January 2013 - Devel - libvirt List Archives

[libvirt] persistent virtual networks
by Gene Czarcinski 05 Feb '13

05 Feb '13

I seem to remember that, if you use net-define, the network will be persistent and, if you use net-create, the network will not be persistent. I am now running libvirt-1.0.2 on Fedora 18. When I use net-define a network from a template and then do a net-list --all, the network is NOT marked as persistent. If I restart libvirtd.service, then the defined networks are persistent. I noticed that the libvirt-1.0.2 announcement included: - network: bridge: Fix regression when defining persistent networks (Peter Krempa) Do I have a bug to report? Gene

3 2

[libvirt] [PATCH 0/4] FreeBSD: netdev support
by Roman Bogorodskiy 05 Feb '13

05 Feb '13

Add support for some essential netdev functions. Roman Bogorodskiy (4): FreeBSD: implement virNetDevExists() and virNetDevSetName() FreeBSD: implement virNetDevTapCreate() and virNetDevTapDelete(). FreeBSD: implement virNetDevSetMAC(). FreeBSD: implement virNetDevGetMTU(). src/util/virnetdev.c | 113 ++++++++++++++++++++++++++++++++++++-- src/util/virnetdevtap.c | 140 ++++++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 247 insertions(+), 6 deletions(-) -- 1.8.0

4 36

[libvirt] [PATCH 00/10] -numa option parsing fixes & improvements
by Eduardo Habkost 05 Feb '13

05 Feb '13

This series contains the following: * Patches 1-7 are multiple bug fixes to the current code * Patch 8 introduce a feature that libvirt requires since a long time, and even tries to use it today (in a way that doesn't work, using the "-numa node,cpus=1,2,3,4" format): having non-contiguous CPU ranges assigned to a NUMA node. The last 2 patches I am sending as RFCs: * Patch 9 makes the "-numa" option deprecated and introduces a "-numa-node" command-line and config file option. * Patch 10 adds a small hack to the (now deprecated) "-numa" option, that makes the "cpus=1,2,3,4" format currently used by libvirt work. Eduardo Habkost (10): vl.c: Fix off-by-one bug when handling "-numa node" argument vl.c: Abort on unknown -numa option type vl.c: Isolate code specific to "-numa node" option type vl.c: Check for NUMA node limit inside numa_node_add() vl.c: Extract -numa "cpus" parsing to separate function vl.c: handle invalid NUMA CPU ranges properly vl.c: numa_add_node(): Validate nodeid before using it vl.c: Support multiple CPU ranges on -numa option vl.c: Introduce QemuOpts-friendly "-numa-node" config option vl.c: Handle legacy "-numa node,cpus=A,B,C,D" format qemu-config.c | 25 +++++++ qemu-options.hx | 50 +++++++++++++- vl.c | 204 ++++++++++++++++++++++++++++++++++++++++++++++---------- 3 files changed, 240 insertions(+), 39 deletions(-) -- 1.7.11.7

3 15

[libvirt] [PATCH] complete virterror->virerror name change
by Serge Hallyn 05 Feb '13

05 Feb '13

Without these two string changes in generator.py, the virGetLastError wrapper does not get created in /usr/share/pyshared/libvirt.py. Noticed when running tests with virt-install. Signed-off-by: Serge Hallyn <serge.hallyn(a)ubuntu.com> --- python/generator.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/python/generator.py b/python/generator.py index 5d27f66..71ca883 100755 --- a/python/generator.py +++ b/python/generator.py @@ -123,7 +123,7 @@ class docParser(xml.sax.handler.ContentHandler): self.function_return_field = attrs['field'] elif tag == 'enum': if (attrs['file'] == "libvirt" or - attrs['file'] == "virterror"): + attrs['file'] == "virerror"): enum(attrs['type'],attrs['name'],attrs['value']) elif attrs['file'] == "libvirt-lxc": lxc_enum(attrs['type'],attrs['name'],attrs['value']) @@ -137,7 +137,7 @@ class docParser(xml.sax.handler.ContentHandler): if self.function != None: if (self.function_module == "libvirt" or self.function_module == "virevent" or - self.function_module == "virterror"): + self.function_module == "virerror"): function(self.function, self.function_descr, self.function_return, self.function_args, self.function_file, self.function_module, -- 1.8.0

4 6

[libvirt] [PATCH 0/9] Glue domain and storage
by Osier Yang 05 Feb '13

05 Feb '13

This is the 4th part to implement NPIV migration support [1]. Part 1: https://www.redhat.com/archives/libvir-list/2013-January/msg00859.html Part 2: (Already ACKed by Michal) https://www.redhat.com/archives/libvir-list/2013-January/msg00859.html Part 3: https://www.redhat.com/archives/libvir-list/2013-January/msg01012.html To achieve the NPIV migration support, Part 2, Part 3 and this set are necessary, Part 1 is optional. However, this set can be reviewd independantly. ============ This introduces new XMLs to specify the disk source with storage like <disk type='volume' device='disk'> <driver name='qemu' type='raw' cache='none'/> <source pool="$pool_name" volume='$vol_name'/> <seclabel relabel='no'/> </source> <target dev='vdb' bus='virtio'/> </disk> And before domain starting, the source represented by storage is translated into the real underlying source. Diff with RFC: * The XMLs are more simpler - only using pool name and volume name to specify disk source. * Support network pool (rbd, and sheepdog) * Support startupPolicy for volume type disk * Support seclabels for volume type disk * Fix bugs on disk source formating Osier Yang (9): rng: Add definition for network disk source Introduce new XMLs to specify disk source using libvirt storage New files to glue domain and storage together Implement translateDiskSourcePool qemu: Translate the source when starting domain Support startupPolicy for 'volume' disk conf: Fix bugs on disk source formating Support seclabels for volume type disk Support network pool for volume disk docs/formatdomain.html.in | 49 +++-- docs/schemas/domaincommon.rng | 111 +++++--- src/Makefile.am | 3 +- src/conf/domain_conf.c | 278 ++++++++++++++------ src/conf/domain_conf.h | 9 + src/conf/domain_storage.c | 44 +++ src/conf/domain_storage.h | 37 +++ src/libvirt_private.syms | 5 + src/qemu/qemu_driver.c | 7 + src/storage/storage_driver.c | 144 ++++++++++ .../qemuxml2argv-disk-source-pool.xml | 33 +++ tests/qemuxml2xmltest.c | 1 + 12 files changed, 573 insertions(+), 148 deletions(-) create mode 100644 src/conf/domain_storage.c create mode 100644 src/conf/domain_storage.h create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-source-pool.xml [1] https://www.redhat.com/archives/libvir-list/2012-November/msg00826.html Osier

3 29

[libvirt] [PATCH 00/17] virsh: Rework string option retrieval
by Peter Krempa 04 Feb '13

04 Feb '13

This patchset fixes retrieval of string arguments in virsh to report sane errors in all the places it's used. A new string option retrieval function is added that supports error reporting and changes appropriate places to use it. Along with that this series cleans up a few places with old code and bad style. *** BLURB HERE *** Peter Krempa (17): virsh-snapshot: Fix coding style and remove dead code virsh-volume: Rename helper function makeCloneXML to vshMakeCloneXML virsh-pool: Rename helper function buildPoolXML to vshBuildPoolXML virsh: Add helper to request string arguments with error reporting virsh-pool: Update pool commands to use vshCommandOptStringReq virsh-volume: Update volume commands to use vshCommandOptStringReq virsh-domain-monitor: Update domain commands to use vshCommandOptStringReq virsh-domain: Update domain commands to use vshCommandOptStringReq virsh-host: Update host commands to use vshCommandOptStringReq virsh-network: Update network commands to use vshCommandOptStringReq virsh-interface: Update interface commands to use vshCommandOptStringReq virsh-nodedev: Update node device commands to use vshCommandOptStringReq virsh-nodedev: Refactor error paths, error messages and whitespace virsh-snapshot: Refactor cmdSnapshotDumpXML virsh-snapshot: Update snapshot commands to use vshCommandOptStringReq virsh-nwfilter: Update nwfilter commands to use vshCommandOptStringReq virsh-secret: Refactor error paths tools/virsh-domain-monitor.c | 44 +++---- tools/virsh-domain.c | 299 +++++++++++++++++-------------------------- tools/virsh-host.c | 9 +- tools/virsh-interface.c | 8 +- tools/virsh-network.c | 19 +-- tools/virsh-nodedev.c | 44 ++++--- tools/virsh-nwfilter.c | 4 +- tools/virsh-pool.c | 56 ++++---- tools/virsh-secret.c | 30 +++-- tools/virsh-snapshot.c | 55 +++----- tools/virsh-volume.c | 73 +++++------ tools/virsh.c | 51 ++++++++ tools/virsh.h | 4 + 13 files changed, 320 insertions(+), 376 deletions(-) -- 1.8.1.1

2 44

[libvirt] [PATCH] rpc: Fix crash on error paths of message dispatching
by Peter Krempa 04 Feb '13

04 Feb '13

When reading and dispatching of a message failed the message was freed but wasn't removed from the message queue. After that when the connection was about to be closed the pointer for the message was still present in the queue and it was passed to virNetMessageFree which tried to call the callback function from an uninitialized pointer. This patch removes the message from the queue before it's freed. * rpc/virnetserverclient.c: virNetServerClientDispatchRead: - avoid use after free of RPC messages --- src/rpc/virnetserverclient.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/rpc/virnetserverclient.c b/src/rpc/virnetserverclient.c index af0560e..446e1e9 100644 --- a/src/rpc/virnetserverclient.c +++ b/src/rpc/virnetserverclient.c @@ -987,6 +987,7 @@ readmore: /* Decode the header so we can use it for routing decisions */ if (virNetMessageDecodeHeader(msg) < 0) { + virNetMessageQueueServe(&client->rx); virNetMessageFree(msg); client->wantClose = true; return; @@ -996,6 +997,7 @@ readmore: * file descriptors */ if (msg->header.type == VIR_NET_CALL_WITH_FDS && virNetMessageDecodeNumFDs(msg) < 0) { + virNetMessageQueueServe(&client->rx); virNetMessageFree(msg); client->wantClose = true; return; /* Error */ @@ -1005,6 +1007,7 @@ readmore: for (i = msg->donefds ; i < msg->nfds ; i++) { int rv; if ((rv = virNetSocketRecvFD(client->sock, &(msg->fds[i]))) < 0) { + virNetMessageQueueServe(&client->rx); virNetMessageFree(msg); client->wantClose = true; return; -- 1.8.1.1

5 9

[libvirt] [PATCH] hyperv: Avoid Coverity warning about resource leak in hypervOpen
by Matthias Bolte 02 Feb '13

02 Feb '13

Coverity tagged the esxhypervPrivate pointer a potentially leaked because of the conditional free call. But this is a false positive, there is not actual leak. Avoid the conditial code to make this more obvious to Coverity. --- This patch is meant as a replacement for this patch https://www.redhat.com/archives/libvir-list/2013-January/msg00542.html src/hyperv/hyperv_driver.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c index 601a85a..0d2eaae 100644 --- a/src/hyperv/hyperv_driver.c +++ b/src/hyperv/hyperv_driver.c @@ -200,14 +200,11 @@ hypervOpen(virConnectPtr conn, virConnectAuthPtr auth, unsigned int flags) } conn->privateData = priv; - + priv = NULL; result = VIR_DRV_OPEN_SUCCESS; cleanup: - if (result == VIR_DRV_OPEN_ERROR) { - hypervFreePrivate(&priv); - } - + hypervFreePrivate(&priv); VIR_FREE(username); VIR_FREE(password); hypervFreeObject(priv, (hypervObject *)computerSystem); -- 1.7.9.5

2 3

[libvirt] [PATCH] esx: Avoid Coverity warning about resource leak in esxOpen
by Matthias Bolte 02 Feb '13

02 Feb '13

Commit 4445e16bfa8056980ac643fabf17186f9e685925 changed the signature of esxConnectToHost and esxConnectToVCenter by replacing the esxPrivate pointer with virConnectPtr. The esxPrivate pointer was then retrieved again from virConnectPtr's privateData. This resulted in a NULL pointer dereference, because the privateData pointer was not yet initialized at the point where esxConnectToHost and esxConnectToVCenter are called. This was fixed in commit b126715a48cd0cbe32ec6468c267cd8cf2961c55 that moved the initialization of privateData before the problematic calls. Coverity tagged the esxPrivate pointer a potentially leaked because of the conditional free call. But this is a false positive, there is not actual leak. Avoid this warning from Coverity by making the call to esxFreePrivate unconditional and changing esxConnectToHost and esxConnectToVCenter back to take a esxPrivate pointer directly. This allows to assign esxPrivate to the virConnectPtr's privateData pointer as one of the last steps in esxOpen making it more obvious that it is not initialized during the earlier steps of esxOpen. --- This patch is meant as a replacement this patch: https://www.redhat.com/archives/libvir-list/2013-January/msg00530.html src/esx/esx_driver.c | 23 ++++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-) diff --git a/src/esx/esx_driver.c b/src/esx/esx_driver.c index 1366c81..dad10a1 100644 --- a/src/esx/esx_driver.c +++ b/src/esx/esx_driver.c @@ -650,7 +650,8 @@ esxCapsInit(esxPrivate *priv) static int -esxConnectToHost(virConnectPtr conn, +esxConnectToHost(esxPrivate *priv, + virConnectPtr conn, virConnectAuthPtr auth, char **vCenterIpAddress) { @@ -663,7 +664,6 @@ esxConnectToHost(virConnectPtr conn, esxVI_String *propertyNameList = NULL; esxVI_ObjectContent *hostSystem = NULL; esxVI_Boolean inMaintenanceMode = esxVI_Boolean_Undefined; - esxPrivate *priv = conn->privateData; esxVI_ProductVersion expectedProductVersion = STRCASEEQ(conn->uri->scheme, "esx") ? esxVI_ProductVersion_ESX : esxVI_ProductVersion_GSX; @@ -785,7 +785,8 @@ esxConnectToHost(virConnectPtr conn, static int -esxConnectToVCenter(virConnectPtr conn, +esxConnectToVCenter(esxPrivate *priv, + virConnectPtr conn, virConnectAuthPtr auth, const char *hostname, const char *hostSystemIpAddress) @@ -796,7 +797,6 @@ esxConnectToVCenter(virConnectPtr conn, char *unescapedPassword = NULL; char *password = NULL; char *url = NULL; - esxPrivate *priv = conn->privateData; if (hostSystemIpAddress == NULL && (priv->parsedUri->path == NULL || STREQ(priv->parsedUri->path, "/"))) { @@ -1008,8 +1008,6 @@ esxOpen(virConnectPtr conn, virConnectAuthPtr auth, priv->supportsLongMode = esxVI_Boolean_Undefined; priv->usedCpuTimeCounterId = -1; - conn->privateData = priv; - /* * Set the port dependent on the transport protocol if no port is * specified. This allows us to rely on the port parameter being @@ -1036,7 +1034,7 @@ esxOpen(virConnectPtr conn, virConnectAuthPtr auth, if (STRCASEEQ(conn->uri->scheme, "esx") || STRCASEEQ(conn->uri->scheme, "gsx")) { /* Connect to host */ - if (esxConnectToHost(conn, auth, + if (esxConnectToHost(priv, conn, auth, &potentialVCenterIpAddress) < 0) { goto cleanup; } @@ -1075,7 +1073,7 @@ esxOpen(virConnectPtr conn, virConnectAuthPtr auth, } } - if (esxConnectToVCenter(conn, auth, + if (esxConnectToVCenter(priv, conn, auth, vCenterIpAddress, priv->host->ipAddress) < 0) { goto cleanup; @@ -1085,7 +1083,7 @@ esxOpen(virConnectPtr conn, virConnectAuthPtr auth, priv->primary = priv->host; } else { /* VPX */ /* Connect to vCenter */ - if (esxConnectToVCenter(conn, auth, + if (esxConnectToVCenter(priv, conn, auth, conn->uri->server, NULL) < 0) { goto cleanup; @@ -1101,13 +1099,12 @@ esxOpen(virConnectPtr conn, virConnectAuthPtr auth, goto cleanup; } + conn->privateData = priv; + priv = NULL; result = VIR_DRV_OPEN_SUCCESS; cleanup: - if (result == VIR_DRV_OPEN_ERROR) { - esxFreePrivate(&priv); - } - + esxFreePrivate(&priv); VIR_FREE(potentialVCenterIpAddress); return result; -- 1.7.9.5

3 6

[libvirt] [PATCH] maint: update to latest gnulib
by Eric Blake 02 Feb '13

02 Feb '13

CVE-2013-0242 in glibc's regex() can cause a DoS in any daemon that runs a regex search on user input while in a multibyte locale. I'm not sure how hard it would be to trigger such a setup for libvirtd, but rather than risk things, we can avoid the issue: gnulib has worked around the problem, and by updating to the latest gnulib, we can avoid the bug even on platforms where glibc has yet to be patched. * .gnulib: Update to latest, for various fixes, including regex. * bootstrap: Resync from upstream. --- * .gnulib 61c7b1e...a0b25da (45): > regex: avoid infinite configure test > openpty: fix bug where HAVE_OPENPTY wasn't defined > sys_time: port to Solaris 2.6 > regex: test for buffer overrun > regex: fix buffer overrun in regexp matcher > mountlist: don't consider "devtmpfs" as dummy > futimens-tests, utimens-tests: Depend on gettext. > test-getpeername: fix typo > bootstrap: remove the need for a sorted .gitignore > readlinkat: don't depend on gl_FUNC_OPENAT > statat: new module, split out from fstatat > autoupdate > Fix typo in previous change, by including <unistd.h>. > tests: don't assume fd 99 is closed > Fix ChangeLog typo. > stpncpy: port to OS X 10.8 > In intro, say only latest mingw is tested. > unistd: port to recent mingw > Fix typo in comment. > largefile: port better to Mac OS X 10.5 > doc: clarify -Werror > stdint: fix build with Android's Bionic fox x86 > net_if-tests: port to Solaris 7 + GCC 3.4.6 > net_if-tests: port to older Solaris > system-quote-tests: port to older Solaris > c-xvasprintf etc.: fix link errors on older Solaris > locale: port to Solaris 2.6 and 7 + GNU gettext > autoupdate > stdlib: port to Solaris 2.6 > Fix misspellings of "occurred". > autoupdate > autoupdate > doc: update main copyright year > doc: improve ISO 8601 discussion > doc: avoid small caps > regex: conform to strict C > gnulib-tool: fix incompatibility with autopoint 0.18.2 > autoupdate > fprintftime: bring back and reword fwrite comment > autoupdate > stdio: remove now-unnecessary stdio.c > unicodeio: depend on stdio, not ignore-value > fprintftime: depend on stdio, not ignore-value > stdint-tests: Fix expanded-before-required-warning. * modules/stdint-tests (Depends-on): Use AC_REQUIRE. > fwrite: silence __wur only for older glibc versions .gnulib | 2 +- bootstrap | 44 ++++++++++++++++++++++---------------------- 2 files changed, 23 insertions(+), 23 deletions(-) diff --git a/.gnulib b/.gnulib index 61c7b1e..a0b25da 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit 61c7b1e32e11e9e40b4d59ab888a807620befcd3 +Subproject commit a0b25daf001cb1a0352f329e2b5e1640acc88541 diff --git a/bootstrap b/bootstrap index 012907a..bee7765 100755 --- a/bootstrap +++ b/bootstrap @@ -1,6 +1,6 @@ #! /bin/sh # Print a version string. -scriptversion=2012-12-28.10; # UTC +scriptversion=2013-01-20.16; # UTC # Bootstrap this package from checked-out sources. @@ -306,34 +306,34 @@ if test -n "$checkout_only_file" && test ! -r "$checkout_only_file"; then die "Bootstrapping from a non-checked-out distribution is risky." fi -# Ensure that lines starting with ! sort last, per gitignore conventions -# for whitelisting exceptions after a more generic blacklist pattern. -sort_patterns() { - sort -u "$@" | sed '/^!/ { - H - d - } - $ { - P - x - s/^\n// - }' | sed '/^$/d' +# Strip blank and comment lines to leave significant entries. +gitignore_entries() { + sed '/^#/d; /^$/d' "$@" } -# If $STR is not already on a line by itself in $FILE, insert it, -# sorting the new contents of the file and replacing $FILE with the result. -insert_sorted_if_absent() { +# If $STR is not already on a line by itself in $FILE, insert it at the start. +# Entries are inserted at the start of the ignore list to ensure existing +# entries starting with ! are not overridden. Such entries support +# whitelisting exceptions after a more generic blacklist pattern. +insert_if_absent() { file=$1 str=$2 test -f $file || touch $file - echo "$str" | sort_patterns - $file | cmp -s - $file > /dev/null \ - || { echo "$str" | sort_patterns - $file > $file.bak \ - && mv $file.bak $file; } \ - || die "insert_sorted_if_absent $file $str: failed" + test -r $file || die "Error: failed to read ignore file: $file" + duplicate_entries=$(gitignore_entries $file | sort | uniq -d) + if [ "$duplicate_entries" ] ; then + die "Error: Duplicate entries in $file: " $duplicate_entries + fi + linesold=$(gitignore_entries $file | wc -l) + linesnew=$(echo "$str" | gitignore_entries - $file | sort -u | wc -l) + if [ $linesold != $linesnew ] ; then + { echo "$str" | cat - $file > $file.bak && mv $file.bak $file; } \ + || die "insert_if_absent $file $str: failed" + fi } # Adjust $PATTERN for $VC_IGNORE_FILE and insert it with -# insert_sorted_if_absent. +# insert_if_absent. insert_vc_ignore() { vc_ignore_file="$1" pattern="$2" @@ -344,7 +344,7 @@ insert_vc_ignore() { # .gitignore entry. pattern=$(echo "$pattern" | sed s,^,/,);; esac - insert_sorted_if_absent "$vc_ignore_file" "$pattern" + insert_if_absent "$vc_ignore_file" "$pattern" } # Die if there is no AC_CONFIG_AUX_DIR($build_aux) line in configure.ac. -- 1.8.1

2 2