- Devel - libvirt List Archives

[libvirt] [PATCHv2] LXC: Fix handling of RAM filesystem size units
by Ján Tomko 09 Oct '13

09 Oct '13

Since 76b644c when the support for RAM filesystems was introduced, libvirt accepted the following XML: <source usage='1024' unit='KiB'/> This was parsed correctly and internally stored in bytes, but it was formatted as (with an extra 's'): <source usage='1024' units='KiB'/> When read again, this was treated as if the units were missing, meaning libvirt was unable to parse its own XML correctly. The usage attribute was documented as being in KiB, but it was not scaled if the unit was missing. Transient domains still worked, because this was balanced by an extra 'k' in the mount options. This patch: Outputs the units via the 'unit' attribute (fixing persistent domains), but accepts both 'unit' and 'units' (for compatibility with older libvirt and libvirt-sandbox). Removes the extra 'k' from the tmpfs mount options, which is needed because now we parse our own XML correctly. Changes the default unit to KiB to match documentation, fixing: https://bugzilla.redhat.com/show_bug.cgi?id=1015689 --- v1: https://www.redhat.com/archives/libvir-list/2013-October/msg00361.html v2: keeps the output unit KiB changes the default input unit to KiB accepts 'units' as well as 'unit' in input XML docs/formatdomain.html.in | 6 ++-- docs/schemas/domaincommon.rng | 5 ++++ src/conf/domain_conf.c | 12 ++++++-- src/lxc/lxc_container.c | 2 +- tests/domainschematest | 2 +- tests/lxcxml2xmldata/lxc-filesystem-ram.xml | 41 ++++++++++++++++++++++++++ tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml | 41 ++++++++++++++++++++++++++ tests/lxcxml2xmltest.c | 1 + 8 files changed, 103 insertions(+), 7 deletions(-) create mode 100644 tests/lxcxml2xmldata/lxc-filesystem-ram.xml create mode 100644 tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 3689399..dfd11bf 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -2213,7 +2213,8 @@ <dd> An in-memory filesystem, using memory from the host OS. The source element has a single attribute <code>usage</code> - which gives the memory usage limit in kibibytes. Only used + which gives the memory usage limit in KiB, unless units + are specified by the <code>unit</code> attribute. Only used by LXC driver. <span class="since"> (since 0.9.13)</span></dd> <dt><code>type='bind'</code></dt> @@ -2279,7 +2280,8 @@ <code>name</code> attribute must be used with <code>type='template'</code>, and the <code>dir</code> attribute must be used with <code>type='mount'</code>. The <code>usage</code> attribute - is used with <code>type='ram'</code> to set the memory limit in KB. + is used with <code>type='ram'</code> to set the memory limit KiB, unless + units are specified by the <code>unit</code> attribute. </dd> <dt><code>target</code></dt> diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 5c5301d..582d4c3 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -1736,6 +1736,11 @@ <ref name='unit'/> </attribute> </optional> + <optional> + <attribute name='units'> + <ref name='unit'/> + </attribute> + </optional> <empty/> </element> </interleave> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 0d63845..0870047 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -5917,6 +5917,7 @@ virDomainFSDefParseXML(xmlNodePtr node, char *wrpolicy = NULL; char *usage = NULL; char *unit = NULL; + char *units = NULL; ctxt->node = node; @@ -5973,6 +5974,7 @@ virDomainFSDefParseXML(xmlNodePtr node, else if (def->type == VIR_DOMAIN_FS_TYPE_RAM) { usage = virXMLPropString(cur, "usage"); unit = virXMLPropString(cur, "unit"); + units = virXMLPropString(cur, "units"); } } else if (!target && xmlStrEqual(cur->name, BAD_CAST "target")) { @@ -6042,8 +6044,11 @@ virDomainFSDefParseXML(xmlNodePtr node, usage); goto error; } - if (unit && - virScaleInteger(&def->usage, unit, + /* Older libvirt only parsed 'unit' here, despite + * printing it in the 'units' attribute. + * + * The default was bytes when 'unit' wasn't present */ + if (virScaleInteger(&def->usage, units ? units : unit, 1024, ULLONG_MAX) < 0) goto error; } @@ -6066,6 +6071,7 @@ cleanup: VIR_FREE(wrpolicy); VIR_FREE(usage); VIR_FREE(unit); + VIR_FREE(units); VIR_FREE(format); return def; @@ -14764,7 +14770,7 @@ virDomainFSDefFormat(virBufferPtr buf, break; case VIR_DOMAIN_FS_TYPE_RAM: - virBufferAsprintf(buf, " <source usage='%lld' units='KiB'/>\n", + virBufferAsprintf(buf, " <source usage='%lld' unit='KiB'/>\n", def->usage / 1024); break; } diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index b1f429c..7c722cc 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1428,7 +1428,7 @@ static int lxcContainerMountFSTmpfs(virDomainFSDefPtr fs, VIR_DEBUG("usage=%lld sec=%s", fs->usage, sec_mount_options); if (virAsprintf(&data, - "size=%lldk%s", fs->usage, sec_mount_options) < 0) + "size=%lld%s", fs->usage, sec_mount_options) < 0) goto cleanup; if (virFileMakePath(fs->dst) < 0) { diff --git a/tests/domainschematest b/tests/domainschematest index 0e360ca..9ebf0b9 100755 --- a/tests/domainschematest +++ b/tests/domainschematest @@ -7,7 +7,7 @@ DIRS="" DIRS="$DIRS domainschemadata qemuxml2argvdata sexpr2xmldata" DIRS="$DIRS xmconfigdata xml2sexprdata qemuxml2xmloutdata " -DIRS="$DIRS lxcxml2xmldata" +DIRS="$DIRS lxcxml2xmldata lxcxml2xmloutdata" SCHEMA="domain.rng" check_schema "$DIRS" "$SCHEMA" diff --git a/tests/lxcxml2xmldata/lxc-filesystem-ram.xml b/tests/lxcxml2xmldata/lxc-filesystem-ram.xml new file mode 100644 index 0000000..4b2abb5 --- /dev/null +++ b/tests/lxcxml2xmldata/lxc-filesystem-ram.xml @@ -0,0 +1,41 @@ +<domain type='lxc'> + <name>demo</name> + <uuid>8369f1ac-7e46-e869-4ca5-759d51478066</uuid> + <memory unit='KiB'>500000</memory> + <currentMemory unit='KiB'>500000</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64'>exe</type> + <init>/bin/sh</init> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/libexec/libvirt_lxc</emulator> + <filesystem type='ram'> + <source usage='1048576'/> + <target dir='/mnt/mississippi'/> + </filesystem> + <filesystem type='ram'> + <source usage='1048576' unit='bytes'/> + <target dir='/mnt/titicaca'/> + </filesystem> + <filesystem type='ram'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/orinoco'/> + </filesystem> + <filesystem type='ram'> + <source usage='1048576' units='bytes'/> + <target dir='/mnt/antananarivo'/> + </filesystem> + <filesystem type='ram'> + <source usage='1024' units='KiB'/> + <target dir='/mnt/ouagadougou'/> + </filesystem> + <console type='pty'> + <target type='lxc' port='0'/> + </console> + </devices> +</domain> diff --git a/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml b/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml new file mode 100644 index 0000000..65d07dd --- /dev/null +++ b/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml @@ -0,0 +1,41 @@ +<domain type='lxc'> + <name>demo</name> + <uuid>8369f1ac-7e46-e869-4ca5-759d51478066</uuid> + <memory unit='KiB'>500000</memory> + <currentMemory unit='KiB'>500000</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64'>exe</type> + <init>/bin/sh</init> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/libexec/libvirt_lxc</emulator> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1048576' unit='KiB'/> + <target dir='/mnt/mississippi'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/titicaca'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/orinoco'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/antananarivo'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/ouagadougou'/> + </filesystem> + <console type='pty'> + <target type='lxc' port='0'/> + </console> + </devices> +</domain> diff --git a/tests/lxcxml2xmltest.c b/tests/lxcxml2xmltest.c index 5846ab0..1692e4b 100644 --- a/tests/lxcxml2xmltest.c +++ b/tests/lxcxml2xmltest.c @@ -136,6 +136,7 @@ mymain(void) DO_TEST("systemd"); DO_TEST("hostdev"); DO_TEST("disk-formats"); + DO_TEST_DIFFERENT("filesystem-ram"); virObjectUnref(caps); virObjectUnref(xmlopt); -- 1.8.1.5

3 4

[libvirt] [PATCH 0/3] qemu:Allow users specify -incoming listen address
by Michal Privoznik 09 Oct '13

09 Oct '13

However, if user provides a wrong address, the libvirtd dies horribly: Core was generated by `/home/zippy/tmp/libvirt.git/daemon/.libs/libvirtd --listen'. Program terminated with signal 11, Segmentation fault. #0 0x00007f5c3859a7fc in qemuProcessHandleMonitorEOF (mon=0x7f5c24001870, vm=0x7f5c24001fa0, opaque=0x7f5c300c8600) at qemu/qemu_process.c:300 300 if (priv->beingDestroyed) { (gdb) p priv $1 = (qemuDomainObjPrivatePtr) 0x0 (gdb) p vm $2 = (virDomainObjPtr) 0x7f5c24001fa0 This is supposed to be fixed by my patch: https://www.redhat.com/archives/libvir-list/2013-October/msg00347.html Michal Privoznik (3): Migration: Introduce VIR_MIGRATE_PARAM_LISTEN_ADDRESS qemu: Implement support for VIR_MIGRATE_PARAM_LISTEN_ADDRESS qemu_conf: Introduce "listen_address" include/libvirt/libvirt.h.in | 10 ++++ src/qemu/libvirtd_qemu.aug | 3 ++ src/qemu/qemu.conf | 6 +++ src/qemu/qemu_conf.c | 2 + src/qemu/qemu_conf.h | 2 + src/qemu/qemu_driver.c | 33 +++++++++---- src/qemu/qemu_migration.c | 98 +++++++++++++++++++++++++++----------- src/qemu/qemu_migration.h | 13 +++-- src/qemu/test_libvirtd_qemu.aug.in | 1 + tools/virsh-domain.c | 11 +++++ tools/virsh.pod | 10 +++- 11 files changed, 146 insertions(+), 43 deletions(-) -- 1.8.1.5

2 6

[libvirt] [PATCH] LXC: Fix the output units of ram filesystem size
by Ján Tomko 09 Oct '13

09 Oct '13

Commit 76b644c added support for adding RAM filesystems to LXC domains, with a syntax of: <source usage='10' unit='MiB'/> where default unit would be KiB per documentation, however the XML parser treated sizes without units as bytes. When formatting the XML, this was divided by 1024, but the KiB units were put inside the 'units' attribute, as opposed to the 'unit' attribute the parser looks for. The code generating the mount options assumed the size in the domain definition to be in KiB, despite it being parsed as B. This worked as long as exaclty one re-format of the XML happened (for domains that were just created). Change the XML output to bytes and fix the documentation. https://bugzilla.redhat.com/show_bug.cgi?id=1015689 --- docs/formatdomain.html.in | 5 ++-- src/conf/domain_conf.c | 4 ++-- src/lxc/lxc_container.c | 2 +- tests/lxcxml2xmldata/lxc-filesystem-ram.xml | 33 ++++++++++++++++++++++++++ tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml | 33 ++++++++++++++++++++++++++ tests/lxcxml2xmltest.c | 1 + 6 files changed, 73 insertions(+), 5 deletions(-) create mode 100644 tests/lxcxml2xmldata/lxc-filesystem-ram.xml create mode 100644 tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 3689399..bfe0ae4 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -2213,7 +2213,8 @@ <dd> An in-memory filesystem, using memory from the host OS. The source element has a single attribute <code>usage</code> - which gives the memory usage limit in kibibytes. Only used + which gives the memory usage limit in bytes, unless units + are specified by the attribute <code>unit</code>. Only used by LXC driver. <span class="since"> (since 0.9.13)</span></dd> <dt><code>type='bind'</code></dt> @@ -2279,7 +2280,7 @@ <code>name</code> attribute must be used with <code>type='template'</code>, and the <code>dir</code> attribute must be used with <code>type='mount'</code>. The <code>usage</code> attribute - is used with <code>type='ram'</code> to set the memory limit in KB. + is used with <code>type='ram'</code> to set the memory limit in bytes. </dd> <dt><code>target</code></dt> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 0d63845..2c67966 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -14764,8 +14764,8 @@ virDomainFSDefFormat(virBufferPtr buf, break; case VIR_DOMAIN_FS_TYPE_RAM: - virBufferAsprintf(buf, " <source usage='%lld' units='KiB'/>\n", - def->usage / 1024); + virBufferAsprintf(buf, " <source usage='%lld' unit='B'/>\n", + def->usage); break; } diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index b1f429c..7c722cc 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1428,7 +1428,7 @@ static int lxcContainerMountFSTmpfs(virDomainFSDefPtr fs, VIR_DEBUG("usage=%lld sec=%s", fs->usage, sec_mount_options); if (virAsprintf(&data, - "size=%lldk%s", fs->usage, sec_mount_options) < 0) + "size=%lld%s", fs->usage, sec_mount_options) < 0) goto cleanup; if (virFileMakePath(fs->dst) < 0) { diff --git a/tests/lxcxml2xmldata/lxc-filesystem-ram.xml b/tests/lxcxml2xmldata/lxc-filesystem-ram.xml new file mode 100644 index 0000000..404429b --- /dev/null +++ b/tests/lxcxml2xmldata/lxc-filesystem-ram.xml @@ -0,0 +1,33 @@ +<domain type='lxc'> + <name>demo</name> + <uuid>8369f1ac-7e46-e869-4ca5-759d51478066</uuid> + <memory unit='KiB'>500000</memory> + <currentMemory unit='KiB'>500000</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64'>exe</type> + <init>/bin/sh</init> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/libexec/libvirt_lxc</emulator> + <filesystem type='ram'> + <source usage='1048576'/> + <target dir='/mnt/mississippi'/> + </filesystem> + <filesystem type='ram'> + <source usage='1048576' unit='b'/> + <target dir='/mnt/titicaca'/> + </filesystem> + <filesystem type='ram'> + <source usage='1024' unit='KiB'/> + <target dir='/mnt/orinoco'/> + </filesystem> + <console type='pty'> + <target type='lxc' port='0'/> + </console> + </devices> +</domain> diff --git a/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml b/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml new file mode 100644 index 0000000..431b8cb --- /dev/null +++ b/tests/lxcxml2xmloutdata/lxc-filesystem-ram.xml @@ -0,0 +1,33 @@ +<domain type='lxc'> + <name>demo</name> + <uuid>8369f1ac-7e46-e869-4ca5-759d51478066</uuid> + <memory unit='KiB'>500000</memory> + <currentMemory unit='KiB'>500000</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64'>exe</type> + <init>/bin/sh</init> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/libexec/libvirt_lxc</emulator> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1048576' unit='B'/> + <target dir='/mnt/mississippi'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1048576' unit='B'/> + <target dir='/mnt/titicaca'/> + </filesystem> + <filesystem type='ram' accessmode='passthrough'> + <source usage='1048576' unit='B'/> + <target dir='/mnt/orinoco'/> + </filesystem> + <console type='pty'> + <target type='lxc' port='0'/> + </console> + </devices> +</domain> diff --git a/tests/lxcxml2xmltest.c b/tests/lxcxml2xmltest.c index 5846ab0..1692e4b 100644 --- a/tests/lxcxml2xmltest.c +++ b/tests/lxcxml2xmltest.c @@ -136,6 +136,7 @@ mymain(void) DO_TEST("systemd"); DO_TEST("hostdev"); DO_TEST("disk-formats"); + DO_TEST_DIFFERENT("filesystem-ram"); virObjectUnref(caps); virObjectUnref(xmlopt); -- 1.8.1.5

3 3

[libvirt] [PATCH] storage: Use bool instead of int
by Ján Tomko 09 Oct '13

09 Oct '13

Commit 532fef3 added two-state 'need_alloc' and exposed 'want_sparse' which also only has two states. Change their type from int to bool. --- src/storage/storage_backend.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c index 662af32..4eec0f3 100644 --- a/src/storage/storage_backend.c +++ b/src/storage/storage_backend.c @@ -129,7 +129,7 @@ virStorageBackendCopyToFD(virStorageVolDefPtr vol, virStorageVolDefPtr inputvol, int fd, unsigned long long *total, - int want_sparse) + bool want_sparse) { int inputfd = -1; int amtread = -1; @@ -270,7 +270,7 @@ virStorageBackendCreateBlockFrom(virConnectPtr conn ATTRIBUTE_UNUSED, if (inputvol) { int res = virStorageBackendCopyToFD(vol, inputvol, - fd, &remain, 0); + fd, &remain, false); if (res < 0) goto cleanup; } @@ -315,7 +315,7 @@ static int createRawFile(int fd, virStorageVolDefPtr vol, virStorageVolDefPtr inputvol) { - int need_alloc = 1; + bool need_alloc = true; int ret = 0; unsigned long long remain; @@ -338,7 +338,7 @@ createRawFile(int fd, virStorageVolDefPtr vol, * available, and since we're going to copy data from another * file it doesn't make sense to write the file twice. */ if (fallocate(fd, 0, 0, vol->allocation) == 0) { - need_alloc = 0; + need_alloc = false; } else if (errno != ENOSYS && errno != EOPNOTSUPP) { ret = -errno; virReportSystemError(errno, @@ -354,8 +354,8 @@ createRawFile(int fd, virStorageVolDefPtr vol, /* allow zero blocks to be skipped if we've requested sparse * allocation (allocation < capacity) or we have already * been able to allocate the required space. */ - int want_sparse = (need_alloc == 0) || - (vol->allocation < inputvol->capacity); + bool want_sparse = !need_alloc || + (vol->allocation < inputvol->capacity); ret = virStorageBackendCopyToFD(vol, inputvol, fd, &remain, want_sparse); if (ret < 0) { -- 1.8.1.5

2 2

[libvirt] [PATCH 0/2] change the minimum blkio weight according to kernel version
by Chen Hanxiao 09 Oct '13

09 Oct '13

From: Chen Hanxiao <chenhanxiao(a)cn.fujitsu.com> Since 2.6.39, kernel changed the minimum weight of device blkio. We need to follow them according to different kernel version. Chen Hanxiao (2): [libvirt][PATCH]cgroup: introduce kernel version check function for cgroup [libvirt][PATCH]blkio: change the minimum weight according to kernel version src/util/vircgroup.c | 76 ++++++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 65 insertions(+), 11 deletions(-) -- 1.8.2.1

2 6

[libvirt] [PATCH] tests: Add qemu test for multiple timers
by Cole Robinson 08 Oct '13

08 Oct '13

The following XML is the recommended default clock configuration for qemu: <clock offset='utc'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='no'/> </clock> However we weren't testing any of those timer elements. --- .../qemuxml2argv-clock-hpet-off.args | 5 ++++ .../qemuxml2argv-clock-hpet-off.xml | 30 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 2 ++ 3 files changed, 37 insertions(+) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.xml diff --git a/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.args b/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.args new file mode 100644 index 0000000..5d11d91 --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.args @@ -0,0 +1,5 @@ +LC_ALL=C PATH=/bin HOME=/home/test USER=test LOGNAME=test QEMU_AUDIO_DRV=none \ +/usr/bin/qemu -S -M \ +pc -m 214 -smp 1 -nographic -monitor unix:/tmp/test-monitor,server,nowait \ +-rtc base=utc,driftfix=slew -no-kvm-pit-reinjection -no-hpet -no-acpi -boot c \ +-usb -hda /dev/HostVG/QEMUGuest1 -net none -serial none -parallel none diff --git a/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.xml b/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.xml new file mode 100644 index 0000000..3aed49f --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-clock-hpet-off.xml @@ -0,0 +1,30 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory unit='KiB'>219100</memory> + <currentMemory unit='KiB'>219100</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='i686' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'> + <timer name='rtc' tickpolicy='catchup'/> + <timer name='pit' tickpolicy='delay'/> + <timer name='hpet' present='no'/> + </clock> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <disk type='block' device='disk'> + <source dev='/dev/HostVG/QEMUGuest1'/> + <target dev='hda' bus='ide'/> + <address type='drive' controller='0' bus='0' target='0' unit='0'/> + </disk> + <controller type='usb' index='0'/> + <controller type='ide' index='0'/> + <memballoon model='virtio'/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 58165fd..286f1a3 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -421,6 +421,8 @@ mymain(void) DO_TEST("clock-variable", QEMU_CAPS_RTC); */ DO_TEST("clock-france", QEMU_CAPS_RTC); + DO_TEST("clock-hpet-off", QEMU_CAPS_RTC, QEMU_CAPS_NO_HPET, + QEMU_CAPS_NO_KVM_PIT); DO_TEST("cpu-kvmclock", QEMU_CAPS_ENABLE_KVM); DO_TEST("cpu-host-kvmclock", QEMU_CAPS_ENABLE_KVM, QEMU_CAPS_CPU_HOST); DO_TEST("kvmclock", QEMU_CAPS_KVM); -- 1.8.3.1

2 1

[libvirt] [PATCH] Ignore thin pool LVM devices
by Dusty Mabe 08 Oct '13

08 Oct '13

For BZ 924672 the problem stems from the fact that thin pool logical volume devices show up in /sbin/lvs output just like normal logical volumes do. Since the thin pool devices show up libvirt assumes they are just normal logical volumes and assumes there will be a corresponding /dev/vgname/lvname device that has been created. This is not the case and you will receive the following error when starting the storage pool: error: cannot stat file '/dev/vgvirt/thinpool': No such file or directory This patch modifies virStorageBackendLogicalMakeVol() to ignore thin pool backing devices. --- src/storage/storage_backend_logical.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/storage/storage_backend_logical.c b/src/storage/storage_backend_logical.c index a1a37a1..74d26bd 100644 --- a/src/storage/storage_backend_logical.c +++ b/src/storage/storage_backend_logical.c @@ -85,6 +85,10 @@ virStorageBackendLogicalMakeVol(virStoragePoolObjPtr pool, if (attrs[4] != 'a') return 0; + /* BZ 924672 - Skip thin pools(t) and thin pool data(T) */ + if (attrs[0] == 't' || attrs[0] == 'T') + return 0; + /* See if we're only looking for a specific volume */ if (data != NULL) { vol = data; -- 1.8.3.1

2 2

[libvirt] [PATCH] qemu_migration: Avoid crashing if domain dies too quickly
by Michal Privoznik 08 Oct '13

08 Oct '13

I've noticed a SIGSEGV-ing libvirtd on the destination when the qemu died too quickly = in Prepare phase. What is happening here is: 1) [Thread 3493] We are in qemuMigrationPrepareAny() and calling qemuProcessStart() which subsequently calls qemuProcessWaitForMonitor() and qemuConnectMonitor(). So far so good. The qemuMonitorOpen() succeeds, however switching monitor to QMP mode fails as qemu died meanwhile. That is qemuMonitorSetCapabilities() returns -1. 2013-10-08 15:54:10.629+0000: 3493: debug : qemuMonitorSetCapabilities:1356 : mon=0x14a53da0 2013-10-08 15:54:10.630+0000: 3493: debug : qemuMonitorJSONCommandWithFd:262 : Send command '{"execute":"qmp_capabilities","id":"libvirt-1"}' for write with FD -1 2013-10-08 15:54:10.630+0000: 3493: debug : virEventPollUpdateHandle:147 : EVENT_POLL_UPDATE_HANDLE: watch=17 events=13 ... 2013-10-08 15:54:10.631+0000: 3493: debug : qemuMonitorSend:956 : QEMU_MONITOR_SEND_MSG: mon=0x14a53da0 msg={"execute":"qmp_capabilities","id":"libvirt-1"} fd=-1 2013-10-08 15:54:10.631+0000: 3262: debug : virEventPollRunOnce:641 : Poll got 1 event(s) 2) [Thread 3262] The event loop is trying to do the talking to monitor. However, qemu is dead already, remember? 2013-10-08 15:54:13.436+0000: 3262: error : qemuMonitorIORead:551 : Unable to read from monitor: Connection reset by peer 2013-10-08 15:54:13.516+0000: 3262: debug : virFileClose:90 : Closed fd 25 ... 2013-10-08 15:54:13.533+0000: 3493: debug : qemuMonitorSend:968 : Send command resulted in error internal error: early end of file from monitor: possible problem: 3) [Thread 3493] qemuProcessStart() failed. No big deal. Go to the 'endjob' label and subsequently to the 'cleanup'. Since the domain is not persistent and ret is -1, the qemuDomainRemoveInactive() is called. This has an (unpleasant) effect of virObjectUnref()-in the @vm object. Unpleasant because the event loop which is about to trigger EOF callback still holds a pointer to the @vm (not the reference). See the valgrind output below. 4) [Thread 3262] So the even loop starts triggering EOF: 2013-10-08 15:54:13.542+0000: 3262: debug : qemuMonitorIO:729 : Triggering EOF callback 2013-10-08 15:54:13.543+0000: 3262: debug : qemuProcessHandleMonitorEOF:294 : Received EOF on 0x14549110 'migt10' And the monitor is cleaned up. This results in calling qemuProcessHandleMonitorEOF with the @vm pointer passed. The pointer is kept in qemuMonitor struct. ==3262== Thread 1: ==3262== Invalid read of size 4 ==3262== at 0x77ECCAA: pthread_mutex_lock (in /lib64/libpthread-2.15.so) ==3262== by 0x52FAA06: virMutexLock (virthreadpthread.c:85) ==3262== by 0x52E3891: virObjectLock (virobject.c:320) ==3262== by 0x11626743: qemuProcessHandleMonitorEOF (qemu_process.c:296) ==3262== by 0x11642593: qemuMonitorIO (qemu_monitor.c:730) ==3262== by 0x52BD526: virEventPollDispatchHandles (vireventpoll.c:501) ==3262== by 0x52BDD49: virEventPollRunOnce (vireventpoll.c:648) ==3262== by 0x52BBC68: virEventRunDefaultImpl (virevent.c:274) ==3262== by 0x542D3D9: virNetServerRun (virnetserver.c:1112) ==3262== by 0x11F368: main (libvirtd.c:1513) ==3262== Address 0x14549128 is 24 bytes inside a block of size 136 free'd ==3262== at 0x4C2AF5C: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==3262== by 0x529B1FF: virFree (viralloc.c:580) ==3262== by 0x52E3703: virObjectUnref (virobject.c:270) ==3262== by 0x531557E: virDomainObjListRemove (domain_conf.c:2355) ==3262== by 0x1160E899: qemuDomainRemoveInactive (qemu_domain.c:2061) ==3262== by 0x1163A0C6: qemuMigrationPrepareAny (qemu_migration.c:2450) ==3262== by 0x1163A923: qemuMigrationPrepareDirect (qemu_migration.c:2626) ==3262== by 0x11682D71: qemuDomainMigratePrepare3Params (qemu_driver.c:10309) ==3262== by 0x53B0976: virDomainMigratePrepare3Params (libvirt.c:7266) ==3262== by 0x1502D3: remoteDispatchDomainMigratePrepare3Params (remote.c:4797) ==3262== by 0x12DECA: remoteDispatchDomainMigratePrepare3ParamsHelper (remote_dispatch.h:5741) ==3262== by 0x54322EB: virNetServerProgramDispatchCall (virnetserverprogram.c:435) The mon->vm is set in qemuMonitorOpenInternal() which is the correct place to increase @vm ref counter. The correct place to decrease the ref counter is then qemuMonitorDispose(). However, in order to avoid cyclic calling of dispose functions, we must use hack, and set priv->mon to NULL. The qemuDomainObjPrivateFree won't call us again then. This describes the last two chunks of the commit. The others are there for handling virObjectRef() and virObjectUnref() correctly. So far it doesn't work well with NULL virClassPtr :) Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Yet another one of those patches, where chasing it down took a half of a day and description what went wrong is longer than a few lines of fix. src/qemu/qemu_capabilities.c | 14 ++++++++++---- src/qemu/qemu_monitor.c | 12 ++++++++++++ 2 files changed, 22 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 7c39c1c..17095b4 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -2587,7 +2587,8 @@ virQEMUCapsInitQMP(virQEMUCapsPtr qemuCaps, char *monpath = NULL; char *pidfile = NULL; pid_t pid = 0; - virDomainObj vm; + virDomainObjPtr vm = NULL; + virDomainXMLOptionPtr xmlopt = NULL; /* the ".sock" sufix is important to avoid a possible clash with a qemu * domain called "capabilities" @@ -2650,10 +2651,13 @@ virQEMUCapsInitQMP(virQEMUCapsPtr qemuCaps, goto cleanup; } - memset(&vm, 0, sizeof(vm)); - vm.pid = pid; + if (!(xmlopt = virDomainXMLOptionNew(NULL, NULL, NULL)) || + !(vm = virDomainObjNew(xmlopt))) + goto cleanup; + + vm->pid = pid; - if (!(mon = qemuMonitorOpen(&vm, &config, true, &callbacks, NULL))) { + if (!(mon = qemuMonitorOpen(vm, &config, true, &callbacks, NULL))) { ret = 0; goto cleanup; } @@ -2673,6 +2677,8 @@ cleanup: virCommandFree(cmd); VIR_FREE(monarg); VIR_FREE(monpath); + virObjectUnref(vm); + virObjectUnref(xmlopt); if (pid != 0) { char ebuf[1024]; diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index a601ee0..2cf68da 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -251,10 +251,21 @@ static char * qemuMonitorEscapeNonPrintable(const char *text) static void qemuMonitorDispose(void *obj) { qemuMonitorPtr mon = obj; + qemuDomainObjPrivatePtr priv = mon->vm->privateData; VIR_DEBUG("mon=%p", mon); if (mon->cb && mon->cb->destroy) (mon->cb->destroy)(mon, mon->vm, mon->callbackOpaque); + /* HACK: priv->mon is still non-NULL. If we are the last to hold the + * reference to mon->vm, the qemuDomainObjPrivateFree() will call + * qemuMonitorClose() and hence us again. IOW, dispose callback which calls + * dispose callback which calls dispose callback ... + * Moreover, not all domains have privateData - for instance a dummy domain + * object created in virQEMUCapsInitQMP(). */ + if (priv) + priv->mon = NULL; + virObjectUnref(mon->vm); + virCondDestroy(&mon->notify); VIR_FREE(mon->buffer); virJSONValueFree(mon->options); @@ -781,6 +792,7 @@ qemuMonitorOpenInternal(virDomainObjPtr vm, } mon->fd = fd; mon->hasSendFD = hasSendFD; + virObjectRef(vm); mon->vm = vm; mon->json = json; if (json) -- 1.8.1.5

2 1

Re: [libvirt] [virt-tools-list] [PATCH] build: fix linker error on FreeBSD
by Eric Blake 08 Oct '13

08 Oct '13

On 10/08/2013 10:21 AM, Giuseppe Scrivano wrote: > Commit 2d74822a9eb4856c7f5216bb92bcb76630660f72 renamed > "freebsdNodeGetCPUCount" to "appleFreebsdNodeGetCPUCount", leaving one > call to "freebsdNodeGetCPUCount". Fix this other case. ACK, although it looks like you intended this to go to libvir-list (now cc'd) > > Signed-off-by: Giuseppe Scrivano <gscrivan(a)redhat.com> > --- > src/nodeinfo.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/src/nodeinfo.c b/src/nodeinfo.c > index 320d8f8..702ad7c 100644 > --- a/src/nodeinfo.c > +++ b/src/nodeinfo.c > @@ -1056,8 +1056,8 @@ nodeGetCPUCount(void) > > VIR_FREE(cpupath); > return ncpu; > -#elif defined(__FreeBSD__) > - return freebsdNodeGetCPUCount(); > +#elif defined(__FreeBSD__) || defined(__APPLE__) > + return appleFreebsdNodeGetCPUCount (); > #else > virReportError(VIR_ERR_NO_SUPPORT, "%s", > _("host cpu counting not implemented on this platform")); > -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

2 3

[libvirt] [PATCH] Fix typo breaking cgroups for NBD backed filesystems
by Daniel P. Berrange 08 Oct '13

08 Oct '13

From: "Daniel P. Berrange" <berrange(a)redhat.com> A typo in the setup of NBD backed filesystems meant the /dev/nbdN device would not be added to the cgroups device ACL. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/lxc/lxc_controller.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/lxc/lxc_controller.c b/src/lxc/lxc_controller.c index e7e7acc..6cb36ff 100644 --- a/src/lxc/lxc_controller.c +++ b/src/lxc/lxc_controller.c @@ -423,7 +423,7 @@ static int virLXCControllerSetupNBDDeviceFS(virDomainFSDefPtr fs) * We now change it into a block device type, so that * the rest of container setup 'just works' */ - fs->type = VIR_DOMAIN_DISK_TYPE_BLOCK; + fs->type = VIR_DOMAIN_FS_TYPE_BLOCK; VIR_FREE(fs->src); fs->src = dev; -- 1.8.3.1

2 1