Daniel Veillard [Fri, 14 Nov 2008 08:42:47 +0000 (08:42 +0000)]
KVM/QEmu migration support
* qemud/remote.c qemud/remote_dispatch_localvars.h
qemud/remote_dispatch_proc_switch.h qemud/remote_dispatch_prototypes.h
qemud/remote_protocol.c qemud/remote_protocol.h
qemud/remote_protocol.x src/driver.h src/libvirt.c
src/libvirt_internal.h src/libvirt_sym.version src/lxc_driver.c
src/openvz_driver.c src/qemu_conf.h src/qemu_driver.c
src/remote_internal.c src/test.c src/virsh.c: large patch to
add migration support for KVM/QEmu, based on the work of Rich Jones
and Chris Lalancette. This introduce a new version of the prepare
and finish steps of the migration, so changes the driver API and
a lot of code.
Daniel
Jim Meyering [Fri, 7 Nov 2008 16:44:38 +0000 (16:44 +0000)]
tweak lstat.c to avoid mingw link failure
* gnulib/lib/lstat.c: Include <sys/stat.h> *before* the use of stat in
orig_stat. Otherwise, on mingw (which lacks lstat), any program using
the lstat module would not get the redefinition-to-stat provided by
gnulib's sys/stat.h. Reported by Daniel P. Berrange.
Call vgscan before attempting to scan logical volumes. This makes sure we've
"touched" any logical volumes, so they appear in the subsequent "pvs" call.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Give iSCSI and disk storage backend drivers the
ability to resolve any kind of volume path to the pool target volume
path. For instance, if the pool was defined with a
<target><path>/dev/disk/by-id</path></target> section, and one of the
volumes is /dev/disk/by-id/scsi-S_beaf11, then you would be able to
call virStorageVolLookupByPath("/dev/sdc"), and get the correct volume
back.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Daniel Veillard [Fri, 31 Oct 2008 10:13:45 +0000 (10:13 +0000)]
* python/Makefile.am python/generator.py python/libvir.c
python/libvir.py python/libvirt_wrap.h python/types.c:
adds support for events from the python bindings, also
improves the generator allowing to embbed per function
definition files, patch by Ben Guthro
* examples/domain-events/events-python/event-test.py: also
adds a programming example
Daniel
Jim Meyering [Tue, 28 Oct 2008 19:09:04 +0000 (19:09 +0000)]
remove src/socketcompat.h; no longer needed
* src/socketcompat.h: Remove file.
* src/remote_internal.c: Don't include socketcompat.h.
* qemud/qemud.h: Likewise.
* qemud/remote_protocol.x: Likewise, but do include <arpa/inet.h>
for htonl.
* src/remote_internal.c: Likewise, but do include <errno.h>.
* qemud/remote_protocol.c: Regenererate.
* qemud/remote_protocol.h: Regenererate.
* src/Makefile.am (REMOTE_DRIVER_SOURCES): Remove socketcompat.h.
Jim Meyering [Tue, 28 Oct 2008 17:48:06 +0000 (17:48 +0000)]
avoid compiler warning when all storage backends are disabled
* src/storage_backend.c (backends): Add a NULL terminator.
(virStorageBackendForType): Use NULL terminator rather than
warning-provoking (possibly 0) array size.
Chris Lalancette [Thu, 23 Oct 2008 13:21:01 +0000 (13:21 +0000)]
Brown paper bag bug. I didn't preserve the
<sources> tags properly, and I didn't put the <source> tag into the
proper place, so I totally changed the output of the discovery XML.
Fix this up to do the right thing.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Daniel Veillard [Thu, 23 Oct 2008 13:18:18 +0000 (13:18 +0000)]
Massive patch adding event APIs by Ben Guthro
* include/libvirt/libvirt.h include/libvirt/libvirt.h.in
src/libvirt.c src/libvirt_sym.version: new libvirt event entry
points, big patch provided by Ben Guthro
* Makefile.am configure.in src/driver.h src/event.c src/event.h
src/internal.h src/libvirt.c src/libvirt_sym.version src/lxc_driver.c
src/openvz_driver.c src/qemu_conf.h src/qemu_driver.c
src/remote_internal.c src/storage_backend_fs.c src/test.c
qemud/event.c qemud/event.h qemud/mdns.c qemud/qemud.c
qemud/qemud.h qemud/remote.c qemud/remote_dispatch_localvars.h
qemud/remote_dispatch_proc_switch.h qemud/remote_dispatch_prototypes.h
qemud/remote_protocol.c qemud/remote_protocol.h
qemud/remote_protocol.x proxy/Makefile.am python/generator.py:
Not much is left untouched by the patch adding the events support
* docs/libvirt-api.xml docs/libvirt-refs.xml
docs/html/libvirt-libvirt.html: regenerated the docs
* examples/domain-events/events-c/Makefile.am
examples/domain-events/events-c/event-test.c: a test example
* AUTHORS: added Ben Guthro
daniel
Chris Lalancette [Thu, 23 Oct 2008 11:39:53 +0000 (11:39 +0000)]
An updated patch for adding <device> tags to logical
findPoolSources. Given danpb's last feedback, I completely removed the XML
parsing and did it all with structures. The result should (hopefully) be a lot
easier on the eyes, and is a little more generic.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Chris Lalancette [Thu, 23 Oct 2008 11:32:22 +0000 (11:32 +0000)]
As suggested by danpb, to fix up the regression caused by last week's VIR_ENUM
cleanup patch, add a ".defaultFormat" member to .poolOptions. In
storage_conf.c, if virXPathString(/pool/source/format/@type) returns NULL, then
set the pool type to .defaultFormat; otherwise, lookup the type via
formatFromString.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
To make this more idempotent, do the <device name='/dev/sdb'/> form by default,
and only do the <device>...</device> form if .nfreeExtent is defined for the
storage pool.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Chris Lalancette [Tue, 21 Oct 2008 17:18:45 +0000 (17:18 +0000)]
Because of my patch last week that converted the various virStorage*FromString
and virStorage*ToString implementations to the generic VIR_ENUM_IMPL, there were
a couple of places that didn't properly set errors when they failed. This patch
fixes these places up.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Chris Lalancette [Tue, 21 Oct 2008 17:15:53 +0000 (17:15 +0000)]
While doing on this patch series, I mistakenly added a bogus piece of
storage XML to libvirtd, which was saved in /etc/libvirt/storage. On subsequent
stop/start of libvirtd, because of another bug , an
error wasn't being set properly in an error path, so libvirtd was SEGV'ing in
storage_conf.c:virStoragePoolObjLoad when trying to dereference the NULL err
returned from virGetLastError(). Make this more robust against errors by always
doing "err ? err->message : NULL" in the printf. I looked around the tree and
found a couple of other places that weren't guarded, so this patch fixes them as
well.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Daniel Veillard [Tue, 21 Oct 2008 14:26:59 +0000 (14:26 +0000)]
Document somme commit rules, and add application monitoring links to docs
* HACKING: update with some rules for commiters
* docs/apps.html docs/apps.html.in: add a section on monitoring
support
daniel
Jim Meyering [Fri, 17 Oct 2008 10:03:15 +0000 (10:03 +0000)]
generate .gitignore files from .cvsignore ones
* Makefile.maint (sync-vcs-ignore-files): New target.
Prompted by a patch from James Morris.
http://thread.gmane.org/gmane.comp.emulators.libvirt/8619/focus=8773
Add all (now-generated) .gitignore files.
* .gitignore: New file.
* build-aux/.gitignore: New file.
* docs/.gitignore: New file.
* docs/devhelp/.gitignore: New file.
* docs/examples/.gitignore: New file.
* docs/examples/python/.gitignore: New file.
* gnulib/lib/.gitignore: New file.
* gnulib/lib/arpa/.gitignore: New file.
* gnulib/lib/netinet/.gitignore: New file.
* gnulib/lib/sys/.gitignore: New file.
* gnulib/tests/.gitignore: New file.
* include/.gitignore: New file.
* include/libvirt/.gitignore: New file.
* po/.gitignore: New file.
* proxy/.gitignore: New file.
* python/.gitignore: New file.
* python/tests/.gitignore: New file.
* qemud/.gitignore: New file.
* src/.gitignore: New file.
* tests/.gitignore: New file.
* tests/confdata/.gitignore: New file.
* tests/sexpr2xmldata/.gitignore: New file.
* tests/virshdata/.gitignore: New file.
* tests/xencapsdata/.gitignore: New file.
* tests/xmconfigdata/.gitignore: New file.
* tests/xml2sexprdata/.gitignore: New file.
Jim Meyering [Fri, 17 Oct 2008 09:59:02 +0000 (09:59 +0000)]
generate .gitignore files from .cvsignore ones
* Makefile.maint (sync-vcs-ignore-files): New target.
Prompted by a patch from James Morris.
http://thread.gmane.org/gmane.comp.emulators.libvirt/8619/focus=8773
Add all (now-generated) .gitignore files.
* .gitignore: New file.
* build-aux/.gitignore: New file.
* docs/.gitignore: New file.
* docs/devhelp/.gitignore: New file.
* docs/examples/.gitignore: New file.
* docs/examples/python/.gitignore: New file.
* gnulib/lib/.gitignore: New file.
* gnulib/lib/arpa/.gitignore: New file.
* gnulib/lib/netinet/.gitignore: New file.
* gnulib/lib/sys/.gitignore: New file.
* gnulib/tests/.gitignore: New file.
* include/.gitignore: New file.
* include/libvirt/.gitignore: New file.
* po/.gitignore: New file.
* proxy/.gitignore: New file.
* python/.gitignore: New file.
* python/tests/.gitignore: New file.
* qemud/.gitignore: New file.
* src/.gitignore: New file.
* tests/.gitignore: New file.
* tests/confdata/.gitignore: New file.
* tests/sexpr2xmldata/.gitignore: New file.
* tests/virshdata/.gitignore: New file.
* tests/xencapsdata/.gitignore: New file.
* tests/xmconfigdata/.gitignore: New file.
* tests/xml2sexprdata/.gitignore: New file.
Daniel Veillard [Fri, 17 Oct 2008 09:29:29 +0000 (09:29 +0000)]
qemudDomainBlockStats use qemudDiskDeviceName
* src/qemu_driver.c: use qemudDiskDeviceName to determine the block
device name in qemudDomainBlockStats(), patch by Guido Günther
daniel
Jim Meyering [Thu, 16 Oct 2008 17:18:48 +0000 (17:18 +0000)]
build: when po-check fails, say why and suggest a fix
* Makefile.maint (po-check): Before, when this check failed, it just
spat out a diff mentioning two temporary files. Now, it tells you
what's wrong and suggests a fix with a patch using the name of the
affected file (rather than temporary file names) in the diff output.
Chris Lalancette [Thu, 16 Oct 2008 15:06:03 +0000 (15:06 +0000)]
Add support for detecting the partition table type when scanning
iSCSI volumes. This is implemented in the
virStorageBackendUpdateVolInfoFD function, so all future callers will
automatically benefit. This is a somewhat large patch because the
conversion of the virStorageBackendPartTableTypeToString necessitated
a change to the formatToString and formatFromString function pointers,
which caused fallout in other places in the storage stuff. The good
news is that most of these callers are now converted over to the
VIR_ENUM_IMPL, which means a lot of redundant code is now gone.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
Chris Lalancette [Thu, 16 Oct 2008 12:04:06 +0000 (12:04 +0000)]
"make syntax-check" was complaining that network_driver.c was missing from
POTFILES.in. Add it there, and then fix up one warning about included
c-ctypes.h that wasn't being used.
Signed-off-by: Chris Lalancette <clalance@redhat.com>