- Devel - libvirt List Archives

[libvirt] [PATCH] Keep track of guest paused state after disk IO errors
by Daniel P. Berrange 19 Mar '10

19 Mar '10

With the QMP mode monitor, it is possible to get a notification when the guest is paused indirectly (eg as result of a disk IO error). This patch enables such reporting and when receiving an error updates libvirt's view of the guest to indicate that it is now paused. It also emits an event VIR_DOMAIN_EVENT_SUSPENDED with a detail of: VIR_DOMAIN_EVENT_SUSPENDED_IOERROR NB this patch does not make the error behaviour configurable. It just copes with the pause transition, if QEMU is setup to do that by default (current upstream now works this way for disks). * include/libvirt/libvirt.h.in: Add VIR_DOMAIN_EVENT_SUSPENDED_IOERROR * src/qemu/qemu_driver.c: Update VM state to paused when IO error occurrs * src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h, src/qemu/qemu_monitor_json.c: Wire up handlers for disk IO errors --- include/libvirt/libvirt.h.in | 1 + src/qemu/qemu_driver.c | 42 ++++++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_monitor.c | 18 ++++++++++++++++++ src/qemu/qemu_monitor.h | 16 ++++++++++++++++ src/qemu/qemu_monitor_json.c | 34 ++++++++++++++++++++++++++++++++++ 5 files changed, 111 insertions(+), 0 deletions(-) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index 260505e..b7a6922 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -1361,6 +1361,7 @@ typedef enum { typedef enum { VIR_DOMAIN_EVENT_SUSPENDED_PAUSED = 0, /* Normal suspend due to admin pause */ VIR_DOMAIN_EVENT_SUSPENDED_MIGRATED = 1, /* Suspended for offline migration */ + VIR_DOMAIN_EVENT_SUSPENDED_IOERROR = 2, /* Suspended due to a disk I/O error */ } virDomainEventSuspendedDetailType; /** diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 8766ca2..ae19097 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -824,9 +824,51 @@ cleanup: return ret; } + +static int +qemuHandleDiskIOError(qemuMonitorPtr mon ATTRIBUTE_UNUSED, + virDomainObjPtr vm, + int action, + const char *diskalias ATTRIBUTE_UNUSED) +{ + struct qemud_driver *driver = qemu_driver; + virDomainEventPtr event = NULL; + + VIR_DEBUG("Received IO error on %p '%s': action=%d disk=%s", vm, vm->def->name, action, diskalias); + + if (action != QEMU_MONITOR_DISK_IO_ERROR_STOP) + return 0; + + virDomainObjLock(vm); + + if (action == QEMU_MONITOR_DISK_IO_ERROR_STOP && + vm->state == VIR_DOMAIN_RUNNING) { + VIR_DEBUG("Transitioned guest %s to paused state", vm->def->name); + + vm->state = VIR_DOMAIN_PAUSED; + event = virDomainEventNewFromObj(vm, + VIR_DOMAIN_EVENT_SUSPENDED, + VIR_DOMAIN_EVENT_SUSPENDED_IOERROR); + + if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) + VIR_WARN("Unable to save status on vm %s after IO error", vm->def->name); + } + + virDomainObjUnlock(vm); + + if (event) { + qemuDriverLock(driver); + qemuDomainEventQueue(driver, event); + qemuDriverUnlock(driver); + } + return 0; +} + + static qemuMonitorCallbacks monitorCallbacks = { .eofNotify = qemuHandleMonitorEOF, .diskSecretLookup = findVolumeQcowPassphrase, + .diskIOError = qemuHandleDiskIOError, }; static int diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index b88532c..7205001 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -791,6 +791,24 @@ int qemuMonitorEmitStop(qemuMonitorPtr mon) } +int qemuMonitorEmitDiskIOError(qemuMonitorPtr mon, + int action, + const char *deviceAlias) +{ + int ret = -1; + VIR_DEBUG("mon=%p action=%d deviceAlias=%s", mon, action, deviceAlias); + + qemuMonitorRef(mon); + qemuMonitorUnlock(mon); + if (mon->cb && mon->cb->diskIOError) + ret = mon->cb->diskIOError(mon, mon->vm, action, deviceAlias); + qemuMonitorLock(mon); + qemuMonitorUnref(mon); + + return ret; +} + + int qemuMonitorSetCapabilities(qemuMonitorPtr mon) { int ret; diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index 0ac3957..75a2853 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -60,6 +60,14 @@ struct _qemuMonitorMessage { void *passwordOpaque; }; +typedef enum { + QEMU_MONITOR_DISK_IO_ERROR_STOP, + QEMU_MONITOR_DISK_IO_ERROR_REPORT, + QEMU_MONITOR_DISK_IO_ERROR_IGNORE, + + QEMU_MONITOR_DISK_IO_ERROR_LAST +} qemuMonitorDiskIOErrorAction; + typedef struct _qemuMonitorCallbacks qemuMonitorCallbacks; typedef qemuMonitorCallbacks *qemuMonitorCallbacksPtr; struct _qemuMonitorCallbacks { @@ -86,6 +94,11 @@ struct _qemuMonitorCallbacks { virDomainObjPtr vm); int (*domainStop)(qemuMonitorPtr mon, virDomainObjPtr vm); + + int (*diskIOError)(qemuMonitorPtr mon, + virDomainObjPtr vm, + int actOBion, + const char *diskAlias); }; @@ -122,6 +135,9 @@ int qemuMonitorEmitShutdown(qemuMonitorPtr mon); int qemuMonitorEmitReset(qemuMonitorPtr mon); int qemuMonitorEmitPowerdown(qemuMonitorPtr mon); int qemuMonitorEmitStop(qemuMonitorPtr mon); +int qemuMonitorEmitDiskIOError(qemuMonitorPtr mon, + int action, + const char *deviceAlias); int qemuMonitorStartCPUs(qemuMonitorPtr mon, virConnectPtr conn); diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 7b45594..8b3cda1 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -49,6 +49,7 @@ static void qemuMonitorJSONHandleShutdown(qemuMonitorPtr mon, virJSONValuePtr da static void qemuMonitorJSONHandleReset(qemuMonitorPtr mon, virJSONValuePtr data); static void qemuMonitorJSONHandlePowerdown(qemuMonitorPtr mon, virJSONValuePtr data); static void qemuMonitorJSONHandleStop(qemuMonitorPtr mon, virJSONValuePtr data); +static void qemuMonitorJSONHandleDiskIOError(qemuMonitorPtr mon, virJSONValuePtr data); struct { const char *type; @@ -58,6 +59,7 @@ struct { { "RESET", qemuMonitorJSONHandleReset, }, { "POWERDOWN", qemuMonitorJSONHandlePowerdown, }, { "STOP", qemuMonitorJSONHandleStop, }, + { "BLOCK_IO_ERROR", qemuMonitorJSONHandleDiskIOError, }, }; @@ -496,6 +498,38 @@ static void qemuMonitorJSONHandleStop(qemuMonitorPtr mon, virJSONValuePtr data A } +VIR_ENUM_DECL(qemuMonitorDiskIOErrorAction) +VIR_ENUM_IMPL(qemuMonitorDiskIOErrorAction, QEMU_MONITOR_DISK_IO_ERROR_LAST, + "stop", "report", "ignore"); + + +static void qemuMonitorJSONHandleDiskIOError(qemuMonitorPtr mon, virJSONValuePtr data) +{ + const char *device; + const char *action; + int actionID; + + /* Throughout here we try our best to carry on upon errors, + since its imporatant to get as much info as possible out + to the application */ + + if ((action = virJSONValueObjectGetString(data, "action")) == NULL) { + VIR_WARN0("Missing action in disk io error event"); + action = "ignore"; + } + + if ((device = virJSONValueObjectGetString(data, "device")) == NULL) { + VIR_WARN0("missing device in disk io error event"); + } + + if ((actionID = qemuMonitorDiskIOErrorActionTypeFromString(action)) < 0) { + VIR_WARN("unknown disk io error action '%s'", action); + actionID = QEMU_MONITOR_DISK_IO_ERROR_IGNORE; + } + + qemuMonitorEmitDiskIOError(mon, actionID, device); +} + int qemuMonitorJSONSetCapabilities(qemuMonitorPtr mon) { -- 1.6.2.5

2 1

[libvirt] [PATCH] qemu: Fix FD leak in qemudStartVMDaemon
by Matthias Bolte 19 Mar '10

19 Mar '10

The logfile FD is dup2'ed in __virExec in the child. The FD needs to be closed in the parent, otherwise it leaks. --- src/qemu/qemu_driver.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index c8f3a15..fbb1275 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2963,6 +2963,9 @@ static int qemudStartVMDaemon(virConnectPtr conn, if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) goto abort; + if (logfile != -1) + close(logfile); + return 0; cleanup: -- 1.6.3.3

2 4

[libvirt] [PATCH 14/14] Extension for iptables rules
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds support for L3/L4 filtering using iptables. This adds support for 'tcp', 'udp', 'icmp', 'igmp', 'sctp' etc. filtering. As mentioned in the introduction, a .c file provided by this patch is #include'd into a .c file. This will need work, but should be alright for review. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 2

[libvirt] [PATCH 5/14] Implementation of the public API
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds the implementation of the public API for the network filtering (ACL) extensions to libvirt.c . Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 2

[libvirt] [PATCH 2/14] Add build support
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds build support for the network filtering framework. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 2

[libvirt] [PATCH] util: ensure virMutexInit is not recursive
by Eric Blake 18 Mar '10

18 Mar '10

POSIX states that creation of a mutex with default attributes is unspecified whether the mutex is recursive or non-recursive. We specifically want non-recursive (deadlock is desirable in flushing out coding bugs that used our mutex incorrectly). * src/util/threads-pthread.c (virMutexInit): Specifically request non-recursive mutex, rather than relying on unspecified default. --- src/util/threads-pthread.c | 7 +++++-- 1 files changed, 5 insertions(+), 2 deletions(-) diff --git a/src/util/threads-pthread.c b/src/util/threads-pthread.c index ad42483..2f0746b 100644 --- a/src/util/threads-pthread.c +++ b/src/util/threads-pthread.c @@ -1,7 +1,7 @@ /* * threads-pthread.c: basic thread synchronization primitives * - * Copyright (C) 2009 Red Hat, Inc. + * Copyright (C) 2009-2010 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -36,7 +36,10 @@ void virThreadOnExit(void) int virMutexInit(virMutexPtr m) { int ret; - if ((ret = pthread_mutex_init(&m->lock, NULL)) != 0) { + pthread_mutexattr_t attr; + pthread_mutexattr_init(&attr); + pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_NORMAL); + if ((ret = pthread_mutex_init(&m->lock, &attr)) != 0) { errno = ret; return -1; } -- 1.6.6.1

3 2

[libvirt] [PATCH] maint: make Red Hat copyright notices consistent
by Eric Blake 18 Mar '10

18 Mar '10

Spell out 'Red Hat, Inc.': git grep -i 'Copyright.*Red Hat' | grep -v Inc Include (C) consistently: git grep -i 'Copyright [^(].*Red Hat' * src/lxc/lxc_container.c: Update copyright formatting. * src/node_device/node_device_udev.c: Likewise. * src/node_device/node_device_udev.h: Likewise. * src/xen/xend_internal.h: Likewise. * src/xen/xm_internal.c: Likewise. * src/xen/xm_internal.h: Likewise. * tests/xmconfigtest.c: Likewise. * tests/object-locking.ml: Likewise. --- src/lxc/lxc_container.c | 4 ++-- src/node_device/node_device_udev.c | 2 +- src/node_device/node_device_udev.h | 2 +- src/xen/xend_internal.h | 2 +- src/xen/xm_internal.c | 2 +- src/xen/xm_internal.h | 2 +- tests/object-locking.ml | 2 +- tests/xmconfigtest.c | 2 +- 8 files changed, 9 insertions(+), 9 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 68b4656..706c796 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1,6 +1,6 @@ /* - * Copyright IBM Corp. 2008 - * Copyright Red Hat 2008-2009 + * Copyright (C) 2008-2010 Red Hat, Inc. + * Copyright (C) 2008 IBM Corp. * * lxc_container.c: file description * diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index 11e27e0..e3114fa 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1,7 +1,7 @@ /* * node_device_udev.c: node device enumeration - libudev implementation * - * Copyright (C) 2009-2010 Red Hat + * Copyright (C) 2009-2010 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public diff --git a/src/node_device/node_device_udev.h b/src/node_device/node_device_udev.h index 8367494..cdaa142 100644 --- a/src/node_device/node_device_udev.h +++ b/src/node_device/node_device_udev.h @@ -1,7 +1,7 @@ /* * node_device_udev.h: node device enumeration - libudev implementation * - * Copyright (C) 2009 Red Hat + * Copyright (C) 2009-2010 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public diff --git a/src/xen/xend_internal.h b/src/xen/xend_internal.h index 483253f..028a293 100644 --- a/src/xen/xend_internal.h +++ b/src/xen/xend_internal.h @@ -1,12 +1,12 @@ /* * xend_internal.h * + * Copyright (C) 2006-2008, 2010 Red Hat, Inc. * Copyright (C) 2005,2006 * * Anthony Liguori <aliguori(a)us.ibm.com> * Daniel Veillard <veillard(a)redhat.com> * - * Copyright 2006-2008 Red Hat * * This file is subject to the terms and conditions of the GNU Lesser General * Public License. See the file COPYING in the main directory of this archive diff --git a/src/xen/xm_internal.c b/src/xen/xm_internal.c index 3d4624d..74bf0b6 100644 --- a/src/xen/xm_internal.c +++ b/src/xen/xm_internal.c @@ -1,7 +1,7 @@ /* * xm_internal.h: helper routines for dealing with inactive domains * - * Copyright (C) 2006-2007, 2009, 2010 Red Hat + * Copyright (C) 2006-2007, 2009-2010 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or diff --git a/src/xen/xm_internal.h b/src/xen/xm_internal.h index 37132ef..3ad3456 100644 --- a/src/xen/xm_internal.h +++ b/src/xen/xm_internal.h @@ -1,7 +1,7 @@ /* * xm_internal.h: helper routines for dealing with inactive domains * - * Copyright (C) 2006-2007 Red Hat + * Copyright (C) 2006-2007, 2010 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or diff --git a/tests/object-locking.ml b/tests/object-locking.ml index a4cd77d..1fddd51 100644 --- a/tests/object-locking.ml +++ b/tests/object-locking.ml @@ -1,7 +1,7 @@ (* * Analyse libvirt driver API methods for mutex locking mistakes * - * Copyright 2008-2009 Red Hat, Inc + * Copyright (C) 2008-2010 Red Hat, Inc * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public diff --git a/tests/xmconfigtest.c b/tests/xmconfigtest.c index acb6d18..a199c72 100644 --- a/tests/xmconfigtest.c +++ b/tests/xmconfigtest.c @@ -1,7 +1,7 @@ /* * xmconfigtest.c: Test backend for xm_internal config file handling * - * Copyright (C) 2007 Red Hat + * Copyright (C) 2007, 2010 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public -- 1.6.6.1

2 7

Re: [libvirt] virtio serial config
by Matthew Booth 18 Mar '10

18 Mar '10

On 18/03/10 14:31, Matthew Whitehead wrote: > Matthew, > Dan B referred me to you. I want to create 4 (or more) virtual serial devices between two KVM virtual hosts. The output of one serial device would be the input to the other. Virt-manager doesn't provide this, so I'm prepared to hack the config files manually. Unfortunately, I can't find documentation on the syntax. > > Can you assist? I'm normally the one complaining about lack of documentation :) That's quite an oversight. The basic syntax for a virtio-serial channel is: <channel type='pty'> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> You'll want 4 <channel/> entries per guest. Give them all different names. The interesting bit is going to be plugging the character devices together on the host. I'm pretty sure you'll have to use pipes. Note that afaik the libvirt documentation on this is wrong and misleading. Don't read it, just test instead :) IIRC, the way pipes actually work is that it expects to be able to use 2 pipes based on the path you specify, called foo.in and foo.out. I seem to recall you need to create these manually too. The following advice is completely untested. It may also be responsible for global warming and the rise of reality TV. Host: # mkfifo /tmp/guesta-virtioserial{1,2}.{in,out} # ln /tmp/guesta-virtioserial1.in /tmp/guestb-virtioserial1.out # ln /tmp/guesta-virtioserial1.out /tmp/guestb-virtioserial1.in # ln /tmp/guesta-virtioserial2.in /tmp/guestb-virtioserial2.out # ln /tmp/guesta-virtioserial2.out /tmp/guestb-virtioserial2.in Guest A: <channel type='pipe'> <source path="/tmp/guesta-virtioserial1"/> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> <channel type='pipe'> <source path="/tmp/guesta-virtioserial2"/> <target type='virtio' name='org.linux-kvm.port.2'/> </channel> Guest B: <channel type='pipe'> <source path="/tmp/guestb-virtioserial1"/> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> <channel type='pipe'> <source path="/tmp/guestb-virtioserial2"/> <target type='virtio' name='org.linux-kvm.port.2'/> </channel> At this point, both virtio-serial devices in each guest should be full-duplex. Matt -- Matthew Booth, RHCA, RHCSS Red Hat Engineering, Virtualisation Team M: +44 (0)7977 267231 GPG ID: D33C3490 GPG FPR: 3733 612D 2D05 5458 8A8A 1600 3441 EA19 D33C 3490

1 0

[libvirt] [APPENDIX] Sample filters
by Stefan Berger 18 Mar '10

18 Mar '10

Attached are some sample filter templates. Some of these should probably become available through libvirt repository later on as 'standard filters'. One may copy the XML files into /etc/libvirt/nwfilter or use 'virsh nwfilter-define <filename>' to make them known to libvirt. Using 'virsh nwfilter-list' one can then list the filters. The simpleloop.xml filter will not be accepted since it would directly introduce a loop. More complex loops are also detected. Regards, Stefan, Gerhard

1 0

[libvirt] [PATCH 13/14] Add IPv6 support for ebtables layer
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds IPv6 support for the ebtables layer. Since the parser etc. are all parameterized, it was fairly easy to add this... Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0