Matthias Bolte [Sun, 1 May 2011 19:57:43 +0000 (21:57 +0200)]
esx: Fix race condition in esxVI_EnsureSession
When the session has expired then multiple threads can race while
reestablishing it.
This race condition is not that critical as it requires a special usage
pattern to be triggered. It can only happen when an application doesn't
do API calls for quite some time (the session expires after 30 min
inactivity) and then multiple threads doing simultaneous API calls and
end up doing simultaneous calls to esxVI_EnsureSession.
Jiri Denemark [Mon, 9 May 2011 11:57:09 +0000 (13:57 +0200)]
virsh: Correctly initialize libvirt
virsh didn't call virInitialize(), which (among other things)
initializes virLastErr thread local variable. As a result of that, virsh
could just segfault in virEventRegisterDefaultImpl() since that is the
first call that touches (resets) virLastErr.
I have no idea what lucky coincidence made this bug visible but I was
able to reproduce it in 100% cases but only in one specific environment
which included building in sandbox.
By running the doTunnelSendAll code in a separate thread, the
main thread can do qemuMigrationWaitForCompletion as with
normal migration. This in turn ensures that job signals work
correctly and that progress monitoring can be done
* src/qemu/qemu_migration.c: Run tunnelled migration in
separate thread
Close all sockets before cancelling QEMU migration
Cancelling the QEMU migration may cause QEMU to flush pending
data on the migration socket. This may in turn block QEMU if
nothing reads from the other end of the socket. Closing the
socket before cancelling QEMU migration avoids this possible
deadlock.
* src/qemu/qemu_migration.c: Close sockets before cancelling
migration on failure
Ensure we always read a full buffer in tunnelled migration
The 'nbytes' variable was not re-initialized to the
buffer size on each iteration of the tunnelled migration
loop. While saferead() will ensure a full read, except
on EOF, it is clearer to use the real buffer size
* src/qemu/qemu_migration.c: Always read full buffer of data
Refactor migration completion loop to allow code reuse
The qemuMigrationWaitForCompletion method contains a loop which
repeatedly queries QEMU to check migration progress, and also
processes job signals (pause, setspeed, setbandwidth, cancel).
The tunnelled migration loop does not currently support this
functionality, but should. Refactor the code to allow it to
be used with tunnelled migration.
Implement the v3 migration protocol, which has two extra
steps, 'begin' on the source host and 'confirm' on the
source host. All other methods also gain both input and
output cookies to allow bi-directional data passing at
all stages.
The QEMU peer2peer migration method gains another impl
to provide the v3 migration. This finally allows migration
cookies to work with tunnelled migration, which is required
for Spice seamless migration & the lock manager transfer
* src/qemu/qemu_driver.c: Wire up migrate v3 APIs
* src/qemu/qemu_migration.c, src/qemu/qemu_migration.h: Add
begin & confirm methods, and peer2peer impl of v3
Merge the doNonTunnelMigrate2 and doTunnelMigrate2 methods
into one doPeer2PeerMigrate2 method, since they are substantially
the same. With the introduction of v3 migration, this will be
even more important, to avoid massive code duplication.
To facilitate the introduction of the v3 migration protocol,
the doTunnelMigrate method is refactored into two pieces. One
piece is intended to mirror the flow of virDomainMigrateVersion2,
while the other is the helper for setting up sockets and processing
the data.
Previously socket setup would be done before the 'prepare' step,
so errors could be dealt with immediately, avoiding need to shut
off the destination QEMU. In the new split, socket setup is done
after the 'prepare' step. This is not a serious problem, since
the control flow already requires calling 'finish' to tear down
the destination QEMU upon several errors.
Wire up SPICE client relocation with QEMU migration
Use the graphics information from the QEMU migration cookie to
issue a 'client_migrate_info' monitor command to QEMU. This causes
the SPICE client to automatically reconnect to the target host
when migration completes
* src/qemu/qemu_migration.c: Set data for SPICE client relocation
before starting migration on src
* src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h,
src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_json.h,
src/qemu/qemu_monitor_text.c, src/qemu/qemu_monitor_text.h: Add
new qemuMonitorGraphicsRelocate() command
Pass graphics setup from dst back to src via migration cookies
Extend the QEMU migration cookie structure to allow information
about the destination host graphics setup to be passed by to
the source host. This will enable seamless migration of any
connected graphics clients
* src/qemu/qemu_migration.c: Add graphics info to migration
cookies
* daemon/libvirtd.c: Always initialize gnutls to enable
x509 cert parsing in QEMU
The migration protocol has support for a 'cookie' parameter which
is an opaque array of bytes as far as libvirt is concerned. Drivers
may use this for passing around arbitrary extra data they might
need during migration. The QEMU driver needs to do a few things:
- Pass hostname/uuid to allow strict protection against localhost
migration attempts
- Pass SPICE/VNC server port from the target back to the source to
allow seamless relocation of client sessions
- Pass lock driver state from source to destination
This patch introduces the basic glue for handling cookies
but only includes the host/guest UUID & name.
* src/libvirt_private.syms: Export virXMLParseStrHelper
* src/qemu/qemu_migration.c, src/qemu/qemu_migration.h: Parsing
and formatting of migration cookies
* src/qemu/qemu_driver.c: Pass in cookie parameters where possible
* src/remote/remote_protocol.h, src/remote/remote_protocol.x: Change
cookie max length to 16384 bytes
Migration just seems to go from bad to worse. We already had to
introduce a second migration protocol when adding the QEMU driver,
since the one from Xen was insufficiently flexible to cope with
passing the data the QEMU driver required.
It turns out that this protocol still has some flaws that we
need to address. The current sequence is
* Src: DumpXML
- Generate XML to pass to dst
* Dst: Prepare
- Get ready to accept incoming VM
- Generate optional cookie to pass to src
* Src: Perform
- Start migration and wait for send completion
- Kill off VM if successful, resume if failed
* Dst: Finish
- Wait for recv completion and check status
- Kill off VM if unsuccessful
The problems with this are:
- Since the first step is a generic 'DumpXML' call, we can't
add in other migration specific data. eg, we can't include
any VM lease data from lock manager plugins
- Since the first step is a generic 'DumpXML' call, we can't
emit any 'migration begin' event on the source, or have
any hook that runs right at the start of the process
- Since there is no final step on the source, if the Finish
method fails to receive all migration data & has to kill
the VM, then there's no way to resume the original VM
on the source
This patch attempts to introduce a version 3 that uses the
improved 5 step sequence
* Src: Begin
- Generate XML to pass to dst
- Generate optional cookie to pass to dst
* Dst: Prepare
- Get ready to accept incoming VM
- Generate optional cookie to pass to src
* Src: Perform
- Start migration and wait for send completion
- Generate optional cookie to pass to dst
* Dst: Finish
- Wait for recv completion and check status
- Kill off VM if failed, resume if success
- Generate optional cookie to pass to src
* Src: Confirm
- Kill off VM if success, resume if failed
The API is designed to allow both input and output cookies
in all methods where applicable. This lets us pass around
arbitrary extra driver specific data between src & dst during
migration. Combined with the extra 'Begin' method this lets
us pass lease information from source to dst at the start of
migration
Moving the killing of the source VM out of Perform and
into Confirm, means we can now recover if the dst host
can't successfully Finish receiving migration data.
Automatically generate the hvsupport.html.in file from source files
The hvsupport.html.in file is forever out of date. By annotating
the driver struct tables in each driver with version information,
we can auto-generate the hvsupport.html.in file. Annotating the
drivers will be mandatory for new patches, ensuring hvsupport.html.in
is never out of date again.
Convert all driver struct intializers to C99 style
Change all the driver struct initializers to use the
C99 style, leaving out unused fields. This will make
it possible to add new APIs without changing every
driver. eg change:
Jiri Denemark [Wed, 4 May 2011 09:07:01 +0000 (11:07 +0200)]
Implement domain state reason
Only in drivers which use virDomainObj, drivers that query hypervisor
for domain status need to be updated separately in case their hypervisor
supports this functionality.
The reason is also saved into domain state XML so if a domain is not
running (i.e., no state XML exists) the reason will be lost by libvirtd
restart. I think this is an acceptable limitation.
Matthias Bolte [Sun, 1 May 2011 06:05:58 +0000 (08:05 +0200)]
esx: Generate implicit _this macros
Several vSphere API methods are called on global objects like the
FileManager, the PerformanceManager or the SearchIndex. The generator
input file allows to mark such methods and the generator generates
such method in a way that automatically handles marked parameter. This
is done by some special macros. Those were manually written and this
patch moves them to the generator.
Cole Robinson [Fri, 13 May 2011 15:00:10 +0000 (11:00 -0400)]
remote: Use virDomainEventState helpers
One functionality change here is that we no longer force enable the event
timeout for every queued event, only enable it for the first event after
the queue has been flushed. This is how other drivers have already done it,
and I haven't encountered problems in practice.
v3:
Adjust for new virDomainEventStateNew argument
Cole Robinson [Thu, 12 May 2011 20:12:05 +0000 (16:12 -0400)]
domain: Require <init> for container guests
Use capabilities to allow a driver to register a default <init> if none
is specified in the XML. Openvz was already open-coding this to be /sbin/init
LXC currently falls over if no init is specified, so an explicit error is
an improvement IMO.
(Side note: I don't think we can set a default value for LXC. If we use
/sbin/init but the user doesn't specify a separate root FS for their guest,
the container will rerun the host's init which can be traumatic :). For
virt-install I'm thinking of defaulting to /sbin/init if a root FS has
been specified, otherwise require the user to manually specify <init>)
virFDStream: Add option for delete file after it's opening
This is needed if we want to transfer a temporary file. If the
transfer is done with iohelper, we might run into a race condition,
where we unlink() file before iohelper is executed.
* src/fdstream.c, src/fdstream.h,
src/util/iohelper.c: Add new option
* src/lxc/lxc_driver.c, src/qemu/qemu_driver.c,
src/storage/storage_driver.c, src/uml/uml_driver.c,
src/xen/xen_driver.c: Expand existing function calls
Osier Yang [Thu, 12 May 2011 10:48:23 +0000 (18:48 +0800)]
docs: Fix documentation for cputune parameters
This re-adds the example section originally written by Osier Yang,
and indicates the version in which the cputune parameters became
available in libvirt.
Signed-off-by: Igor Serebryany <igor47@moomers.org>
Eric Blake [Wed, 11 May 2011 22:30:36 +0000 (16:30 -0600)]
nwfilter: drop unused flag argument
The public API and RPC over-the-wire format have no flags argument,
so neither should the internal callback API. This simplifies the
RPC generator.
* src/driver.h (virDrvNWFilterDefineXML): Drop argument that does
not match public API.
* src/nwfilter/nwfilter_driver.c (nwfilterDefine): Likewise.
* src/libvirt.c (virNWFilterDefineXML): Likewise.
* daemon/remote_generator.pl: Drop special case.
Eric Blake [Wed, 11 May 2011 15:08:44 +0000 (09:08 -0600)]
maint: omit translation for all VIR_INFO
We were 31/73 on whether to translate; since less than 50% translated
and since VIR_INFO is less than VIR_WARN which also doesn't translate,
this makes sense.
Eric Blake [Wed, 11 May 2011 20:04:00 +0000 (14:04 -0600)]
build: drop need for VIR_DOMAIN_DEBUG0()
This one's tricker than the VIR_DEBUG0() removal, but the end
result is still C99 compliant, and reasonable with enough comments.
* src/libvirt.c (VIR_ARG10, VIR_HAS_COMMA)
(VIR_DOMAIN_DEBUG_EXPAND, VIR_DOMAIN_DEBUG_PASTE): New macros.
(VIR_DOMAIN_DEBUG): Rewrite to handle one argument, moving
multi-argument guts to...
(VIR_DOMAIN_DEBUG_1): New macro.
(VIR_DOMAIN_DEBUG0): Rename to VIR_DOMAIN_DEBUG_0.
glibc 2.13.90 has obsoleted its rpc implementation in favour of
the one provided by the TI-RPC library:
> * The RPC implementation in libc is obsoleted. Old programs keep working
> but new programs cannot be linked with the routines in libc anymore.
> Programs in need of RPC functionality must be linked against TI-RPC.
> The TI-RPC implemtation is IPv6 enabled and there are other benefits.
>
> Visible changes of this change include (obviously) the inability to
> link
> programs using RPC functions without referencing the TI-RPC library,
> the
> removal of the RPC headers from the glibc headers, and the lack of
> symbols defined in <rpc/netdb.h> when <netdb.h> is installed.
> Implemented by Ulrich Drepper.
(from glibc NEWS)
Thus with recent glibc, we need to try linking with -ltirpc when looking
for the XDR functions. The daemon also needs to use XDR_CFLAGS to be able
to find the XDR headers stored in /usr/include/tirpc.
When using TI-RPC, there are some warnings about redundant declarations, but
the fix probably belongs in other modules:
/usr/include/tirpc/rpc/rpcent.h:68:13: warning: redundant redeclaration of
'setrpcent' [-Wredundant-decls]
/usr/include/rpc/netdb.h:53:13: note: previous declaration of 'setrpcent'
was here
/usr/include/tirpc/rpc/rpcent.h:69:13: warning: redundant redeclaration of
'endrpcent' [-Wredundant-decls]
/usr/include/rpc/netdb.h:54:13: note: previous declaration of 'endrpcent'
was here
/usr/include/tirpc/rpc/rpc.h:84:12: warning: redundant redeclaration of
'bindresvport' [-Wredundant-decls]
/usr/include/netinet/in.h:440:12: note: previous declaration of
'bindresvport' was here
Lai Jiangshan [Mon, 9 May 2011 09:24:09 +0000 (17:24 +0800)]
libvirt,logging: cleanup VIR_XXX0()
These VIR_XXXX0 APIs make us confused, use the non-0-suffix APIs instead.
How do these coversions works? The magic is using the gcc extension of ##.
When __VA_ARGS__ is empty, "##" will swallow the "," in "fmt," to
avoid compile error.
Eric Blake [Wed, 4 May 2011 23:03:52 +0000 (17:03 -0600)]
node_device: avoid null dereference on error
If we plow on after udev_device_get_syspath fails, we will hit a NULL
dereference. Clang found one due to strdup later in udevSetParent,
but in fact we hit a NULL dereference sooner because of the use of
STREQ within virNodeDeviceFindBySysfsPath.
projects / libvirt.git / log