MIG_CMD_PACKAGED is a migration command that wraps a chunk of migration
stream inside a package whose length can be determined purely by reading
its header.  The destination guarantees that the whole MIG_CMD_PACKAGED
is read off the stream prior to parsing the contents.

This is used by postcopy to load device state (from the package)
while leaving the main stream free to receive memory pages.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The state of the postcopy process is managed via a series of messages;
   * Add wrappers and handlers for sending/receiving these messages
   * Add state variable that track the current state of postcopy
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The 'postcopy ram' capability allows postcopy migration of RAM;
note that the migration starts off in precopy mode until
postcopy mode is triggered (see the migrate_start_postcopy
patch later in the series).
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Postcopy needs to have two migration streams loading concurrently;
one from memory (with the device state) and the other from the fd
with the memory transactions.

Split the core of qemu_loadvm_state out so we can use it for both.

Allow the inner loadvm loop to quit and cause the parent loops to
exit as well.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Open a return path, and handle messages that are received upon it.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add 'migration_is_setup_or_active' utility function to check state.
(It gets postcopy added to it's list later on in the series)
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add migrate_send_rp_message to send a message from destination to source along the return path.
  (It uses a mutex to let it be called from multiple threads)
Add migrate_send_rp_shut to send a 'shut' message to indicate
  the destination is finished with the RP.
Add migrate_send_rp_ack to send a 'PONG' message in response to a PING
  Use it in the MSG_RP_PING handler
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add two src->dest commands:
   * OPEN_RETURN_PATH - To request that the destination open the return path
   * PING - Request an acknowledge from the destination
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Create QEMU_VM_COMMAND section type for sending commands from
source to destination.  These commands are not intended to convey
guest state but to control the migration process.

For use in postcopy.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The destination sets the fd to non-blocking on incoming migrations;
this also affects the return path from the destination, and thus we
need to make sure we can safely write to the return path.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Postcopy needs a method to send messages from the destination back to
the source, this is the 'return path'.

Wire it up for 'socket' QEMUFile's.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Postcopy uses the userfaultfd.h feature in the Linux kernel; include
the header.

(In early versions of the patch series we had this, and then we dropped
this by only including it if the kernel headers defined the syscall
number; however 1842bdfd added the syscall definition to our
headers, which means we can't tell if the kernel has it or not)
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

In postcopy we're going to need to perform the complete phase
for postcopiable devices at a different point, start out by
renaming all of the 'complete's to make the difference obvious.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Suspend to file is very much like a migrate, and it makes life
easier if we have the Migration state available, so initialise it
in the savevm.c code for suspending.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewd-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The main RAM load loop has a call to host_from_stream_offset for
each page type that actually loads data with the same test;
factor it out before the switch.

The host = NULL is to silence a bogus gcc warning of
an unitialised in the RAM_SAVE_COMPRESS_PAGE case, it
doesn't seem to realise that host is always initialised by the if at
the top in the cases the switch takes.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Useful for debugging the migration bitmap and other bitmaps
of the same format (including the sentmap in postcopy).

The bitmap is printed to stderr.
Lines that are all the expected value are excluded so the output
can be quite compact for many bitmaps.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add QEMU_MADV_NOHUGEPAGE as an OS-independent version of
MADV_NOHUGEPAGE.

We include sys/mman.h before making the test to ensure
that we pick up the system defines.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add a wrapper to change the blocking status on a QEMUFile
rather than having to use qemu_set_block(qemu_get_fd(f));
it seems best to avoid exposing the fd since not all QEMUFile's
really have one.  With this wrapper we could move the implementation
down to be different on different transports.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

qemu_get_buffer always copies the data it reads to a users buffer,
however in many cases the file buffer inside qemu_file could be given
back to the caller, avoiding the copy.  This isn't always possible
depending on the size and alignment of the data.

Thus 'qemu_get_buffer_in_place' either copies the data to a supplied
buffer or updates a pointer to the internal buffer if convenient.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

'file' becomes confusing when you have flows in each direction;
rename to make it clear.
This leaves just the main forward direction ms->file, which is used
in a lot of places and is probably not worth renaming given the churn.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Add a function to find a RAMBlock by name; use it in two
of the places that already open code that loop; we've
got another use later in postcopy.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Postcopy sends RAMBlock names and offsets over the wire (since it can't
rely on the order of ramaddr being the same), and it starts out with
HVA fault addresses from the kernel.

qemu_ram_block_from_host translates a HVA into a RAMBlock, an offset
in the RAMBlock and the global ram_addr_t value.

Rewrite qemu_ram_addr_from_host to use qemu_ram_block_from_host.

Provide qemu_ram_get_idstr since its the actual name text sent on the
wire.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The HOST_PAGE_ALIGN macros don't work until the page size variables
have been set up; later in postcopy I use those macros in the RAM
code, and it can be triggered using -object.

Fix this by initialising page_size_init() earlier - it's currently
initialised inside the accelerators, move it up into vl.c.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The vmstate_configuration is currently written
in 'qemu_savevm_state_begin', move it to
'qemu_savevm_state_header' since it's got a hard
requirement that it must be the 1st thing after
the header.
(In postcopy some 'command' sections get sent
early before the saving of the main sections
and hence before qemu_savevm_state_begin).
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

The migration code generally is built target-independent, however
there are a few places where knowing the target page size would
avoid artificially moving stuff into migration/ram.c.

Provide 'qemu_target_page_bits()' that returns TARGET_PAGE_BITS
to other bits of code so that they can stay target-independent.
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NAmit Shah <amit.shah@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

QAPI patches

# gpg: Signature made Tue 10 Nov 2015 07:12:25 GMT using RSA key ID EB918653
# gpg: Good signature from "Markus Armbruster <armbru@redhat.com>"
# gpg:                 aka "Markus Armbruster <armbru@pond.sub.org>"

* remotes/armbru/tags/pull-qapi-2015-11-10:
  qapi-introspect: Document lack of sorting
  qapi: Provide nicer array names in introspection
  qapi: More tests of input arrays
  qapi: Test failure in middle of array parse
  qapi: More tests of alternate output
  qapi: Simplify error cleanup in test-qmp-*
  qapi: Simplify non-error testing in test-qmp-*
  qapi: Plug leaks in test-qmp-*
  qapi: Share test_init code in test-qmp-input*
  qobject: Protect against use-after-free in qobject_decref()
  qapi: Strengthen test of TestStructList
  qapi: Use generated TestStruct machinery in tests
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

qapi-code-gen.txt already claims that types, commands, and
events share a common namespace; set this in stone by further
documenting that our introspection output will never have
collisions with the same name tied to more than one meta-type.

Our largest QMP enum currently has 125 values, our largest
object type has 27 members, and the mean for each is less than
10.  These sizes are small enough that the per-element overhead
of O(log n) binary searching probably outweighs the speed
possible with direct O(n) linear searching (a better algorithm
with more overhead will only beat a leaner naive algorithm only
as you scale to larger input sizes).

Arguably, the overall SchemaInfo array could be sorted by name;
there, we currently have 531 entities, large enough for a binary
search to be faster than linear.  However, remember that we have
mutually-recursive types, which means there is no topological
ordering that will allow clients to learn all information about
that type in a single linear pass; thus clients will want to do
random access over the data, and they will probably read the
introspection output into a hashtable for O(1) lookup rather
than O(log n) binary searching, at which point, pre-sorting our
introspection output doesn't help the client.

It doesn't help that sorting can be subjective if you introduce
locales into the mix (I'm not experienced enough with Python
to know for sure, but at least it looks like it defaults to
sorting in the C locale even when run under a different locale).
And while our current introspection output is deterministic
(because we visit entities in a sorted order), we may want
to change that order in the future (such as using OrderedDict
to stick to .json declaration order).

For these reasons, we simply document that clients should not
rely on any particular order of items in introspection output.
And since it is now a documented part of the contract, we have
the freedom to later rearrange output if needed, without
worrying about breaking well-written clients.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-13-git-send-email-eblake@redhat.com>
[Commit message tweaked]
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

For the sake of humans reading introspection output, it is nice
to have the name of implicit array types be recognizable as
arrays of the underlying type.  However, while this patch allows
humans to skip from a command with return type "[123]" straight
to the definition of type "123" without having to first inspect
type "[123]", document that this shortcut should not be taken by
client apps.

This makes the resulting introspection string slightly larger by
default (just over 200 bytes), but it's in the noise (less than
0.3% of the overall 70k size of 'query-qmp-capabilities').
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-12-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Our testsuite had no coverage of empty arrays, nor of what
happens when the input does not match the expected type.
Useful to have, especially if we start changing the visitor
contracts.

I did not think it worth duplicating these additions to
test-qmp-input-strict; since all strict mode does is add
the ability to reject JSON input that has more keys than
what the visitor expects, yet the additions in this patch
error out earlier than that point regardless of whether
strict mode was requested.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-11-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Our generated list visitors have the same problem as has been
mentioned elsewhere (see commit 2f52e205): they allocate data
even on failure. An upcoming patch will correct things to
provide saner guarantees, but first we need to expose the
behavior in the testsuite to ensure we aren't introducing any
memory usage bugs.

There are more test cases throughout the test-qmp-input-* tests
that already deal with partial allocation; a later commit will
clean up all visit_type_FOO(), without marking all of the tests
with FIXME at this time.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-10-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

The testsuite was only covering that we could output the 'int'
branch of an alternate (no additional allocation/cleanup required).
Add a test of the 'str' branch, to make sure that things still
work even when a branch involves allocation.

Update to modern style of g_new0() over g_malloc0() while
touching it.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-9-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

We have several tests that perform multiple sub-actions that are
expected to fail.  Asserting that an error occurred, then clearing
it up to prepare for the next action, turned into enough
boilerplate that it was sometimes forgotten (for example, a number
of tests added to test-qmp-input-visitor.c in d88f5fd1 leaked err).
Worse, if an error is not reset to NULL, we risk invalidating
later use of that error (passing a non-NULL err into a function
is generally a bad idea).  Encapsulate the boilerplate into a
single helper function error_free_or_abort(), and consistently
use it.

The new function is added into error.c for use everywhere,
although it is anticipated that testsuites will be the main
client.
Signed-off-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Commit b553a042 inadvertently disabled optimization
for all non-fortify builds. Fix this bug so we only do an
unoptimized build if we want debug.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Message-id: 1447082049-25099-1-git-send-email-peter.maydell@linaro.org

Signed integer overflow in C is undefined behaviour, and the compiler
is at liberty to assume it can never happen and optimize accordingly.
In particular, the subtractions in hpet_time_after() and hpet_time_after64()
were causing OSX clang to optimize the code such that it was prone to
hangs and complaints about the main loop stalling (presumably because
we were spending all our time trying to service very high frequency
HPET timer callbacks). The clang sanitizer confirms the UB:

hw/timer/hpet.c:119:26: runtime error: signed integer overflow: -2146967296 - 2147003978 cannot be represented in type 'int'

Fix this by doing the subtraction as an unsigned operation and then
converting to signed for the comparison.
Reported-by: NAaron Elkins <threcius@yahoo.com>
Cc: qemu-stable@nongnu.org
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Message-id: 1447080991-24995-1-git-send-email-peter.maydell@linaro.org

By using &error_abort, we can avoid a local err variable in
situations where we expect success.  It also has the nice
effect that if the test breaks, the error message from
error_abort tends to be nicer than that of g_assert().

This patch has an additional bonus of fixing several call sites that
were passing &err to two different functions without checking it in
between.  In general that is unsafe practice; because if the first
function sets an error, the second function could abort() if it tries to
set a different error. We got away with it because we were asserting
that err was NULL through the entire chain, but switching to
&error_abort avoids the questionable practice up front.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Make valgrind happy with the current state of the tests, so that
it is easier to see if future patches introduce new memory problems
without being drowned in noise.  Many of the leaks were due to
calling a second init without tearing down the data from an earlier
visit.  But since teardown is already idempotent, and we already
register teardown as part of input_visitor_test_add(), it is nicer
to just make init() safe to call multiple times than it is to have
to make all tests call teardown.

Another common leak was forgetting to clean up an error object,
after testing that an error was raised.

Another leak was in test_visitor_in_struct_nested(), failing to
clean the base member of UserDefTwo.  Cleaning that up left
check_and_free_str() as dead code (since using the qapi_free_*
takes care of recursion, and we don't want double frees).

A final leak was in test_visitor_out_any(), which was reassigning
the qobj local variable to a subset of the overall structure
needing freeing; it did not result in a use-after-free, but
was not cleaning up all the qdict.

test-qmp-event and test-qmp-commands were already clean.
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-6-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Rather than duplicate the body of two functions just to
decide between qobject_from_jsonv() and qobject_from_json(),
exploit the fact that qobject_from_jsonv() intentionally
takes 'va_list *' instead of the more common 'va_list', and
that qobject_from_json() just calls qobject_from_jsonv(,NULL).
For each file, our two existing init functions then become
thin wrappers around a new internal function, and future
updates to initialization don't have to be duplicated.
Suggested-by: NMarkus Armbruster <armbru@redhat.com>
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-5-git-send-email-eblake@redhat.com>
[Two old comment typos fixed]
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Adding an assertion to qobject_decref() will ensure that a
programming error causing use-after-free will result in
immediate failure (provided no other thread has started
using the memory) instead of silently attempting to wrap
refcnt around and leaving the problem to potentially bite
later at a harder point to diagnose.
Suggested-by: NMarkus Armbruster <armbru@redhat.com>
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-4-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>

Make each list element different, to ensure that order is
preserved, and use the generated free function instead of
hand-rolling our own to ensure (under valgrind) that the
list is properly cleaned.
Suggested-by: NMarkus Armbruster <armbru@redhat.com>
Signed-off-by: NEric Blake <eblake@redhat.com>
Message-Id: <1446791754-23823-3-git-send-email-eblake@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>