December 2012 - Devel - libvirt List Archives

[libvirt] [PATCH] conf: cpu: Fix to escape cpu model, vendor_id and vendor
by Ken ICHIKAWA 20 Dec '12

20 Dec '12

This patch fixes a problem that encoded xml special charas of cpu model, vendor_id and vendor are saved as decoded xml special charas. If I define model name including encoded xml special charas in domain XML like below: <domain> <cpu mode='custom' match='exact'> <model fallback='allow'>><</model> </cpu> </domain> domain xml definition file is saved as <domain> <cpu mode='custom' match='exact'> <model fallback='allow'>><</model> </cpu> </domain> After that, if libvirtd is restarted, dumpxml command, edit command and etc. are disabled with below error until I stop libvirtd and fix the xml file manually. 2012-12-20 04:56:54.371+0000: 11893: error : catchXMLError:691 : /etc/libvirt/qemu/f17.xml:25: StartTag: invalid element name <model fallback='allow'>><</model> ------------------------------^ Also vendor_id and vendor have same problem. Signed-off-by: Ken ICHIKAWA <ichikawa.ken(a)jp.fujitsu.com> --- src/conf/cpu_conf.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/conf/cpu_conf.c b/src/conf/cpu_conf.c index 3973c83..e9f9431 100644 --- a/src/conf/cpu_conf.c +++ b/src/conf/cpu_conf.c @@ -609,17 +609,17 @@ virCPUDefFormatBuf(virBufferPtr buf, } virBufferAsprintf(buf, " fallback='%s'", fallback); if (def->vendor_id) - virBufferAsprintf(buf, " vendor_id='%s'", def->vendor_id); + virBufferEscapeString(buf, " vendor_id='%s'", def->vendor_id); } if (formatModel && def->model) { - virBufferAsprintf(buf, ">%s</model>\n", def->model); + virBufferEscapeString(buf, ">%s</model>\n", def->model); } else { virBufferAddLit(buf, "/>\n"); } } if (formatModel && def->vendor) - virBufferAsprintf(buf, "<vendor>%s</vendor>\n", def->vendor); + virBufferEscapeString(buf, "<vendor>%s</vendor>\n", def->vendor); if (def->sockets && def->cores && def->threads) { virBufferAddLit(buf, "<topology"); -- 1.7.11.7

2 1

[libvirt] [PATCH v2.1 00/11] Rework storage migration
by Michal Privoznik 20 Dec '12

20 Dec '12

This patch set re-implements migration with storage for enough new qemu. Currently, you can migrate a domain to a host without need for shared storage. This is done by setting 'blk' or 'inc' attribute (representing VIR_MIGRATE_NON_SHARED_DISK and VIR_MIGRATE_NON_SHARED_INC flags respectively) of 'migrate' monitor command. However, the qemu implementation is buggy and applications are advised to switch to new impementation which, moreover, offers some nice features, like migrating only explicitly specified disks. The new functionality is controlled via 'nbd-server-*' and 'drive-mirror' commands. The flow is meant to look like this: 1) User invokes libvirt's migrate functionality. 2) libvirt checks that no block jobs are active on the source. 3) libvirt starts the destination QEMU and sets up the NBD server using the nbd-server-start and nbd-server-add commands. 4) libvirt starts drive-mirror with a destination pointing to the remote NBD server, for example nbd:host:port:exportname=diskname (where diskname is the -drive id specified on the destination). 5) once all mirroring jobs reach steady state, libvirt invokes the migrate command. 6) once migration completed, libvirt invokes the nbd-server-stop command on the destination QEMU. If we just skip the 2nd step and there is an active block-job, qemu will fail in step 4. No big deal. Since we try to NOT break migration and keep things compatible, this feature is enabled iff both sides support it. Since there's obvious need for some data transfer between src and dst, I've put it into qemuCookieMigration: 1) src -> dest: (QEMU_MIGRATION_PHASE_BEGIN3 -> QEMU_MIGRATION_PHASE_PREPARE) <nbd> <disk size='17179869184'/> </nbd> Hey destination, I know how to use this cool new feature. Moreover, these are the disks I'll send you. Each one of them is X bytes big. It's one of the prerequisite - the file (disk->src) on dst exists and has at least the same size as on dst. 2) dst -> src: (QEMU_MIGRATION_PHASE_PREPARE -> QEMU_MIGRATION_PHASE_PERFORM3) <nbd port='X'/> Okay, I (destination) support this feature as well. I've created all files as you (src) told me to and you can start rolling data. I am listening on port X. 3) src -> dst: (QEMU_MIGRATION_PHASE_PERFORM3 -> QEMU_MIGRATION_PHASE_FINISH3) <nbd port='-1'/> Migration completed, destination, you may shut the NBD server down. If either src or dst doesn't support NBD, it is not used and whole process fall backs to old implementation. diff to v1: -Eric's and Daniel's suggestions worked in. To point out the bigger ones: don't do NBD style when TUNNELLED requested, added 'b:writable' to 'nbd-server-add' -drop '/qemu-migration/nbd/disk/@src' attribute from migration cookie. As pointed out by Jirka, disk->src can be changed during migration (e.g. by migration hook or by passed xml). So I've tried (as suggested on the list) passing disk alias. However, since qemu hasn't been started on destination yet, the aliases hasn't been generated yet. So we have to rely on ordering completely. diff to v2: -rebase to reflect changes made by offline migration patch -send initial nbd cookie only if needed The patches 1,3 and 5 has been ACKed already. Michal Privoznik (11): qemu: Introduce NBD_SERVER capability Introduce NBD migration cookie qemu: Introduce nbd-server-start command qemu: Introduce nbd-server-add command qemu: Introduce nbd-server-stop command qemu_migration: Introduce qemuMigrationStartNBDServer qemu_migration: Move port allocation to a separate func qemu_migration: Implement qemuMigrationStartNBDServer() qemu_migration: Implement qemuMigrationDriveMirror qemu_migration: Check size prerequisites qemu_migration: Stop NBD server at Finish phase src/qemu/qemu_capabilities.c | 3 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_migration.c | 624 ++++++++++++++++++++++++++++++++++++++++--- src/qemu/qemu_monitor.c | 63 +++++ src/qemu/qemu_monitor.h | 7 + src/qemu/qemu_monitor_json.c | 95 +++++++ src/qemu/qemu_monitor_json.h | 7 + 7 files changed, 768 insertions(+), 32 deletions(-) -- 1.8.0.2

2 14

[libvirt] [PATCH] maint: formatting cleanups in buf.c
by Eric Blake 19 Dec '12

19 Dec '12

* src/util/buf.c: Use consistent formatting. --- Pushing under the trivial rule. src/util/buf.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/src/util/buf.c b/src/util/buf.c index e5a6ab7..e5406da 100644 --- a/src/util/buf.c +++ b/src/util/buf.c @@ -106,8 +106,8 @@ virBufferGetIndent(const virBufferPtr buf, bool dynamic) /** * virBufferGrow: - * @buf: the buffer - * @len: the minimum free size to allocate on top of existing used space + * @buf: the buffer + * @len: the minimum free size to allocate on top of existing used space * * Grow the available space of a buffer to at least @len bytes. * @@ -282,8 +282,8 @@ virBufferUse(const virBufferPtr buf) /** * virBufferAsprintf: * @buf: the buffer to append to - * @format: the format - * @...: the variable list of arguments + * @format: the format + * @...: the variable list of arguments * * Do a formatted print to an XML buffer. Auto indentation may be applied. */ @@ -299,8 +299,8 @@ virBufferAsprintf(virBufferPtr buf, const char *format, ...) /** * virBufferVasprintf: * @buf: the buffer to append to - * @format: the format - * @argptr: the variable list of arguments + * @format: the format + * @argptr: the variable list of arguments * * Do a formatted print to an XML buffer. Auto indentation may be applied. */ @@ -458,7 +458,7 @@ virBufferEscapeSexpr(virBufferPtr buf, virBufferEscape(buf, '\\', "\\'", format, str); } -/* Work around spurious strchr() diagnostics given by -Wlogical-op * +/* Work around spurious strchr() diagnostics given by -Wlogical-op * for gcc < 4.6. Doing it via a local pragma keeps the damage * smaller than disabling it on the package level. Unfortunately, the * affected GCCs don't allow diagnostic push/pop which would have @@ -522,7 +522,7 @@ virBufferEscape(virBufferPtr buf, char escape, const char *toescape, /** * virBufferURIEncodeString: * @buf: the buffer to append to - * @str: the string argument which will be URI-encoded + * @str: the string argument which will be URI-encoded * * Append the string to the buffer. The string will be URI-encoded * during the append (ie any non alpha-numeric characters are replaced @@ -630,7 +630,7 @@ virBufferEscapeShell(virBufferPtr buf, const char *str) /** * virBufferStrcat: * @buf: the buffer to append to - * @...: the variable list of strings, the last argument must be NULL + * @...: the variable list of strings, the last argument must be NULL * * Concatenate strings to an XML buffer. Auto indentation may be applied * after each string argument. -- 1.8.0.2

1 0

[libvirt] [PATCH] build: make broken -Wlogical-op test be gcc-only
by Eric Blake 19 Dec '12

19 Dec '12

Commit 8b8fcdea introduced a check for broken gcc -Wlogical-op, but did not guard the check against non-gcc compilers, which might lead to spurious failures when another compiler encounters an unknown pragma. Additionally, all of our compiler warning logic should belong in a single file, and use cache variables to allow overriding the decision at configure time if necessary. * configure.ac (BROKEN_GCC_WLOGICALOP): Move... * m4/virt-compile-warnings.m4 (LIBVIRT_COMPILE_WARNINGS): ...here, and update to modern autoconf idioms. --- Fixes the issues mentioned here: https://www.redhat.com/archives/libvir-list/2012-December/msg01197.html https://www.redhat.com/archives/libvir-list/2012-December/msg01211.html configure.ac | 23 ----------------------- m4/virt-compile-warnings.m4 | 22 ++++++++++++++++++++++ 2 files changed, 22 insertions(+), 23 deletions(-) diff --git a/configure.ac b/configure.ac index 20caa92..3c97e4f 100644 --- a/configure.ac +++ b/configure.ac @@ -255,29 +255,6 @@ AC_CHECK_TYPE([struct ifreq], #include <net/if.h> ]]) -dnl Check whether strchr(s, char variable) causes a bogus compile warning -dnl which is the case with a certain range of GCC versions -AC_MSG_CHECKING([whether GCC -Wlogical-op is broken]) - -save_CFLAGS="$CFLAGS" -CFLAGS="-O2 -Wlogical-op -Werror" - -AC_TRY_COMPILE([#include <string.h>], - [const char *haystack; - char needle; - return strchr(haystack, needle) == haystack;], - [gcc_false_strchr_warning=no], - [gcc_false_strchr_warning=yes]) - -CFLAGS="$save_CFLAGS" - -if test "x$gcc_false_strchr_warning" = xyes; then - AC_DEFINE_UNQUOTED([BROKEN_GCC_WLOGICALOP], 1, - [GCC -Wlogical-op is reporting false positive on strchr]) -fi - -AC_MSG_RESULT([$gcc_false_strchr_warning]) - dnl Our only use of libtasn1.h is in the testsuite, and can be skipped dnl if the header is not present. Assume -ltasn1 is present if the dnl header could be found. diff --git a/m4/virt-compile-warnings.m4 b/m4/virt-compile-warnings.m4 index 4a0f289..a08fcac 100644 --- a/m4/virt-compile-warnings.m4 +++ b/m4/virt-compile-warnings.m4 @@ -95,6 +95,22 @@ AC_DEFUN([LIBVIRT_COMPILE_WARNINGS],[ dontwarn="$dontwarn -Wmissing-declarations" fi + dnl Check whether strchr(s, char variable) causes a bogus compile + dnl warning, which is the case with GCC < 4.6 on some glibc + AC_CACHE_CHECK([whether GCC -Wlogical-op gives bogus warnings], + [lv_cv_gcc_wlogical_op_broken], [ + save_CFLAGS="$CFLAGS" + CFLAGS="-O2 -Wlogical-op -Werror" + AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[ + #include <string.h> + ]], [[ + const char *haystack; + char needle; + return strchr(haystack, needle) == haystack;]])], + [lv_cv_gcc_wlogical_op_broken=no], + [lv_cv_gcc_wlogical_op_broken=yes]) + CFLAGS="$save_CFLAGS"]) + # We might fundamentally need some of these disabled forever, but # ideally we'd turn many of them on dontwarn="$dontwarn -Wfloat-equal" @@ -196,4 +212,10 @@ AC_DEFUN([LIBVIRT_COMPILE_WARNINGS],[ WARN_PYTHON_CFLAGS=$WARN_CFLAGS AC_SUBST(WARN_PYTHON_CFLAGS) WARN_CFLAGS=$save_WARN_CFLAGS + + if test "$gl_cv_warn_c__Wlogical_op" = yes && + test "$lv_cv_gcc_wlogical_op_broken" = yes; then + AC_DEFINE_UNQUOTED([BROKEN_GCC_WLOGICALOP], 1, + [Define to 1 if gcc -Wlogical-op reports false positives on strchr]) + fi ]) -- 1.8.0.2

2 2

[libvirt] [PATCH] build: use strchr now that we can work around broken gcc
by Eric Blake 19 Dec '12

19 Dec '12

Revert the complex workaround of commit 39d91e9, now that we have a nicer framework for shutting up broken gcc. * src/util/buf.c (virBufferEscape): Simplify. --- src/util/buf.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/src/util/buf.c b/src/util/buf.c index 030dc97..e5a6ab7 100644 --- a/src/util/buf.c +++ b/src/util/buf.c @@ -458,2 +458,2 @@ virBufferEscapeSexpr(virBufferPtr buf, virBufferEscape(buf, '\\', "\\'", format, str); } +/* Work around spurious strchr() diagnostics given by -Wlogical-op + * for gcc < 4.6. Doing it via a local pragma keeps the damage + * smaller than disabling it on the package level. Unfortunately, the + * affected GCCs don't allow diagnostic push/pop which would have + * further reduced the impact. */ +#if BROKEN_GCC_WLOGICALOP +# pragma GCC diagnostic ignored "-Wlogical-op" +#endif + /** * virBufferEscape: * @buf: the buffer to append to @@ -499,11 +508,7 @@ virBufferEscape(virBufferPtr buf, char escape, const char *toescape, cur = str; out = escaped; while (*cur != 0) { - /* strchr work-around for gcc 4.3 & 4.4 bug with -Wlogical-op - * http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36513 - */ - char needle[2] = { *cur, 0 }; - if (strstr(toescape, needle)) + if (strchr(toescape, *cur)) *out++ = escape; *out++ = *cur; cur++; -- 1.8.0.2

2 1

[libvirt] [PATCH] build: Check for broken GCC -Wlogical-op was too optimistic
by Viktor Mihajlovski 19 Dec '12

19 Dec '12

The check for a broken GCC -Wlogical-op implemententation was written under the wrong assumption that unknown warning options would be gracefully ignored. Thanks to Eric Blake and Daniel Berrange for pointing that out. Signed-off-by: Viktor Mihajlovski <mihajlov(a)linux.vnet.ibm.com> --- configure.ac | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/configure.ac b/configure.ac index 20caa92..828ae01 100644 --- a/configure.ac +++ b/configure.ac @@ -260,7 +260,13 @@ dnl which is the case with a certain range of GCC versions AC_MSG_CHECKING([whether GCC -Wlogical-op is broken]) save_CFLAGS="$CFLAGS" -CFLAGS="-O2 -Wlogical-op -Werror" +for w_opt in $WARN_CFLAGS; do + case $w_opt in + "-Wlogical-op" | "-Werror" ) + CFLAGS="$CFLAGS $w_opt" + ;; + esac +done AC_TRY_COMPILE([#include <string.h>], [const char *haystack; -- 1.7.9.5

2 1

[libvirt] virtulazation error
by Gangadahr Jena 19 Dec '12

19 Dec '12

hi every one when i am start windows os using virtual macine manager on redhat 6 os it showing this type of error-Error starting domain: unsupported configuration: Only PCI device addresses with function=0 are supported with this QEMU binary

2 1

[libvirt] [PATCH 0/2] FreeBSD: nodeinfo support and CPU affinity stubs
by Roman Bogorodskiy 19 Dec '12

19 Dec '12

With this set of changes it's now possible to connect to libvirtd using virsh and start/shutdown domains with the obvious limitation that networking is not supported at this point. CPU affinity most likely could be implemented using cpu_setaffinity(2) and cpu_getaffinity(2) on FreeBSD, but networking support looks more important now. Roman Bogorodskiy (2): FreeBSD: add nodeinfo support. FreeBSD: stub out CPU affinity functions. src/nodeinfo.c | 75 +++++++++++++++++++++++++++++++++++++++++++++++++- src/util/processinfo.c | 22 +++++++++++++++ 2 files changed, 96 insertions(+), 1 deletion(-) -- 1.8.0

3 16

[libvirt] [PATCHv3 0/3] sysinfo: s390 bug fix and tests
by Viktor Mihajlovski 19 Dec '12

19 Dec '12

Primarily this fixes a memory issue in virSysinfoRead on s390. As a side effect it became necessary to work around a GCC bug. Finally I have followed Daniel Berrange's suggestion to provide tests for sysinfo. V2: Turned single patch into series V3: Fixed -Werror induced build break with current GCC Viktor Mihajlovski (3): build: Check for broken GCC -Wlogical-op in configure S390: Fix virSysinfoRead memory corruption tests: Add tests for sysinfo cfg.mk | 2 +- configure.ac | 23 ++++ src/libvirt_private.syms | 1 + src/util/sysinfo.c | 195 +++++++++++++++++---------------- tests/Makefile.am | 6 + tests/sysinfodata/dmidecode.sh | 3 + tests/sysinfodata/s390cpuinfo.data | 12 ++ tests/sysinfodata/s390sysinfo.data | 114 +++++++++++++++++++ tests/sysinfodata/s390sysinfo.expect | 25 +++++ tests/sysinfodata/x86sysinfo.data | 83 ++++++++++++++ tests/sysinfodata/x86sysinfo.expect | 53 +++++++++ tests/sysinfotest.c | 200 ++++++++++++++++++++++++++++++++++ 12 files changed, 622 insertions(+), 95 deletions(-) create mode 100755 tests/sysinfodata/dmidecode.sh create mode 100644 tests/sysinfodata/s390cpuinfo.data create mode 100644 tests/sysinfodata/s390sysinfo.data create mode 100644 tests/sysinfodata/s390sysinfo.expect create mode 100644 tests/sysinfodata/x86sysinfo.data create mode 100644 tests/sysinfodata/x86sysinfo.expect create mode 100644 tests/sysinfotest.c -- 1.7.9.5

3 13

[libvirt] [RFC] Support for qemu's virtio-rng
by Peter Krempa 19 Dec '12

19 Dec '12

Commit 16c915ba42b45df7a64a6908287f03bfa3764bed in the qemu.git tree introduced support for emulating a virtio-rng device used to pass-through entropy to the guests. Qemu provides two backends to gather entropy from: 1) Default chardev backend This backend supports reading non-blocking character devices that provide entropy such as /dev/random (by default) or others according to configuration. This backend also supports basic rate limiting of the entropy read from the source. Unfortunately as the rate limiting is done just on a per-guest basis this cannot protect the host from being starved of entropy by multiple guests although their individual rates are limited. 2) EGD backend This backend uses the EGD protocol to communicate with a daemon that servers entropy. The EGD protocol is a simple protocol that was introduced by the entropy gathering daemon (a substitute for /dev/random on machines that don't support it). This protocol can be used for centralized management of entropy allocation to the hosts. This by itself isn't really useful as implementations of EGD daemons are trying to create the entropy instead of using available sources and multiplexing/rate-limiting them. For this functionality we will (probably) need a separate server to serve the entropy to the guests but more on this topic later. To represent the configuration of the virtio-rng device to the guest I propose the following domain XML snipped to be used: For the first use case I propose: <domain> [...] <devices> [...] <rng model='virtio'> <rate bytes='1024' period='1000'/> <source type='char'>/dev/urandom</source> -- or -- <source type='char'/> -- for the default source -- </rng> </devices> </domain> For the EGD backend used manually (without any managed server) I propose: <rng model='virtio'> <rate bytes='1024' period='1000'/> <source type='egd'> <remote type='unix'>/path/to/socket</remote> -- or -- <remote type='tcp' port='12345'>host.addr</remote> -- optionally or even more options, but they don't seem useful </source> </rng> Now these solutions are mere envelopes on top of the qemu functionality and don't really let us do any advanced configuration or management and they don't protect the host or guests by themselves from starving others from entropy. The solution for the problem described above is to have a central point where guests can request entropy and this central point will do all the rate limiting. To do this we will need to add a daemon that will talk the EGD protocol. This daemon will need to read entropy on request from the kernel software entropy device (/dev/random) or possibly a hardware RNG and distribute it to the hosts. To enable concurrent access and avoid starvation the daemon will implement a "shaping protocol" based on the hierarchical token bucket approach used to rate-limit network flows. For implementation details on this one I'll send a separate RFC. To allow configuration of the groups and sources I propose to create a separate driver with infrastructure similar to the network driver. This will allow to start multiple entropy distribution daemons with different sources and define the hierarchical classes to allow guest grouping. With this infrastructure you will be able to specify the rng device as follows: <rng model='virtio'> <source type='pool' pool='default'>classname</source> </rng> This will auto-configure all parameters according to the configuration of the pool and ensure that te guest gets classified correctly. As the start I'll implement the two basic options and then I'll propose another RFC how I will approach the second part more closely. Peter

2 1