May 2013 - Devel - libvirt List Archives

[libvirt] [PATCH] Forbid use of ':' in RBD pool names
by Daniel P. Berrange 14 May '13

14 May '13

From: "Daniel P. Berrange" <berrange(a)redhat.com> The QEMU command line syntax for RBD disks is file=rbd:pool/image:opt1=val1:opt2=val2... There is no way to escape the ':' if it appears in the pool or image name. Thus it must be explicitly forbidden if it occurs in the libvirt XML. People are known to be abusing the lack of escaping in current libvirt to pass arbitrary args to QEMU. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/qemu/qemu_command.c | 6 ++++ ...qemuxml2argv-disk-drive-network-rbd-invalid.xml | 37 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 5 +++ 3 files changed, 48 insertions(+) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-invalid.xml diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index eddc263..c0cb250 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -2383,6 +2383,12 @@ qemuBuildRBDString(virConnectPtr conn, char *secret = NULL; size_t secret_size; + if (strchr(disk->src, ':')) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, + _("':' not allowed in RBD source volume name")); + return -1; + } + virBufferEscape(opt, ',', ",", "rbd:%s", disk->src); if (disk->auth.username) { virBufferEscape(opt, '\\', ":", ":id=%s", disk->auth.username); diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-invalid.xml b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-invalid.xml new file mode 100644 index 0000000..e8d3280 --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-invalid.xml @@ -0,0 +1,37 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory unit='KiB'>219136</memory> + <currentMemory unit='KiB'>219136</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='i686' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <disk type='block' device='disk'> + <driver name='qemu' type='raw'/> + <source dev='/dev/HostVG/QEMUGuest1'/> + <target dev='hda' bus='ide'/> + <address type='drive' controller='0' bus='0' target='0' unit='0'/> + </disk> + <disk type='network' device='disk'> + <driver name='qemu' type='raw'/> + <source protocol='rbd' name='poolname/imagename:rbd_cache=1:rbd_cache_size=67108864:rbd_cache_max_dirty=0'> + <host name='mon1.example.org' port='6321'/> + <host name='mon2.example.org' port='6322'/> + <host name='mon3.example.org' port='6322'/> + </source> + <target dev='vda' bus='virtio'/> + </disk> + <controller type='usb' index='0'/> + <controller type='ide' index='0'/> + <controller type='pci' index='0' model='pci-root'/> + <memballoon model='virtio'/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 98ceb83..579c016 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -160,6 +160,9 @@ static int testCompareXMLToArgvFiles(const char *xml, VIR_NETDEV_VPORT_PROFILE_OP_NO_OP))) { if (flags & FLAG_EXPECT_FAILURE) { ret = 0; + if (virTestGetDebug() > 1) + fprintf(stderr, "Got expected error: %s\n", + virGetLastErrorMessage()); virResetLastError(); } goto out; @@ -528,6 +531,8 @@ mymain(void) QEMU_CAPS_DRIVE, QEMU_CAPS_DRIVE_FORMAT); DO_TEST("disk-drive-network-rbd-ipv6", QEMU_CAPS_DRIVE, QEMU_CAPS_DRIVE_FORMAT); + DO_TEST_FAILURE("disk-drive-network-rbd-invalid", + QEMU_CAPS_DRIVE, QEMU_CAPS_DRIVE_FORMAT); DO_TEST("disk-drive-no-boot", QEMU_CAPS_DRIVE, QEMU_CAPS_DEVICE, QEMU_CAPS_BOOTINDEX); DO_TEST("disk-usb", NONE); -- 1.8.2.1

3 3

[libvirt] [PATCHv2] storage: Ensure 'qemu-img resize' size arg is a 512 multiple
by Christophe Fergeau 14 May '13

14 May '13

qemu-img resize will fail with "The new size must be a multiple of 512" if libvirt doesn't round it first. This fixes rhbz#951495 Signed-off-by: Christophe Fergeau <cfergeau(a)redhat.com> --- Changes since v1: - move rounding code to a new VIR_ROUND_UP() macro - add a note in virStorageVolResize API doc that the requested capacity may be rounded up src/internal.h | 3 +++ src/libvirt.c | 3 ++- src/storage/storage_backend_fs.c | 4 ++++ 3 files changed, 9 insertions(+), 1 deletion(-) diff --git a/src/internal.h b/src/internal.h index d819aa3..28540f4 100644 --- a/src/internal.h +++ b/src/internal.h @@ -339,6 +339,9 @@ /* divide value by size, rounding up */ # define VIR_DIV_UP(value, size) (((value) + (size) - 1) / (size)) +/* round up value to the closest multiple of size */ +# define VIR_ROUND_UP(value, size) (VIR_DIV_UP(value, size) * (size)) + # if WITH_DTRACE_PROBES # ifndef LIBVIRT_PROBES_H diff --git a/src/libvirt.c b/src/libvirt.c index 2b3515e..f2360f8 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -14124,7 +14124,8 @@ error: * Changes the capacity of the storage volume @vol to @capacity. The * operation will fail if the new capacity requires allocation that would * exceed the remaining free space in the parent pool. The contents of - * the new capacity will appear as all zero bytes. + * the new capacity will appear as all zero bytes. The capacity value will + * be rounded to the granularity supported by the hypervisor. * * Normally, the operation will attempt to affect capacity with a minimum * impact on allocation (that is, the default operation favors a sparse diff --git a/src/storage/storage_backend_fs.c b/src/storage/storage_backend_fs.c index 9b83e57..1379f17 100644 --- a/src/storage/storage_backend_fs.c +++ b/src/storage/storage_backend_fs.c @@ -1219,6 +1219,10 @@ virStorageBackendFilesystemResizeQemuImg(const char *path, return -1; } + /* Round capacity as qemu-img resize errors out on sizes which are not + * a multiple of 512 */ + capacity = VIR_ROUND_UP(capacity, 512); + cmd = virCommandNew(img_tool); virCommandAddArgList(cmd, "resize", path, NULL); virCommandAddArgFormat(cmd, "%llu", capacity); -- 1.8.2.1

2 1

[libvirt] [PATCH] virsh: Pretty the output of qemu-agent-command
by Osier Yang 14 May '13

14 May '13

This adds a new option "--pretty" for qemu-agent-command, to pretty-format the returned JSON string. --- tools/virsh-domain.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 546b182..bb94ac1 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -7605,6 +7605,10 @@ static const vshCmdOptDef opts_qemu_agent_command[] = { .type = VSH_OT_BOOL, .help = N_("execute command without timeout") }, + {.name = "pretty", + .type = VSH_OT_BOOL, + .help = N_("pretty-print the output") + }, {.name = "cmd", .type = VSH_OT_ARGV, .flags = VSH_OFLAG_REQ, @@ -7626,6 +7630,7 @@ cmdQemuAgentCommand(vshControl *ctl, const vshCmd *cmd) const vshCmdOpt *opt = NULL; virBuffer buf = VIR_BUFFER_INITIALIZER; bool pad = false; + virJSONValuePtr pretty = NULL; dom = vshCommandOptDomain(ctl, cmd, NULL); if (dom == NULL) @@ -7670,6 +7675,17 @@ cmdQemuAgentCommand(vshControl *ctl, const vshCmd *cmd) } result = virDomainQemuAgentCommand(dom, guest_agent_cmd, timeout, flags); + if (vshCommandOptBool(cmd, "pretty")) { + char *tmp; + pretty = virJSONValueFromString(result); + if (pretty && (tmp = virJSONValueToString(pretty, true))) { + VIR_FREE(result); + result = tmp; + } else { + vshResetLibvirtError(); + } + } + vshPrint(ctl, "%s\n", result); ret = true; -- 1.8.1.4

2 2

[libvirt] [libvirt PATCH] virsh: Use vshPrint instead of printf
by Osier Yang 14 May '13

14 May '13

--- tools/virsh-domain.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 86786ee..546b182 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -7670,7 +7670,7 @@ cmdQemuAgentCommand(vshControl *ctl, const vshCmd *cmd) } result = virDomainQemuAgentCommand(dom, guest_agent_cmd, timeout, flags); - if (result) printf("%s\n", result); + vshPrint(ctl, "%s\n", result); ret = true; -- 1.8.1.4

2 2

[libvirt] [PATCH] Fix invalid argument reference in virnetdev.h
by Martin Kletzander 14 May '13

14 May '13

Commit ccff335f added ATTRIBUTE_NONNULL for an attribute which is not a pointer and made files including virnetdev.h not compilable, so fix that. --- Pushed under the build-breaker rule. src/util/virnetdev.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/util/virnetdev.h b/src/util/virnetdev.h index 0b394ad..bc0777c 100644 --- a/src/util/virnetdev.h +++ b/src/util/virnetdev.h @@ -48,7 +48,7 @@ int virNetDevAddRoute(const char *ifname, virSocketAddrPtr gateway, unsigned int metric) ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) ATTRIBUTE_NONNULL(4) - ATTRIBUTE_NONNULL(5) ATTRIBUTE_RETURN_CHECK; + ATTRIBUTE_RETURN_CHECK; int virNetDevClearIPv4Address(const char *ifname, virSocketAddr *addr, unsigned int prefix) -- 1.8.2.1

1 0

[libvirt] [libvirt PATCH 0/8] Fixes for cgroup setting
by Osier Yang 14 May '13

14 May '13

1/8 ~ 5/8 are refactorings, real fixes are 6/8, 7/8. 8/8 is RFC. Osier Yang (8): qemu: Abstract the code for blkio controller setting into a helper qemu: Abstract code for memory controller setting into a helper qemu: Abstract code for devices controller setting into a helper qemu: Abstract code for cpuset controller setting into a helper qemu: Abstract code for the cpu controller setting into a helper qemu: Set cpuset.cpus for domain process qemu: Prohibit getting the numa parameters if mode is not strict qemu: Set cpuset.mems even if the numatune mode is not strict src/qemu/qemu_cgroup.c | 564 +++++++++++++++++++++++++++++-------------------- src/qemu/qemu_driver.c | 11 +- 2 files changed, 348 insertions(+), 227 deletions(-) -- 1.8.1.4

2 14

[libvirt] libxl requires libuuid?
by Eric Blake 14 May '13

14 May '13

I got this compile error on a relatively new rawhide machine install: CC libvirt_driver_libxl_impl_la-libxl_conf.lo In file included from /usr/include/libxl.h:259:0, from libxl/libxl_conf.c:29: /usr/include/libxl_uuid.h:27:23: fatal error: uuid/uuid.h: No such file or directory #include <uuid/uuid.h> ^ compilation terminated. Is this a case of the libxl development package not pulling in enough pre-reqs? Does libvirt need to work around this, or do I just file a bug against libxl? -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

2 2

[libvirt] [PATCH] storage: Ensure 'qemu-img resize' size arg is a 512 multiple
by Christophe Fergeau 14 May '13

14 May '13

qemu-img resize will fail with "The new size must be a multiple of 512" if libvirt doesn't round it first. This fixes rhbz#951495 Signed-off-by: Christophe Fergeau <cfergeau(a)redhat.com> --- src/storage/storage_backend_fs.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/storage/storage_backend_fs.c b/src/storage/storage_backend_fs.c index 9b83e57..99973b0 100644 --- a/src/storage/storage_backend_fs.c +++ b/src/storage/storage_backend_fs.c @@ -1219,6 +1219,10 @@ virStorageBackendFilesystemResizeQemuImg(const char *path, return -1; } + /* Round capacity up to the closest 512 multiple as qemu-img errors out + * on sizes which are not a multiple of 512 */ + capacity = (capacity + 511) / 512 * 512; + cmd = virCommandNew(img_tool); virCommandAddArgList(cmd, "resize", path, NULL); virCommandAddArgFormat(cmd, "%llu", capacity); -- 1.8.2.1

4 4

[libvirt] [PATCH] docs: Fix the wrong links in secret documentation
by Osier Yang 14 May '13

14 May '13

docs/formatsecret.html.in: (s/domain\.html/formatdomain\.html/g) --- Pushed under trivial rule. --- docs/formatsecret.html.in | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/formatsecret.html.in b/docs/formatsecret.html.in index cfea459..50c9533 100644 --- a/docs/formatsecret.html.in +++ b/docs/formatsecret.html.in @@ -64,7 +64,7 @@ a single <code>name</code> element that specifies a usage name for the secret. The Ceph secret can then be used by UUID or by this usage name via the <code><auth></code> element of - a <a href="domain.html#elementsDisks">disk + a <a href="formatdomain.html#elementsDisks">disk device</a>. Since 0.9.7. @@ -76,7 +76,7 @@ a single <code>target</code> element that specifies a usage name for the secret. The iSCSI secret can then be used by UUID or by this usage name via the <code><auth></code> element of - a <a href="domain.html#elementsDisks">disk + a <a href="formatdomain.html#elementsDisks">disk device</a>. Since 1.0.4. -- 1.8.1.4

1 0

[libvirt] [PATCH] docs: Add the missed usage type 'iscsi'
by Osier Yang 14 May '13

14 May '13

Pushed under trivial rule. --- docs/formatsecret.html.in | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/formatsecret.html.in b/docs/formatsecret.html.in index f816b50..cfea459 100644 --- a/docs/formatsecret.html.in +++ b/docs/formatsecret.html.in @@ -41,8 +41,8 @@ <dd> Specifies what this secret is used for. A mandatory <code>type</code> attribute specifies the usage category, currently - only <code>volume</code> and <code>ceph</code> are defined. - Specific usage categories are described below. + only <code>volume</code>, <code>ceph</code> and <code>iscsi</code> + are defined. Specific usage categories are described below. </dd> </dl> -- 1.8.1.4

1 0