We're only adding only info about L3 caches, we can add more
later (just by changing one line), but for now that's more than enough
without overwhelming anyone.

XML snippet of how this should look like (also seen as part of the commit):

  <cache>
    <bank id='0' level='3' type='both' size='8192' unit='KiB' cpus='0-7'/>
  </cache>
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

It is no longer needed thanks to the great virfilewrapper.c.  And this
way we don't have to add a new set of functions for each prefixed
path.

While on that, add two functions that weren't there before, string and
scaled integer reading ones.  Also increase the length of the string
being read by one to accompany for the optional newline at the
end (i.e. change INT_STRLEN_BOUND to INT_BUFSIZE_BOUND).
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

When adding a nwfilter onto the list in
virNWFilterObjListAssignDef() this array is re-allocated to match
demand for new size. However, it is never freed leading to a
leak:

==26535== 136 bytes in 1 blocks are definitely lost in loss record 1,079 of 1,250
==26535==    at 0x4C2E2BE: realloc (vg_replace_malloc.c:785)
==26535==    by 0x54BA28E: virReallocN (viralloc.c:245)
==26535==    by 0x54BA384: virExpandN (viralloc.c:294)
==26535==    by 0x54BA657: virInsertElementsN (viralloc.c:436)
==26535==    by 0x55DB011: virNWFilterObjListAssignDef (virnwfilterobj.c:362)
==26535==    by 0x55DB530: virNWFilterObjListLoadConfig (virnwfilterobj.c:503)
==26535==    by 0x55DB635: virNWFilterObjListLoadAllConfigs (virnwfilterobj.c:539)
==26535==    by 0x2AC5A28B: nwfilterStateInitialize (nwfilter_driver.c:250)
==26535==    by 0x5621C64: virStateInitialize (libvirt.c:770)
==26535==    by 0x124379: daemonRunStateInit (libvirtd.c:881)
==26535==    by 0x554AC78: virThreadHelper (virthread.c:206)
==26535==    by 0x8F5F493: start_thread (in /lib64/libpthread-2.23.so)
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

... with VIR_NET_GENERATED_MACV???_PREFIX, which is defined in
util/virnetdevmacvlan.h.

Since VIR_NET_GENERATED_PREFIX is used for plain tap devices, it is
renamed to VIR_NET_GENERATED_TAP_PREFIX and moved to virnetdev.h

The parser had been clearing out *all* suggested device names for
type='direct' (aka macvtap) interfaces. All of the code implementing
macvtap allows for a user-specified device name, so we should allow
it. In the case that an interface name starts with "macvtap" or
"macvlan" though, we do still clear it out, just as we do with "vnet"
(which is the prefix used for automatically generated tap device
names), since those are the prefixes for the names we autogenerate for
macvtap and macvlan devices.

Resolves: https://bugzilla.redhat.com/1335798

This patch changes following output:

    ...
    <graphics type='vnc' port='-1' autoport='yes'>
      <listen type='none'/>
    </graphics>
    ...

into this output:

    ...
    <graphics type='vnc'>
      <listen type='none'/>
    </graphics>
    ...
Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1438682Signed-off-by: NPavel Hrdina <phrdina@redhat.com>
Acked-by: NAndrea Bolognani <abologna@redhat.com>

This patch introduces

    <cache level='N' mode='emulate'/>
    <cache mode='passthrough'/>
    <cache mode='disable'/>

sub element of /domain/cpu. Currently only a single <cache> element is
allowed.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

The type of this parameter is virCPUType so calling it 'mode' is pretty
strange, 'type' is a much better name.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

There's no need to separate, so just have one.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than overloading one function - split apart the logic to have
separate interfaces and local/private structures to manage the data
for which the helper is collecting.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Makes it a bit more clear what it is.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than 'nuuids' it should be 'maxuuids' and rather than 'got'
it should be 'nuuids'.  Alter the logic of the list traversal to
utilize those names.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than using "ret = -1" and cleanup processing, alter the return
path on failure to goto error and then just return the data.got.

In the error path, we no longer check for ret < 0, we just can free
anything added to the array and return -1 directly.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than dereferencing obj->def->X, create a local 'def' variable
that will dereference the def and use directly.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

When processing a virSecretPtr use 'secret' as a variable name.

When processing a virSecretObjPtr use 'obj' as a variable name.

When processing a virSecretDefPtr use 'def' as a variable name,
unless a distinction needs to be made with a 'newdef' such as
virSecretObjListAddLocked (which also used the VIR_STEAL_PTR macro
for the configFile and base64File).
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than have the caller check if !obj before calling, just check
in the function for !obj and return.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Make various virSecretObjList*Locked functions static and make
virSecretObjNew static since they're only called within virtsecretobj.c.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Only save the config when using a generated UUID if we were able to
create an object for the def. There could have been "other reasons"
for the assignment to fail, so saving the config could be incorrect.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than "wait" for the first config file to be created, force creation
of the configDir during driver state initialization.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Essentially virNWFilterSaveDef executed in a different order the same
sequence of calls, so let's just make one point of reference.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Remove open coded helper.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Move from virnwfilterobj.h to virnwfilterobj.c.

Create the virNWFilterObjListNew() API in order to allocate.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Prefix should have been virNWFilterObjList since the API is operating on
the list of filters.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Perform the object initialization in a helper rather than inline.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Move the structure to virnwfilterobj.c and create necessary accessor API's
for the various fields.

Also make virNWFilterObjFree static since there's no external callers.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

It is what it is anyway, so let's describe it that way too.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

It was only ever set to false, which is ironically the default.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than dereferencing obj->def->XXX or nwfilters->objs[i]->X
create local virNWFilterObjPtr and virNWFilterDefPtr variables.

Future adjustments will be privatizing the object more, so this just
prepares the code for that reality.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

When processing a virNWFilterPtr use 'nwfilter' as a variable name.

When processing a virNWFilterObjPtr use 'obj' as a variable name.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Format the attributes in a separate buffer and only
print the element if it's not empty.

Make adding new attributes easier.

Eliminate the big condition at the start.
Instead use a buffer and only format the element if the buffer
is non-empty.

Instead of figuring out upfront whether <input> will be a single
or a pair element, format the subelements into a separate buffer
and close <input/> early if this buffer is empty.

Shorten the time needed to keep the list lock and alter the cleanup
path to be more of an error path.

Utilize the the virObjectListFree function to handle the calls for
virObjectUnref on each list element and the VIR_FREE of the list
instead of open coding it.

Change the name of the virHashForEach callback to match the name
of the Export function with the Callback added onto it.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Rather than have caller need to do it, have the object returned locked.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Alter the prototypes to use the newer formatting style
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>

Commit id 'bb1fba62' neglected to set when creating the function.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>