the kernel and grepping it, eg: # virt-dmesg --dump-kernel F14x64 | strings | grep '^<[0-9]>' | less There's an accompanying tool called virt-uname which does what you would think: # virt-uname F14x64 Linux f14x64.home.annexia.org 2.6.35.13-91.fc14.x86_64 #1 SMP Tue May 3 13:23:06 UTC 2011 x86_64 (none) Please let me know if you find (Linux) guests for which virt-dmesg doesn't work properly (particularly the --dump-kernel option which ought to work for any Linux guest). Describe the guest as accurately as you can and provide the output of 'virt-dmesg -v Guest'. Rich. [1] Sadly because of some regressions in libvirt, it *won't* work in Fedora 15 or earlier unless you are able to upgrade to libvirt >= 0.9.1-3. -- 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 1

[libvirt] [PATCH 1/3] virsh: clarify snapshot vs. save
by Eric Blake 20 Jun '11

20 Jun '11

* tools/virsh.c (info_snapshot_create, info_save): Clarify description. * tools/virsh.pod (save): Likewise. --- tools/virsh.c | 6 +++--- tools/virsh.pod | 8 +++++++- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 1c0b1dc..6b5f3d9 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -1469,7 +1469,7 @@ cmdStart(vshControl *ctl, const vshCmd *cmd) */ static const vshCmdInfo info_save[] = { {"help", N_("save a domain state to a file")}, - {"desc", N_("Save a running domain.")}, + {"desc", N_("Save the RAM state of a running domain.")}, {NULL, NULL} }; @@ -10947,8 +10947,8 @@ cmdQuit(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) * "snapshot-create" command */ static const vshCmdInfo info_snapshot_create[] = { - {"help", N_("Create a snapshot")}, - {"desc", N_("Snapshot create")}, + {"help", N_("Create a snapshot from XML")}, + {"desc", N_("Create a snapshot (disk and RAM) from XML")}, {NULL, NULL} }; diff --git a/tools/virsh.pod b/tools/virsh.pod index c6879cb..33f72ed 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -594,7 +594,8 @@ the state file was created. =item B<save> I<domain-id> I<state-file> -Saves a running domain to a state file so that it can be restored +Saves a running domain (RAM, but not disk state) to a state file so that +it can be restored later. Once saved, the domain will no longer be running on the system, thus the memory allocated for the domain will be free for other domains to use. B<virsh restore> restores from this state file. @@ -603,6 +604,11 @@ This is roughly equivalent to doing a hibernate on a running computer, with all the same limitations. Open network connections may be severed upon restore, as TCP timeouts may have expired. +Domain saved state files assume that disk images will be unchanged +between the creation and restore point. For a more complete system +restore point, where the disk state is saved alongside the memory +state, see the B<snapshot> family of commands. + =item B<schedinfo> optional I<--set> B<parameter=value> I<domain-id> I<--config> I<--live> I<--current> -- 1.7.1

3 10

[libvirt] [PATCH v2]: set and restore MAC address of a NIC when using PASSTHROUGH mode
by Gerhard Stenzel 20 Jun '11

20 Jun '11

This is a rework of the patch Dirk sent out last week. The attached patch addresses the problem that when a PASSTHROUGH mode DIRECT NIC connection is made the MAC address of the NIC is not automatically set and reset to the configured VM MAC and back again. The patch fixes this problem by setting and resetting the MAC while remembering the previous setting while the VM is running. This also works if libvirtd is restarted while the VM is running. The patch passes make syntax-check Signed-off-by: Dirk Herrendoerfer <d.herrendoerfer at herrendoerfer.name> Signed-off-by: Gerhard Stenzel <gerhard.stenzel(a)de.ibm.com> --- src/libvirt_macvtap.syms | 2 + src/qemu/qemu_command.c | 48 +++++++++++++++ src/qemu/qemu_process.c | 45 ++++++++++++++ src/util/macvtap.c | 149 ++++++++++++++++++++++++++++++++++++++++++++++ src/util/macvtap.h | 6 ++ 5 files changed, 250 insertions(+), 0 deletions(-) Index: libvirt/src/libvirt_macvtap.syms =================================================================== --- libvirt.orig/src/libvirt_macvtap.syms +++ libvirt/src/libvirt_macvtap.syms @@ -5,6 +5,8 @@ # macvtap.h delMacvtap; +getMacaddr; openMacvtapTap; +setMacaddr; vpAssociatePortProfileId; vpDisassociatePortProfileId; Index: libvirt/src/qemu/qemu_command.c =================================================================== --- libvirt.orig/src/qemu/qemu_command.c +++ libvirt/src/qemu/qemu_command.c @@ -125,6 +125,77 @@ qemuPhysIfaceConnect(virDomainDefPtr def net->model && STREQ(net->model, "virtio")) vnet_hdr = 1; + /** Note: When using PASSTHROUGH mode with MACVTAP devices the link + * device's MAC address must be set to the VMs MAC address. In + * order to not confuse the first switch or bridge in line this MAC + * address must be reset when the VM is shut down. + * This is especially important when using SRIOV capable cards that + * emulate their switch in firmware. + */ + if (net->data.direct.mode == VIR_DOMAIN_NETDEV_MACVTAP_MODE_PASSTHRU) { + unsigned char oldmac[6]; + rc = getMacaddr(&oldmac, net->data.direct.linkdev); + if (rc) { + virReportSystemError(rc, + _("Getting MAC address from '%s' " + "to '%02x:%02x:%02x:%02x:%02x:%02x' failed."), + net->data.direct.linkdev, + oldmac[0], oldmac[1], oldmac[2], + oldmac[3], oldmac[4], oldmac[5]); + } else { + char *oldmacpath = NULL; + char *oldmacname = NULL; + char *newmacname = NULL; + char *basedir = "/var/run/libvirt/network"; + + if (virAsprintf(&oldmacname,"%02x:%02x:%02x:%02x:%02x:%02x", + oldmac[0], oldmac[1], oldmac[2], + oldmac[3], oldmac[4], oldmac[5]) < 0) { + virReportOOMError(); + } + if (virAsprintf(&oldmacpath,"%s/%s", basedir, oldmacname) < 0) { + virReportOOMError(); + } + + if (virAsprintf(&newmacname, "%s/%s@%02x:%02x:%02x:%02x:%02x:%02x", + basedir, + net->data.direct.linkdev, + net->mac[0],net->mac[1],net->mac[2], + net->mac[3],net->mac[4],net->mac[5]) < 0) { + virReportOOMError(); + } + + rc = symlink (oldmacname, newmacname); + if (rc) { + virReportSystemError(errno, + _("MAC link file creation failed for %s."), + net->data.direct.linkdev); + } + rc = creat(oldmacpath, S_IWUSR); + if (rc) { + virReportSystemError(errno, + _("MAC link file creation failed for %s."), + oldmacpath); + } + rc = VIR_CLOSE(rc); + if (rc) { + virReportSystemError(errno, + _("MAC link file closing failed for %s."), + oldmacpath); + } + } + + rc = setMacaddr(net->mac, net->data.direct.linkdev); + if (rc) { + virReportSystemError(errno, + _("Setting MAC address on '%s' to " + "'%02x:%02x:%02x:%02x:%02x:%02x' failed."), + net->data.direct.linkdev, + net->mac[0], net->mac[1], net->mac[2], + net->mac[3], net->mac[4], net->mac[5]); + } + } + rc = openMacvtapTap(net->ifname, net->mac, net->data.direct.linkdev, net->data.direct.mode, vnet_hdr, def->uuid, &net->data.direct.virtPortProfile, &res_ifname, Index: libvirt/src/qemu/qemu_process.c =================================================================== --- libvirt.orig/src/qemu/qemu_process.c +++ libvirt/src/qemu/qemu_process.c @@ -2708,6 +2708,48 @@ void qemuProcessStop(struct qemud_driver if (net->type == VIR_DOMAIN_NET_TYPE_DIRECT) { delMacvtap(net->ifname, net->mac, net->data.direct.linkdev, &net->data.direct.virtPortProfile); + + if (net->data.direct.mode == VIR_DOMAIN_NETDEV_MACVTAP_MODE_PASSTHRU) { + char *newmacname = NULL; + char *oldmacname = NULL; + char *oldmacpath = NULL; + char *basedir = "/var/run/libvirt/network"; + + if (virAsprintf(&newmacname, "%s/%s@%02x:%02x:%02x:%02x:%02x:%02x", + basedir, + net->data.direct.linkdev, + net->mac[0],net->mac[1],net->mac[2], + net->mac[3],net->mac[4],net->mac[5]) < 0) { + virReportOOMError(); + } + + ret = virFileResolveLink(newmacname, &oldmacpath); + if (ret < 0) { + virReportSystemError(errno, + _("Reading MAC address from '%s' failed !"), + newmacname); + } else { + unsigned char oldmac[6]; + + oldmacname = basename(oldmacpath); + + if (virParseMacAddr(oldmacname, &oldmac[0]) != 0) { + virReportSystemError(VIR_ERR_INTERNAL_ERROR, + _("Cannot parse MAC address from '%s'"), + oldmacname); + } + + VIR_WARN(("Resetting MAC address on '%s' " + "to '%02x:%02x:%02x:%02x:%02x:%02x'."), + net->data.direct.linkdev, + oldmac[0],oldmac[1],oldmac[2], + oldmac[3],oldmac[4],oldmac[5]); + /*reset mac and remove link file-ignore results*/ + ret = setMacaddr(oldmac, net->data.direct.linkdev); + ret = unlink(oldmacpath); + ret = unlink(newmacname); + } + } VIR_FREE(net->ifname); } } Index: libvirt/src/util/macvtap.c =================================================================== --- libvirt.orig/src/util/macvtap.c +++ libvirt/src/util/macvtap.c @@ -87,6 +87,7 @@ # define LLDPAD_PID_FILE "/var/run/lldpad.pid" +#define MACADDRSIZE 6 enum virVirtualPortOp { ASSOCIATE = 0x1, @@ -191,6 +192,149 @@ err_exit: # if WITH_MACVTAP +/** + * getMacaddr: + * Get the MAC address of a network device + * + * @macaddress: Pointer where the MAC address will be stored + * @srcdev: The interface name of the NIC to get the MAC from + * + * Returns zero in case of success, + * negative value otherwise with error reported. + * + */ +int +getMacaddr(const unsigned char *macaddress, const char *srcdev ) +{ + int sockfd; + int io; + struct ifreq ifr; + + strcpy(ifr.ifr_name, srcdev); + + sockfd = socket(AF_INET, SOCK_STREAM, 0); + if(sockfd < 0){ + return -1; + } + + io = ioctl(sockfd, SIOCGIFHWADDR, (char *)&ifr); + if(io < 0){ + return -1; + } + + memcpy(macaddress, ifr.ifr_ifru.ifru_hwaddr.sa_data, MACADDRSIZE); + + return 0; +} + +/** + * setMacaddr: + * Set the MAC address of a network device + * + * @macaddress: MAC address to assign to the NIC + * @srcdev: The interface name of the NIC + * + * Returns zero in case of success, + * negative value otherwise with error reported. + * + */ +int +setMacaddr(const unsigned char *macaddress, const char *srcdev ) +{ + int rc = 0; + struct nlmsghdr *resp; + struct nlmsgerr *err; + struct ifinfomsg ifinfo = { .ifi_family = AF_UNSPEC }; + int ifindex; + unsigned char *recvbuf = NULL; + unsigned int recvbuflen; + struct nl_msg *nl_msg; + + if (ifaceGetIndex(true, srcdev, &ifindex) != 0) + return -1; + + nl_msg = nlmsg_alloc_simple(RTM_SETLINK, NLM_F_REQUEST); + + if (!nl_msg) { + virReportOOMError(); + return -1; + } + + if (nlmsg_append(nl_msg, &ifinfo, sizeof(ifinfo), NLMSG_ALIGNTO) < 0) + goto buffer_too_small; + + if (nla_put_u32(nl_msg, IFLA_LINK, ifindex) < 0) + goto buffer_too_small; + + if (nla_put(nl_msg, IFLA_ADDRESS, MACADDRSIZE, macaddress) < 0) + goto buffer_too_small; + + if (srcdev && + nla_put(nl_msg, IFLA_IFNAME, strlen(srcdev)+1, srcdev) < 0) + goto buffer_too_small; + + if (nlComm(nl_msg, &recvbuf, &recvbuflen, 0) < 0) { + rc = -1; + goto err_exit; + } + + if (recvbuflen < NLMSG_LENGTH(0) || recvbuf == NULL) + goto malformed_resp; + + resp = (struct nlmsghdr *)recvbuf; + + switch (resp->nlmsg_type) { + case NLMSG_ERROR: + err = (struct nlmsgerr *)NLMSG_DATA(resp); + if (resp->nlmsg_len < NLMSG_LENGTH(sizeof(*err))) + goto malformed_resp; + + switch (err->error) { + + case 0: + break; + + case -EEXIST: + rc = -1; + break; + + default: + macvtapError(VIR_ERR_INTERNAL_ERROR, "%s", + _("error setting device mac address")); + rc = -1; + } + break; + + case NLMSG_DONE: + break; + + default: + goto malformed_resp; + } + +err_exit: + nlmsg_free(nl_msg); + + VIR_FREE(recvbuf); + + return rc; + +malformed_resp: + nlmsg_free(nl_msg); + + macvtapError(VIR_ERR_INTERNAL_ERROR, "%s", + _("malformed netlink response message")); + VIR_FREE(recvbuf); + return -1; + +buffer_too_small: + nlmsg_free(nl_msg); + + macvtapError(VIR_ERR_INTERNAL_ERROR, "%s", + _("allocated netlink buffer is too small")); + return -1; +} + static int link_add(const char *type, const unsigned char *macaddress, int macaddrsize, Index: libvirt/src/util/macvtap.h =================================================================== --- libvirt.orig/src/util/macvtap.h +++ libvirt/src/util/macvtap.h @@ -75,6 +75,10 @@ enum virVMOperationType { # include "internal.h" +int getMacaddr(const unsigned char *macaddress, const char *srcdev ); + +int setMacaddr(const unsigned char *macaddress, const char *srcdev ); + int openMacvtapTap(const char *ifname, const unsigned char *macaddress, const char *linkdev, =================================================================== Best regards, Gerhard Stenzel ------------------------------------------------------------------------------------- IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschaeftsfuehrung: Dirk Wittkopp Sitz der Gesellschaft: Boeblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

3 3

[libvirt] [PATCH] virsh: Fix compile error due to const mismatch in cmdVcpupin
by Matthias Bolte 20 Jun '11

20 Jun '11

Broken by commit c4a8ca71b10eb adding the reset option. Instead of altering a const string move the while loop into an else clause. --- tools/virsh.c | 110 ++++++++++++++++++++++++++++----------------------------- 1 files changed, 54 insertions(+), 56 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index e04c9fc..0e9efad 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -3000,7 +3000,7 @@ cmdVcpupin(vshControl *ctl, const vshCmd *cmd) int cpumaplen; int i, cpu, lastcpu, maxcpu; bool unuse = false; - char *cur; + const char *cur; int config = vshCommandOptBool(cmd, "config"); int live = vshCommandOptBool(cmd, "live"); int current = vshCommandOptBool(cmd, "current"); @@ -3061,72 +3061,70 @@ cmdVcpupin(vshControl *ctl, const vshCmd *cmd) cpumap = vshCalloc(ctl, 0, cpumaplen); /* Parse cpulist */ - cur = cpulist; - if (*cur == 0) { + if (*cpulist == '\0') { goto parse_error; - } else if (*cur == 'r') { + } else if (*cpulist == 'r') { for (cpu = 0; cpu < maxcpu; cpu++) VIR_USE_CPU(cpumap, cpu); - *cur = 0; - } - - while (*cur != 0) { - - /* the char '^' denotes exclusive */ - if (*cur == '^') { - cur++; - unuse = true; - } - - /* parse physical CPU number */ - if (!c_isdigit(*cur)) - goto parse_error; - cpu = virParseNumber(&cur); - if (cpu < 0) { - goto parse_error; - } - if (cpu >= maxcpu) { - vshError(ctl, _("Physical CPU %d doesn't exist."), cpu); - goto parse_error; - } - virSkipSpaces(&cur); - - if ((*cur == ',') || (*cur == 0)) { - if (unuse) { - VIR_UNUSE_CPU(cpumap, cpu); - } else { - VIR_USE_CPU(cpumap, cpu); + } else { + cur = cpulist; + while (*cur != '\0') { + /* the char '^' denotes exclusive */ + if (*cur == '^') { + cur++; + unuse = true; } - } else if (*cur == '-') { - /* the char '-' denotes range */ - if (unuse) { + + /* parse physical CPU number */ + if (!c_isdigit(*cur)) goto parse_error; - } - cur++; - virSkipSpaces(&cur); - /* parse the end of range */ - lastcpu = virParseNumber(&cur); - if (lastcpu < cpu) { + cpu = virParseNumber(&cur); + if (cpu < 0) { goto parse_error; } - if (lastcpu >= maxcpu) { - vshError(ctl, _("Physical CPU %d doesn't exist."), maxcpu); + if (cpu >= maxcpu) { + vshError(ctl, _("Physical CPU %d doesn't exist."), cpu); goto parse_error; } - for (i = cpu; i <= lastcpu; i++) { - VIR_USE_CPU(cpumap, i); - } virSkipSpaces(&cur); - } - if (*cur == ',') { - cur++; - virSkipSpaces(&cur); - unuse = false; - } else if (*cur == 0) { - break; - } else { - goto parse_error; + if ((*cur == ',') || (*cur == '\0')) { + if (unuse) { + VIR_UNUSE_CPU(cpumap, cpu); + } else { + VIR_USE_CPU(cpumap, cpu); + } + } else if (*cur == '-') { + /* the char '-' denotes range */ + if (unuse) { + goto parse_error; + } + cur++; + virSkipSpaces(&cur); + /* parse the end of range */ + lastcpu = virParseNumber(&cur); + if (lastcpu < cpu) { + goto parse_error; + } + if (lastcpu >= maxcpu) { + vshError(ctl, _("Physical CPU %d doesn't exist."), maxcpu); + goto parse_error; + } + for (i = cpu; i <= lastcpu; i++) { + VIR_USE_CPU(cpumap, i); + } + virSkipSpaces(&cur); + } + + if (*cur == ',') { + cur++; + virSkipSpaces(&cur); + unuse = false; + } else if (*cur == '\0') { + break; + } else { + goto parse_error; + } } } -- 1.7.0.4

1 2

[libvirt] [PATCH] Fix compile warnings in virsh vcpupin
by Daniel P. Berrange 20 Jun '11

20 Jun '11

The 'char *cur' variable was being assigned from a 'const char *' string, thus discarding constness. As well as causing a compile warning, it masked a piece of code which attempts to assign to the previously const string. * tools/virsh.c: Fix const-ness of 'cur' variable in vcpupin --- tools/virsh.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 3df0683..ed26b93 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -3000,7 +3000,7 @@ cmdVcpupin(vshControl *ctl, const vshCmd *cmd) int cpumaplen; int i, cpu, lastcpu, maxcpu; bool unuse = false; - char *cur; + const char *cur; int config = vshCommandOptBool(cmd, "config"); int live = vshCommandOptBool(cmd, "live"); int current = vshCommandOptBool(cmd, "current"); @@ -3067,7 +3067,7 @@ cmdVcpupin(vshControl *ctl, const vshCmd *cmd) } else if (*cur == 'r') { for (cpu = 0; cpu < maxcpu; cpu++) VIR_USE_CPU(cpumap, cpu); - *cur = 0; + cur = ""; } while (*cur != 0) { -- 1.7.4.4

2 1

[libvirt] nwfilter: limit VM traffic to specific MAC
by Shahar Havivi 20 Jun '11

20 Jun '11

Hi, I am trying to add custom filter to block VM traffic to other VMs by limiting the traffic only to the gateways MAC address. The filter XML: <filter name='rhev' chain='root'> <uuid>cd4e5890-ccc9-1b0f-303f-e7fe7123646d</uuid> <filterref filter='allow-dhcp'/> <rule action='drop' direction='out' priority='500'> <mac match='no' dstmacaddr='$MAC'/> </rule> </filter> The MAC is not the interface MAC address it's the gateways MAC that pass as a parameter (I use the gateway address hardcoded as well). The VM is getting DHCP ip but cannot get any traffic, I notice that when I edit (comment and uncomment) the drop rule, the filter is working fine, ie no traffic other then the gateway. 1. Am I doing something wrong? 1. What is the table name that libvirt use for ebtables? Shahar.

2 3

[libvirt] Availability of patchchecker
by Daniel Veillard 20 Jun '11

20 Jun '11

As some may have noticed, I'm often behind on patch review and it happens I notice patches way too late i.e. just before a release. For some time I though that having something automatic to remind me about unreviewed/unapplied patches would be a good way to fight this. I know that Eric, Matthias, Dan ... scans mostly everything but well are all getting quite busy and sometime things get forgotten and that's bad for the person who spent time on it and it's bad for the community. Asking to resend old patches tend to inflate the problem it's not a solution and humanly it's poor taste. So I spent most part of the week writing patchchecker, it's an automated tool, working out of the public mail archives and a git checkout, and was tested only with libvirt but should work also with other projects using MHonarc for mail archives and git (or could be made more flexible, yadda yadda ...) http://www.libvirt.org/git/?p=patchchecker.git git clone git://libvirt.org/patchchecker.git Check the README, there is an indexer to run regulary to extract from the mail archives, and a command to try to show intereting things, right now it reports only patches which didn't get any comments or acks. It detects patchset but doesn't fully use them yet. It's crude but it's automatic ! Next step is to so a bit of XML output and XSLT to export the output as a web page on libvirt.org , I will also need to add support for patch versioning i.e. being able to obsolete an older version with a newer mail (or patch set). I guess discussion and patches should be held and sent to this list, which already raises a limitation of this tool: it currently support only one git checkout, and if you send a patch for it it will complain if it doesn't get applied after being ACK'ed :-) BTW one of the results is that we should try to use the formal ACK keyword all the time (capital) when accepting and commiting a patch and I may occasionally send reply mail to threads with ACK or something like COMMITED if the algorithm of the checker get stuck on one patch. I really expect the amount of manual intervention to be limited to this I want the tool to help me, not give me/us more work :-) Hopefully it may prove useful to a few of us, and possibly be adapted to other projects (as usual I take patches :-) Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

3 3

[libvirt] [PATCH] remote: Generate virDomainGetBlockPullInfo
by Matthias Bolte 20 Jun '11

20 Jun '11

It was already generatable but skipped. --- daemon/remote.c | 36 ------------------------------------ src/remote/remote_driver.c | 32 -------------------------------- src/remote/remote_protocol.x | 4 ++-- 3 files changed, 2 insertions(+), 70 deletions(-) diff --git a/daemon/remote.c b/daemon/remote.c index 1914b25..8e0f47d 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -1737,42 +1737,6 @@ cleanup: return rv; } -static int -remoteDispatchDomainGetBlockPullInfo(struct qemud_server *server ATTRIBUTE_UNUSED, - struct qemud_client *client ATTRIBUTE_UNUSED, - virConnectPtr conn, - remote_message_header *hdr ATTRIBUTE_UNUSED, - remote_error * rerr, - remote_domain_get_block_pull_info_args *args, - remote_domain_get_block_pull_info_ret *ret) -{ - virDomainPtr dom = NULL; - virDomainBlockPullInfo tmp; - int rv = -1; - - if (!conn) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not open")); - goto cleanup; - } - - if (!(dom = get_nonnull_domain(conn, args->dom))) - goto cleanup; - - if (virDomainGetBlockPullInfo(dom, args->path, &tmp, args->flags) < 0) - goto cleanup; - ret->cur = tmp.cur; - ret->end = tmp.end; - rv = 0; - -cleanup: - if (rv < 0) - remoteDispatchError(rerr); - if (dom) - virDomainFree(dom); - return rv; -} - - /*-------------------------------------------------------------*/ static int diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index f4b43e0..5db1064 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -2671,38 +2671,6 @@ done: return rv; } -static int remoteDomainGetBlockPullInfo(virDomainPtr domain, - const char *path, - virDomainBlockPullInfoPtr info, - unsigned int flags) -{ - int rv = -1; - remote_domain_get_block_pull_info_args args; - remote_domain_get_block_pull_info_ret ret; - struct private_data *priv = domain->conn->privateData; - - remoteDriverLock(priv); - - make_nonnull_domain(&args.dom, domain); - args.path = (char *)path; - args.flags = flags; - - if (call(domain->conn, priv, 0, REMOTE_PROC_DOMAIN_GET_BLOCK_PULL_INFO, - (xdrproc_t)xdr_remote_domain_get_block_pull_info_args, - (char *)&args, - (xdrproc_t)xdr_remote_domain_get_block_pull_info_ret, - (char *)&ret) == -1) - goto done; - - info->cur = ret.cur; - info->end = ret.end; - rv = 0; - -done: - remoteDriverUnlock(priv); - return rv; -} - /*----------------------------------------------------------------------*/ static virDrvOpenStatus ATTRIBUTE_NONNULL (1) diff --git a/src/remote/remote_protocol.x b/src/remote/remote_protocol.x index 2b9784b..b136e41 100644 --- a/src/remote/remote_protocol.x +++ b/src/remote/remote_protocol.x @@ -1003,7 +1003,7 @@ struct remote_domain_get_block_pull_info_args { unsigned int flags; }; -struct remote_domain_get_block_pull_info_ret { +struct remote_domain_get_block_pull_info_ret { /* insert@2 */ unsigned hyper cur; unsigned hyper end; }; @@ -2409,7 +2409,7 @@ enum remote_procedure { REMOTE_PROC_DOMAIN_BLOCK_PULL_ALL = 230, /* autogen autogen */ REMOTE_PROC_DOMAIN_BLOCK_PULL_ABORT = 231, /* autogen autogen */ - REMOTE_PROC_DOMAIN_GET_BLOCK_PULL_INFO = 232, /* skipgen skipgen */ + REMOTE_PROC_DOMAIN_GET_BLOCK_PULL_INFO = 232, /* autogen autogen */ REMOTE_PROC_DOMAIN_EVENT_BLOCK_PULL = 233 /* skipgen skipgen */ /* -- 1.7.0.4

2 2

[libvirt] [PATCH] remote: Handle functions thar return optional strings
by Matthias Bolte 20 Jun '11

20 Jun '11

--- daemon/remote.c | 64 ------------------------------------------ daemon/remote_generator.pl | 28 ++++++++++++++++++ src/remote/remote_protocol.x | 2 +- 3 files changed, 29 insertions(+), 65 deletions(-) diff --git a/daemon/remote.c b/daemon/remote.c index 8e0f47d..f2e193d 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -1485,70 +1485,6 @@ cleanup: } static int -remoteDispatchDomainScreenshot(struct qemud_server *server ATTRIBUTE_UNUSED, - struct qemud_client *client, - virConnectPtr conn, - remote_message_header *hdr, - remote_error *rerr, - remote_domain_screenshot_args *args, - remote_domain_screenshot_ret *ret) -{ - int rv = -1; - struct qemud_client_stream *stream = NULL; - virDomainPtr dom = NULL; - char *mime, **mime_p; - - if (!conn) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not open")); - goto cleanup; - } - - ret->mime = NULL; - - if (!(dom = get_nonnull_domain (conn, args->dom))) - goto cleanup; - - if (!(stream = remoteCreateClientStream(conn, hdr))) - goto cleanup; - - if (!(mime = virDomainScreenshot(dom, stream->st, args->screen, args->flags))) - goto cleanup; - - if (remoteAddClientStream(client, stream, 1) < 0) { - virStreamAbort(stream->st); - goto cleanup; - } - - if (VIR_ALLOC(mime_p) < 0) { - virReportOOMError(); - goto cleanup; - } - - *mime_p = strdup(mime); - if (*mime_p == NULL) { - virReportOOMError(); - VIR_FREE(mime_p); - goto cleanup; - } - - ret->mime = mime_p; - - rv = 0; - -cleanup: - if (rv < 0) - remoteDispatchError(rerr); - VIR_FREE(mime); - if (dom) - virDomainFree(dom); - if (stream && rv != 0) { - virStreamAbort(stream->st); - remoteFreeClientStream(client, stream); - } - return rv; -} - -static int remoteDispatchNodeGetCPUStats (struct qemud_server *server ATTRIBUTE_UNUSED, struct qemud_client *client ATTRIBUTE_UNUSED, virConnectPtr conn, diff --git a/daemon/remote_generator.pl b/daemon/remote_generator.pl index bcf5fd4..d9d9ded 100755 --- a/daemon/remote_generator.pl +++ b/daemon/remote_generator.pl @@ -351,6 +351,7 @@ elsif ($opt_b) { my @optionals_list = (); my @getters_list = (); my @args_list = (); + my @prepare_ret_list = (); my @ret_list = (); my @free_list = (); my @free_list_on_error = ("remoteDispatchError(rerr);"); @@ -551,6 +552,27 @@ elsif ($opt_b) { $single_ret_var = $1; $single_ret_by_ref = 0; $single_ret_check = " == NULL"; + } elsif ($ret_member =~ m/^remote_string (\S+);/) { + push(@vars_list, "char *$1 = NULL"); + push(@vars_list, "char **$1_p = NULL"); + push(@ret_list, "ret->$1 = $1_p;"); + push(@free_list, " VIR_FREE($1);"); + push(@free_list_on_error, "VIR_FREE($1_p);"); + push(@prepare_ret_list, + "if (VIR_ALLOC($1_p) < 0) {\n" . + " virReportOOMError();\n" . + " goto cleanup;\n" . + " }\n" . + " \n" . + " *$1_p = strdup($1);\n" . + " if (*$1_p == NULL) {\n" . + " virReportOOMError();\n" . + " goto cleanup;\n" . + " }\n"); + + $single_ret_var = $1; + $single_ret_by_ref = 0; + $single_ret_check = " == NULL"; } elsif ($ret_member =~ m/^remote_nonnull_(domain|network|storage_pool|storage_vol|interface|node_device|secret|nwfilter|domain_snapshot) (\S+);/) { my $type_name = name_to_ProcName($1); @@ -864,6 +886,12 @@ elsif ($opt_b) { print "\n"; } + if (@prepare_ret_list) { + print " "; + print join("\n ", @prepare_ret_list); + print "\n"; + } + if (@ret_list) { print " "; print join("\n ", @ret_list); diff --git a/src/remote/remote_protocol.x b/src/remote/remote_protocol.x index b136e41..0aa3dfb 100644 --- a/src/remote/remote_protocol.x +++ b/src/remote/remote_protocol.x @@ -2386,7 +2386,7 @@ enum remote_procedure { REMOTE_PROC_STORAGE_VOL_DOWNLOAD = 209, /* autogen autogen | readstream@1 */ REMOTE_PROC_DOMAIN_INJECT_NMI = 210, /* autogen autogen */ - REMOTE_PROC_DOMAIN_SCREENSHOT = 211, /* skipgen autogen | readstream@1 */ + REMOTE_PROC_DOMAIN_SCREENSHOT = 211, /* autogen autogen | readstream@1 */ REMOTE_PROC_DOMAIN_GET_STATE = 212, /* skipgen skipgen */ REMOTE_PROC_DOMAIN_MIGRATE_BEGIN3 = 213, /* skipgen skipgen */ REMOTE_PROC_DOMAIN_MIGRATE_PREPARE3 = 214, /* skipgen skipgen */ -- 1.7.0.4

2 2

[libvirt] [PATCH] remote generator: Allow to annotate arrays with typecasts
by Matthias Bolte 20 Jun '11

20 Jun '11

Removes special case code from the generator and handle additional methods. The generated version of remoteDispatchDomainPinVcpu(Flags) has no length check, but this check was useless anyway as it was applied to data that was already deserialized from its XDR form. --- daemon/remote.c | 82 -------------------------------------- daemon/remote_generator.pl | 90 ++++++++++++++++++++++++++++++------------ src/remote/remote_protocol.x | 17 +++++--- 3 files changed, 74 insertions(+), 115 deletions(-) diff --git a/daemon/remote.c b/daemon/remote.c index f2e193d..37fbed0 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -1269,88 +1269,6 @@ cleanup: } static int -remoteDispatchDomainPinVcpu(struct qemud_server *server ATTRIBUTE_UNUSED, - struct qemud_client *client ATTRIBUTE_UNUSED, - virConnectPtr conn, - remote_message_header *hdr ATTRIBUTE_UNUSED, - remote_error *rerr, - remote_domain_pin_vcpu_args *args, - void *ret ATTRIBUTE_UNUSED) -{ - virDomainPtr dom = NULL; - int rv = -1; - - if (!conn) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not open")); - goto cleanup; - } - - if (!(dom = get_nonnull_domain(conn, args->dom))) - goto cleanup; - - if (args->cpumap.cpumap_len > REMOTE_CPUMAP_MAX) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("cpumap_len > REMOTE_CPUMAP_MAX")); - goto cleanup; - } - - if (virDomainPinVcpu(dom, args->vcpu, - (unsigned char *) args->cpumap.cpumap_val, - args->cpumap.cpumap_len) < 0) - goto cleanup; - - rv = 0; - -cleanup: - if (rv < 0) - remoteDispatchError(rerr); - if (dom) - virDomainFree(dom); - return rv; -} - -static int -remoteDispatchDomainPinVcpuFlags(struct qemud_server *server ATTRIBUTE_UNUSED, - struct qemud_client *client ATTRIBUTE_UNUSED, - virConnectPtr conn, - remote_message_header *hdr ATTRIBUTE_UNUSED, - remote_error *rerr, - remote_domain_pin_vcpu_flags_args *args, - void *ret ATTRIBUTE_UNUSED) -{ - virDomainPtr dom = NULL; - int rv = -1; - - if (!conn) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not open")); - goto cleanup; - } - - if (!(dom = get_nonnull_domain(conn, args->dom))) - goto cleanup; - - if (args->cpumap.cpumap_len > REMOTE_CPUMAP_MAX) { - virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("cpumap_len > REMOTE_CPUMAP_MAX")); - goto cleanup; - } - - if (virDomainPinVcpuFlags(dom, - args->vcpu, - (unsigned char *) args->cpumap.cpumap_val, - args->cpumap.cpumap_len, - args->flags) < 0) - goto cleanup; - - rv = 0; - -cleanup: - if (rv < 0) - remoteDispatchError(rerr); - if (dom) - virDomainFree(dom); - return rv; -} - -static int remoteDispatchDomainGetMemoryParameters(struct qemud_server *server ATTRIBUTE_UNUSED, struct qemud_client *client diff --git a/daemon/remote_generator.pl b/daemon/remote_generator.pl index d9d9ded..146c56b 100755 --- a/daemon/remote_generator.pl +++ b/daemon/remote_generator.pl @@ -403,20 +403,32 @@ elsif ($opt_b) { " virDomainSnapshotFree(snapshot);\n" . " if (dom)\n" . " virDomainFree(dom);"); - } elsif ($args_member =~ m/^(remote_string|remote_nonnull_string|remote_uuid|opaque) (\S+)<\S+>;/) { + } elsif ($args_member =~ m/^(?:remote_string|remote_uuid) (\S+)<\S+>;/) { if (! @args_list) { push(@args_list, "conn"); } - if ($call->{ProcName} eq "SecretSetValue") { - push(@args_list, "(const unsigned char *)args->$2.$2_val"); - } elsif ($call->{ProcName} eq "CPUBaseline") { - push(@args_list, "(const char **)args->$2.$2_val"); - } else { - push(@args_list, "args->$2.$2_val"); + push(@args_list, "args->$1.$1_val"); + push(@args_list, "args->$1.$1_len"); + } elsif ($args_member =~ m/^(?:opaque|remote_nonnull_string) (\S+)<\S+>;(.*)$/) { + if (! @args_list) { + push(@args_list, "conn"); + } + + my $cast = ""; + my $arg_name = $1; + my $annotation = $2; + + if ($annotation ne "") { + if ($annotation =~ m/\s*\/\*\s*(.*)\s*\*\//) { + $cast = $1; + } else { + die "malformed cast annotation for argument: $args_member"; + } } - push(@args_list, "args->$2.$2_len"); + push(@args_list, "${cast}args->$arg_name.${arg_name}_val"); + push(@args_list, "args->$arg_name.${arg_name}_len"); } elsif ($args_member =~ m/^(?:unsigned )?int (\S+)<\S+>;/) { if (! @args_list) { push(@args_list, "conn"); @@ -998,37 +1010,63 @@ elsif ($opt_k) { } elsif ($args_member =~ m/^remote_string (\S+);/) { push(@args_list, "const char *$1"); push(@setters_list, "args.$1 = $1 ? (char **)&$1 : NULL;"); - } elsif ($args_member =~ m/^remote_nonnull_string (\S+)<(\S+)>;/) { - push(@args_list, "const char **$1"); - push(@args_list, "unsigned int ${1}len"); - push(@setters_list, "args.$1.${1}_val = (char **)$1;"); - push(@setters_list, "args.$1.${1}_len = ${1}len;"); - push(@args_check_list, { name => "\"$1\"", arg => "${1}len", limit => $2 }); + } elsif ($args_member =~ m/^remote_nonnull_string (\S+)<(\S+)>;(.*)$/) { + my $type_name = "const char **"; + my $arg_name = $1; + my $limit = $2; + my $annotation = $3; + + if ($annotation ne "") { + if ($annotation =~ m/\s*\/\*\s*$(.*)$\s*\*\//) { + $type_name = $1; + } else { + die "malformed cast annotation for argument: $args_member"; + } + } + + push(@args_list, "$type_name$arg_name"); + push(@args_list, "unsigned int ${arg_name}len"); + push(@setters_list, "args.$arg_name.${arg_name}_val = (char **)$arg_name;"); + push(@setters_list, "args.$arg_name.${arg_name}_len = ${arg_name}len;"); + push(@args_check_list, { name => "\"$arg_name\"", arg => "${arg_name}len", limit => $2 }); } elsif ($args_member =~ m/^remote_nonnull_string (\S+);/) { push(@args_list, "const char *$1"); push(@setters_list, "args.$1 = (char *)$1;"); - } elsif ($args_member =~ m/^(remote_string|opaque) (\S+)<(\S+)>;/) { - my $type_name = $1; - my $arg_name = $2; - my $limit = $3; + } elsif ($args_member =~ m/^opaque (\S+)<(\S+)>;(.*)$/) { + my $type_name = "const char *"; + my $arg_name = $1; + my $limit = $2; + my $annotation = $3; + + if ($annotation ne "") { + if ($annotation =~ m/\s*\/\*\s*$(.*)$\s*\*\//) { + $type_name = $1; + } else { + die "malformed cast annotation for argument: $args_member"; + } + } + + push(@args_list, "$type_name$arg_name"); if ($call->{ProcName} eq "SecretSetValue") { - push(@args_list, "const unsigned char *$arg_name"); + # SPECIAL: virSecretSetValue uses size_t instead of int push(@args_list, "size_t ${arg_name}len"); - } elsif ($call->{ProcName} eq "DomainPinVcpu") { - push(@args_list, "unsigned char *$arg_name"); - push(@args_list, "int ${arg_name}len"); - } elsif ($call->{ProcName} eq "DomainPinVcpuFlags") { - push(@args_list, "unsigned char *$arg_name"); - push(@args_list, "int ${arg_name}len"); } else { - push(@args_list, "const char *$arg_name"); push(@args_list, "int ${arg_name}len"); } push(@setters_list, "args.$arg_name.${arg_name}_val = (char *)$arg_name;"); push(@setters_list, "args.$arg_name.${arg_name}_len = ${arg_name}len;"); push(@args_check_list, { name => "\"$arg_name\"", arg => "${arg_name}len", limit => $limit }); + } elsif ($args_member =~ m/^remote_string (\S+)<(\S+)>;/) { + my $arg_name = $1; + my $limit = $2; + + push(@args_list, "const char *$arg_name"); + push(@args_list, "int ${arg_name}len"); + push(@setters_list, "args.$arg_name.${arg_name}_val = (char *)$arg_name;"); + push(@setters_list, "args.$arg_name.${arg_name}_len = ${arg_name}len;"); + push(@args_check_list, { name => "\"$arg_name\"", arg => "${arg_name}len", limit => $limit }); } elsif ($args_member =~ m/^((?:unsigned )?int) (\S+)<(\S+)>;/) { my $type_name = $1; my $arg_name = $2; diff --git a/src/remote/remote_protocol.x b/src/remote/remote_protocol.x index 0aa3dfb..a77fd81 100644 --- a/src/remote/remote_protocol.x +++ b/src/remote/remote_protocol.x @@ -357,7 +357,10 @@ struct remote_node_get_memory_stats { * If the 'remote_CALL_ret' maps to a struct in the public API then it is * also filled via call-by-reference and must be annotated with a * insert@<offset> comment to indicate the offset in the parameter list of - * the function to be called. */ + * the function to be called. + * + * Dynamic opaque and remote_nonnull_string arrays can be annotated with an + * optional typecast */ struct remote_open_args { /* NB. "name" might be NULL although in practice you can't @@ -890,13 +893,13 @@ struct remote_domain_get_vcpus_flags_ret { struct remote_domain_pin_vcpu_args { remote_nonnull_domain dom; unsigned int vcpu; - opaque cpumap<REMOTE_CPUMAP_MAX>; + opaque cpumap<REMOTE_CPUMAP_MAX>; /* (unsigned char *) */ }; struct remote_domain_pin_vcpu_flags_args { remote_nonnull_domain dom; unsigned int vcpu; - opaque cpumap<REMOTE_CPUMAP_MAX>; + opaque cpumap<REMOTE_CPUMAP_MAX>; /* (unsigned char *) */ unsigned int flags; }; @@ -1701,7 +1704,7 @@ struct remote_secret_get_xml_desc_ret { struct remote_secret_set_value_args { remote_nonnull_secret secret; - opaque value<REMOTE_SECRET_VALUE_MAX>; + opaque value<REMOTE_SECRET_VALUE_MAX>; /* (const unsigned char *) */ unsigned int flags; }; @@ -1819,7 +1822,7 @@ struct remote_cpu_compare_ret { struct remote_cpu_baseline_args { - remote_nonnull_string xmlCPUs<REMOTE_CPU_BASELINE_MAX>; + remote_nonnull_string xmlCPUs<REMOTE_CPU_BASELINE_MAX>; /* (const char **) */ unsigned int flags; }; @@ -2182,7 +2185,7 @@ enum remote_procedure { REMOTE_PROC_DOMAIN_LOOKUP_BY_NAME = 23, /* autogen autogen */ REMOTE_PROC_DOMAIN_LOOKUP_BY_UUID = 24, /* autogen autogen */ REMOTE_PROC_NUM_OF_DEFINED_DOMAINS = 25, /* autogen autogen */ - REMOTE_PROC_DOMAIN_PIN_VCPU = 26, /* skipgen autogen */ + REMOTE_PROC_DOMAIN_PIN_VCPU = 26, /* autogen autogen */ REMOTE_PROC_DOMAIN_REBOOT = 27, /* autogen autogen */ REMOTE_PROC_DOMAIN_RESUME = 28, /* autogen autogen */ REMOTE_PROC_DOMAIN_SET_AUTOSTART = 29, /* autogen autogen */ @@ -2401,7 +2404,7 @@ enum remote_procedure { REMOTE_PROC_INTERFACE_CHANGE_ROLLBACK = 222, /* autogen autogen */ REMOTE_PROC_DOMAIN_GET_SCHEDULER_PARAMETERS_FLAGS = 223, /* skipgen autogen */ REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224, /* skipgen skipgen */ - REMOTE_PROC_DOMAIN_PIN_VCPU_FLAGS = 225, /* skipgen autogen */ + REMOTE_PROC_DOMAIN_PIN_VCPU_FLAGS = 225, /* autogen autogen */ REMOTE_PROC_DOMAIN_SEND_KEY = 226, /* autogen autogen */ REMOTE_PROC_NODE_GET_CPU_STATS = 227, /* skipgen skipgen */ REMOTE_PROC_NODE_GET_MEMORY_STATS = 228, /* skipgen skipgen */ -- 1.7.0.4

2 2

[libvirt] [PATCH] Fix preprocessor indentation in nodeinfo.c
by Matthias Bolte 20 Jun '11

20 Jun '11

--- I'm pushing this under the build breaker rule. src/nodeinfo.c | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index c6dbb84..9e0f906 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -685,23 +685,23 @@ int nodeGetMemoryStats(virConnectPtr conn ATTRIBUTE_UNUSED, return -1; } } else { -#if HAVE_NUMACTL +# if HAVE_NUMACTL if (numa_available() < 0) { -#endif +# endif nodeReportError(VIR_ERR_NO_SUPPORT, "%s", _("NUMA not supported on this host")); return -1; -#if HAVE_NUMACTL +# if HAVE_NUMACTL } -#endif +# endif -#if HAVE_NUMACTL +# if HAVE_NUMACTL if (cellNum > numa_max_node()) { nodeReportError(VIR_ERR_INVALID_ARG, "%s", _("Invalid cell number")); return -1; } -#endif +# endif if (virAsprintf(&meminfo_path, "%s/node%d/meminfo", NODE_SYS_PATH, cellNum) < 0) { -- 1.7.0.4

2 2

Re: [libvirt] [libvirt-php] - zend_mm_heap corrupted
by Michal Novotny 20 Jun '11

20 Jun '11

Hi Henrik, what exactly did you to do to get the error page? I never run into those in any version I'm having in libvirt-php git repository since I always check but it may be Apache & PHP version connected and I'm running it as a module. Could you please provide me more information about this or at least provide me the results of `make check` command in the libvirt-php directory as cloned from git? Please note that to try to `make check` you need to install it first so you need to do `make`, then `(sudo) make install` and then `make check`. The `make check` command is running the tests and you shouldn't be seeing any segmentation fault there otherwise there's some bug in the code which obviously seems to be reproducible only on your environment since it's not reproducible on my environment at all. All the tests pass in my case: $ make check ... Making check in tests make[1]: Entering directory `tests' ./runtests.sh Test test-connect.phpt has been completed successfully Test test-version-check.phpt has been completed successfully Test test-version-get.phpt has been completed successfully Test test-domain-define-undefine.phpt has been completed successfully Test test-domain-define-create-destroy.phpt has been completed successfully Test test-domain-create.phpt has been completed successfully Test test-domain-create-and-get-xpath.phpt has been completed successfully Test test-domain-create-and-coredump.phpt has been completed successfully Test test-domain-snapshot.phpt has been completed successfully All tests passed successfully ... $ Could you please provide me with the Apache logs as well and also the Apache and PHP version numbers? You mention Drupal, did you integrate libvirt-php functionality into Drupal? The backtraces using xdebug should be sufficient. Please provide me with them. Thanks, Michal On 06/18/2011 06:51 PM, Henrik Jönsson wrote: > > I basicly just got an erronous page load and checked the apache logs. > After some further testing it seems the problem only occurs when ran > within Drupal and not when ran as a standalone script. I will do some > further testing in a different cms to see if it is Drupal specific. I > can also do backtraces with xdebug and coredump apache if that will help. > > Best Regards > > /Henrik > Den 15 jun 2011 09:03 skrev "Michal Novotny" <minovotn(a)redhat.com > <mailto:minovotn@redhat.com>>: > > > > Hi Henrik, > > thanks a lot for your observations. How did you find out about the PHP > > memory usage and about zend_mm_heap corrupted at the first place? > > > > This could help me debug this issue. > > > > Thanks, > > Michal > > > > On 06/12/2011 06:05 PM, Henrik Jönsson wrote: > > > [libvirt-php] > > > > > > Hello! > > > > > > I am using Libvirt-php for a private project, i do however have a > > > problem that seems to occur when i call any function that returns a > > > domain resource. I have gotten 3 different errors so far. First i had > > > a problem with php memory usage going extreme(2gb), then i have had > > > some apache segmentation fault ending up in apache killing the > > > process. I have also had zend_mm_heap corrupted which i think is the > > > most common one. > > > > > > I have tried with Ubuntu 10.04, 10.10 and 11.04 repository versions of > > > Apache. Other than that i am using libvirt version 0.9.1 and > > > libvirt-php version 0.4.1. This is within the contexts of Drupal as i > > > am using the code in a Drupal module. > > > > > > I have checked out the source code for libvirt-php but since i am no > > > C/C++ developer i have a hard time figuring out what the issue might > > > be. Some help would be greatly appreciated. :) > > > > > > Best Regards! > > > > > > /Henrik > > > > > > > > > -- > > > libvir-list mailing list > > > libvir-list(a)redhat.com <mailto:libvir-list@redhat.com> > > > https://www.redhat.com/mailman/listinfo/libvir-list > > > > > > -- > > Michal Novotny <minovotn(a)redhat.com <mailto:minovotn@redhat.com>>, RHCE > > Virtualization Team (xen userspace), Red Hat > > > -- Michal Novotny <minovotn(a)redhat.com>, RHCE Virtualization Team (xen userspace), Red Hat

1 0

[libvirt] [PATCH] Move XenAPI driver to correct spec file section
by Matthias Bolte 20 Jun '11

20 Jun '11

The XenAPI driver works like the ESX and PHyp driver by using its own HTTPS based remote protocol. --- libvirt.spec.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/libvirt.spec.in b/libvirt.spec.in index 97ebd65..75b145a 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -44,13 +44,13 @@ %define with_lxc 0%{!?_without_lxc:%{server_drivers}} %define with_vbox 0%{!?_without_vbox:%{server_drivers}} %define with_uml 0%{!?_without_uml:%{server_drivers}} -%define with_xenapi 0%{!?_without_xenapi:%{server_drivers}} %define with_libxl 0%{!?_without_libxl:%{server_drivers}} %define with_vmware 0%{!?_without_vmware:%{server_drivers}} # Then the hypervisor drivers that talk a native remote protocol %define with_phyp 0%{!?_without_phyp:1} %define with_esx 0%{!?_without_esx:1} +%define with_xenapi 0%{!?_without_xenapi:1} # Then the secondary host drivers %define with_network 0%{!?_without_network:%{server_drivers}} -- 1.7.0.4

2 2