- Devel - libvirt List Archives

[libvirt] [RH-BZ #595428] 'virsh list' should output more information 'xm list'
by Maxim Sditanov 02 Apr '12

02 Apr '12

Is this new feature request still https://bugs.gentoo.org/show_bug.cgi?id=366561 actual? Because virsh still show not enough information: $./virsh -c qemu:///system list --all Id Name State ---------------------------------------------------- 1 " running - winxp shut off And it will be good idea to show more detailed information or add additional flag, something like --detail. ./virsh -c qemu:///system list --all --detailed Id Name State Memory VCPU Uptime CPU IOPS ---------------------------------- ------------------------------------------------------------- 1 " running 768M 2 3hours 5% 12.1 - winxp shut off 512M 1 --- With best wishes, Maxim Sditanov

1 0

[libvirt] call for volunteer: FLOSS weekly podcast opportunity
by Eric Blake 02 Apr '12

02 Apr '12

The FLOSS weekly podcast is interested in doing an interview about libvirt development on June 6, at 9:30am Los Angeles time (UTC-7). For more information on past podcasts, see: http://twit.tv/show/floss-weekly With prodding from Justin Clift, I have volunteered to take on the task, but the show host, Randal, prefers to have two developers per show, so I'm looking for anyone else who has developed for or with libvirt, who would be interested in sharing the opportunity. You don't have to have a lot of commits under your belt, so much as having enough active interest in the project and what it enables you to do. In fact, having someone not directly tied to Red Hat as the second contact will probably provide a nicer balance to the interview. -- Eric Blake eblake(a)redhat.com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

1 0

[libvirt] [PATCH] Fix typos and spacing in messages
by Yuri Chornoivan 02 Apr '12

02 Apr '12

2 1

[libvirt] [PATCH] qemu: fix memory leak in virDomainGetVcpus
by Laine Stump 02 Apr '12

02 Apr '12

https://bugzilla.redhat.com/show_bug.cgi?id=808979 The leak is really in virProcessInfoGetAffinity, as shown in the valgrind output given in the above bug report - it calls CPU_ALLOC(), but then fails to call CPU_FREE(). This leak has existed in every version of libvirt since 0.7.5. --- src/util/processinfo.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/util/processinfo.c b/src/util/processinfo.c index b1b1737..af19723 100644 --- a/src/util/processinfo.c +++ b/src/util/processinfo.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2009-2010 Red Hat, Inc. + * Copyright (C) 2009-2010, 2012 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 @@ -140,6 +140,7 @@ realloc: for (i = 0 ; i < maxcpu ; i++) if (CPU_ISSET_S(i, masklen, mask)) VIR_USE_CPU(map, i); + CPU_FREE(mask); # else /* Legacy method uses a fixed size cpu mask, only allows upto 1024 cpus */ cpu_set_t mask; -- 1.7.7.6

2 2

[libvirt] [PATCH] python: improve conversion validation
by Eric Blake 01 Apr '12

01 Apr '12

Laszlo Ersek pointed out that in trying to convert a long to an unsigned int, we used: long long_val = ...; if ((unsigned int)long_val == long_val) According to C99 integer promotion rules, the if statement is equivalent to: (unsigned long)(unsigned int)long_val == (unsigned long)long_val since you get an unsigned comparison if at least one side is unsigned, using the largest rank of the two sides; but on 32-bit platforms, where unsigned long and unsigned int are the same size, this comparison is always true and ends up converting negative long_val into posigive unsigned int values, rather than rejecting the negative value as we had originally intended (python longs are unbounded size, and we don't want to do silent modulo arithmetic when converting to C code). Fix this by using direct comparisons, rather than casting. * python/typewrappers.c (libvirt_intUnwrap, libvirt_uintUnwrap) (libvirt_ulongUnwrap, libvirt_ulonglongUnwrap): Fix conversion checks. --- python/typewrappers.c | 27 ++++++++++++++++++++------- 1 files changed, 20 insertions(+), 7 deletions(-) diff --git a/python/typewrappers.c b/python/typewrappers.c index af209e6..4ae3ee1 100644 --- a/python/typewrappers.c +++ b/python/typewrappers.c @@ -132,7 +132,7 @@ libvirt_intUnwrap(PyObject *obj, int *val) if ((long_val == -1) && PyErr_Occurred()) return -1; - if ((int)long_val == long_val) { + if (long_val <= INT_MAX && long_val >= INT_MIN) { *val = long_val; } else { PyErr_SetString(PyExc_OverflowError, @@ -151,7 +151,7 @@ libvirt_uintUnwrap(PyObject *obj, unsigned int *val) if ((long_val == -1) && PyErr_Occurred()) return -1; - if ((unsigned int)long_val == long_val) { + if (long_val >= 0 && long_val <= UINT_MAX) { *val = long_val; } else { PyErr_SetString(PyExc_OverflowError, @@ -183,7 +183,13 @@ libvirt_ulongUnwrap(PyObject *obj, unsigned long *val) if ((long_val == -1) && PyErr_Occurred()) return -1; - *val = long_val; + if (long_val >= 0) { + *val = long_val; + } else { + PyErr_SetString(PyExc_OverflowError, + "negative Python int cannot be converted to C unsigned long"); + return -1; + } return 0; } @@ -207,16 +213,23 @@ int libvirt_ulonglongUnwrap(PyObject *obj, unsigned long long *val) { unsigned long long ullong_val = -1; + long long llong_val; /* The PyLong_AsUnsignedLongLong doesn't check the type of * obj, only accept argument of PyLong_Type, so we check it instead. */ - if (PyInt_Check(obj)) - ullong_val = PyInt_AsLong(obj); - else if (PyLong_Check(obj)) + if (PyInt_Check(obj)) { + llong_val = PyInt_AsLong(obj); + if (llong_val < 0) + PyErr_SetString(PyExc_OverflowError, + "negative Python int cannot be converted to C unsigned long long"); + else + ullong_val = llong_val; + } else if (PyLong_Check(obj)) { ullong_val = PyLong_AsUnsignedLongLong(obj); - else + } else { PyErr_SetString(PyExc_TypeError, "an integer is required"); + } if ((ullong_val == -1) && PyErr_Occurred()) return -1; -- 1.7.1

3 3

Re: [libvirt] Start of freeze for libvirt-0.9.11 and availability of rc1
by Maxim Sditanov 01 Apr '12

01 Apr '12

2012/4/1 Maxim Sditanov <feniksa(a)rambler.ru> > 2012/4/1 Feniks Gordon Freeman <feniksa(a)rambler.ru> > >> 2012/4/1 Feniks Gordon Freeman <feniksa(a)rambler.ru> >> >>> 2012/4/1 Laine Stump <laine(a)laine.org> >>> >>>> On 03/31/2012 07:56 AM, Daniel Veillard wrote: >>>> > I just made the second release candidate >>>> > >>>> > ftp://libvirt.org/libvirt/libvirt-0.9.11-rc2.tar.gz >>>> > >>>> > along with the rpms and tagged the git tree. >>>> > >>>> > It seems to work okay in my limited testing. If all goes well I would >>>> > probably push on Tuesday, >>>> > >>>> >>>> A gentoo user reported in irc (#virt on irc.oftc.net) earlier today >>>> that >>>> builds were failing for him on a fresh checkout from git: >>>> >>>> <feniksa> hi, when i compile libvirt on my gentoo i get warning: >>>> >>>> cc1: warning: unrecognized command line option >>>> "-Wno-suggest-attribute=const" >>>> cc1: warning: unrecognized command line option >>>> "-Wno-suggest-attribute=pure" >>>> >>>> <feniksa> gcc (Gentoo 4.5.3-r2 p1.1, pie-0.4.7) 4.5.3 >>>> <feniksa> last part of conf: http://pastebin.com/gnT4Y2Ba >>>> and compile log http://pastebin.com/FcSwshzS >>>> <feniksa> laine: 2 weeks ago i didn't see this warning. Now i pull >>>> from master sources and get this warning. But, this is only because i am >>>> using gcc 4.5.3 (this is latest gcc version in gentoo OS). >>>> >>>> I haven't looked at any of this code before, but it looks to me like >>>> ql_WARN_ADD() in gnulib must not be doing the right thing, and is adding >>>> the option even though the compiler apparently doesn't support it. This >>>> probably didn't crop up before because full warnings weren't turned on >>>> in the build until March 27 in commit 851117bd. That commit does things >>>> different for builds of git trees vs. tarballs, so I don't know if this >>>> failure will affect builds of release tarballs (I asked feniksa to try a >>>> build of the rc2 tarball, but he hasn't responded yet). >>>> >>> I compiled rc2 (ftp://libvirt.org/libvirt/libvirt-0.9.11-rc2.tar.gz) Same problem present in release candidate rc2 cc1: warning: unrecognized command line option "-Wno-suggest-attribute=const" cc1: warning: unrecognized command line option "-Wno-suggest-attribute=pure" -- With best wishes, Maxim Sditanov

1 0

[libvirt] [PATCH] libvirt:qemu: fix max_balloon less than cur_balloon
by Zhou Peng 31 Mar '12

31 Mar '12

In qemuDomainGetXMLDesc, if cur_balloon is bigger than max_balloon, max_balloon should be updated too, otherwise the currentMemory is bigger than memory in dumped xml, which will produce invalid checkpoint by virsh-save. The bug appears like this below: $virsh save vm-num foo.ckp $virsh restore foo.ckp error: XML error: current memory '824320k' exceeds maximum '824288k' Signed-off-by: Zhou Peng <ailvpeng25(a)gmail.com> --- src/qemu/qemu_driver.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 0d3b0bd..96d3219 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4485,7 +4485,13 @@ endjob: if (err < 0) goto cleanup; if (err > 0) + { vm->def->mem.cur_balloon = balloon; + if (balloon > vm->def->mem.max_balloon) + { + vm->def->mem.max_balloon = balloon; + } + } /* err == 0 indicates no balloon support, so ignore it */ } } -- 1.7.7.6

3 12

[libvirt] [PATCH] Fix client only RPM build & other misc RPM problems
by Daniel P. Berrange 31 Mar '12

31 Mar '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> Pushing under build break rule - some misc problems discovered while doing the RPM refactoring * libvirt.spec.in: Remove obsolete --with-remote-pid-file arg. Add missing %{without_libxl} statement. Fix handling of docs in client only build. Put systemtap files in -client RPM instead of -daemon RPM * examples/xml/nwfilter/Makefile.am: Don't install examples if nwfilter is disabled. --- examples/xml/nwfilter/Makefile.am | 2 ++ libvirt.spec.in | 30 +++++++++--------------------- 2 files changed, 11 insertions(+), 21 deletions(-) diff --git a/examples/xml/nwfilter/Makefile.am b/examples/xml/nwfilter/Makefile.am index 12a3619..c503682 100644 --- a/examples/xml/nwfilter/Makefile.am +++ b/examples/xml/nwfilter/Makefile.am @@ -22,6 +22,7 @@ FILTERS = \ EXTRA_DIST=$(FILTERS) +if WITH_NWFILTER confdir = $(sysconfdir)/libvirt NWFILTER_DIR = "$(DESTDIR)$(sysconfdir)/libvirt/nwfilter" @@ -37,3 +38,4 @@ uninstall-local:: rm -f "$(NWFILTER_DIR)/$$f"; \ done -test -z $(shell ls $(NWFILTER_DIR)) || rmdir $(NWFILTER_DIR) +endif diff --git a/libvirt.spec.in b/libvirt.spec.in index d91653c..b787fde 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -973,6 +973,7 @@ autoreconf -if %{?_without_openvz} \ %{?_without_lxc} \ %{?_without_vbox} \ + %{?_without_libxl} \ %{?_without_xenapi} \ %{?_without_sasl} \ %{?_without_avahi} \ @@ -1008,8 +1009,7 @@ autoreconf -if %{with_packager_version} \ --with-qemu-user=%{qemu_user} \ --with-qemu-group=%{qemu_group} \ - %{init_scripts} \ - --with-remote-pid-file=%{_localstatedir}/run/libvirtd.pid + %{init_scripts} make %{?_smp_mflags} gzip -9 ChangeLog @@ -1059,16 +1059,6 @@ rm -f $RPM_BUILD_ROOT%{_datadir}/augeas/lenses/tests/test_libvirtd_lxc.aug rm -rf $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-python-%{version} %endif -%if %{client_only} -rm -rf $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-%{version} -%endif - -%if ! %{with_libvirtd} -rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/libvirt/nwfilter -mv $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-%{version}/html \ - $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-devel-%{version}/ -%endif - %if ! %{with_qemu} rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/libvirt/qemu.conf rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/logrotate.d/libvirtd.qemu @@ -1081,11 +1071,8 @@ rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/logrotate.d/libvirtd.lxc rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/logrotate.d/libvirtd.uml %endif -mv $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-%{version} $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-docs-%{version} - -%if ! %{with_nwfilter} -rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/libvirt/nwfilter -%endif +mv $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-%{version} \ + $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-docs-%{version} %clean rm -fr %{buildroot} @@ -1305,10 +1292,6 @@ fi %else rm -f $RPM_BUILD_ROOT%{_sysconfdir}/sysctl.d/libvirtd %endif -%if %{with_dtrace} -%{_datadir}/systemtap/tapset/libvirt_probes.stp -%{_datadir}/systemtap/tapset/libvirt_functions.stp -%endif %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/qemu/ %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/lxc/ %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/uml/ @@ -1495,6 +1478,11 @@ rm -f $RPM_BUILD_ROOT%{_sysconfdir}/sysctl.d/libvirtd %{_bindir}/virt-host-validate %{_libdir}/lib*.so.* +%if %{with_dtrace} +%{_datadir}/systemtap/tapset/libvirt_probes.stp +%{_datadir}/systemtap/tapset/libvirt_functions.stp +%endif + %dir %{_datadir}/libvirt/ %dir %{_datadir}/libvirt/schemas/ -- 1.7.7.6

1 0

[libvirt] [PATCH] openvz: Allow missing micro tag in the version
by Osier Yang 30 Mar '12

30 Mar '12

No reason to cause the openvz driver to fail on startup if the micro tag in missed. This solves the problem: https://www.redhat.com/archives/libvirt-users/2012-March/msg00202.html --- src/openvz/openvz_conf.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/openvz/openvz_conf.c b/src/openvz/openvz_conf.c index 28f86ff..da6869b 100644 --- a/src/openvz/openvz_conf.c +++ b/src/openvz/openvz_conf.c @@ -95,11 +95,13 @@ openvzExtractVersionInfo(const char *cmdstr, int *retversion) tmp = help; - /* expected format: vzctl version <major>.<minor>.<micro> */ + /* expected format: vzctl version <major>.<minor>.<micro>, + * missing <micro> is allowed. + */ if ((tmp = STRSKIP(tmp, "vzctl version ")) == NULL) goto cleanup; - if (virParseVersionString(tmp, &version, false) < 0) + if (virParseVersionString(tmp, &version, true) < 0) goto cleanup; if (retversion) -- 1.7.7.3

1 0

Re: [libvirt] Using Libvirt to change the bridge a virtual network card of a running vm is connected to
by Hendrik Schwartke 30 Mar '12

30 Mar '12

I fixed the patch and added it to https://bugzilla.redhat.com/show_bug.cgi?id=784767.

4 13