Commit Graph

3780 Commits

Author SHA1 Message Date
Daniel P. Berrange
746c336495 Fix passing of address family to virSocketParseAddr
The virSocketParseAddr function was accepting any AF_* constant
and using that to set the ai_flags field in struct addrinfo.
This is invalid, since address families must go in the ai_family
field of the struct.

* src/util/network.c: Fix handling of address family
* src/conf/network_conf.c, src/network/bridge_driver.c: Pass
  AF_UNSPEC instead of relying on it being 0.
2010-10-22 11:16:37 +01:00
Daniel P. Berrange
af3d4eec0d Include length with virSocketAddr data
Some operations on socket addresses need to know the length of
the sockaddr struct for the particular address family. This
info was being discarded when passing around virSocketAddr
instances. Turn it from a union into a struct containing
union+socklen_t fields, so length is always kept around.

* src/util/network.h: Add socklen_t field to virSocketAddr
* src/util/network.c, src/network/bridge_driver.c,
  src/conf/domain_conf.c: Update to take account of new
  struct definition.
2010-10-22 11:15:36 +01:00
Daniel P. Berrange
9e42b40a95 Remove useless code in error path of getnameinfo()
If getnameinfo() with NI_NUMERICHOST set fails, there are no
grounds to expect inet_ntop to succeed, since these calls
are functionally equivalent. Remove useless inet_ntop code
in the getnameinfo() error path.

* daemon/remote.c, src/remote/remote_driver.c: Remove
  calls to inet_ntop
2010-10-22 11:15:14 +01:00
Eric Blake
134bcb62db maint: sort private sym lists
* src/libvirt_private.syms: Sort by header name, then within
header, and drop duplicate virNetworkDefParseNode,
virFileLinkPointsTo and virXPathBoolean.
2010-10-21 08:28:01 -06:00
Daniel P. Berrange
aa1e3f6706 Enable JSON and netdev features in QEMU >= 0.13
The QEMU 0.13 release is finally out and from testing in RHEL-6
we know that its JSON and netdev features are now good enough
for us to use by default.

* src/qemu/qemu_conf.c: Enable JSON + netdev for QEMU >= 0.13
2010-10-21 11:16:22 +01:00
Eric Blake
f1eb9ed954 audit: simplify declaration
* src/util/virtaudit.c (virAuditSend): one less ifdef, since gcc
does not care if an ATTRIBUTE_UNUSED var gets used in some paths.
2010-10-20 12:21:52 -06:00
Matthias Bolte
a1109a7c7c qemu: Exit on first error in qemuDomainGetMemoryParameters
There is no point in trying to fill params beyond the first error,
because when qemuDomainGetMemoryParameters returns -1 then the caller
cannot detect which values in params are valid.
2010-10-20 19:33:11 +02:00
Matthias Bolte
916f95b7aa Rename VIR_DOMAIN_SWAP_HARD_LIMIT to VIR_DOMAIN_MEMORY_SWAP_HARD_LIMIT
To get them under the common VIR_DOMAIN_MEMORY_* prefix.
2010-10-20 19:33:11 +02:00
Matthias Bolte
e05cdac855 Fix formatting of the memtune XML element
Also output the min_guarantee element when set.
2010-10-20 19:33:11 +02:00
John Morrissey
c08c7b0143 Add process= support for 'qemu-kvm -name'
This sets the process name to the same value as the Windows title,
but since the name is limited to 16 chars only this is kept as a
configuration option and turned off by default
* src/qemu/qemu.conf src/qemu/qemu_conf.[ch]: hceck for support in the
  QEmu help output, add the option in qemu conf file and augment
  qemudBuildCommandLine to add it if switched on
* src/qemu/libvirtd_qemu.aug src/qemu/test_libvirtd_qemu.aug: augment
  the augeas lenses accordingly
* tests/qemuhelptest.c: cope with the extra flag being detected now
2010-10-20 10:30:30 +02:00
Stefan Berger
c2fbdf1088 nwfilter: avoid dir. enforcement for certain types of rules
Avoid the enforcement of direction if
- icmp rules specify the type/code information
- the 'skipMatch' variable is set to 'true'
2010-10-19 19:25:37 -04:00
Guido Günther
679b464bd4 Don't fail lxc domain start when memory controller support is missing
Debian stock kernel has CONFIG_CGROUP_MEM_RES_CTLR disabled due to the
overhead [1]. Allow to start containers if the corresponding files in
the cgroup filesystem are missing. This fixes Debian bug #566180 [2].

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=534964
[2] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=566180
2010-10-19 21:29:12 +02:00
Matthias Bolte
83e5711418 Fix compile errors in remote.c and newly added audit code 2010-10-19 19:01:26 +02:00
Daniel P. Berrange
9bd3cce0d2 Fix symbol exports & remove duplicated libvirt_util.la linkage
The libvirt_util.la library was mistakenly linked into libvirtd
directly. Since libvirt_util.la is already linked to libvirt.so,
this resulted in libvirtd getting two copies of the code and
more critically 2 copies of static global variables.

Testing in turn exposed a issue with loadable modules. The
gnulib replacement functions are not exported to loadable
modules. Rather than trying to figure out the name sof all
gnulib functions & export them, just linkage all loadable
modules against libgnu.la statically.

* daemon/Makefile.am: Remove linkage of libvirt_util.la
  and libvirt_driver.la
* src/Makefile.am: Link driver modules against libgnu.la
* src/libvirt.c: Don't try to load modules which were
  compiled out
* src/libvirt_private.syms: Export all other internal
  symbols that are required  by drivers
2010-10-19 17:31:31 +01:00
Miloslav Trmač
b8e2de8899 Audit SELinux label assignment.
A more natural auditing point would perhaps be
SELinuxSetSecurityProcessLabel, but this happens in the child after root
permissions are dropped, so the kernel would refuse the audit record.
2010-10-19 17:31:31 +01:00
Miloslav Trmač
a8b5f9bd27 Audit VM start/stop/suspend/resume
Most operations are audited at the libvirtd level; auditing in
src/libvirt.c would result in two audit entries per operation (one in
the client, one in libvirtd).

The only exception is a domain stopping of its own will (e.g. because
the user clicks on "shutdown" inside the interface).  There can often be
no client connected at the time the domain stops, so libvirtd does not
have any virConnectPtr object on which to attach an event watch.  This
patch therefore adds auditing directly inside the qemu driver (other
drivers are not supported).
2010-10-19 17:31:31 +01:00
Daniel P. Berrange
8f680ad3b8 Basic framework for auditing integration
Integrate with libaudit.so for auditing of important operations.
libvirtd gains a couple of config entries for auditing. By
default it will enable auditing, if its enabled on the host.
It can be configured to force exit if auditing is disabled
on the host. It will can also send audit messages via libvirt
internal logging API

Places requiring audit reporting can use the VIR_AUDIT
macro to report data. This is a no-op unless auditing is
enabled

* autobuild.sh, mingw32-libvirt.spec.in: Disable audit
  on mingw
* configure.ac: Add check for libaudit
* daemon/libvirtd.aug, daemon/libvirtd.conf,
  daemon/test_libvirtd.aug, daemon/libvirtd.c: Add config
  options to enable auditing
* include/libvirt/virterror.h, src/util/virterror.c: Add
  VIR_FROM_AUDIT source
* libvirt.spec.in: Enable audit
* src/util/virtaudit.h, src/util/virtaudit.c: Simple internal
  API for auditing messages
2010-10-19 17:31:31 +01:00
Eric Blake
b013788742 vcpu: remove dead xen code
* src/xen/xen_driver.h (xenUnifiedDriver): Remove now-unused
domainGetMaxVcpus, domainSetVcpus.
* src/xen/proxy_internal.c (xenProxyDriver): Likewise.
* src/xen/xen_hypervisor.c (xenHypervisorDriver): Likewise.
* src/xen/xen_inotify.c (xenInotifyDriver): Likewise.
* src/xen/xend_internal.c (xenDaemonDriver)
(xenDaemonDomainSetVcpus): Likewise.
* src/xen/xm_internal.c (xenXMDriver, xenXMDomainSetVcpus):
Likewise.
* src/xen/xs_internal.c (xenStoreDriver): Likewise.
2010-10-19 10:07:10 -06:00
Eric Blake
e443a00312 vcpu: improve support for setting xen vcpu counts
Tested with RHEL 5.6 (xendConfigVersion 2, where xend_internal
controls live domains and xm_internal controls inactive domains).
Hopefully this works with xendConfigVersion 3 (where xend_internal
controls everything).

* src/xen/xen_driver.c (xenUnifiedDomainSetVcpusFlags): Support
more flags.
(xenUnifiedGetMaxVcpus): Export.
* src/xen/xm_internal.h (xenXMDomainSetVcpusFlags): New prototype.
* src/xen/xend_internal.h (xenDaemonDomainSetVcpusFlags): Likewise.
* src/xen/xen_driver.h (xenUnifiedGetMaxVcpus): Likewise.
* src/xen/xm_internal.c (xenXMDomainSetVcpusFlags): New function.
* src/xen/xend_internal.c (xenDaemonDomainSetVcpusFlags): Likewise.
2010-10-19 10:07:02 -06:00
Eric Blake
290ea33111 vcpu: improve support for getting xen vcpu counts
* src/xen/xen_driver.c (xenUnifiedDomainGetVcpusFlags): Support
more flags.
* src/xen/xend_internal.h (xenDaemonDomainGetVcpusFlags): New
prototype.
* src/xen/xm_internal.h (xenXMDomainGetVcpusFlags): Likewise.
* src/xen/xend_internal.c (virDomainGetVcpusFlags): New function.
* src/xen/xm_internal.c (xenXMDomainGetVcpusFlags): Likewise.
2010-10-19 10:06:55 -06:00
Eric Blake
0fab10e5ed vcpu: improve vcpu support in xen command line
This patch series focuses on xendConfigVersion 2 (xm_internal) and 3
(xend_internal), but leaves out changes for xenapi drivers.

See this link for more details about vcpu_avail for xm usage.
http://lists.xensource.com/archives/html/xen-devel/2009-11/msg01061.html

This relies on the fact that def->maxvcpus can be at most 32 with xen.

* src/xen/xend_internal.c (xenDaemonParseSxpr)
(sexpr_to_xend_domain_info, xenDaemonFormatSxpr): Use vcpu_avail
when current vcpus is less than maximum.
* src/xen/xm_internal.c (xenXMDomainConfigParse)
(xenXMDomainConfigFormat): Likewise.
* tests/xml2sexprdata/xml2sexpr-pv-vcpus.sexpr: New file.
* tests/sexpr2xmldata/sexpr2xml-pv-vcpus.sexpr: Likewise.
* tests/sexpr2xmldata/sexpr2xml-pv-vcpus.xml: Likewise.
* tests/xmconfigdata/test-paravirt-vcpu.cfg: Likewise.
* tests/xmconfigdata/test-paravirt-vcpu.xml: Likewise.
* tests/xml2sexprtest.c (mymain): New test.
* tests/sexpr2xmltest.c (mymain): Likewise.
* tests/xmconfigtest.c (mymain): Likewise.
2010-10-19 10:06:45 -06:00
Eric Blake
28a3605906 vcpu: complete vcpu support in qemu driver
* src/qemu/qemu_driver.c (qemudDomainSetVcpusFlags)
(qemudDomainGetVcpusFlags): Support all feasible flag
combinations.
2010-10-19 10:06:38 -06:00
Eric Blake
d67c189e80 vcpu: improve vcpu support in qemu command line
* src/qemu/qemu_conf.c (qemuParseCommandLineSmp): Distinguish
between vcpus and maxvcpus, for new enough qemu.
* tests/qemuargv2xmltest.c (mymain): Add new test.
* tests/qemuxml2argvtest.c (mymain): Likewise.
* tests/qemuxml2xmltest.c (mymain): Likewise.
* tests/qemuxml2argvdata/qemuxml2argv-smp.args: New file.
2010-10-19 10:06:33 -06:00
Eric Blake
6c9e6b9564 vcpu: support all flags in test driver
* src/test/test_driver.c (testDomainGetVcpusFlags)
(testDomainSetVcpusFlags): Support all flags.
(testDomainUpdateVCPUs): Update cpu count here.
2010-10-19 10:06:25 -06:00
Eric Blake
4617eedfae vcpu: support maxvcpu in domain_conf
Although this patch adds a distinction between maximum vcpus and
current vcpus in the XML, the values should be identical for all
drivers at this point.  Only in subsequent per-driver patches will
a distinction be made.

In general, virDomainGetInfo should prefer the current vcpus.

* src/conf/domain_conf.h (_virDomainDef): Adjust vcpus to unsigned
short, to match virDomainGetInfo limit.  Add maxvcpus member.
* src/conf/domain_conf.c (virDomainDefParseXML)
(virDomainDefFormat): parse and print out vcpu details.
* src/xen/xend_internal.c (xenDaemonParseSxpr)
(xenDaemonFormatSxpr): Manage both vcpu numbers, and require them
to be equal for now.
* src/xen/xm_internal.c (xenXMDomainConfigParse)
(xenXMDomainConfigFormat): Likewise.
* src/phyp/phyp_driver.c (phypDomainDumpXML): Likewise.
* src/openvz/openvz_conf.c (openvzLoadDomains): Likewise.
* src/openvz/openvz_driver.c (openvzDomainDefineXML)
(openvzDomainCreateXML, openvzDomainSetVcpusInternal): Likewise.
* src/vbox/vbox_tmpl.c (vboxDomainDumpXML, vboxDomainDefineXML):
Likewise.
* src/xenapi/xenapi_driver.c (xenapiDomainDumpXML): Likewise.
* src/xenapi/xenapi_utils.c (createVMRecordFromXml): Likewise.
* src/esx/esx_vmx.c (esxVMX_ParseConfig, esxVMX_FormatConfig):
Likewise.
* src/qemu/qemu_conf.c (qemuBuildSmpArgStr)
(qemuParseCommandLineSmp, qemuParseCommandLine): Likewise.
* src/qemu/qemu_driver.c (qemudDomainHotplugVcpus): Likewise.
* src/opennebula/one_conf.c (xmlOneTemplate): Likewise.
2010-10-19 10:05:51 -06:00
Eric Blake
50c51f13e2 vcpu: make old API trivially wrap to new API
Note - this wrapping is completely mechanical; the old API will
function identically, since the new API validates that the exact
same flags are provided by the old API.  On a per-driver basis,
it may make sense to have the old API pass a different set of flags,
but that should be done in the per-driver patch that implements
the full range of flag support in the new API.

* src/esx/esx_driver.c (esxDomainSetVcpus, escDomainGetMaxVpcus):
Move guts...
(esxDomainSetVcpusFlags, esxDomainGetVcpusFlags): ...to new
functions.
(esxDriver): Trivially support the new API.
* src/openvz/openvz_driver.c (openvzDomainSetVcpus)
(openvzDomainSetVcpusFlags, openvzDomainGetMaxVcpus)
(openvzDomainGetVcpusFlags, openvzDriver): Likewise.
* src/phyp/phyp_driver.c (phypDomainSetCPU)
(phypDomainSetVcpusFlags, phypGetLparCPUMAX)
(phypDomainGetVcpusFlags, phypDriver): Likewise.
* src/qemu/qemu_driver.c (qemudDomainSetVcpus)
(qemudDomainSetVcpusFlags, qemudDomainGetMaxVcpus)
(qemudDomainGetVcpusFlags, qemuDriver): Likewise.
* src/test/test_driver.c (testSetVcpus, testDomainSetVcpusFlags)
(testDomainGetMaxVcpus, testDomainGetVcpusFlags, testDriver):
Likewise.
* src/vbox/vbox_tmpl.c (vboxDomainSetVcpus)
(vboxDomainSetVcpusFlags, virDomainGetMaxVcpus)
(virDomainGetVcpusFlags, virDriver): Likewise.
* src/xen/xen_driver.c (xenUnifiedDomainSetVcpus)
(xenUnifiedDomainSetVcpusFlags, xenUnifiedDomainGetMaxVcpus)
(xenUnifiedDomainGetVcpusFlags, xenUnifiedDriver): Likewise.
* src/xenapi/xenapi_driver.c (xenapiDomainSetVcpus)
(xenapiDomainSetVcpusFlags, xenapiDomainGetMaxVcpus)
(xenapiDomainGetVcpusFlags, xenapiDriver): Likewise.
(xenapiError): New helper macro.
2010-10-19 10:03:33 -06:00
Eric Blake
eb826444f9 vcpu: implement the remote protocol
Done by editing the first three files, then running
'make -C src rpcgen', then editing src/remote_protocol-structs
to match.

* daemon/remote.c (remoteDispatchDomainSetVcpusFlags)
(remoteDispatchDomainGetVcpusFlags): New functions.
* src/remote/remote_driver.c (remoteDomainSetVcpusFlags)
(remoteDomainGetVcpusFlags, remote_driver): Client side
serialization.
* src/remote/remote_protocol.x
(remote_domain_set_vcpus_flags_args)
(remote_domain_get_vcpus_flags_args)
(remote_domain_get_vcpus_flags_ret)
(REMOTE_PROC_DOMAIN_SET_VCPUS_FLAGS)
(REMOTE_PROC_DOMAIN_GET_VCPUS_FLAGS): Define wire format.
* daemon/remote_dispatch_args.h: Regenerate.
* daemon/remote_dispatch_prototypes.h: Likewise.
* daemon/remote_dispatch_table.h: Likewise.
* src/remote/remote_protocol.c: Likewise.
* src/remote/remote_protocol.h: Likewise.
* src/remote_protocol-structs: Likewise.
2010-10-19 10:02:33 -06:00
Eric Blake
9d2c607992 vcpu: implement the public APIs
Factors common checks (such as nonzero vcpu count) up front, but
drivers will still need to do additional flag checks.

* src/libvirt.c (virDomainSetVcpusFlags, virDomainGetVcpusFlags):
New functions.
(virDomainSetVcpus, virDomainGetMaxVcpus): Refer to new API.
2010-10-19 10:02:06 -06:00
Eric Blake
dd255d6405 vcpu: define internal driver API
* src/driver.h (virDrvDomainSetVcpusFlags)
(virDrvDomainGetVcpusFlags): New typedefs.
(_virDriver): New callback members.
* src/esx/esx_driver.c (esxDriver): Add stub for driver.
* src/lxc/lxc_driver.c (lxcDriver): Likewise.
* src/opennebula/one_driver.c (oneDriver): Likewise.
* src/openvz/openvz_driver.c (openvzDriver): Likewise.
* src/phyp/phyp_driver.c (phypDriver): Likewise.
* src/qemu/qemu_driver.c (qemuDriver): Likewise.
* src/remote/remote_driver.c (remote_driver): Likewise.
* src/test/test_driver.c (testDriver): Likewise.
* src/uml/uml_driver.c (umlDriver): Likewise.
* src/vbox/vbox_tmpl.c (Driver): Likewise.
* src/xen/xen_driver.c (xenUnifiedDriver): Likewise.
* src/xenapi/xenapi_driver.c (xenapiDriver): Likewise.
2010-10-19 10:00:47 -06:00
Eric Blake
ea3f5c6809 vcpu: add new public API
API agreed on in
https://www.redhat.com/archives/libvir-list/2010-September/msg00456.html,
but modified for enum names to be consistent with virDomainDeviceModifyFlags.

* include/libvirt/libvirt.h.in (virDomainVcpuFlags)
(virDomainSetVcpusFlags, virDomainGetVcpusFlags): New
declarations.
* src/libvirt_public.syms: Export new symbols.
2010-10-19 10:00:17 -06:00
Stefan Berger
7d79da247a nwfilter: changes to rules in VM->host table
In the table built for traffic coming from the VM going to the host make the following changes:

- don't ACCEPT the packets but do a 'RETURN' and let the host-specific firewall rules in subsequent rules evaluate whether the traffic is allowed to enter

- use the '-m state' in the rules as everywhere else
2010-10-19 11:35:58 -04:00
Matthias Bolte
1c61648961 esx: Handle non-UTF-8 encoded VMX files
ESX(i) uses UTF-8, but a Windows based GSX server writes
Windows-1252 encoded VMX files.

Add a test case to ensure that libxml2 provides Windows-1252
to UTF-8 conversion.
2010-10-19 16:29:12 +02:00
Stefan Berger
f04de501bc Introduce VIR_CLOSE to be used rather than close()
Since bugs due to double-closed file descriptors are difficult to track down in a multi-threaded system, I am introducing the VIR_CLOSE(fd) macro to help avoid mistakes here.

There are lots of places where close() is being used. In this patch I am only cleaning up usage of close() in src/conf where the problems were.

I also dare to declare close() as being deprecated in libvirt code base (HACKING).
2010-10-19 10:23:51 -04:00
Dan Kenigsberg
b2c9a87940 root_squash: virFileOperation may fail with EPERM too
Over root-squashing nfs, when virFileOperation() is called as uid==0,
it may fail with EACCES, but also with EPERM, due to
virFileOperationNoFork()'s failed attemp to chown a writable file.

qemudDomainSaveFlag() should expect this case, too.
2010-10-19 15:26:32 +02:00
Dan Kenigsberg
b7bd75c4c0 Run initgroups() in qemudOpenAsUID()
qemudOpenAsUID is intended to open a file with the credentials of a
specified uid. Current implementation fails if the file is accessible to
one of uid's groups but not owned by uid.

This patch replaces the supplementary group list that the child process
inherited from libvirtd with the default group list of uid.
2010-10-19 15:22:57 +02:00
Nikunj A. Dadhania
61dfbf8c3a Update docs for memory parameters and memtune command
* docs/formatdomain.html.in: Add memtune element details, added min_guarantee
* src/libvirt.c: Update virDomainGetMemoryParameters api description, make
  it more clear that the user first needs to call the api to get the number
  of parameters supported and then call again to get the values.
* tools/virsh.pod: Add usage of new command memtune in virsh manpage
2010-10-19 14:27:12 +02:00
Matthias Bolte
6c4b04142b esx: Fix check in esxDomainGetInfo's perf metric handling 2010-10-16 11:39:36 +02:00
Matthias Bolte
b1933e9e6a esx: Explictly declare VMX file content as UTF-8 2010-10-15 17:51:28 +02:00
Laine Stump
d3fec47f81 Fix warning about a non-literal format string in qemu_driver.c 2010-10-15 11:34:38 -04:00
Matthias Bolte
2dd86bbe5a esx: Handle name escaping properly
VMware uses a mix of percent-, pipe- and base64-encoding in
different combinations in different places.

Add a testcase for this.
2010-10-14 22:43:16 +02:00
Stefan Berger
4cfcde2d83 nwfilter: prevent filters with different name but same UUID
Patch to prevent multiple nwfilters with different name but same UUID.
2010-10-14 11:53:08 -04:00
Harsh Prateek Bora
75a6a9a8e0 new attribute accessmode to filesystem element
This introduces new attribute to filesystem element
to support customizable access mode for mount type.
Valid accessmode are: passthrough, mapped and squash.

Usage:
        <filesystem type='mount' accessmode='passthrough'>
          <source dir='/export/to/guest'/>
          <target dir='mount_tag'/>
        </filesystem>

passthrough is the default model if not specified, that's
also the current behaviour.
2010-10-14 15:08:24 +02:00
Stefan Berger
2b3df906f3 nwfilter: cut off connections after changing filters
The following filter transition from a filter allowing incoming TCP connections

  <rule action='accept' direction='in' priority='401'>
    <tcp/>
  </rule>
  <rule action='accept' direction='out' priority='500'>
    <tcp/>
  </rule>

to one that does not allow them

  <rule action='drop' direction='in' priority='401'>
    <tcp/>
  </rule>
  <rule action='accept' direction='out' priority='500'>
    <tcp/>
  </rule>

did previously not cut off existing (ssh) connections but only prevented newly initiated ones. The attached patch allows to cut off existing connections as well, thus enforcing what the filter is showing.

I had only tested with a configuration where the physical interface is connected to the bridge where the filters are applied. This patch now also solves a filtering problem where the physical interface is not connected to the bridge, but the bridge is given an IP address and the host routes between bridge and physical interface. Here the filters drop non-allowed traffic on the outgoing side on the host.
2010-10-14 08:54:03 -04:00
Jiri Denemark
f93924f465 qemu: Prohibit migration of guests with host devices
Explicitly raising a nice error in the case user tries to migrate a
guest with assigned host devices is much better than waiting for a
mysterious error with no clue for the reason.
2010-10-14 09:36:54 +02:00
Jiri Denemark
b2de33e2a7 cpu: Use vendor in baseline CPU only if all hosts use it
When only some host CPUs given to cpuBaseline contain <vendor> element,
baseline CPU should not contain it. Otherwise the result would not be
compatible with the host CPUs without vendor. CPU vendors are still
taken into account when computing baseline CPU, it's just removed from
the result.
2010-10-14 09:36:54 +02:00
Jiri Denemark
ac7afbeb9e cpu: Fix vendor for recent CPU models
Recent CPU models were specified using invalid vendor element
<vendor>NAME</vendor>, which was silently ignored due to a bug in the
code which was parsing it.
2010-10-14 09:36:54 +02:00
Eric Blake
450cbebe1c memory: fix remote protocol compilation
'make -C src rpcgen' is supposed to be idempotent.  But commit
f928f43b7b mistakently manually edited a generated file rather
than fixing the upstream file.

* src/remote/remote_protocol.x (remote_memory_param_value): Use
correct spelling of enum values.
* src/remote/remote_protocol.c: Regenerate.
2010-10-13 11:09:40 -06:00
Daniel P. Berrange
f98a6cd6ae Enable support for nested SVM
This enables support for nested SVM using the regular CPU
model/features block. If the CPU model or features include
'svm', then the '-enable-nesting' flag will be added to the
QEMU command line. Latest out of tree patches for nested
'vmx', no longer require the '-enable-nesting' flag. They
instead just look at the cpu features. Several of the models
already include svm support, but QEMU was just masking out
the svm bit silently. So this will enable SVM on such
models

* src/qemu/qemu_conf.h: flag for -enable-nesting
* src/qemu/qemu_conf.c: Use -enable-nesting if VMX or SVM are in
  the CPUID
* src/cpu/cpu.h, src/cpu/cpu.c: API to check for a named feature
* src/cpu/cpu_x86.c: x86 impl of feature check
* src/libvirt_private.syms: Add cpuHasFeature
* src/qemuhelptest.c: Add nesting flag where required
2010-10-13 16:45:31 +01:00
Daniel P. Berrange
3a092f3899 Fix Xen SEXPR generation to properly quote strings containing ()
* src/xen/sexpr.c: Ensure () are escaped in sexpr2string
* tests/sexpr2xmldata/sexpr2xml-boot-grub.sexpr,
  tests/sexpr2xmldata/sexpr2xml-boot-grub.xml,
  tests/xml2sexprdata/xml2sexpr-boot-grub.sexpr,
  tests/xml2sexprdata/xml2sexpr-boot-grub.xml: Data files to
  check escaping
* tests/sexpr2xmltest.c, tests/xml2sexprtest.c: Add boot-grub
  escaping test case
2010-10-13 16:42:48 +01:00
Stefan Berger
4435f3c477 nwfilter: resolve deadlock between VM ops and filter update
This is from a bug report and conversation on IRC where Soren reported that while a filter update is occurring on one or more VMs (due to a rule having been edited for example), a deadlock can occur when a VM referencing a filter is started.

The problem is caused by the two locking sequences of

qemu driver, qemu domain, filter             # for the VM start operation
filter, qemu_driver, qemu_domain            # for the filter update operation

that obviously don't lock in the same order. The problem is the 2nd lock sequence. Here the qemu_driver lock is being grabbed in qemu_driver:qemudVMFilterRebuild()

The following solution is based on the idea of trying to re-arrange the 2nd sequence of locks as follows:

qemu_driver, filter, qemu_driver, qemu_domain

and making the qemu driver recursively lockable so that a second lock can occur, this would then lead to the following net-locking sequence

qemu_driver, filter, qemu_domain

where the 2nd qemu_driver lock has been ( logically ) eliminated.

The 2nd part of the idea is that the sequence of locks (filter, qemu_domain) and (qemu_domain, filter) becomes interchangeable if all code paths where filter AND qemu_domain are locked have a preceding qemu_domain lock that basically blocks their concurrent execution

So, the following code paths exist towards qemu_driver:qemudVMFilterRebuild where we now want to put a qemu_driver lock in front of the filter lock.

-> nwfilterUndefine()   [ locks the filter ]
    -> virNWFilterTestUnassignDef()
        -> virNWFilterTriggerVMFilterRebuild()
            -> qemudVMFilterRebuild()

-> nwfilterDefine()
    -> virNWFilterPoolAssignDef() [ locks the filter ]
        -> virNWFilterTriggerVMFilterRebuild()
            -> qemudVMFilterRebuild()

-> nwfilterDriverReload()
    -> virNWFilterPoolLoadAllConfigs()
        ->virNWFilterPoolObjLoad()
            -> virNWFilterPoolAssignDef() [ locks the filter ]
                -> virNWFilterTriggerVMFilterRebuild()
                    -> qemudVMFilterRebuild()

-> nwfilterDriverStartup()
    -> virNWFilterPoolLoadAllConfigs()
        ->virNWFilterPoolObjLoad()
            -> virNWFilterPoolAssignDef() [ locks the filter ]
                -> virNWFilterTriggerVMFilterRebuild()
                    -> qemudVMFilterRebuild()

Qemu is not the only driver using the nwfilter driver, but also the UML driver calls into it. Therefore qemuVMFilterRebuild() can be exchanged with umlVMFilterRebuild() along with the driver lock of qemu_driver that can now be a uml_driver. Further, since UML and Qemu domains can be running on the same machine, the triggering of a rebuild of the filter can touch both types of drivers and their domains.

In the patch below I am now extending each nwfilter callback driver with functions for locking and unlocking the (VM) driver (UML, QEMU) and introduce new functions for locking all registered callback drivers and unlocking them. Then I am distributing the lock-all-cbdrivers/unlock-all-cbdrivers call into the above call paths. The last shown callpath starting with nwfilterDriverStart() is problematic since it is initialize before the Qemu and UML drives are and thus a lock in the path would result in a NULL pointer attempted to be locked -- the call to virNWFilterTriggerVMFilterRebuild() is never called, so we never lock either the qemu_driver or the uml_driver in that path. Therefore, only the first 3 paths now receive calls to lock and unlock all callback drivers. Now that the locks are distributed where it matters I can remove the qemu_driver and uml_driver lock from qemudVMFilterRebuild() and umlVMFilterRebuild() and not requiring the recursive locks.

For now I want to put this out as an RFC patch. I have tested it by 'stretching' the critical section after the define/undefine functions each lock the filter so I can (easily) concurrently execute another VM operation (suspend,start). That code is in this patch and if you want you can de-activate it. It seems to work ok and operations are being blocked while the update is being done.
I still also want to verify the other assumption above that locking filter and qemu_domain always has a preceding qemu_driver lock.
2010-10-13 10:33:26 -04:00