Commit Graph

17326 Commits

Author SHA1 Message Date
John Ferlan
42957661dc hostdev: Introduce virDomainHostdevSubsysSCSIHost
Split virDomainHostdevSubsysSCSI further. In preparation for having
either SCSI or iSCSI data, create a union in virDomainHostdevSubsysSCSI
to contain just a virDomainHostdevSubsysSCSIHost to describe the
'scsi_host' host device
2014-07-24 06:39:28 -04:00
John Ferlan
5805621cd9 hostdev: Introduce virDomainHostdevSubsysSCSI
Create a separate typedef for the hostdev union data describing SCSI
Then adjust the code to use the new pointer
2014-07-24 06:39:27 -04:00
John Ferlan
1c8da0d44e hostdev: Introduce virDomainHostdevSubsysPCI
Create a separate typedef for the hostdev union data describing PCI.
Then adjust the code to use the new pointer
2014-07-24 06:39:27 -04:00
John Ferlan
7540d07f09 hostdev: Introduce virDomainHostdevSubsysUSB
Create a separate typedef for the hostdev union data describing USB.
Then adjust the code to use the new pointer
2014-07-24 06:39:27 -04:00
Peter Krempa
185e07a5f8 qemu: snapshot: Use storage driver to pre-create snapshot file
Move the last operation done on local files to the storage driver API.
2014-07-24 09:59:00 +02:00
Peter Krempa
a2a67eff18 storage: Implement virStorageFileCreate for local and gluster files
Add backends for this frontend function so that we can use it in the
snapshot creation code.
2014-07-24 09:59:00 +02:00
Peter Krempa
24e5cafba6 qemu: Implement DAC driver chown callback to co-operate with storage drv
Use the storage driver to chown remote images.
2014-07-24 09:59:00 +02:00
Peter Krempa
0a515a3ba3 security: DAC: Plumb usage of chown callback
Use the callback to set disk and storage image labels by modifying the
existing functions and adding wrappers to avoid refactoring a lot of the
code.
2014-07-24 09:59:00 +02:00
Peter Krempa
7490a6d272 security: DAC: Introduce callback to perform image chown
To integrate the security driver with the storage driver we need to
pass a callback for a function that will chown storage volumes.

Introduce and document the callback prototype.
2014-07-24 09:58:59 +02:00
Peter Krempa
9f28599d51 security: DAC: Remove superfluous link resolution
When restoring security labels in the dac driver the code would resolve
the file path and use the resolved one to be chown-ed. The setting code
doesn't do that. Remove the unnecessary code.
2014-07-24 09:58:59 +02:00
Peter Krempa
222860cd36 storage: Add witness for checking storage volume use in security driver
With my intended use of storage driver assist to chown files on remote
storage we will need a witness that will tell us whether the given
storage volume supports operations needed by the storage driver.
2014-07-24 09:58:59 +02:00
Peter Krempa
50f09651df storage: Implement storage driver helper to chown disk images
Gluster storage works on a similar principle to NFS where it takes the
uid and gid of the actual process and uses it to access the storage
volume on the remote server. This introduces a need to chown storage
files on gluster via native API.
2014-07-24 09:58:59 +02:00
Michal Privoznik
12926a7c39 qemuConnectGetDomainCapabilities: Use wiser defaults
Up to now, users have to pass two arguments at least: domain virt type
('qemu' vs 'kvm') and one of emulatorbin or architecture. This is not
much user friendly. Nowadays users mostly use KVM and share the host
architecture with the guest. So now, the API (and subsequently virsh
command) can be called with all NULLs  (without any arguments).

Before this patch:
 # virsh domcapabilities
 error: failed to get emulator capabilities
 error: virttype_str in qemuConnectGetDomainCapabilities must not be NULL

 # virsh domcapabilities kvm
 error: failed to get emulator capabilities
 error: invalid argument: at least one of emulatorbin or architecture fields must be present

After:

 # virsh domcapabilities
 <domainCapabilities>
   <path>/usr/bin/qemu-system-x86_64</path>
   <domain>kvm</domain>
   <machine>pc-i440fx-2.1</machine>
   <arch>x86_64</arch>
   <vcpu max='255'/>
 </domainCapabilities>

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-07-24 09:19:09 +02:00
Martin Kletzander
dc8b7ce7bc numatune: finish the split from domain_conf and remove all dependencies
This patch adds back the virDomainDef typedef into domain_conf and
makes all the numatune_conf functions independent of any virDomainDef
definitions.

Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
2014-07-24 08:40:25 +02:00
Nehal J Wani
6675a0ab65 leaseshelper: avoid mem leak after storing lease entries
Contents of existing lease file were being stored in a variable
which was never freed.
2014-07-23 19:27:39 -06:00
Eric Blake
e1f7914d3e maint: simplify some syntax check exemptions
Commit 5028160 accidentally weakened the strtol prohibitions to
skip ALL files under src/util instead of the former situation of
just protecting util/virsexpr.c; even though NONE of the files
in that directory need any protection.

Shorten some long lines while at it.

* cfg.mk (exclude_file_name_regexp--sc_prohibit_strtol): No need
to exclude all of util.
(exclude_file_name_regexp--sc_prohibit_sprintf): Reduce long line.
(exclude_file_name_regexp--sc_prohibit_raw_allocation): Likewise.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-07-23 14:00:15 -06:00
Eric Blake
55d54dd938 conf: avoid memory leaks while parsing seclabel
Our seclabel parsing was repeatedly assigning malloc'd data into a
temporary variable, without first freeing the previous use.  Among
other leaks flagged by valgrind:

==9312== 8 bytes in 1 blocks are definitely lost in loss record 88 of 821
==9312==    at 0x4A0645D: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==9312==    by 0x8C40369: strdup (strdup.c:42)
==9312==    by 0x50EA799: virStrdup (virstring.c:676)
==9312==    by 0x50FAEB9: virXPathString (virxml.c:90)
==9312==    by 0x50FAF1E: virXPathStringLimit (virxml.c:112)
==9312==    by 0x510F516: virSecurityLabelDefParseXML (domain_conf.c:4571)
==9312==    by 0x510FB20: virSecurityLabelDefsParseXML (domain_conf.c:4720)

While it was multiple problems, it looks like commit da78351 (thankfully
unreleased) was to blame for all of them.

* src/conf/domain_conf.c (virSecurityLabelDefParseXML): Plug leaks
detected by valgrind.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-07-23 13:52:57 -06:00
Eric Blake
ee70839bbf nwfilter: plug memory leak with firewall
Introduced in commit 70571ccc (v1.2.4). Caught by valgrind:

==9816== 170 (32 direct, 138 indirect) bytes in 1 blocks are definitely lost in loss record 646 of 821
==9816==    at 0x4A081D4: calloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==9816==    by 0x50836FB: virAlloc (viralloc.c:144)
==9816==    by 0x50AEC2B: virFirewallNew (virfirewall.c:204)
==9816==    by 0x1E2308ED: ebiptablesDriverProbeStateMatch (nwfilter_ebiptables_driver.c:3715)
==9816==    by 0x1E2309AD: ebiptablesDriverInit (nwfilter_ebiptables_driver.c:3742)

* src/nwfilter/nwfilter_ebiptables_driver.c
(ebiptablesDriverProbeStateMatch): Properly clean up.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-07-23 13:15:14 -06:00
Eric Blake
60e4944059 metadata: track title edits across libvirtd restart
https://bugzilla.redhat.com/show_bug.cgi?id=1122205

Although the edits were changing in-memory XML, it was not flushed
to disk; so unless some other action changes XML, a libvirtd restart
would lose the changed information.

* src/conf/domain_conf.c (virDomainObjSetMetadata): Add parameter,
to save live status across restarts.
(virDomainSaveXML): Allow for test driver.
* src/conf/domain_conf.h (virDomainObjSetMetadata): Adjust
signature.
* src/bhyve/bhyve_driver.c (bhyveDomainSetMetadata): Adjust caller.
* src/lxc/lxc_driver.c (lxcDomainSetMetadata): Likewise.
* src/qemu/qemu_driver.c (qemuDomainSetMetadata): Likewise.
* src/test/test_driver.c (testDomainSetMetadata): Likewise.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-07-23 10:07:34 -06:00
John Ferlan
b6938a7c88 docs: Point to list of valid pool target volume formats
https://bugzilla.redhat.com/show_bug.cgi?id=1092886

Rather than point off to some nefarious "pool-specific docs" page when
describing the "format" field for the target pool provide a link to the
storage driver page which describes the various valid formats for each
pool type.  Also make it a bit more clear that if a valid format isn't
specified, then the type field is ignored.
2014-07-23 11:05:56 -04:00
Jiri Denemark
b983742029 spec: Enable sanlock on qemu_kvm_arches for RHEL
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
2014-07-23 16:43:30 +02:00
Cédric Bosdonnat
741049950b Added myself in AUTHORS.in 2014-07-23 16:23:46 +02:00
Cédric Bosdonnat
94157da601 Documented VIR_DOMAIN_FEATURE_CAPABILITIES use of virDomainCapabilitesPolicy 2014-07-23 16:21:59 +02:00
Ján Tomko
3227e17d82 Introduce virTristateSwitch enum
For the values "default", "on", "off"

Replaces
virDeviceAddressPCIMulti
virDomainFeatureState
virDomainIoEventFd
virDomainVirtioEventIdx
virDomainDiskCopyOnRead
virDomainMemDump
virDomainPCIRombarMode
virDomainGraphicsSpicePlaybackCompression
2014-07-23 12:59:40 +02:00
Ján Tomko
bb018ce6c8 Introduce virTristateBool enum type
Replace all three-state (default/yes/no) enums with it:
virDomainBIOSUseserial
virDomainBootMenu
virDomainPMState
virDomainGraphicsSpiceClipboardCopypaste
virDomainGraphicsSpiceAgentFileTransfer
virNetworkDNSForwardPlainNames
2014-07-23 12:37:39 +02:00
Peter Krempa
0e6cacc4b1 Fix build after 47e5b5ae32
The patch described above introduced two problems caught by the compiler
and thus breaking the build.

One of the problems was comparison of unsigned with < 0 and the second
one jumped a variable init.
2014-07-23 10:19:20 +02:00
Chen Hanxiao
1ce7c1d20c LXC: show used memory as 0 when domain is not active
Before:
virsh # dominfo chx3
State:          shut off
Max memory:     92160 KiB
Used memory:    92160 KiB

After:
virsh # dominfo container1
State:          shut off
Max memory:     92160 KiB
Used memory:    0 KiB

Similar to qemu cases.

Signed-off-by: Chen Hanxiao <chenhanxiao@cn.fujitsu.com>
2014-07-23 15:12:52 +08:00
Cédric Bosdonnat
26d67015b8 lxc domain from xml: convert lxc.cap.drop 2014-07-23 15:12:37 +08:00
Cédric Bosdonnat
47e5b5ae32 lxc: allow to keep or drop capabilities
Added <capabilities> in the <features> section of LXC domains
configuration. This section can contain elements named after the
capabilities like:

  <mknod state="on"/>, keep CAP_MKNOD capability
  <sys_chroot state="off"/> drop CAP_SYS_CHROOT capability

Users can restrict or give more capabilities than the default using
this mechanism.
2014-07-23 15:12:37 +08:00
Chen Hanxiao
a86b6215a7 LXC: create a bind mount for sysfs when enable userns but disable netns
kernel commit 7dc5dbc879bd0779924b5132a48b731a0bc04a1e
forbid us doing a fresh mount for sysfs
when enable userns but disable netns.
This patch will create a bind mount in this senario.

Signed-off-by: Chen Hanxiao <chenhanxiao@cn.fujitsu.com>
2014-07-23 15:09:09 +08:00
Michal Privoznik
692016f790 tests: Remove stale scsihostdata dir
In the fbd91d49 commit, new scsihostdata dir is added to EXTRA_DIST in
the tests/Makefile.am. However, the directory itself is not created
anywhere, nor in the commit.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-07-22 11:10:57 +02:00
Peter Krempa
1e833899ce qemu: snapshot: Forbid taking/reverting snapshots in PMSUSPENDED state
Qemu doesn't currently support them and behaves strangely. Just forbid
them.

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1079162
2014-07-22 10:22:35 +02:00
Peter Krempa
c71045a9cb qemu: snapshot: Forbid taking snapshot in invalid state
Similarly to 49a3a649a8 forbid creating
snapshots in domain states impossible to reach in qemu.
2014-07-22 10:22:35 +02:00
Michal Privoznik
d82170d757 domtop: Fix build on mingw
Firstly, there's no sigaction() nor struct sigaction on mingw. We have
to use the one implemented by gnulib (and hence link with gnulib).
Then, for some reason one header file from windows defines ERROR
symbol. Yes it does. Sigh.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-07-22 09:53:18 +02:00
Eric Blake
72823b4443 build: fix build without numactl
Under ./configure --without-numactl but with numactl-devel installed,
the build fails with:

../../src/util/virnuma.c: In function 'virNumaNodeIsAvailable':
../../src/util/virnuma.c:407:5: error: implicit declaration of function 'numa_bitmask_isbitset' [-Werror=implicit-function-declaration]
     return numa_bitmask_isbitset(numa_nodes_ptr, node);
     ^

and other failures, all because the configure results for particular
functions were used without regard to whether libnuma was even being
linked in.

* src/util/virnuma.c (virNumaGetPages): Fix message typo.
(virNumaNodeIsAvailable): Correct build when not using numactl.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-07-21 12:50:00 -06:00
Roman Bogorodskiy
53939d58cb storage: logical: drop useless if
virStorageBackendLogicalCreateVol contains a piece like:

    if (vol->target.path != NULL) {
        /* A target path passed to CreateVol has no meaning */
        VIR_FREE(vol->target.path);
    }

The 'if' is useless here, but 'syntax-check' doesn't catch that
because of the comment, so drop the 'if'.
2014-07-21 21:34:14 +04:00
Roman Bogorodskiy
b5f57be2a2 Fix build on non-Linux platforms
Commit ef48a1b introduced virFindSCSIHostByPCI for Linux and
a stub for other platforms that returns -1 while the function
should return 'char *', so use 'return NULL' instead.

Commit fbd91d4 introduced virReadSCSIUniqueId with the third
argument 'int *result', however the stub for non-Linux patform
uses 'unsigned int *result', so change it to 'int *result'.

Pushed under the build breaker rule.
2014-07-21 21:26:00 +04:00
John Ferlan
ea37fb34a9 getAdapterName: Lookup stable scsi_host
If a parentaddr was provided in the XML, have getAdapterName lookup
the stable address.  This allows virStorageBackendSCSICheckPool() and
virStorageBackendSCSIRefreshPool() to automagically find the scsi_host
by its PCI address and unique_id
2014-07-21 12:55:11 -04:00
John Ferlan
ef48a1b613 scsi_host: Introduce virFindSCSIHostByPCI
Introduce a new function to parse the provided scsi_host parent address
and unique_id value in order to find the /sys/class/scsi_host directory
which will allow a stable SCSI host address

Add a test to scsihosttest to lookup the host# name by using the PCI address
and unique_id value
2014-07-21 12:55:11 -04:00
John Ferlan
f3271f4cb3 Add unique_id to nodedev output
Add an optional unique_id parameter to nodedev.  Allows for easier lookup
and display of the unique_id value in order to document for use with
scsi_host code.
2014-07-21 12:55:11 -04:00
John Ferlan
fbd91d496e virutil: Introduce virReadSCSIUniqueId
Introduce a new function to read the current scsi_host entry and return
the value found in the 'unique_id' file.

Add a 'scsihosttest' test (similar to the fchosttest, but incorporating some
of the concepts of the mocked pci test library) in order to read the
unique_id file like would be found in the /sys/class/scsi_host tree.
2014-07-21 12:55:11 -04:00
John Ferlan
aa9dac09b3 scsi_backend: Use existing LINUX_SYSFS_SCSI_HOST_PREFIX definition
Rather than supplying the path again in the formatting of the sysfs
scsi_host directory.
2014-07-21 12:55:10 -04:00
Osier Yang
a4bd62adc1 storage: Introduce parentaddr into virStoragePoolSourceAdapter
Between reboots and kernel reloads, the SCSI host number used for SCSI
storage pools may change requiring modification to the storage pool XML
in order to use a specific SCSI host adapter.

This patch introduces the "parentaddr" element and "unique_id" attribute
for the SCSI host adapter in order to uniquely identify the adapter
between reboots and kernel reloads. For now the goal is to only parse
and format the XML. Both will be required to be provided in order to
uniquely identify the desired SCSI host.

The new XML is expected to be as follows:

  <adapter type='scsi_host'>
    <parentaddr unique_id='3'>
      <address domain='0x0000' bus='0x00' slot='0x1f' func='0x2'/>
    </parentaddr>
  </adapter>

where "parentaddr" is the parent device of the SCSI host using the PCI
address on which the device resides and the value from the unique_id file
for the device. Both the PCI address and unique_id values will be used
to traverse the /sys/class/scsi_host/ directories looking at each link
to match the PCI address reformatted to the directory link format where
"domain🚌slot:function" is found.  Then for each matching directory
the unique_id file for the scsi_host will be used to match the unique_id
value in the xml.

For a PCI address listed above, this will be formatted to "0000:00:1f.2"
and the links in /sys/class/scsi_host will be used to find the host#
to be used for the 'scsi_host' device. Each entry is a link to the
/sys/bus/pci/devices directories, e.g.:

%  ls -al /sys/class/scsi_host/host2
lrwxrwxrwx. 1 root root 0 Jun  1 00:22 /sys/class/scsi_host/host2 -> ../../devices/pci0000:00/0000:00:1f.2/ata3/host2/scsi_host/host2

% cat /sys/class/scsi_host/host2/unique_id
3

The "parentaddr" and "name" attributes are mutually exclusive to identify
the SCSI host number. Use of the "parentaddr" element will be the preferred
mechanism.

This patch only supports to parse and format the XMLs. Later patches will
add code to find out the scsi host number.
2014-07-21 12:55:10 -04:00
Osier Yang
53f620568e virStoragePoolSourceAdapter: Refine the SCSI_HOST adapter name
Preparation for future patches by creating a scsi_host union. For now,
just the 'name' will be present.
2014-07-21 12:55:10 -04:00
John Ferlan
8d854e5b5b getAdapterName: check for SCSI_HOST
Rather than assume that NOT FC_HOST is SCSI_HOST, let's call them out
specifically. Makes it easier to find SCSI_HOST code/structs and ensures
something isn't missed in the future
2014-07-21 12:55:10 -04:00
Peter Krempa
b325be128a schema: pool: netfs: Don't enforce slash in glusterfs pool source
Gluster volumes don't start with a leading slash. Our schema for netfs
gluster pools enforces it though. Luckily mount.glusterfs skips it.
Allow a slashless volume name for glusterfs netfs mounts in the schema.

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1101999
2014-07-21 18:28:17 +02:00
Peter Krempa
bbfc826787 doc: Explicitly specify how to override spice channel mode
Be more clear that the "<channel mode=" attribute overrides the default
set by "defaultMode".

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1033704
2014-07-21 17:20:33 +02:00
Peter Krempa
6b1f9feccf node_device: HAL: Ignore return value of virStrToLong_ui
Commit 5df813177c forgot to adjust a few
callers of virStrToLong_ui to ignore the returned value in some ancient
parts of the code.
2014-07-21 16:32:53 +02:00
Peter Krempa
5df813177c util: Check return value from virStrToLong* functions
We do so in the vast majority of places, so there's no problem of adding
the attribute to enforce it by the complier and fix a few leftover
places.

This was originally pointed out by Coverity as a recent change triggered
it's warning that our code checked the vast majority of returns from
virStrToLong_ui.
2014-07-21 15:20:59 +02:00
Peter Krempa
ba2b4da02d examples: domtop: Avoid leaking memory
Use the virTypedParamsFree unconditionally as it handles NULL well and
has the benefit of freeing a typed parameter array even if it wasn't yet
assigned, but only allocated.
2014-07-21 15:20:59 +02:00